Clinical Review

Each year in the United States, more than 44 million young people participate in sports activities.¹ Yet the number of concussions incurred annually by children and adolescents engaged in sports and recreational play has been underestimated for years, and largely unknown.^1,2

Some estimates were based solely on the number of young athletes treated in emergency departments or sports concussion clinics. Others focused only on team players of middle school or high school age, excluding younger children who were hit in the head on playgrounds or during other recreational activities. What’s more, large numbers of concussions—as many as 4 in 10 incurred by high school athletes—were never reported to a coach or medical professional.³

In a new study published in the journal Pediatrics in June, researchers used national databases and current literature to provide what they believe to be “the most accurate and precise estimate of youth concussion” thus far: Between 1.1 and 1.9 million sports- and recreation-related concussions occur among US youth ages 18 or younger annually.¹

Standardized protocols for managing sport-related concussions have been adopted in most clinical settings. But use among primary care physicians is inconsistent.

Among young people playing team sports, concussions are between 2 and 7 times more likely to occur during competitive games than in practice sessions.^4-7 Boys on football and ice hockey teams have the highest rates of concussion in young athletes.For overall number of concussions, however, girls on soccer teams are second only to football players.⁴ Female soccer players are more likely than male soccer players to sustain concussions during equal number of hours of play.^4,7

An increase in incidence. The incidence of concussion among young athletes appears to have increased in the past decade, a likely result of greater involvement in team sports, an increasing focus on safeguarding young people from the potential dangers associated with a blow to the brain, and better diagnostic techniques.^4,8-10 And a recent study based on data from electronic medical records at a large regional pediatric health care network found that more than three-quarters of young people with sports-related concussions were first seen in a primary care setting.²

With this in mind, we present a comprehensive update of the evidence regarding the diagnosis and management of sport-related concussion. The recommendations we include are consistent with professional association guidelines.^8-10 Although we focus on concussion in children and adolescents involved in athletic activities, the principles generally apply to patients of all ages and to concussions that may not be sports related.

Removal from play: A vital first step

Whenever you conduct a physical exam for a young athlete, remind him or her—and the patient’s parents—that after a blow to the head, immediate removal from play is critical. Concussion is caused by a direct or indirect force to the brain that results in a transient disturbance in brain function,^8-10 manifested by alterations in neurocognitive and motor function. While the signs and symptoms (TABLE 1)^8-10 resolve within 10 days of injury in about 90% of cases, those who incur additional head impact within 24 hours have a higher symptom burden and prolonged recovery period.¹¹ Even without repetitive impact, younger athletes may take longer to recover.^8-10

The initial assessment

A child or adolescent who sustains a suspected concussion should be seen by a physician within 24 to 48 hours. Whether the initial assessment occurs in your office or on the sidelines of a game, it is important to confirm the time the incident occurred and the mechanism of injury.

Concussion is diagnosed by a combination of history, physical exam, and objective testing when symptoms or exam findings associated with mild brain trauma—headache, dizziness, light and/or noise sensitivity, among others—closely follow a head injury.^8-10 Certain maneuvers—assessing eye movements by asking the athlete to look in various directions, for instance, then to follow a pen or finger as you move it closer to his or her face—may provoke dizziness, headache, or other symptoms of concussion that were not apparent initially.

The differential diagnosis includes cervical musculoskeletal injury, craniofacial injury, epidural and subdural hematoma, heat-related illness, uncomplicated headache and migraine, upper respiratory infection, and vertigo.^8-10

Tools aid in diagnosis

Many clinical assessment tools exist to aid in the diagnosis of concussion (TABLE 2).^8-10,12-14 Any one of these tools, many of which use combinations of symptom checklists, balance exams, and cognitive assessments, may be included in your evaluation. No single tool has been found to be superior to any other.^8-10 A combination of tools may improve diagnostic accuracy, but assessment tools should not be the sole basis used to diagnose or rule out concussion.

Reserve neuroimaging, such as CT and MRI, for patients with more serious clinical findings or symptoms that persist longer than expected.

Any child or adolescent who had a blow to the head and at least one sign or symptom of concussion should be evaluated as soon as possible and assessed again later that day or the next day if any reason for concern remains.

Neuropsychological (NP) testing may involve computerized tests developed specifically for athletes. Patients may be required to react to objects that appear on a screen, for example, in a way that tests memory, performance, and reaction time. Because cognitive recovery often lags behind symptom resolution, NP testing may identify subtle brain deficits even in athletes who are asymptomatic at rest or with exercise. In general, NP testing has a sensitivity of 71% to 88% for athletes with concussion,¹⁰ but it is most beneficial when baseline test results are available. Interpretation of NP testing should be done only by qualified clinicians.

While NP testing may provide additional prognostic information, it should not alter the management of athletes who are symptomatic either at rest or with exercise.¹⁵ Nor is NP testing vital, as concussion can be accurately diagnosed and adequately managed without it.

Neuroimaging, including computed tomography (CT) and magnetic resonance imaging (MRI), is often used unnecessarily in the initial assessment of a patient who sustained a possible concussion.^8-10 In fact, neuroimaging should be reserved for cases in which it is necessary to rule out more serious pathology: intracranial or subdural hematoma or a craniofacial injury, for example, in patients with clinical findings that are red flags. These red flags include focal neurologic deficits, continuing nausea/vomiting, or persistent disorientation (TABLE 3),^8-10 or symptoms that worsen or persist beyond a few weeks. In such cases, further evaluation—with MRI of the brain, formal NP testing, and/or referral to a neurologist, physiatrist, or other physician who specializes in concussion care—is indicated.

Concussion management: Rest is key

While there is a dearth of high-quality studies on the management of sport-related concussion across all age groups, standardized protocols for both children and adults have been adopted in most clinical settings.^8-10,16,17 The protocols provide a framework for an individualized treatment plan. Yet their use among primary care physicians is inconsistent.^18-20

Traditionally, concussion management begins with relative physical and cognitive rest to allow the brain time to recover.^8-10 Recent randomized controlled trials have challenged this premise by suggesting that mild to moderate physical activity for post-concussion patients who are mildly symptomatic does not adversely affect recovery.^21,22 These studies have significant limitations, however, and further research is needed to provide specific guidance on this aspect of concussion management before it is adopted.

Physical restrictions include organized sports, recreational activity, recess, and physical education classes. Walking is permitted unless it exacerbates symptoms. These restrictions should continue until the patient is symptom-free.

Recent trials suggest that mild to moderate physical activity for mildly symptomatic post-concussion patients does not adversely affect recovery.

Cognitive restrictions include modifications at school and at home. Once an athlete is able to concentrate and tolerate visual and auditory stimuli, he or she may return to school. But classroom modifications should be considered, possibly including shortened school days, extra time for testing and homework, help with note taking, and restrictions from classes likely to provoke symptoms, such as computer science or music. Limiting use of mobile devices, television viewing, noisy environments, and other possible provocations may help speed symptom resolution. These restrictions, too, should remain in place until the patient is symptom-free.

Driving is often not addressed by physicians managing the care of athletes with concussion, but evidence suggests it should be. A study of patients presenting to the emergency department found that within 24 hours of a concussion diagnosis, individuals had an impaired response to traffic hazards.^23,24 And Canadian clinical practice guidelines recommend that athletes with mild traumatic brain injury (TBI) avoid driving within the first 24 hours.²⁵

While American guidelines are silent on the question of driving for this patient population, we recommend that athletes with concussion be restricted from driving and engaging in other risky complex tasks, such as welding or shop class, for at least 24 hours. For many athletes diagnosed with concussion, driving restrictions of longer duration may be necessary based on their symptom profile and neurocognitive test results. Continued dizziness or visual deficits would pose a greater risk than fatigue or short-term memory loss, for example.

Overseeing the return to play

Return-to-activity progression follows a stepwise protocol, with 6 steps that the injured athlete must complete before resuming full activity (FIGURE 1A).^8-10 This stepwise progression begins only when athletes are symptom free, even during provocative maneuvers; have had a normal neurologic exam, are back to school full time with no restriction; are off any medications prescribed for concussion symptoms (TABLE 4),^8-10 and when neurocognitive testing, if performed, is back to baseline. If an athlete develops symptoms at any stage of the progression, rest is required until he or she remains asymptomatic for at least 24 hours. The progression is then restarted at the last stage at which the patient was symptom free.

Some individualization, of course, is recommended here, too. Younger athletes and those with a prior history of concussion may require 10 days or more to complete all the steps, allowing an extra day at various steps. Neurologic maturation affects recovery time, and for younger individuals, a more conservative return-to-play protocol based on initial concussion symptom duration has been proposed (FIGURE 1B).¹⁶

Return to activity is often supervised by a certified athletic trainer at the athlete’s school. In the event that no athletic trainer is available, patients may be referred to physical therapists with experience in monitoring injured athletes.²⁶ Anyone involved in the patient’s care, including the athlete himself, may use a symptom checklist to monitor recovery.

Allowing asymptomatic athletes to engage in non-contact sports activity less than 7 to 10 days after concussion can help them avoid injury when they are cleared for full play.

Although there is no evidence that the ongoing use of a symptom checklist affects the course of recovery, its use is often helpful in identifying specific symptoms that can be managed by means other than physical and cognitive rest—a sleep hygiene program for an individual with lingering difficulty sleeping, for example, or the continued application of ice, heat, and massage for persistent neck pain.

Checklist monitoring may be especially helpful for athletes whose symptoms extend beyond 10 days or who have multiple symptoms. Final clearance once all the steps have been completed requires follow-up with a health care provider.

Is a symptom-free waiting period necessary?

There is no evidence suggesting a need for a symptom-free waiting period before starting the return-to-play protocol.^10,27 Because a repeat concussion is most likely within 7 to 10 days of the initial injury,^8,9 however, most athletes should not return to contact play during that time frame, regardless of symptom resolution.

It is helpful to have asymptomatic athletes participate in non-contact activity before the 7 to 10 days are up, however. Doing so can help prevent deconditioning and injury upon return to contact sport, as there is evidence of increased risk of lower-extremity injury in the 90 days after concussion.²⁸

What to tell athletes—and parents—about repetitive head trauma

There is growing concern about the long-term risks of concussion and repetitive head impact that may manifest as chronic traumatic encephalopathy (CTE) and chronic neurocognitive impairment (CNI) later in life. Indeed, some data strongly suggest—but do not definitively prove—a relationship between repetitive head injury and chronic neurodegenerative disease.^8-10 You can tell worried patients or parents, however, that the majority of research on CTE and CNI has been based on professional football players.

Studies of long-term effects of soccer heading have shown conflicting results, with some finding cognitive impairment, altered postural control, and anatomic changes of the brain, while others found no effect on encephalopathy, concussion symptoms, or neurocognitive performance.^29-36Here, too, most studies showing negative effects of soccer heading involved professional athletes.

Repetitive sub-concussive impact in high school football athletes has been found to induce biochemical changes to the brain,³⁷ but the long-term effects are unknown. And, while concussion in high school athletes has been associated with short-term cognitive impairment, altered neurochemistry, and evidence of increased symptoms on baseline neurocognitive testing,^8-10,38 no studies have linked concussion during middle school or high school with CNI. What’s more, a long-term (50-year) follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease compared with age-matched controls.³⁹

A 50-year follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease when compared with age-matched controls.

A new study of high school and college football players (mean age: 17.4 years) presented at the American Academy of Neurology 2016 Sports Concussion Conference in Chicago in July, however, found significant alterations in white matter 6 months post injury.⁴⁰ The researchers compared 17 athletes with sport-related concussion with matched controls, using diffusion tensor imaging and diffusion kurtosis tensor imaging as biomarkers of brain recovery. The concussed athletes underwent MRI and symptom assessment at 24 hours, 8 days, and 6 months. The controls followed identical protocols.

At the 6-month assessment, there were no differences between the concussed group and the controls in terms of self-reported concussion symptoms, cognition, or balance. However, the concussed athletes had widespread decreased mean diffusivity compared with the controls. Despite the lack of clinical symptoms, the concussed athletes showed significant alterations in white matter “that were related to initial symptom severity ratings,” the authors concluded. These findings have implications both for determination of recovery from concussion and concussion management, they added.⁴⁰

Although there is no way to eliminate all concussions, limited evidence suggests that improving athletic technique, limiting contact at practices, better enforcement of game rules, and rule changes regarding physical contact may decrease concussion risk.^41-43 Many youth sports organizations have developed policies placing restrictions on head impact during practices and games. Studies are ongoing, too, to see if better headgear—or requiring helmets for soccer players—makes a difference.

CORRESPONDENCE
Ryan A. Sprouse, MD, CAQSM, 203 East Fourth Avenue, Ranson, WV 25438; [email protected].

Each year in the United States, more than 44 million young people participate in sports activities.¹ Yet the number of concussions incurred annually by children and adolescents engaged in sports and recreational play has been underestimated for years, and largely unknown.^1,2

Some estimates were based solely on the number of young athletes treated in emergency departments or sports concussion clinics. Others focused only on team players of middle school or high school age, excluding younger children who were hit in the head on playgrounds or during other recreational activities. What’s more, large numbers of concussions—as many as 4 in 10 incurred by high school athletes—were never reported to a coach or medical professional.³

In a new study published in the journal Pediatrics in June, researchers used national databases and current literature to provide what they believe to be “the most accurate and precise estimate of youth concussion” thus far: Between 1.1 and 1.9 million sports- and recreation-related concussions occur among US youth ages 18 or younger annually.¹

Standardized protocols for managing sport-related concussions have been adopted in most clinical settings. But use among primary care physicians is inconsistent.

Among young people playing team sports, concussions are between 2 and 7 times more likely to occur during competitive games than in practice sessions.^4-7 Boys on football and ice hockey teams have the highest rates of concussion in young athletes.For overall number of concussions, however, girls on soccer teams are second only to football players.⁴ Female soccer players are more likely than male soccer players to sustain concussions during equal number of hours of play.^4,7

An increase in incidence. The incidence of concussion among young athletes appears to have increased in the past decade, a likely result of greater involvement in team sports, an increasing focus on safeguarding young people from the potential dangers associated with a blow to the brain, and better diagnostic techniques.^4,8-10 And a recent study based on data from electronic medical records at a large regional pediatric health care network found that more than three-quarters of young people with sports-related concussions were first seen in a primary care setting.²

With this in mind, we present a comprehensive update of the evidence regarding the diagnosis and management of sport-related concussion. The recommendations we include are consistent with professional association guidelines.^8-10 Although we focus on concussion in children and adolescents involved in athletic activities, the principles generally apply to patients of all ages and to concussions that may not be sports related.

Removal from play: A vital first step

Whenever you conduct a physical exam for a young athlete, remind him or her—and the patient’s parents—that after a blow to the head, immediate removal from play is critical. Concussion is caused by a direct or indirect force to the brain that results in a transient disturbance in brain function,^8-10 manifested by alterations in neurocognitive and motor function. While the signs and symptoms (TABLE 1)^8-10 resolve within 10 days of injury in about 90% of cases, those who incur additional head impact within 24 hours have a higher symptom burden and prolonged recovery period.¹¹ Even without repetitive impact, younger athletes may take longer to recover.^8-10

The initial assessment

A child or adolescent who sustains a suspected concussion should be seen by a physician within 24 to 48 hours. Whether the initial assessment occurs in your office or on the sidelines of a game, it is important to confirm the time the incident occurred and the mechanism of injury.

Concussion is diagnosed by a combination of history, physical exam, and objective testing when symptoms or exam findings associated with mild brain trauma—headache, dizziness, light and/or noise sensitivity, among others—closely follow a head injury.^8-10 Certain maneuvers—assessing eye movements by asking the athlete to look in various directions, for instance, then to follow a pen or finger as you move it closer to his or her face—may provoke dizziness, headache, or other symptoms of concussion that were not apparent initially.

The differential diagnosis includes cervical musculoskeletal injury, craniofacial injury, epidural and subdural hematoma, heat-related illness, uncomplicated headache and migraine, upper respiratory infection, and vertigo.^8-10

Tools aid in diagnosis

Many clinical assessment tools exist to aid in the diagnosis of concussion (TABLE 2).^8-10,12-14 Any one of these tools, many of which use combinations of symptom checklists, balance exams, and cognitive assessments, may be included in your evaluation. No single tool has been found to be superior to any other.^8-10 A combination of tools may improve diagnostic accuracy, but assessment tools should not be the sole basis used to diagnose or rule out concussion.

Reserve neuroimaging, such as CT and MRI, for patients with more serious clinical findings or symptoms that persist longer than expected.

Any child or adolescent who had a blow to the head and at least one sign or symptom of concussion should be evaluated as soon as possible and assessed again later that day or the next day if any reason for concern remains.

Neuropsychological (NP) testing may involve computerized tests developed specifically for athletes. Patients may be required to react to objects that appear on a screen, for example, in a way that tests memory, performance, and reaction time. Because cognitive recovery often lags behind symptom resolution, NP testing may identify subtle brain deficits even in athletes who are asymptomatic at rest or with exercise. In general, NP testing has a sensitivity of 71% to 88% for athletes with concussion,¹⁰ but it is most beneficial when baseline test results are available. Interpretation of NP testing should be done only by qualified clinicians.

While NP testing may provide additional prognostic information, it should not alter the management of athletes who are symptomatic either at rest or with exercise.¹⁵ Nor is NP testing vital, as concussion can be accurately diagnosed and adequately managed without it.

Neuroimaging, including computed tomography (CT) and magnetic resonance imaging (MRI), is often used unnecessarily in the initial assessment of a patient who sustained a possible concussion.^8-10 In fact, neuroimaging should be reserved for cases in which it is necessary to rule out more serious pathology: intracranial or subdural hematoma or a craniofacial injury, for example, in patients with clinical findings that are red flags. These red flags include focal neurologic deficits, continuing nausea/vomiting, or persistent disorientation (TABLE 3),^8-10 or symptoms that worsen or persist beyond a few weeks. In such cases, further evaluation—with MRI of the brain, formal NP testing, and/or referral to a neurologist, physiatrist, or other physician who specializes in concussion care—is indicated.

Concussion management: Rest is key

While there is a dearth of high-quality studies on the management of sport-related concussion across all age groups, standardized protocols for both children and adults have been adopted in most clinical settings.^8-10,16,17 The protocols provide a framework for an individualized treatment plan. Yet their use among primary care physicians is inconsistent.^18-20

Traditionally, concussion management begins with relative physical and cognitive rest to allow the brain time to recover.^8-10 Recent randomized controlled trials have challenged this premise by suggesting that mild to moderate physical activity for post-concussion patients who are mildly symptomatic does not adversely affect recovery.^21,22 These studies have significant limitations, however, and further research is needed to provide specific guidance on this aspect of concussion management before it is adopted.

Physical restrictions include organized sports, recreational activity, recess, and physical education classes. Walking is permitted unless it exacerbates symptoms. These restrictions should continue until the patient is symptom-free.

Recent trials suggest that mild to moderate physical activity for mildly symptomatic post-concussion patients does not adversely affect recovery.

Cognitive restrictions include modifications at school and at home. Once an athlete is able to concentrate and tolerate visual and auditory stimuli, he or she may return to school. But classroom modifications should be considered, possibly including shortened school days, extra time for testing and homework, help with note taking, and restrictions from classes likely to provoke symptoms, such as computer science or music. Limiting use of mobile devices, television viewing, noisy environments, and other possible provocations may help speed symptom resolution. These restrictions, too, should remain in place until the patient is symptom-free.

Driving is often not addressed by physicians managing the care of athletes with concussion, but evidence suggests it should be. A study of patients presenting to the emergency department found that within 24 hours of a concussion diagnosis, individuals had an impaired response to traffic hazards.^23,24 And Canadian clinical practice guidelines recommend that athletes with mild traumatic brain injury (TBI) avoid driving within the first 24 hours.²⁵

While American guidelines are silent on the question of driving for this patient population, we recommend that athletes with concussion be restricted from driving and engaging in other risky complex tasks, such as welding or shop class, for at least 24 hours. For many athletes diagnosed with concussion, driving restrictions of longer duration may be necessary based on their symptom profile and neurocognitive test results. Continued dizziness or visual deficits would pose a greater risk than fatigue or short-term memory loss, for example.

Overseeing the return to play

Return-to-activity progression follows a stepwise protocol, with 6 steps that the injured athlete must complete before resuming full activity (FIGURE 1A).^8-10 This stepwise progression begins only when athletes are symptom free, even during provocative maneuvers; have had a normal neurologic exam, are back to school full time with no restriction; are off any medications prescribed for concussion symptoms (TABLE 4),^8-10 and when neurocognitive testing, if performed, is back to baseline. If an athlete develops symptoms at any stage of the progression, rest is required until he or she remains asymptomatic for at least 24 hours. The progression is then restarted at the last stage at which the patient was symptom free.

Some individualization, of course, is recommended here, too. Younger athletes and those with a prior history of concussion may require 10 days or more to complete all the steps, allowing an extra day at various steps. Neurologic maturation affects recovery time, and for younger individuals, a more conservative return-to-play protocol based on initial concussion symptom duration has been proposed (FIGURE 1B).¹⁶

Return to activity is often supervised by a certified athletic trainer at the athlete’s school. In the event that no athletic trainer is available, patients may be referred to physical therapists with experience in monitoring injured athletes.²⁶ Anyone involved in the patient’s care, including the athlete himself, may use a symptom checklist to monitor recovery.

Allowing asymptomatic athletes to engage in non-contact sports activity less than 7 to 10 days after concussion can help them avoid injury when they are cleared for full play.

Although there is no evidence that the ongoing use of a symptom checklist affects the course of recovery, its use is often helpful in identifying specific symptoms that can be managed by means other than physical and cognitive rest—a sleep hygiene program for an individual with lingering difficulty sleeping, for example, or the continued application of ice, heat, and massage for persistent neck pain.

Checklist monitoring may be especially helpful for athletes whose symptoms extend beyond 10 days or who have multiple symptoms. Final clearance once all the steps have been completed requires follow-up with a health care provider.

Is a symptom-free waiting period necessary?

There is no evidence suggesting a need for a symptom-free waiting period before starting the return-to-play protocol.^10,27 Because a repeat concussion is most likely within 7 to 10 days of the initial injury,^8,9 however, most athletes should not return to contact play during that time frame, regardless of symptom resolution.

It is helpful to have asymptomatic athletes participate in non-contact activity before the 7 to 10 days are up, however. Doing so can help prevent deconditioning and injury upon return to contact sport, as there is evidence of increased risk of lower-extremity injury in the 90 days after concussion.²⁸

What to tell athletes—and parents—about repetitive head trauma

There is growing concern about the long-term risks of concussion and repetitive head impact that may manifest as chronic traumatic encephalopathy (CTE) and chronic neurocognitive impairment (CNI) later in life. Indeed, some data strongly suggest—but do not definitively prove—a relationship between repetitive head injury and chronic neurodegenerative disease.^8-10 You can tell worried patients or parents, however, that the majority of research on CTE and CNI has been based on professional football players.

Studies of long-term effects of soccer heading have shown conflicting results, with some finding cognitive impairment, altered postural control, and anatomic changes of the brain, while others found no effect on encephalopathy, concussion symptoms, or neurocognitive performance.^29-36Here, too, most studies showing negative effects of soccer heading involved professional athletes.

Repetitive sub-concussive impact in high school football athletes has been found to induce biochemical changes to the brain,³⁷ but the long-term effects are unknown. And, while concussion in high school athletes has been associated with short-term cognitive impairment, altered neurochemistry, and evidence of increased symptoms on baseline neurocognitive testing,^8-10,38 no studies have linked concussion during middle school or high school with CNI. What’s more, a long-term (50-year) follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease compared with age-matched controls.³⁹

A 50-year follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease when compared with age-matched controls.

A new study of high school and college football players (mean age: 17.4 years) presented at the American Academy of Neurology 2016 Sports Concussion Conference in Chicago in July, however, found significant alterations in white matter 6 months post injury.⁴⁰ The researchers compared 17 athletes with sport-related concussion with matched controls, using diffusion tensor imaging and diffusion kurtosis tensor imaging as biomarkers of brain recovery. The concussed athletes underwent MRI and symptom assessment at 24 hours, 8 days, and 6 months. The controls followed identical protocols.

At the 6-month assessment, there were no differences between the concussed group and the controls in terms of self-reported concussion symptoms, cognition, or balance. However, the concussed athletes had widespread decreased mean diffusivity compared with the controls. Despite the lack of clinical symptoms, the concussed athletes showed significant alterations in white matter “that were related to initial symptom severity ratings,” the authors concluded. These findings have implications both for determination of recovery from concussion and concussion management, they added.⁴⁰

Although there is no way to eliminate all concussions, limited evidence suggests that improving athletic technique, limiting contact at practices, better enforcement of game rules, and rule changes regarding physical contact may decrease concussion risk.^41-43 Many youth sports organizations have developed policies placing restrictions on head impact during practices and games. Studies are ongoing, too, to see if better headgear—or requiring helmets for soccer players—makes a difference.

CORRESPONDENCE
Ryan A. Sprouse, MD, CAQSM, 203 East Fourth Avenue, Ranson, WV 25438; [email protected].

Each year in the United States, more than 44 million young people participate in sports activities.¹ Yet the number of concussions incurred annually by children and adolescents engaged in sports and recreational play has been underestimated for years, and largely unknown.^1,2

Some estimates were based solely on the number of young athletes treated in emergency departments or sports concussion clinics. Others focused only on team players of middle school or high school age, excluding younger children who were hit in the head on playgrounds or during other recreational activities. What’s more, large numbers of concussions—as many as 4 in 10 incurred by high school athletes—were never reported to a coach or medical professional.³

In a new study published in the journal Pediatrics in June, researchers used national databases and current literature to provide what they believe to be “the most accurate and precise estimate of youth concussion” thus far: Between 1.1 and 1.9 million sports- and recreation-related concussions occur among US youth ages 18 or younger annually.¹

Standardized protocols for managing sport-related concussions have been adopted in most clinical settings. But use among primary care physicians is inconsistent.

Among young people playing team sports, concussions are between 2 and 7 times more likely to occur during competitive games than in practice sessions.^4-7 Boys on football and ice hockey teams have the highest rates of concussion in young athletes.For overall number of concussions, however, girls on soccer teams are second only to football players.⁴ Female soccer players are more likely than male soccer players to sustain concussions during equal number of hours of play.^4,7

An increase in incidence. The incidence of concussion among young athletes appears to have increased in the past decade, a likely result of greater involvement in team sports, an increasing focus on safeguarding young people from the potential dangers associated with a blow to the brain, and better diagnostic techniques.^4,8-10 And a recent study based on data from electronic medical records at a large regional pediatric health care network found that more than three-quarters of young people with sports-related concussions were first seen in a primary care setting.²

With this in mind, we present a comprehensive update of the evidence regarding the diagnosis and management of sport-related concussion. The recommendations we include are consistent with professional association guidelines.^8-10 Although we focus on concussion in children and adolescents involved in athletic activities, the principles generally apply to patients of all ages and to concussions that may not be sports related.

Removal from play: A vital first step

Whenever you conduct a physical exam for a young athlete, remind him or her—and the patient’s parents—that after a blow to the head, immediate removal from play is critical. Concussion is caused by a direct or indirect force to the brain that results in a transient disturbance in brain function,^8-10 manifested by alterations in neurocognitive and motor function. While the signs and symptoms (TABLE 1)^8-10 resolve within 10 days of injury in about 90% of cases, those who incur additional head impact within 24 hours have a higher symptom burden and prolonged recovery period.¹¹ Even without repetitive impact, younger athletes may take longer to recover.^8-10

The initial assessment

A child or adolescent who sustains a suspected concussion should be seen by a physician within 24 to 48 hours. Whether the initial assessment occurs in your office or on the sidelines of a game, it is important to confirm the time the incident occurred and the mechanism of injury.

Concussion is diagnosed by a combination of history, physical exam, and objective testing when symptoms or exam findings associated with mild brain trauma—headache, dizziness, light and/or noise sensitivity, among others—closely follow a head injury.^8-10 Certain maneuvers—assessing eye movements by asking the athlete to look in various directions, for instance, then to follow a pen or finger as you move it closer to his or her face—may provoke dizziness, headache, or other symptoms of concussion that were not apparent initially.

The differential diagnosis includes cervical musculoskeletal injury, craniofacial injury, epidural and subdural hematoma, heat-related illness, uncomplicated headache and migraine, upper respiratory infection, and vertigo.^8-10

Tools aid in diagnosis

Many clinical assessment tools exist to aid in the diagnosis of concussion (TABLE 2).^8-10,12-14 Any one of these tools, many of which use combinations of symptom checklists, balance exams, and cognitive assessments, may be included in your evaluation. No single tool has been found to be superior to any other.^8-10 A combination of tools may improve diagnostic accuracy, but assessment tools should not be the sole basis used to diagnose or rule out concussion.

Reserve neuroimaging, such as CT and MRI, for patients with more serious clinical findings or symptoms that persist longer than expected.

Any child or adolescent who had a blow to the head and at least one sign or symptom of concussion should be evaluated as soon as possible and assessed again later that day or the next day if any reason for concern remains.

Neuropsychological (NP) testing may involve computerized tests developed specifically for athletes. Patients may be required to react to objects that appear on a screen, for example, in a way that tests memory, performance, and reaction time. Because cognitive recovery often lags behind symptom resolution, NP testing may identify subtle brain deficits even in athletes who are asymptomatic at rest or with exercise. In general, NP testing has a sensitivity of 71% to 88% for athletes with concussion,¹⁰ but it is most beneficial when baseline test results are available. Interpretation of NP testing should be done only by qualified clinicians.

While NP testing may provide additional prognostic information, it should not alter the management of athletes who are symptomatic either at rest or with exercise.¹⁵ Nor is NP testing vital, as concussion can be accurately diagnosed and adequately managed without it.

Neuroimaging, including computed tomography (CT) and magnetic resonance imaging (MRI), is often used unnecessarily in the initial assessment of a patient who sustained a possible concussion.^8-10 In fact, neuroimaging should be reserved for cases in which it is necessary to rule out more serious pathology: intracranial or subdural hematoma or a craniofacial injury, for example, in patients with clinical findings that are red flags. These red flags include focal neurologic deficits, continuing nausea/vomiting, or persistent disorientation (TABLE 3),^8-10 or symptoms that worsen or persist beyond a few weeks. In such cases, further evaluation—with MRI of the brain, formal NP testing, and/or referral to a neurologist, physiatrist, or other physician who specializes in concussion care—is indicated.

Concussion management: Rest is key

While there is a dearth of high-quality studies on the management of sport-related concussion across all age groups, standardized protocols for both children and adults have been adopted in most clinical settings.^8-10,16,17 The protocols provide a framework for an individualized treatment plan. Yet their use among primary care physicians is inconsistent.^18-20

Traditionally, concussion management begins with relative physical and cognitive rest to allow the brain time to recover.^8-10 Recent randomized controlled trials have challenged this premise by suggesting that mild to moderate physical activity for post-concussion patients who are mildly symptomatic does not adversely affect recovery.^21,22 These studies have significant limitations, however, and further research is needed to provide specific guidance on this aspect of concussion management before it is adopted.

Physical restrictions include organized sports, recreational activity, recess, and physical education classes. Walking is permitted unless it exacerbates symptoms. These restrictions should continue until the patient is symptom-free.

Recent trials suggest that mild to moderate physical activity for mildly symptomatic post-concussion patients does not adversely affect recovery.

Cognitive restrictions include modifications at school and at home. Once an athlete is able to concentrate and tolerate visual and auditory stimuli, he or she may return to school. But classroom modifications should be considered, possibly including shortened school days, extra time for testing and homework, help with note taking, and restrictions from classes likely to provoke symptoms, such as computer science or music. Limiting use of mobile devices, television viewing, noisy environments, and other possible provocations may help speed symptom resolution. These restrictions, too, should remain in place until the patient is symptom-free.

Driving is often not addressed by physicians managing the care of athletes with concussion, but evidence suggests it should be. A study of patients presenting to the emergency department found that within 24 hours of a concussion diagnosis, individuals had an impaired response to traffic hazards.^23,24 And Canadian clinical practice guidelines recommend that athletes with mild traumatic brain injury (TBI) avoid driving within the first 24 hours.²⁵

While American guidelines are silent on the question of driving for this patient population, we recommend that athletes with concussion be restricted from driving and engaging in other risky complex tasks, such as welding or shop class, for at least 24 hours. For many athletes diagnosed with concussion, driving restrictions of longer duration may be necessary based on their symptom profile and neurocognitive test results. Continued dizziness or visual deficits would pose a greater risk than fatigue or short-term memory loss, for example.

Overseeing the return to play

Return-to-activity progression follows a stepwise protocol, with 6 steps that the injured athlete must complete before resuming full activity (FIGURE 1A).^8-10 This stepwise progression begins only when athletes are symptom free, even during provocative maneuvers; have had a normal neurologic exam, are back to school full time with no restriction; are off any medications prescribed for concussion symptoms (TABLE 4),^8-10 and when neurocognitive testing, if performed, is back to baseline. If an athlete develops symptoms at any stage of the progression, rest is required until he or she remains asymptomatic for at least 24 hours. The progression is then restarted at the last stage at which the patient was symptom free.

Some individualization, of course, is recommended here, too. Younger athletes and those with a prior history of concussion may require 10 days or more to complete all the steps, allowing an extra day at various steps. Neurologic maturation affects recovery time, and for younger individuals, a more conservative return-to-play protocol based on initial concussion symptom duration has been proposed (FIGURE 1B).¹⁶

Return to activity is often supervised by a certified athletic trainer at the athlete’s school. In the event that no athletic trainer is available, patients may be referred to physical therapists with experience in monitoring injured athletes.²⁶ Anyone involved in the patient’s care, including the athlete himself, may use a symptom checklist to monitor recovery.

Allowing asymptomatic athletes to engage in non-contact sports activity less than 7 to 10 days after concussion can help them avoid injury when they are cleared for full play.

Although there is no evidence that the ongoing use of a symptom checklist affects the course of recovery, its use is often helpful in identifying specific symptoms that can be managed by means other than physical and cognitive rest—a sleep hygiene program for an individual with lingering difficulty sleeping, for example, or the continued application of ice, heat, and massage for persistent neck pain.

Checklist monitoring may be especially helpful for athletes whose symptoms extend beyond 10 days or who have multiple symptoms. Final clearance once all the steps have been completed requires follow-up with a health care provider.

Is a symptom-free waiting period necessary?

There is no evidence suggesting a need for a symptom-free waiting period before starting the return-to-play protocol.^10,27 Because a repeat concussion is most likely within 7 to 10 days of the initial injury,^8,9 however, most athletes should not return to contact play during that time frame, regardless of symptom resolution.

It is helpful to have asymptomatic athletes participate in non-contact activity before the 7 to 10 days are up, however. Doing so can help prevent deconditioning and injury upon return to contact sport, as there is evidence of increased risk of lower-extremity injury in the 90 days after concussion.²⁸

What to tell athletes—and parents—about repetitive head trauma

There is growing concern about the long-term risks of concussion and repetitive head impact that may manifest as chronic traumatic encephalopathy (CTE) and chronic neurocognitive impairment (CNI) later in life. Indeed, some data strongly suggest—but do not definitively prove—a relationship between repetitive head injury and chronic neurodegenerative disease.^8-10 You can tell worried patients or parents, however, that the majority of research on CTE and CNI has been based on professional football players.

Studies of long-term effects of soccer heading have shown conflicting results, with some finding cognitive impairment, altered postural control, and anatomic changes of the brain, while others found no effect on encephalopathy, concussion symptoms, or neurocognitive performance.^29-36Here, too, most studies showing negative effects of soccer heading involved professional athletes.

Repetitive sub-concussive impact in high school football athletes has been found to induce biochemical changes to the brain,³⁷ but the long-term effects are unknown. And, while concussion in high school athletes has been associated with short-term cognitive impairment, altered neurochemistry, and evidence of increased symptoms on baseline neurocognitive testing,^8-10,38 no studies have linked concussion during middle school or high school with CNI. What’s more, a long-term (50-year) follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease compared with age-matched controls.³⁹

A 50-year follow-up study of individuals who played football in high school found no difference in rates of neurodegenerative disease when compared with age-matched controls.

A new study of high school and college football players (mean age: 17.4 years) presented at the American Academy of Neurology 2016 Sports Concussion Conference in Chicago in July, however, found significant alterations in white matter 6 months post injury.⁴⁰ The researchers compared 17 athletes with sport-related concussion with matched controls, using diffusion tensor imaging and diffusion kurtosis tensor imaging as biomarkers of brain recovery. The concussed athletes underwent MRI and symptom assessment at 24 hours, 8 days, and 6 months. The controls followed identical protocols.

At the 6-month assessment, there were no differences between the concussed group and the controls in terms of self-reported concussion symptoms, cognition, or balance. However, the concussed athletes had widespread decreased mean diffusivity compared with the controls. Despite the lack of clinical symptoms, the concussed athletes showed significant alterations in white matter “that were related to initial symptom severity ratings,” the authors concluded. These findings have implications both for determination of recovery from concussion and concussion management, they added.⁴⁰

Although there is no way to eliminate all concussions, limited evidence suggests that improving athletic technique, limiting contact at practices, better enforcement of game rules, and rule changes regarding physical contact may decrease concussion risk.^41-43 Many youth sports organizations have developed policies placing restrictions on head impact during practices and games. Studies are ongoing, too, to see if better headgear—or requiring helmets for soccer players—makes a difference.

CORRESPONDENCE
Ryan A. Sprouse, MD, CAQSM, 203 East Fourth Avenue, Ranson, WV 25438; [email protected].

IN THIS ARTICLE

• Adverse effects of ciprofloxacin
• Symptoms of common tick-borne diseases
• Symptoms of phase 1 and late-phase disease
• Additional resources

Jane, an 18-year-old college student, presents in early November with a three-week history of worsening cough and sinus congestion. Recently, the cough has been interrupting her sleep and yellow-green nasal drainage and sinus pressure have increased. Ordinarily very fit and athletic, she reports that since she arrived at college two months ago, her body has become “more fragile.”

Further questioning reveals that, over the past two months, the patient’s symptoms have included extreme fatigue, severe unremitting headache, blurred vision, shortness of breath, and a racing heart rate on exertion. Her symptoms make it impossible for her to maintain her demanding exercise routine, a development that compounds her frustration and sadness. She has also been forced to limit her participation in school activities, with significant academic decline as a result.

Aside from depression (well controlled with bupropion HCl extended release, 300 mg/d), Jane’s medical history is unremarkable. She reports having “excellent health” until she arrived at her mid-Atlantic urban college.

A complicated history
Born and raised in Connecticut, Jane is an avid runner who competes in extreme sports. This past summer, she trained for and participated in two “mud run” events (ie, endurance races of several miles with numerous challenges and obstacles) in Connecticut and New York. Training included endurance runs and sprints, as well as crawling through mud-laden fields and woods.

She also did a three-week summer internship on an oyster farm. There, she was required to shuck oysters and stand in brackish water for six-hour shifts to examine oyster beds. In the process, she sustained numerous cuts and bruises on her hands, arms, and legs.

A week or so after returning to college in late August, Jane developed blisters on both heels, which progressed to infected ulcerations. She was evaluated at the university hospital emergency department (ED) and treated with a 21-day course of ciprofloxacin. When left-sided unilateral knee swelling developed about two weeks later, she underwent arthrocentesis at the university health center, but joint aspirate was not sent for analysis. A two-week course of antibiotic therapy was initiated.

From October to her presentation in early November, Jane has experienced intermittent fevers and chills, with a temperature as high as 101°F. In addition, she complains of fasciculations and weakness in her lower limbs; dyspnea, tachycardia, and dizziness during or after any exertion; unremitting posterior neck pain; and a constant, severe headache located primarily in the bitemporal region. She developed bilateral conjunctivitis, which resolved spontaneously in about one week; persistent blurred vision; a transient petechial chest rash; recurring episodes of syncope; pyelonephritis; a persistent vaginal yeast infection; decreased appetite; and a 7-lb weight loss (5% of her total body weight).

Jane’s academic and athletic performance has been severely impaired. Once a long-distance runner, she can no longer walk any distance without frequent rest. In the four months since the mud runs, the patient reports, she has been seen in the student health center four times and in the ED twice. Additionally, she has undergone thorough examinations by clinicians specializing in infectious disease, pulmonology, neurology, and neuro-ophthalmology. She has undergone lab work, including
• Complete blood cell count with differential
• Comprehensive metabolic panel
• Urinalysis and urine culture
• Lyme antibody and blood polymerase chain reaction (PCR)
• HIV testing
• Rheumatoid factor
• Erythrocyte sedimentation rate (ESR)
• C-reactive protein (CRP)
• Epstein-Barr virus IgM
• Cytomegalovirus (CMV) IgM
• Human granulocytic ehrlichiosis (HGE) antibody and human anaplasma phagocytophilum (HGA)
• HGA PCR
• Rickettsia antibody panel
• Babesia microti antibodies
• Pregnancy testing
• Chest x-ray
• Lumbar puncture

All lab results were within normal range. In light of this, several clinicians have told Jane that her illness is “all in her head.”

Continue for the patient investigates >>

The patient investigates
In mid-December, after she has returned home from college, Jane’s symptoms abruptly worsen. She complains of feeling “shakier,” with weakness in her legs and what she calls “brain fog.” Her headache, blurred vision, and dizziness have worsened. Frightened and concerned, she returns to the ED. Results of a thorough evaluation, including lumbar puncture, reveal no abnormality.

Jane has become extremely frail. She is losing weight, her hair has lost its luster, and her nails are cracking and bleeding. She is unable to walk without concern for falling and cannot climb the 20 steps to her bedroom. Once a healthy and vibrant 18-year-old, she now spends most of her time in a lethargic state on a first-floor living room couch.

Frustrated by her unexplained declining health, she begins to research illnesses associated with extreme sports and prolonged marine exposure. She returns to ask about three possible explanations for her condition:
1. Adverse effects of ciprofloxacin use, which include fever or chills, dizziness, racing heartbeat, headache, and nausea.¹
2. A tick-borne disease, possibly contracted during her practice runs in the Connecticut woods (see Table 1).^2-4 Each year, she recalls, she has found and removed four or five embedded ticks. In the northeastern United States, the most common tick-borne diseases are borreliosis, babesiosis, and ehrlichiosis.^5-7
3. Leptospirosis, contracted through the patient’s exposure to mud and brackish water during her summer activities. According to her research, more than 25 outbreaks and 600 cases of leptospirosis (between 1931 and 1998) have been associated with fresh pond, creek, or river water.⁸

Based on Jane’s symptoms and history, and in accord with her research, early-phase leptospirosis is identified as a diagnosis of exclusion (with a possible comorbid tick-borne zoonosis).

Continue for discussion >>

DISCUSSION
Leptospirosis develops when humans come into contact with animal urine infected by leptospires—that is, pathogenic spirochetes excreted via the renal tubules of infected host animals.^9,10 While host animals include dogs, pigs, cattle, reptiles, and amphibians, the animal most commonly associated with human infection is the brown rat (Rattus norvegicus).^11-15

Leptospires enter the human host through mucous membranes, cuts, or abrasions in the skin. Individuals at increased risk for infection include those whose work or other activities expose them “to animal reservoirs or contaminated environments”—including participants in water sports and similar recreation.^11-14 As Mwachui et al explain, “recreational exposure to [Leptospira-]contaminated water has become more important for sport enthusiasts, swimmers and travellers from industrialized countries,” whereas flooding is usually involved in infection in undeveloped countries.¹⁶

The largest outbreak of leptospirosis reported in the US to date occurred in 1998, when heavy rains preceded a triathlon in Springfield, Illinois. When many participants became ill after the event, researchers from the National Center for Infectious Diseases were able to contact and test 834 of the 876 competing athletes; of these, 98 (12%) reported being ill and 52 (11%) tested positive for leptospirosis. Additionally, 14 of the 248 community residents who were sickened (6%) tested positive.¹⁷ According to CDC estimates, between 100 and 200 cases of leptospirosis develop annually in the US, with about half occurring in Hawaii.⁹

Onset of symptoms, which are described as protean and nonspecific, occurs two days to four weeks after exposure, making leptospirosis difficult to diagnosewithout a high degree of suspicion; zoonotic exposure (as with freshwater or mud sports) or a history of travel to Hawaii, Tahiti, Thailand, Indonesia, the Caribbean, and/or Costa Rica may raise suspicion.^12-14,18 In early-phase leptospirosis, symptoms can mimic those of influenza, meningitis, malaria, dengue fever, scrub typhus, rickettsial disease, and typhoid fever (see Table 2).¹⁰ Thus, when a patient presents with these symptoms, it is imperative that the clinician consider leptospirosis.¹⁹Of note: Flu-like symptoms with conjunctival suffusion are considered pathognomonic for leptospirosis.¹⁸

About 10% of patients with early-phase leptospirosis will develop late-phase disease (ie, Weil’s disease), with severe symptoms that include jaundice, meningitis, pulmonary hemorrhage, and acute kidney injury (see Table 3 for a more detailed list).²⁰ The case patient’s history and symptoms were consistent with a diagnosis of early-phase lepto­spirosis.

Epidemiology
In 2015, leptospirosis was estimated to affect more than 1 million persons worldwide, with 58,900 deaths attributed to the disease each year—making leptospirosis the leading cause of death attributable to zoonotic illness.¹¹ Historically, leptospirosis-associated morbidity and mortality have been greatest in resource-poor countries with tropical climates (eg, southern and Southeast Asia, Central America and tropical Latin America, and East Sub-Saharan Africa).^11,12

However, illness resulting from recreational exposures to contaminated water has been linked to increasing travel to exotic destinations, participation in adventure travel, and the growing popularity of extreme sports involving fresh water.⁹ Recreational mud run events, for example, involve swimming in potentially contaminated waters and crawling through flooded farm fields where animal urine can be present—an ideal environment for Leptospira to thrive and for participants to contract the disease.^14,15

Continue for laboratory work-up >>

Laboratory work-up
Diagnosis of leptospirosis is challenging.²¹ Laboratory tests vary, depending on the timing and stage of infection, and are mostly unavailable in resource-poor countries. Test results for the patient with early-phase leptospirosis may demonstrate renal or hepatic abnormalities.¹⁸ However, laboratory confirmation of leptospirosis requires²²
• A fourfold increase in antibody titer between acute and convalescent serum samples, as detected by microscopic agglutination testing (MAT) or
• A high MAT titer (> 1:400 to 1:800), in single or paired samples or
• Isolation of pathogenic Leptospira species from a normally sterile site or
• Detection of DNA from pathogenic Leptospira species by PCR

A positive laboratory result is, of course, confirmatory. However, negative laboratory findings must be viewed with healthy skepticism.¹² A false-negative result may merely indicate the shortcoming of the testing method to accurately assess the presence of Leptospira.

Treatment options
The high mortality rate associated with severe leptospirosis makes early diagnosis and treatment essential.²³ The World Health Organization warns that antibiotic treatment for leptospirosis must be instituted within five days of symptom onset.¹⁰

Treatment options for an ambulatory patient with mild symptoms and no organ involvement include oral doxycycline (100 mg bid for 5-7 d) or oral azithromycin (500 mg/d for 5-7 d). For patients with organ involvement, IV penicillin (1.5 million U every 6 h for 7 d), ceftriaxone (1 g/d for 7 d), or cefotaxime (1 g every 6 h for 7 d) may be considered.^12,20

OUTCOME FOR THE CASE PATIENT
With leptospirosis as the diagnosis of exclusion, Jane was treated successfully with a 21-day course of oral doxycycline (100 mg bid). She has been symptom free since completing the regimen. After undergoing physical therapy and athletic training, she has been able to resume her full exercise regimen, and her recovery is considered complete.

CONCLUSION
The growing popularity of adventure travel and “extreme sports” events, particularly triathlons and mud runs, may precipitate an increase in associated infections with Leptospira and other zoonotic pathogens. For patients with flulike symptoms who routinely engage in such sports—especially those who present with conjunctival suffusion—leptospirosis should be considered in the differential diagnosis.

REFERENCES
1. Owens RC Jr, Ambrose PG. Antimicrobial safety: focus on fluoroquinolones. Clin Infect Dis. 2005;41(suppl 2):S144-S157.
2. CDC. Signs and symptoms of untreated Lyme disease (2015). www.cdc.gov/lyme/signs_symptoms/index.html. Accessed June 7, 2016.
3. CDC. Parasites: babesiosis (2014). www.cdc.gov/parasites/babesiosis/disease.html. Accessed June 7, 2016.
4. CDC. Ehrlichiosis: symptoms, diagnosis, and treatment (2013). www.cdc.gov/Ehrlichiosis/symptoms/index.html. Accessed June 7, 2016.
5. Pritt BS, Mead PS, Johnson DK, et al. Identification of a novel pathogenic Borrelia species causing Lyme borreliosis with unusually high spirochaetaemia: a descriptive study. Lancet Infect Dis. 2016 Feb 5. [Epub ahead of print]
6. Choi E, Pyzocha NJ, Maurer DM. Tick-borne illnesses. Curr Sports Med Rep. 2016;15(2):98-104.
7. Chomel B. Lyme disease. Rev Sci Tech. 2015;34(2):569-576.
8. Levett PN. Leptospirosis. Clin Microbiol Rev. 2001;14(2):296-326.
9. CDC. Leptospirosis: signs and symptoms (2016). www.cdc.gov/leptospirosis/symptoms/index.html. Accessed June 7, 2016.
10. World Health Organization, International Leptospirosis Society. Human Leptospirosis: Guidance for Diagnosis, Surveillance, and Control (2003). http://apps.who.int/iris/bitstream/10665/42667/1/WHO_CDS_CSR_EPH_2002.23.pdf. Accessed June 7, 2016.
11. Costa F, Hagan JE, Calcagno J, et al. Global morbidity and mortality of leptospirosis: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003898.
12. Haake DA, Levett PN. Leptospirosis in humans. Curr Top Microbiol Immunol. 2015;387:65-97.
13. Picardeau M. Diagnosis and epidemiology of leptospirosis. Médecine et Maladies Infectieuses. 2013;43(1):1-9.
14. Picardeau M. Leptospirosis: updating the global picture of an emerging neglected disease. PLoS Negl Trop Dis. 2015;9(9):e0004039.
15. Zavitsanou A, Babatsikou F. Leptospirosis: epidemiology and preventive measures. Health Sci J. 2008;2(2):75-82.
16. Mwachui MA, Crump L, Hartskeerl R, et al. Environmental and behavioural determinants of leptospirosis transmission: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003843.
17. Morgan J, Bornstein SL, Karpati AM, et al. Outbreak of leptospirosis among triathlon participants and community residents in Springfield, Illinois, 1998. Clin Infect Dis. 2002;34(12):1593-1599.
18. Katz AR, Ansdell VE, Effler PV, et al. Assessment of the clinical presentation and treatment of 353 cases of laboratory-confirmed leptospirosis in Hawaii, 1974-1998. Clin Infect Dis. 2001;33(11):1834-1841.
19. Yaakob Y, Rodrigues KF, John DV. Leptospirosis: recent incidents and available diagnostics—a review. Med J Malaysia. 2015;70(6):351-355.
20. Seguro AC, Andrade L. Pathophysiology of leptospirosis. Shock. 2013;39(suppl 1):17-23.
21. Musso D, La Scola B. Laboratory diagnosis of leptospirosis: a challenge. J Microbiol Immunol Infect. 2013;46(4):245-252.
22. Waggoner JJ, Balassiano I, Mohamed-Hadley A, et al. Reverse-transcriptase PCR detection of Leptospira: absence of agreement with single-specimen microscopic agglutination testing. PLoS One. 2015;10(7):e0132988.
23. Iwasaki H, Chagan-Yasutan H, Leano PS, et al. Combined antibody and DNA detection for early diagnosis of leptospirosis after a disaster. Diagn Microbiol Infect Dis. 2016;84(4):287-291

IN THIS ARTICLE

• Adverse effects of ciprofloxacin
• Symptoms of common tick-borne diseases
• Symptoms of phase 1 and late-phase disease
• Additional resources

Jane, an 18-year-old college student, presents in early November with a three-week history of worsening cough and sinus congestion. Recently, the cough has been interrupting her sleep and yellow-green nasal drainage and sinus pressure have increased. Ordinarily very fit and athletic, she reports that since she arrived at college two months ago, her body has become “more fragile.”

Further questioning reveals that, over the past two months, the patient’s symptoms have included extreme fatigue, severe unremitting headache, blurred vision, shortness of breath, and a racing heart rate on exertion. Her symptoms make it impossible for her to maintain her demanding exercise routine, a development that compounds her frustration and sadness. She has also been forced to limit her participation in school activities, with significant academic decline as a result.

Aside from depression (well controlled with bupropion HCl extended release, 300 mg/d), Jane’s medical history is unremarkable. She reports having “excellent health” until she arrived at her mid-Atlantic urban college.

A complicated history
Born and raised in Connecticut, Jane is an avid runner who competes in extreme sports. This past summer, she trained for and participated in two “mud run” events (ie, endurance races of several miles with numerous challenges and obstacles) in Connecticut and New York. Training included endurance runs and sprints, as well as crawling through mud-laden fields and woods.

She also did a three-week summer internship on an oyster farm. There, she was required to shuck oysters and stand in brackish water for six-hour shifts to examine oyster beds. In the process, she sustained numerous cuts and bruises on her hands, arms, and legs.

A week or so after returning to college in late August, Jane developed blisters on both heels, which progressed to infected ulcerations. She was evaluated at the university hospital emergency department (ED) and treated with a 21-day course of ciprofloxacin. When left-sided unilateral knee swelling developed about two weeks later, she underwent arthrocentesis at the university health center, but joint aspirate was not sent for analysis. A two-week course of antibiotic therapy was initiated.

From October to her presentation in early November, Jane has experienced intermittent fevers and chills, with a temperature as high as 101°F. In addition, she complains of fasciculations and weakness in her lower limbs; dyspnea, tachycardia, and dizziness during or after any exertion; unremitting posterior neck pain; and a constant, severe headache located primarily in the bitemporal region. She developed bilateral conjunctivitis, which resolved spontaneously in about one week; persistent blurred vision; a transient petechial chest rash; recurring episodes of syncope; pyelonephritis; a persistent vaginal yeast infection; decreased appetite; and a 7-lb weight loss (5% of her total body weight).

Jane’s academic and athletic performance has been severely impaired. Once a long-distance runner, she can no longer walk any distance without frequent rest. In the four months since the mud runs, the patient reports, she has been seen in the student health center four times and in the ED twice. Additionally, she has undergone thorough examinations by clinicians specializing in infectious disease, pulmonology, neurology, and neuro-ophthalmology. She has undergone lab work, including
• Complete blood cell count with differential
• Comprehensive metabolic panel
• Urinalysis and urine culture
• Lyme antibody and blood polymerase chain reaction (PCR)
• HIV testing
• Rheumatoid factor
• Erythrocyte sedimentation rate (ESR)
• C-reactive protein (CRP)
• Epstein-Barr virus IgM
• Cytomegalovirus (CMV) IgM
• Human granulocytic ehrlichiosis (HGE) antibody and human anaplasma phagocytophilum (HGA)
• HGA PCR
• Rickettsia antibody panel
• Babesia microti antibodies
• Pregnancy testing
• Chest x-ray
• Lumbar puncture

All lab results were within normal range. In light of this, several clinicians have told Jane that her illness is “all in her head.”

Continue for the patient investigates >>

The patient investigates
In mid-December, after she has returned home from college, Jane’s symptoms abruptly worsen. She complains of feeling “shakier,” with weakness in her legs and what she calls “brain fog.” Her headache, blurred vision, and dizziness have worsened. Frightened and concerned, she returns to the ED. Results of a thorough evaluation, including lumbar puncture, reveal no abnormality.

Jane has become extremely frail. She is losing weight, her hair has lost its luster, and her nails are cracking and bleeding. She is unable to walk without concern for falling and cannot climb the 20 steps to her bedroom. Once a healthy and vibrant 18-year-old, she now spends most of her time in a lethargic state on a first-floor living room couch.

Frustrated by her unexplained declining health, she begins to research illnesses associated with extreme sports and prolonged marine exposure. She returns to ask about three possible explanations for her condition:
1. Adverse effects of ciprofloxacin use, which include fever or chills, dizziness, racing heartbeat, headache, and nausea.¹
2. A tick-borne disease, possibly contracted during her practice runs in the Connecticut woods (see Table 1).^2-4 Each year, she recalls, she has found and removed four or five embedded ticks. In the northeastern United States, the most common tick-borne diseases are borreliosis, babesiosis, and ehrlichiosis.^5-7
3. Leptospirosis, contracted through the patient’s exposure to mud and brackish water during her summer activities. According to her research, more than 25 outbreaks and 600 cases of leptospirosis (between 1931 and 1998) have been associated with fresh pond, creek, or river water.⁸

Based on Jane’s symptoms and history, and in accord with her research, early-phase leptospirosis is identified as a diagnosis of exclusion (with a possible comorbid tick-borne zoonosis).

Continue for discussion >>

DISCUSSION
Leptospirosis develops when humans come into contact with animal urine infected by leptospires—that is, pathogenic spirochetes excreted via the renal tubules of infected host animals.^9,10 While host animals include dogs, pigs, cattle, reptiles, and amphibians, the animal most commonly associated with human infection is the brown rat (Rattus norvegicus).^11-15

Leptospires enter the human host through mucous membranes, cuts, or abrasions in the skin. Individuals at increased risk for infection include those whose work or other activities expose them “to animal reservoirs or contaminated environments”—including participants in water sports and similar recreation.^11-14 As Mwachui et al explain, “recreational exposure to [Leptospira-]contaminated water has become more important for sport enthusiasts, swimmers and travellers from industrialized countries,” whereas flooding is usually involved in infection in undeveloped countries.¹⁶

The largest outbreak of leptospirosis reported in the US to date occurred in 1998, when heavy rains preceded a triathlon in Springfield, Illinois. When many participants became ill after the event, researchers from the National Center for Infectious Diseases were able to contact and test 834 of the 876 competing athletes; of these, 98 (12%) reported being ill and 52 (11%) tested positive for leptospirosis. Additionally, 14 of the 248 community residents who were sickened (6%) tested positive.¹⁷ According to CDC estimates, between 100 and 200 cases of leptospirosis develop annually in the US, with about half occurring in Hawaii.⁹

Onset of symptoms, which are described as protean and nonspecific, occurs two days to four weeks after exposure, making leptospirosis difficult to diagnosewithout a high degree of suspicion; zoonotic exposure (as with freshwater or mud sports) or a history of travel to Hawaii, Tahiti, Thailand, Indonesia, the Caribbean, and/or Costa Rica may raise suspicion.^12-14,18 In early-phase leptospirosis, symptoms can mimic those of influenza, meningitis, malaria, dengue fever, scrub typhus, rickettsial disease, and typhoid fever (see Table 2).¹⁰ Thus, when a patient presents with these symptoms, it is imperative that the clinician consider leptospirosis.¹⁹Of note: Flu-like symptoms with conjunctival suffusion are considered pathognomonic for leptospirosis.¹⁸

About 10% of patients with early-phase leptospirosis will develop late-phase disease (ie, Weil’s disease), with severe symptoms that include jaundice, meningitis, pulmonary hemorrhage, and acute kidney injury (see Table 3 for a more detailed list).²⁰ The case patient’s history and symptoms were consistent with a diagnosis of early-phase lepto­spirosis.

Epidemiology
In 2015, leptospirosis was estimated to affect more than 1 million persons worldwide, with 58,900 deaths attributed to the disease each year—making leptospirosis the leading cause of death attributable to zoonotic illness.¹¹ Historically, leptospirosis-associated morbidity and mortality have been greatest in resource-poor countries with tropical climates (eg, southern and Southeast Asia, Central America and tropical Latin America, and East Sub-Saharan Africa).^11,12

However, illness resulting from recreational exposures to contaminated water has been linked to increasing travel to exotic destinations, participation in adventure travel, and the growing popularity of extreme sports involving fresh water.⁹ Recreational mud run events, for example, involve swimming in potentially contaminated waters and crawling through flooded farm fields where animal urine can be present—an ideal environment for Leptospira to thrive and for participants to contract the disease.^14,15

Continue for laboratory work-up >>

Laboratory work-up
Diagnosis of leptospirosis is challenging.²¹ Laboratory tests vary, depending on the timing and stage of infection, and are mostly unavailable in resource-poor countries. Test results for the patient with early-phase leptospirosis may demonstrate renal or hepatic abnormalities.¹⁸ However, laboratory confirmation of leptospirosis requires²²
• A fourfold increase in antibody titer between acute and convalescent serum samples, as detected by microscopic agglutination testing (MAT) or
• A high MAT titer (> 1:400 to 1:800), in single or paired samples or
• Isolation of pathogenic Leptospira species from a normally sterile site or
• Detection of DNA from pathogenic Leptospira species by PCR

A positive laboratory result is, of course, confirmatory. However, negative laboratory findings must be viewed with healthy skepticism.¹² A false-negative result may merely indicate the shortcoming of the testing method to accurately assess the presence of Leptospira.

Treatment options
The high mortality rate associated with severe leptospirosis makes early diagnosis and treatment essential.²³ The World Health Organization warns that antibiotic treatment for leptospirosis must be instituted within five days of symptom onset.¹⁰

Treatment options for an ambulatory patient with mild symptoms and no organ involvement include oral doxycycline (100 mg bid for 5-7 d) or oral azithromycin (500 mg/d for 5-7 d). For patients with organ involvement, IV penicillin (1.5 million U every 6 h for 7 d), ceftriaxone (1 g/d for 7 d), or cefotaxime (1 g every 6 h for 7 d) may be considered.^12,20

OUTCOME FOR THE CASE PATIENT
With leptospirosis as the diagnosis of exclusion, Jane was treated successfully with a 21-day course of oral doxycycline (100 mg bid). She has been symptom free since completing the regimen. After undergoing physical therapy and athletic training, she has been able to resume her full exercise regimen, and her recovery is considered complete.

CONCLUSION
The growing popularity of adventure travel and “extreme sports” events, particularly triathlons and mud runs, may precipitate an increase in associated infections with Leptospira and other zoonotic pathogens. For patients with flulike symptoms who routinely engage in such sports—especially those who present with conjunctival suffusion—leptospirosis should be considered in the differential diagnosis.

REFERENCES
1. Owens RC Jr, Ambrose PG. Antimicrobial safety: focus on fluoroquinolones. Clin Infect Dis. 2005;41(suppl 2):S144-S157.
2. CDC. Signs and symptoms of untreated Lyme disease (2015). www.cdc.gov/lyme/signs_symptoms/index.html. Accessed June 7, 2016.
3. CDC. Parasites: babesiosis (2014). www.cdc.gov/parasites/babesiosis/disease.html. Accessed June 7, 2016.
4. CDC. Ehrlichiosis: symptoms, diagnosis, and treatment (2013). www.cdc.gov/Ehrlichiosis/symptoms/index.html. Accessed June 7, 2016.
5. Pritt BS, Mead PS, Johnson DK, et al. Identification of a novel pathogenic Borrelia species causing Lyme borreliosis with unusually high spirochaetaemia: a descriptive study. Lancet Infect Dis. 2016 Feb 5. [Epub ahead of print]
6. Choi E, Pyzocha NJ, Maurer DM. Tick-borne illnesses. Curr Sports Med Rep. 2016;15(2):98-104.
7. Chomel B. Lyme disease. Rev Sci Tech. 2015;34(2):569-576.
8. Levett PN. Leptospirosis. Clin Microbiol Rev. 2001;14(2):296-326.
9. CDC. Leptospirosis: signs and symptoms (2016). www.cdc.gov/leptospirosis/symptoms/index.html. Accessed June 7, 2016.
10. World Health Organization, International Leptospirosis Society. Human Leptospirosis: Guidance for Diagnosis, Surveillance, and Control (2003). http://apps.who.int/iris/bitstream/10665/42667/1/WHO_CDS_CSR_EPH_2002.23.pdf. Accessed June 7, 2016.
11. Costa F, Hagan JE, Calcagno J, et al. Global morbidity and mortality of leptospirosis: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003898.
12. Haake DA, Levett PN. Leptospirosis in humans. Curr Top Microbiol Immunol. 2015;387:65-97.
13. Picardeau M. Diagnosis and epidemiology of leptospirosis. Médecine et Maladies Infectieuses. 2013;43(1):1-9.
14. Picardeau M. Leptospirosis: updating the global picture of an emerging neglected disease. PLoS Negl Trop Dis. 2015;9(9):e0004039.
15. Zavitsanou A, Babatsikou F. Leptospirosis: epidemiology and preventive measures. Health Sci J. 2008;2(2):75-82.
16. Mwachui MA, Crump L, Hartskeerl R, et al. Environmental and behavioural determinants of leptospirosis transmission: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003843.
17. Morgan J, Bornstein SL, Karpati AM, et al. Outbreak of leptospirosis among triathlon participants and community residents in Springfield, Illinois, 1998. Clin Infect Dis. 2002;34(12):1593-1599.
18. Katz AR, Ansdell VE, Effler PV, et al. Assessment of the clinical presentation and treatment of 353 cases of laboratory-confirmed leptospirosis in Hawaii, 1974-1998. Clin Infect Dis. 2001;33(11):1834-1841.
19. Yaakob Y, Rodrigues KF, John DV. Leptospirosis: recent incidents and available diagnostics—a review. Med J Malaysia. 2015;70(6):351-355.
20. Seguro AC, Andrade L. Pathophysiology of leptospirosis. Shock. 2013;39(suppl 1):17-23.
21. Musso D, La Scola B. Laboratory diagnosis of leptospirosis: a challenge. J Microbiol Immunol Infect. 2013;46(4):245-252.
22. Waggoner JJ, Balassiano I, Mohamed-Hadley A, et al. Reverse-transcriptase PCR detection of Leptospira: absence of agreement with single-specimen microscopic agglutination testing. PLoS One. 2015;10(7):e0132988.
23. Iwasaki H, Chagan-Yasutan H, Leano PS, et al. Combined antibody and DNA detection for early diagnosis of leptospirosis after a disaster. Diagn Microbiol Infect Dis. 2016;84(4):287-291

IN THIS ARTICLE

• Adverse effects of ciprofloxacin
• Symptoms of common tick-borne diseases
• Symptoms of phase 1 and late-phase disease
• Additional resources

Jane, an 18-year-old college student, presents in early November with a three-week history of worsening cough and sinus congestion. Recently, the cough has been interrupting her sleep and yellow-green nasal drainage and sinus pressure have increased. Ordinarily very fit and athletic, she reports that since she arrived at college two months ago, her body has become “more fragile.”

Further questioning reveals that, over the past two months, the patient’s symptoms have included extreme fatigue, severe unremitting headache, blurred vision, shortness of breath, and a racing heart rate on exertion. Her symptoms make it impossible for her to maintain her demanding exercise routine, a development that compounds her frustration and sadness. She has also been forced to limit her participation in school activities, with significant academic decline as a result.

Aside from depression (well controlled with bupropion HCl extended release, 300 mg/d), Jane’s medical history is unremarkable. She reports having “excellent health” until she arrived at her mid-Atlantic urban college.

A complicated history
Born and raised in Connecticut, Jane is an avid runner who competes in extreme sports. This past summer, she trained for and participated in two “mud run” events (ie, endurance races of several miles with numerous challenges and obstacles) in Connecticut and New York. Training included endurance runs and sprints, as well as crawling through mud-laden fields and woods.

She also did a three-week summer internship on an oyster farm. There, she was required to shuck oysters and stand in brackish water for six-hour shifts to examine oyster beds. In the process, she sustained numerous cuts and bruises on her hands, arms, and legs.

A week or so after returning to college in late August, Jane developed blisters on both heels, which progressed to infected ulcerations. She was evaluated at the university hospital emergency department (ED) and treated with a 21-day course of ciprofloxacin. When left-sided unilateral knee swelling developed about two weeks later, she underwent arthrocentesis at the university health center, but joint aspirate was not sent for analysis. A two-week course of antibiotic therapy was initiated.

From October to her presentation in early November, Jane has experienced intermittent fevers and chills, with a temperature as high as 101°F. In addition, she complains of fasciculations and weakness in her lower limbs; dyspnea, tachycardia, and dizziness during or after any exertion; unremitting posterior neck pain; and a constant, severe headache located primarily in the bitemporal region. She developed bilateral conjunctivitis, which resolved spontaneously in about one week; persistent blurred vision; a transient petechial chest rash; recurring episodes of syncope; pyelonephritis; a persistent vaginal yeast infection; decreased appetite; and a 7-lb weight loss (5% of her total body weight).

Jane’s academic and athletic performance has been severely impaired. Once a long-distance runner, she can no longer walk any distance without frequent rest. In the four months since the mud runs, the patient reports, she has been seen in the student health center four times and in the ED twice. Additionally, she has undergone thorough examinations by clinicians specializing in infectious disease, pulmonology, neurology, and neuro-ophthalmology. She has undergone lab work, including
• Complete blood cell count with differential
• Comprehensive metabolic panel
• Urinalysis and urine culture
• Lyme antibody and blood polymerase chain reaction (PCR)
• HIV testing
• Rheumatoid factor
• Erythrocyte sedimentation rate (ESR)
• C-reactive protein (CRP)
• Epstein-Barr virus IgM
• Cytomegalovirus (CMV) IgM
• Human granulocytic ehrlichiosis (HGE) antibody and human anaplasma phagocytophilum (HGA)
• HGA PCR
• Rickettsia antibody panel
• Babesia microti antibodies
• Pregnancy testing
• Chest x-ray
• Lumbar puncture

All lab results were within normal range. In light of this, several clinicians have told Jane that her illness is “all in her head.”

Continue for the patient investigates >>

The patient investigates
In mid-December, after she has returned home from college, Jane’s symptoms abruptly worsen. She complains of feeling “shakier,” with weakness in her legs and what she calls “brain fog.” Her headache, blurred vision, and dizziness have worsened. Frightened and concerned, she returns to the ED. Results of a thorough evaluation, including lumbar puncture, reveal no abnormality.

Jane has become extremely frail. She is losing weight, her hair has lost its luster, and her nails are cracking and bleeding. She is unable to walk without concern for falling and cannot climb the 20 steps to her bedroom. Once a healthy and vibrant 18-year-old, she now spends most of her time in a lethargic state on a first-floor living room couch.

Frustrated by her unexplained declining health, she begins to research illnesses associated with extreme sports and prolonged marine exposure. She returns to ask about three possible explanations for her condition:
1. Adverse effects of ciprofloxacin use, which include fever or chills, dizziness, racing heartbeat, headache, and nausea.¹
2. A tick-borne disease, possibly contracted during her practice runs in the Connecticut woods (see Table 1).^2-4 Each year, she recalls, she has found and removed four or five embedded ticks. In the northeastern United States, the most common tick-borne diseases are borreliosis, babesiosis, and ehrlichiosis.^5-7
3. Leptospirosis, contracted through the patient’s exposure to mud and brackish water during her summer activities. According to her research, more than 25 outbreaks and 600 cases of leptospirosis (between 1931 and 1998) have been associated with fresh pond, creek, or river water.⁸

Based on Jane’s symptoms and history, and in accord with her research, early-phase leptospirosis is identified as a diagnosis of exclusion (with a possible comorbid tick-borne zoonosis).

Continue for discussion >>

DISCUSSION
Leptospirosis develops when humans come into contact with animal urine infected by leptospires—that is, pathogenic spirochetes excreted via the renal tubules of infected host animals.^9,10 While host animals include dogs, pigs, cattle, reptiles, and amphibians, the animal most commonly associated with human infection is the brown rat (Rattus norvegicus).^11-15

Leptospires enter the human host through mucous membranes, cuts, or abrasions in the skin. Individuals at increased risk for infection include those whose work or other activities expose them “to animal reservoirs or contaminated environments”—including participants in water sports and similar recreation.^11-14 As Mwachui et al explain, “recreational exposure to [Leptospira-]contaminated water has become more important for sport enthusiasts, swimmers and travellers from industrialized countries,” whereas flooding is usually involved in infection in undeveloped countries.¹⁶

The largest outbreak of leptospirosis reported in the US to date occurred in 1998, when heavy rains preceded a triathlon in Springfield, Illinois. When many participants became ill after the event, researchers from the National Center for Infectious Diseases were able to contact and test 834 of the 876 competing athletes; of these, 98 (12%) reported being ill and 52 (11%) tested positive for leptospirosis. Additionally, 14 of the 248 community residents who were sickened (6%) tested positive.¹⁷ According to CDC estimates, between 100 and 200 cases of leptospirosis develop annually in the US, with about half occurring in Hawaii.⁹

Onset of symptoms, which are described as protean and nonspecific, occurs two days to four weeks after exposure, making leptospirosis difficult to diagnosewithout a high degree of suspicion; zoonotic exposure (as with freshwater or mud sports) or a history of travel to Hawaii, Tahiti, Thailand, Indonesia, the Caribbean, and/or Costa Rica may raise suspicion.^12-14,18 In early-phase leptospirosis, symptoms can mimic those of influenza, meningitis, malaria, dengue fever, scrub typhus, rickettsial disease, and typhoid fever (see Table 2).¹⁰ Thus, when a patient presents with these symptoms, it is imperative that the clinician consider leptospirosis.¹⁹Of note: Flu-like symptoms with conjunctival suffusion are considered pathognomonic for leptospirosis.¹⁸

About 10% of patients with early-phase leptospirosis will develop late-phase disease (ie, Weil’s disease), with severe symptoms that include jaundice, meningitis, pulmonary hemorrhage, and acute kidney injury (see Table 3 for a more detailed list).²⁰ The case patient’s history and symptoms were consistent with a diagnosis of early-phase lepto­spirosis.

Epidemiology
In 2015, leptospirosis was estimated to affect more than 1 million persons worldwide, with 58,900 deaths attributed to the disease each year—making leptospirosis the leading cause of death attributable to zoonotic illness.¹¹ Historically, leptospirosis-associated morbidity and mortality have been greatest in resource-poor countries with tropical climates (eg, southern and Southeast Asia, Central America and tropical Latin America, and East Sub-Saharan Africa).^11,12

However, illness resulting from recreational exposures to contaminated water has been linked to increasing travel to exotic destinations, participation in adventure travel, and the growing popularity of extreme sports involving fresh water.⁹ Recreational mud run events, for example, involve swimming in potentially contaminated waters and crawling through flooded farm fields where animal urine can be present—an ideal environment for Leptospira to thrive and for participants to contract the disease.^14,15

Continue for laboratory work-up >>

Laboratory work-up
Diagnosis of leptospirosis is challenging.²¹ Laboratory tests vary, depending on the timing and stage of infection, and are mostly unavailable in resource-poor countries. Test results for the patient with early-phase leptospirosis may demonstrate renal or hepatic abnormalities.¹⁸ However, laboratory confirmation of leptospirosis requires²²
• A fourfold increase in antibody titer between acute and convalescent serum samples, as detected by microscopic agglutination testing (MAT) or
• A high MAT titer (> 1:400 to 1:800), in single or paired samples or
• Isolation of pathogenic Leptospira species from a normally sterile site or
• Detection of DNA from pathogenic Leptospira species by PCR

A positive laboratory result is, of course, confirmatory. However, negative laboratory findings must be viewed with healthy skepticism.¹² A false-negative result may merely indicate the shortcoming of the testing method to accurately assess the presence of Leptospira.

Treatment options
The high mortality rate associated with severe leptospirosis makes early diagnosis and treatment essential.²³ The World Health Organization warns that antibiotic treatment for leptospirosis must be instituted within five days of symptom onset.¹⁰

Treatment options for an ambulatory patient with mild symptoms and no organ involvement include oral doxycycline (100 mg bid for 5-7 d) or oral azithromycin (500 mg/d for 5-7 d). For patients with organ involvement, IV penicillin (1.5 million U every 6 h for 7 d), ceftriaxone (1 g/d for 7 d), or cefotaxime (1 g every 6 h for 7 d) may be considered.^12,20

OUTCOME FOR THE CASE PATIENT
With leptospirosis as the diagnosis of exclusion, Jane was treated successfully with a 21-day course of oral doxycycline (100 mg bid). She has been symptom free since completing the regimen. After undergoing physical therapy and athletic training, she has been able to resume her full exercise regimen, and her recovery is considered complete.

CONCLUSION
The growing popularity of adventure travel and “extreme sports” events, particularly triathlons and mud runs, may precipitate an increase in associated infections with Leptospira and other zoonotic pathogens. For patients with flulike symptoms who routinely engage in such sports—especially those who present with conjunctival suffusion—leptospirosis should be considered in the differential diagnosis.

REFERENCES
1. Owens RC Jr, Ambrose PG. Antimicrobial safety: focus on fluoroquinolones. Clin Infect Dis. 2005;41(suppl 2):S144-S157.
2. CDC. Signs and symptoms of untreated Lyme disease (2015). www.cdc.gov/lyme/signs_symptoms/index.html. Accessed June 7, 2016.
3. CDC. Parasites: babesiosis (2014). www.cdc.gov/parasites/babesiosis/disease.html. Accessed June 7, 2016.
4. CDC. Ehrlichiosis: symptoms, diagnosis, and treatment (2013). www.cdc.gov/Ehrlichiosis/symptoms/index.html. Accessed June 7, 2016.
5. Pritt BS, Mead PS, Johnson DK, et al. Identification of a novel pathogenic Borrelia species causing Lyme borreliosis with unusually high spirochaetaemia: a descriptive study. Lancet Infect Dis. 2016 Feb 5. [Epub ahead of print]
6. Choi E, Pyzocha NJ, Maurer DM. Tick-borne illnesses. Curr Sports Med Rep. 2016;15(2):98-104.
7. Chomel B. Lyme disease. Rev Sci Tech. 2015;34(2):569-576.
8. Levett PN. Leptospirosis. Clin Microbiol Rev. 2001;14(2):296-326.
9. CDC. Leptospirosis: signs and symptoms (2016). www.cdc.gov/leptospirosis/symptoms/index.html. Accessed June 7, 2016.
10. World Health Organization, International Leptospirosis Society. Human Leptospirosis: Guidance for Diagnosis, Surveillance, and Control (2003). http://apps.who.int/iris/bitstream/10665/42667/1/WHO_CDS_CSR_EPH_2002.23.pdf. Accessed June 7, 2016.
11. Costa F, Hagan JE, Calcagno J, et al. Global morbidity and mortality of leptospirosis: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003898.
12. Haake DA, Levett PN. Leptospirosis in humans. Curr Top Microbiol Immunol. 2015;387:65-97.
13. Picardeau M. Diagnosis and epidemiology of leptospirosis. Médecine et Maladies Infectieuses. 2013;43(1):1-9.
14. Picardeau M. Leptospirosis: updating the global picture of an emerging neglected disease. PLoS Negl Trop Dis. 2015;9(9):e0004039.
15. Zavitsanou A, Babatsikou F. Leptospirosis: epidemiology and preventive measures. Health Sci J. 2008;2(2):75-82.
16. Mwachui MA, Crump L, Hartskeerl R, et al. Environmental and behavioural determinants of leptospirosis transmission: a systematic review. PLoS Negl Trop Dis. 2015;9(9):e0003843.
17. Morgan J, Bornstein SL, Karpati AM, et al. Outbreak of leptospirosis among triathlon participants and community residents in Springfield, Illinois, 1998. Clin Infect Dis. 2002;34(12):1593-1599.
18. Katz AR, Ansdell VE, Effler PV, et al. Assessment of the clinical presentation and treatment of 353 cases of laboratory-confirmed leptospirosis in Hawaii, 1974-1998. Clin Infect Dis. 2001;33(11):1834-1841.
19. Yaakob Y, Rodrigues KF, John DV. Leptospirosis: recent incidents and available diagnostics—a review. Med J Malaysia. 2015;70(6):351-355.
20. Seguro AC, Andrade L. Pathophysiology of leptospirosis. Shock. 2013;39(suppl 1):17-23.
21. Musso D, La Scola B. Laboratory diagnosis of leptospirosis: a challenge. J Microbiol Immunol Infect. 2013;46(4):245-252.
22. Waggoner JJ, Balassiano I, Mohamed-Hadley A, et al. Reverse-transcriptase PCR detection of Leptospira: absence of agreement with single-specimen microscopic agglutination testing. PLoS One. 2015;10(7):e0132988.
23. Iwasaki H, Chagan-Yasutan H, Leano PS, et al. Combined antibody and DNA detection for early diagnosis of leptospirosis after a disaster. Diagn Microbiol Infect Dis. 2016;84(4):287-291

Nearly 2% of patients who undergo total knee arthroplasty (TKA) or total hip arthroplasty (THA) develop a periprosthetic joint infection (PJI) within 20 years of surgery, and 41% of these infections occur within the first 2 years.¹ PJI is the most common cause of TKA failure and the third leading complication of THA.² The estimated total hospital cost of treating PJI increased from $320 million in 2001 to $566 million in 2009, which can be extrapolated to $1.62 billion in 2020.³ By 2030, the projected increase in demand for TKA and THA will be 673% and 174% of what it was in 2005, respectively.⁴ Treatment of PJI of the knee is estimated to cost 3 to 4 times more than a primary TKA, and the cost of revision THA for PJI is almost $6000 more than that of revision TKA for PJI.³

In this article, we review the numerous preoperative, intraoperative, and postoperative methods of decreasing PJI incidence after total joint arthroplasty (TJA).

Preoperative Risk Prevention

Medical Comorbidities

Preoperative medical optimization is a key element in PJI prevention (Table 1). An American Society of Anesthesiologists classification score of 3 or more has been associated with doubled risk for surgical site infections (SSIs) after THA.5 Autoimmune conditions confer a particularly higher risk. In a retrospective double-cohort study of 924 subjects, Bongartz and colleagues⁶ found that, compared with osteoarthritis, rheumatoid arthritis tripled the risk of PJI. Small case series originally suggested a higher risk of PJI in patients with psoriasis,^7,8 but more recent studies have contradicted that finding.^9,10 Nevertheless, psoriatic plaques have elevated bacterial counts,¹¹ and planned incisions should circumvent these areas.

Diabetes mellitus is a clear risk factor for PJI.^12-16 Regarding whether preoperative glucose control affects risk, findings have been mixed. Mraovic and colleagues¹⁷ showed preoperative hyperglycemia to be an independent risk factor; Jämsen and colleagues,¹⁵ in a single-center analysis of more than 7000 TJAs, suggested preoperative blood glucose levels were not independently associated with PJI; and Iorio and colleagues¹⁶ found no association between surgical infections and hemoglobin A1c levels.

TJA incidence is higher in patients with chronic kidney disease (CKD) than in the general population.¹⁸ Dialysis users have a post-THA PJI rate as high as 13% to 19%.^19,20 Early clinical data suggested that outcomes are improved in dialysis users who undergo renal transplant, but this finding recently has been questioned.^19,21 Deegan and colleagues²² found an increased PJA rate of 3.5% even in low-level CKD (stage 1, 2, or 3), but this may be confounded by the increased association of CKD with other PJI-predisposing comorbidities.

Given a higher incidence of urinary tract infections (UTIs) among patients with PJI, some surgeons think UTIs predispose to PJIs by hematogenous seeding.^12,23,24 Symptomatic UTIs should be cleared before surgery and confirmed on urinalysis. Obstructive symptoms should prompt urologic evaluation. As asymptomatic pyuria and bacteriuria (colony counts, >1 × 105/mL) do not predispose to PJI, patients without symptoms do not require intervention.^25,26 Past history of malignancy may also have a role in PJI. In a case-control study of the Mayo Clinic arthroplasty experience from 1969 to 1991, Berbari and colleagues¹ found an association between malignancy and PJI (odds ratio, 2.4). They theorized the immunosuppressive effects of cancer treatment might be responsible for this increased risk.

Immunocompromising Medications

Immunocompromising medications are modifiable and should be adjusted before surgery. Stopping any disease-modifying antirheumatic drug (DMARD) more than 4 weeks before surgery is not recommended.²⁷

Corticosteroid use can lead to immunosuppression and increased protein catabolism, which impairs soft-tissue healing. To avoid flares or adrenal insufficiency, however, chronic corticosteroid users should continue their regular doses perioperatively.²⁸ On the day of surgery, they should also receive a stress dose of hydrocortisone 50 to 75 mg (for primary arthroplasty) or 100 to 150 mg (for revision arthroplasty), followed by expeditious tapering over 1 to 2 days.²⁹ DMARDs are increasingly used by rheumatologists. One of the most effective DMARDs is methotrexate. Despite its immunocompromising activity, methotrexate should be continued perioperatively, as stopping for even 2 days may increase flare-related complications.³⁰ Hydroxychloroquine can be continued perioperatively and has even been shown, by Johnson and Charnley,³¹ to prevent deep vein thromboses. Sulfasalazine can also be continued perioperatively—but with caution, as it may elevate international normalized ratio (INR) levels in patients receiving warfarin.²⁹ Most other DMARDs should be temporarily discontinued. Leflunomide and interleukin 1 antagonists, such as anakinra, should be stopped 1 to 2 days before surgery and restarted 10 to 14 days after surgery.²⁹ Rituximab should be stopped 1 week before surgery and restarted 10 to 14 days after surgery. Tumor necrosis factor α inhibitors should be discontinued for 2 half-lives before and after surgery.³² Etanercept has a half-life of 3 to 5 days; infliximab, 8 to 10 days; and adalimumab, 10 to 13 days. Most surgeons schedule surgery for the end of a dosing cycle and discontinue these biologic agents for another 10 to 14 days after surgery.

Metabolic Factors

Obese patients are susceptible to longer surgeries, more extensive dissection, poorly vascularized subcutaneous tissue, and higher requirements of weight-adjusted antibiotic dosing.¹³ Body mass index (BMI) of 40 kg/m² or more (morbid obesity) and BMI over 50 kg/m² have been associated with 9 times and 21.3 times increased risk of PJI, respectively.^13,14 Delaying surgery with dietary consultation has been suggested,^33,34 and bariatric surgery before TKA may decrease infection rates by 3.5 times.³⁵

Nutritional markers are considered before arthroplasty. According to most laboratories, a serum transferrin level under 200 mg/dL, albumin level under 3.5 g/dL, and total lymphocyte count under 1500 cells/mm³ indicate malnourishment, which can increase the incidence of wound complications by 5 to 7 times.³⁶ Patients should also have sufficient protein, vitamin, and mineral supplementation, particularly vitamins A and C, zinc, and copper.³⁷Smokers who cease smoking at least 4 to 6 weeks before surgery lower their wound complication rate by up to 26%.^38,39 When nicotine leaves the bloodstream, vasodilation occurs, oxygenation improves, and the immune system recovers.³⁹ Studies have found more SSIs in patients who abuse alcohol,40 and numerous authors have confirmed this finding in the arthroplasty population.^24,41,42 Alcohol inhibits platelet function and may predispose to a postoperative hematoma. In contrast to smoking cessation evidence, evidence regarding alcohol interventions in preventing postoperative infections is less conclusive.^43,44

MRSA Colonization

Methicillin-resistant Staphylococcus aureus (MRSA) is a particularly difficult bacterium to eradicate in PJI. As the mean cost of treating a single case of MRSA-related prosthetic infection is $107,264 vs $68,053 for susceptible strains,45,46 many infection-containment strategies focus on addressing benign MRSA colonization before surgery.

MRSA is present in the nares of 25 million people in the United States. Nasal colonization increases the risk of bacteremia 4-fold⁴⁷ and SSI 2- to 9-fold.^48,49 Nasal swabs are analyzed with either a rapid polymerase chain reaction (PCR) test, which provides results in 2 hours, or a bacterial culture, which provides results in 1 to 4 days. The PCR test is more expensive.

Eradication of MRSA colonization is increasingly prevalent. Several Scandinavian countries have instituted strict practices by which patients are denied elective surgery until negative nasal swabs are obtained.⁴⁹ Nasal decontamination is one method of colonization reduction. Topical mupirocin, which yields eradication in 91% of nasal carriers immediately after treatment and in 87% after 4 weeks,⁵⁰ is effective in reducing SSI rates only when used in conjunction with a body wash, which is used to clean the axilla and groin.⁵¹ There is no consensus on optimal timing, but Bode and colleagues⁵² found a significant decrease in deep SSIs when decontamination occurred just 24 hours before surgery.

Povidone-iodine showers went out of favor with the realization that chlorhexidine gluconate acts longer on the skin surface.^53,54 Preoperative showers involve rinsing with liquid chlorhexidine soap 24 to 48 hours before surgery. However, chlorhexidine binds preferentially to the cotton in washcloths instead of the skin. Edmiston and colleagues^54,55 found that 4% chlorhexidine liquid soaps achieve much lower skin chlorhexidine concentrations than 2% polyester cloths do. Use of these “chlorhexidine wipes” the night before and the day of surgery has decreased PJI after TKA from 2.2% to 0.6%.^56,57

Intraoperative Risk Prevention

Preparation

Which preoperative antibiotic to use is one of the first operative considerations in PJI prophylaxis (Table 2). Cefazolin is recommended as a first-line agent for its excellent soft-tissue penetration, long half-life, and activity against gram-positive bacteria such as skin flora.⁵⁸ Clindamycin may be considered for patients allergic to β-lactam antibiotics. Vancomycin may be considered for adjunctive use with cephalosporins in cases of known MRSA colonization. Vancomycin infusion should be started earlier than infusion with other antibiotics, as vancomycin must be infused slowly and takes longer to become therapeutic.

Antibiotic dosing should be based on local antibiograms, adjusted dosing weight, or BMI.⁵⁹ For revision arthroplasty, preoperative prophylaxis should not be stopped out of fear of affecting operative cultures.⁶⁰ Some surgeons pause antibiotic use if a preoperative joint aspirate has not been obtained. Infusion within 1 hour of incision is part of the pay-for-performance guidelines established by the US Centers for Medicare & Medicaid Services.⁶¹ An antibiotic should be redosed if the operation will take longer than 2 half-lives of the drug.⁵⁹ Surgeons should consider administering a dose every 4 hours or whenever blood loss exceeds 1000 mL.⁶² Engesæter and colleagues⁶³ found that antibiotic prophylaxis was most effective given 4 times perioperatively (1 time before surgery, 3 times after surgery). Postoperative antibiotics should not be administered longer than 24 hours, as prolonged dosing confers no benefit.⁵⁸ Operating room conditions must be optimized for prophylaxis. More people and operating room traffic in nonsterile corridors increase contamination of instruments open to air.⁶⁴ Laminar airflow systems are commonly used. Although there is little dispute that laminar flow decreases the bacterial load of air, there are mixed results regarding its benefit in preventing PJI.^65-68 Skin preparation may address patient risk factors. Hair clipping is preferred to shaving, which may cause microabrasions and increased susceptibility to skin flora.69 Patients should be prepared with antiseptic solution. One randomized controlled trial found that 2% chlorhexidine gluconate mixed with 70% isopropyl alcohol was superior to 10% povidone-iodine in preventing SSIs.⁷⁰ However, a recent cohort study showed a lower rate of superficial wound infections when 1% povidone-iodine (vs 0.5% chlorhexidine) was used with alcohol.⁷¹ This finding may indicate the need for alcohol preparation, higher concentrations of chlorhexidine, or both.

Proper scrubbing and protective gear are needed to reduce surgeon risk factors. Hand washing is a routine part of any surgery. Alcohol-based hand scrubs are as effective as hand scrubbing.65 They reduce local skin flora by 95% immediately and by 99% with repeated applications.⁷² Lidwell and colleagues⁷³ found a 75% reduction in infection when body exhaust suits were used in combination with laminar flow in a multicenter randomized controlled trial of 8052 patients. Sterile draping with impermeable drapes should be done over properly prepared skin. Ioban drapes (3M) are often used as a protective barrier. Interestingly, a Cochrane review found no benefit in using plastic adhesives impregnated with iodine over sterilely prepared skin.⁷⁴

Operative Considerations

Surgical gloves become contaminated in almost one third of cases, half the time during draping.⁷⁵ For this reason, many surgeons change gloves after draping. In addition, double gloving prevents a breech of aseptic technique should the outer glove become perforated.⁷⁶ Demircay and colleagues⁷⁷ assessed double latex gloving in arthroplasty and found the outer and inner gloves perforated in 18.4% and 8.4% of cases, respectively. Punctures are most common along the nondominant index finger, and then the dominant thumb.^77,78 Perforation is more common when 2 latex gloves are worn—vs 1 latex glove plus an outer cloth glove—and the chance of perforation increases with surgery duration. The inner glove may become punctured in up to 100% of operations that last over 3 hours.⁷⁹ Although Dodds and colleagues⁸⁰ found no change in bacterial counts on surgeons’ hands or gloves after perforation, precautions are still recommended. Al-Maiyah and colleagues⁸¹ went as far as to recommend glove changes at 20-minute intervals and before cementation.

Surgical instruments can be sources of contamination. Some authors change the suction tip every hour to minimize the risk of deep wound infection.^82-85 Others change it before femoral canal preparation and prosthesis insertion during THA.⁸⁶ The splash basin is frequently contaminated, and instruments placed in it should not be returned to the operative field.⁸⁷ Hargrove and colleagues⁸⁸ suggested pulsatile lavage decreases PJI more than bulb syringe irrigation does, whereas others argued that high-pressure lavage allows bacteria to penetrate more deeply, which could lead to retention of more bacteria.⁸⁹ Minimizing operating room time was found by Kurtz and colleagues⁹⁰ and Peersman and colleagues⁹¹ to decrease PJI incidence. Carroll and colleagues⁷¹ correlated longer tourniquet use with a higher rate of infection after TKA; proposed mechanisms include local tissue hypoxia and lowered concentrations of prophylactic antibiotics.

Similarly, minimizing blood loss and transfusion needs is another strategy for preventing infection. Allogenic transfusion may increase the risk of PJI 2 times.^23,71,92 The mechanism seems to be immune system modulation by allogenic blood, which impairs microcirculation and oxygen delivery at the surgical site.^23,75 Transfusions should be approached with caution, and consideration given to preoperative optimization and autologous blood donation. Cherian and colleagues⁹³ reviewed different blood management strategies and found preoperative iron therapy, intravenous erythropoietin, and autologous blood donation to be equally effective in reducing the need for allogenic transfusions. Numerous studies of tranexamic acid, thrombin-based hemostatic matrix (Floseal; Baxter Inc), and bipolar sealer with radiofrequency ablation (Aquamantys; Medtronic Inc) have found no alterations in infection rates, but most have used calculated blood loss, not PJI, as the primary endpoint.^94-105 Antibiotic cement also can be used to block infection.^63,106-110 Although liquid gentamicin may weaken bone cement,¹¹¹ most antibiotics, including powdered tobramycin and vancomycin, do not weaken its fatigue strength.^111-114 A recent meta-analysis by Parvizi and colleagues¹¹⁵ revealed that deep infection rates dropped from 2.3% to 1.2% with use of antibiotic cement for primary THAs. Cummins and colleagues,¹¹⁶ however, reported the limited cost-effectiveness of antibiotic cement in primary arthroplasty. Performing povidone-iodine lavage at the end of the case may be a more inexpensive alternative. Brown and colleagues¹¹⁷ found that rinsing with dilute povidone-iodine (.35%) for 3 minutes significantly decreased the incidence of PJI.

Closure techniques and sutures have been a focus of much of the recent literature. Winiarsky and colleagues³⁴ advocated using a longer incision for obese patients and augmenting closure in fattier areas with vertical mattress retention sutures, which are removed after 5 days. A barbed monofilament suture (Quill; Angiotech Inc) is gaining in popularity. Laboratory research has shown that bacteria adhere less to barbed monofilament sutures than to braided sutures.¹¹⁸ Smith and colleagues¹¹⁹ found a statistically nonsignificant higher rate of wound complications with barbed monofilament sutures, whereas Ting and colleagues120 found no difference in complications. These studies were powered to detect differences in time and cost, not postoperative complications. Skin adhesive (Dermabond; Ethicon Inc), also used in closure, may be superior to staples in avoiding superficial skin abscesses.¹²¹ Although expensive, silver-impregnated dressing has antimicrobial activity that reduces PJI incidence by up to 74%.¹²² One brand of this dressing (Aquacel; ConvaTec Inc) has a polyurethane waterproof barrier that allows it to be worn for 7 days.

Three factors commonly mentioned in PJI prevention show little supporting evidence. Drains, which are often used, may create a passage for postoperative infection and are associated with increased transfusion needs.^123,124 Adding antibiotics to irrigation solution¹²⁵ and routinely changing scalpel blades^126-129 also have little supporting evidence. In 2014, the utility of changing scalpel blades after incision was studied by Lee and colleagues,¹³⁰ who reported persistence of Propionibacterium acnes in the dermal layer after skin preparation. Their study, however, was isolated to the upper back region, not the hip or knee.

Postoperative Risk Prevention

Most arthroplasty patients receive anticoagulation after surgery, but it must be used with caution. Large hematomas can predispose to wound complications. Parvizi and colleagues¹³¹ associated wound drainage, hematoma, and subsequent PJI with an INR above 1.5 in the early postoperative period. Therefore, balanced anticoagulation is crucial. Postoperative glucose control is also essential, particularly for patients with diabetes. Although preoperative blood glucose levels may or may not affect PJI risk,^15,17,132 postoperative blood glucose levels of 126 mg/dL or higher are strongly associated with joint infections.¹³³ Even nondiabetic patients with postoperative morning levels over 140 mg/dL are 3 times more likely to develop an infection.¹⁷

Efforts should be made to discharge patients as soon as it is safe to do so. With longer hospital stays, patients are more exposed to nosocomial organisms and increased antibiotic resistance.^5,23,134 Outpatient antibiotics should be considered for dental, gastrointestinal, and genitourinary procedures. Oral antibiotic prophylaxis is controversial, as there is some evidence that dental procedures increase the risk of PJI only minimally.^10,135-138

Conclusion

PJI is a potentially devastating complication of TJA. For this reason, much research has been devoted to proper diagnosis and treatment. Although the literature on PJI prophylaxis is abundant, there is relatively little consensus on appropriate PJI precautions. Preoperative considerations should include medical comorbidities, use of immunocompromising medications, obesity, nutritional factors, smoking, alcohol use, and MRSA colonization. Surgeons must have a consistent intraoperative method of antibiotic administration, skin preparation, scrubbing, draping, gloving, instrument exchange, blood loss management, cementing, and closure. In addition, monitoring of postoperative anticoagulation and blood glucose management is important. Having a thorough understanding of PJI risk factors may help reduce the incidence of this devastating complication.

Nearly 2% of patients who undergo total knee arthroplasty (TKA) or total hip arthroplasty (THA) develop a periprosthetic joint infection (PJI) within 20 years of surgery, and 41% of these infections occur within the first 2 years.¹ PJI is the most common cause of TKA failure and the third leading complication of THA.² The estimated total hospital cost of treating PJI increased from $320 million in 2001 to $566 million in 2009, which can be extrapolated to $1.62 billion in 2020.³ By 2030, the projected increase in demand for TKA and THA will be 673% and 174% of what it was in 2005, respectively.⁴ Treatment of PJI of the knee is estimated to cost 3 to 4 times more than a primary TKA, and the cost of revision THA for PJI is almost $6000 more than that of revision TKA for PJI.³

In this article, we review the numerous preoperative, intraoperative, and postoperative methods of decreasing PJI incidence after total joint arthroplasty (TJA).

Preoperative Risk Prevention

Medical Comorbidities

Preoperative medical optimization is a key element in PJI prevention (Table 1). An American Society of Anesthesiologists classification score of 3 or more has been associated with doubled risk for surgical site infections (SSIs) after THA.5 Autoimmune conditions confer a particularly higher risk. In a retrospective double-cohort study of 924 subjects, Bongartz and colleagues⁶ found that, compared with osteoarthritis, rheumatoid arthritis tripled the risk of PJI. Small case series originally suggested a higher risk of PJI in patients with psoriasis,^7,8 but more recent studies have contradicted that finding.^9,10 Nevertheless, psoriatic plaques have elevated bacterial counts,¹¹ and planned incisions should circumvent these areas.

Diabetes mellitus is a clear risk factor for PJI.^12-16 Regarding whether preoperative glucose control affects risk, findings have been mixed. Mraovic and colleagues¹⁷ showed preoperative hyperglycemia to be an independent risk factor; Jämsen and colleagues,¹⁵ in a single-center analysis of more than 7000 TJAs, suggested preoperative blood glucose levels were not independently associated with PJI; and Iorio and colleagues¹⁶ found no association between surgical infections and hemoglobin A1c levels.

TJA incidence is higher in patients with chronic kidney disease (CKD) than in the general population.¹⁸ Dialysis users have a post-THA PJI rate as high as 13% to 19%.^19,20 Early clinical data suggested that outcomes are improved in dialysis users who undergo renal transplant, but this finding recently has been questioned.^19,21 Deegan and colleagues²² found an increased PJA rate of 3.5% even in low-level CKD (stage 1, 2, or 3), but this may be confounded by the increased association of CKD with other PJI-predisposing comorbidities.

Given a higher incidence of urinary tract infections (UTIs) among patients with PJI, some surgeons think UTIs predispose to PJIs by hematogenous seeding.^12,23,24 Symptomatic UTIs should be cleared before surgery and confirmed on urinalysis. Obstructive symptoms should prompt urologic evaluation. As asymptomatic pyuria and bacteriuria (colony counts, >1 × 105/mL) do not predispose to PJI, patients without symptoms do not require intervention.^25,26 Past history of malignancy may also have a role in PJI. In a case-control study of the Mayo Clinic arthroplasty experience from 1969 to 1991, Berbari and colleagues¹ found an association between malignancy and PJI (odds ratio, 2.4). They theorized the immunosuppressive effects of cancer treatment might be responsible for this increased risk.

Immunocompromising Medications

Immunocompromising medications are modifiable and should be adjusted before surgery. Stopping any disease-modifying antirheumatic drug (DMARD) more than 4 weeks before surgery is not recommended.²⁷

Corticosteroid use can lead to immunosuppression and increased protein catabolism, which impairs soft-tissue healing. To avoid flares or adrenal insufficiency, however, chronic corticosteroid users should continue their regular doses perioperatively.²⁸ On the day of surgery, they should also receive a stress dose of hydrocortisone 50 to 75 mg (for primary arthroplasty) or 100 to 150 mg (for revision arthroplasty), followed by expeditious tapering over 1 to 2 days.²⁹ DMARDs are increasingly used by rheumatologists. One of the most effective DMARDs is methotrexate. Despite its immunocompromising activity, methotrexate should be continued perioperatively, as stopping for even 2 days may increase flare-related complications.³⁰ Hydroxychloroquine can be continued perioperatively and has even been shown, by Johnson and Charnley,³¹ to prevent deep vein thromboses. Sulfasalazine can also be continued perioperatively—but with caution, as it may elevate international normalized ratio (INR) levels in patients receiving warfarin.²⁹ Most other DMARDs should be temporarily discontinued. Leflunomide and interleukin 1 antagonists, such as anakinra, should be stopped 1 to 2 days before surgery and restarted 10 to 14 days after surgery.²⁹ Rituximab should be stopped 1 week before surgery and restarted 10 to 14 days after surgery. Tumor necrosis factor α inhibitors should be discontinued for 2 half-lives before and after surgery.³² Etanercept has a half-life of 3 to 5 days; infliximab, 8 to 10 days; and adalimumab, 10 to 13 days. Most surgeons schedule surgery for the end of a dosing cycle and discontinue these biologic agents for another 10 to 14 days after surgery.

Metabolic Factors

Obese patients are susceptible to longer surgeries, more extensive dissection, poorly vascularized subcutaneous tissue, and higher requirements of weight-adjusted antibiotic dosing.¹³ Body mass index (BMI) of 40 kg/m² or more (morbid obesity) and BMI over 50 kg/m² have been associated with 9 times and 21.3 times increased risk of PJI, respectively.^13,14 Delaying surgery with dietary consultation has been suggested,^33,34 and bariatric surgery before TKA may decrease infection rates by 3.5 times.³⁵

Nutritional markers are considered before arthroplasty. According to most laboratories, a serum transferrin level under 200 mg/dL, albumin level under 3.5 g/dL, and total lymphocyte count under 1500 cells/mm³ indicate malnourishment, which can increase the incidence of wound complications by 5 to 7 times.³⁶ Patients should also have sufficient protein, vitamin, and mineral supplementation, particularly vitamins A and C, zinc, and copper.³⁷Smokers who cease smoking at least 4 to 6 weeks before surgery lower their wound complication rate by up to 26%.^38,39 When nicotine leaves the bloodstream, vasodilation occurs, oxygenation improves, and the immune system recovers.³⁹ Studies have found more SSIs in patients who abuse alcohol,40 and numerous authors have confirmed this finding in the arthroplasty population.^24,41,42 Alcohol inhibits platelet function and may predispose to a postoperative hematoma. In contrast to smoking cessation evidence, evidence regarding alcohol interventions in preventing postoperative infections is less conclusive.^43,44

MRSA Colonization

Methicillin-resistant Staphylococcus aureus (MRSA) is a particularly difficult bacterium to eradicate in PJI. As the mean cost of treating a single case of MRSA-related prosthetic infection is $107,264 vs $68,053 for susceptible strains,45,46 many infection-containment strategies focus on addressing benign MRSA colonization before surgery.

MRSA is present in the nares of 25 million people in the United States. Nasal colonization increases the risk of bacteremia 4-fold⁴⁷ and SSI 2- to 9-fold.^48,49 Nasal swabs are analyzed with either a rapid polymerase chain reaction (PCR) test, which provides results in 2 hours, or a bacterial culture, which provides results in 1 to 4 days. The PCR test is more expensive.

Eradication of MRSA colonization is increasingly prevalent. Several Scandinavian countries have instituted strict practices by which patients are denied elective surgery until negative nasal swabs are obtained.⁴⁹ Nasal decontamination is one method of colonization reduction. Topical mupirocin, which yields eradication in 91% of nasal carriers immediately after treatment and in 87% after 4 weeks,⁵⁰ is effective in reducing SSI rates only when used in conjunction with a body wash, which is used to clean the axilla and groin.⁵¹ There is no consensus on optimal timing, but Bode and colleagues⁵² found a significant decrease in deep SSIs when decontamination occurred just 24 hours before surgery.

Povidone-iodine showers went out of favor with the realization that chlorhexidine gluconate acts longer on the skin surface.^53,54 Preoperative showers involve rinsing with liquid chlorhexidine soap 24 to 48 hours before surgery. However, chlorhexidine binds preferentially to the cotton in washcloths instead of the skin. Edmiston and colleagues^54,55 found that 4% chlorhexidine liquid soaps achieve much lower skin chlorhexidine concentrations than 2% polyester cloths do. Use of these “chlorhexidine wipes” the night before and the day of surgery has decreased PJI after TKA from 2.2% to 0.6%.^56,57

Intraoperative Risk Prevention

Preparation

Which preoperative antibiotic to use is one of the first operative considerations in PJI prophylaxis (Table 2). Cefazolin is recommended as a first-line agent for its excellent soft-tissue penetration, long half-life, and activity against gram-positive bacteria such as skin flora.⁵⁸ Clindamycin may be considered for patients allergic to β-lactam antibiotics. Vancomycin may be considered for adjunctive use with cephalosporins in cases of known MRSA colonization. Vancomycin infusion should be started earlier than infusion with other antibiotics, as vancomycin must be infused slowly and takes longer to become therapeutic.

Antibiotic dosing should be based on local antibiograms, adjusted dosing weight, or BMI.⁵⁹ For revision arthroplasty, preoperative prophylaxis should not be stopped out of fear of affecting operative cultures.⁶⁰ Some surgeons pause antibiotic use if a preoperative joint aspirate has not been obtained. Infusion within 1 hour of incision is part of the pay-for-performance guidelines established by the US Centers for Medicare & Medicaid Services.⁶¹ An antibiotic should be redosed if the operation will take longer than 2 half-lives of the drug.⁵⁹ Surgeons should consider administering a dose every 4 hours or whenever blood loss exceeds 1000 mL.⁶² Engesæter and colleagues⁶³ found that antibiotic prophylaxis was most effective given 4 times perioperatively (1 time before surgery, 3 times after surgery). Postoperative antibiotics should not be administered longer than 24 hours, as prolonged dosing confers no benefit.⁵⁸ Operating room conditions must be optimized for prophylaxis. More people and operating room traffic in nonsterile corridors increase contamination of instruments open to air.⁶⁴ Laminar airflow systems are commonly used. Although there is little dispute that laminar flow decreases the bacterial load of air, there are mixed results regarding its benefit in preventing PJI.^65-68 Skin preparation may address patient risk factors. Hair clipping is preferred to shaving, which may cause microabrasions and increased susceptibility to skin flora.69 Patients should be prepared with antiseptic solution. One randomized controlled trial found that 2% chlorhexidine gluconate mixed with 70% isopropyl alcohol was superior to 10% povidone-iodine in preventing SSIs.⁷⁰ However, a recent cohort study showed a lower rate of superficial wound infections when 1% povidone-iodine (vs 0.5% chlorhexidine) was used with alcohol.⁷¹ This finding may indicate the need for alcohol preparation, higher concentrations of chlorhexidine, or both.

Proper scrubbing and protective gear are needed to reduce surgeon risk factors. Hand washing is a routine part of any surgery. Alcohol-based hand scrubs are as effective as hand scrubbing.65 They reduce local skin flora by 95% immediately and by 99% with repeated applications.⁷² Lidwell and colleagues⁷³ found a 75% reduction in infection when body exhaust suits were used in combination with laminar flow in a multicenter randomized controlled trial of 8052 patients. Sterile draping with impermeable drapes should be done over properly prepared skin. Ioban drapes (3M) are often used as a protective barrier. Interestingly, a Cochrane review found no benefit in using plastic adhesives impregnated with iodine over sterilely prepared skin.⁷⁴

Operative Considerations

Surgical gloves become contaminated in almost one third of cases, half the time during draping.⁷⁵ For this reason, many surgeons change gloves after draping. In addition, double gloving prevents a breech of aseptic technique should the outer glove become perforated.⁷⁶ Demircay and colleagues⁷⁷ assessed double latex gloving in arthroplasty and found the outer and inner gloves perforated in 18.4% and 8.4% of cases, respectively. Punctures are most common along the nondominant index finger, and then the dominant thumb.^77,78 Perforation is more common when 2 latex gloves are worn—vs 1 latex glove plus an outer cloth glove—and the chance of perforation increases with surgery duration. The inner glove may become punctured in up to 100% of operations that last over 3 hours.⁷⁹ Although Dodds and colleagues⁸⁰ found no change in bacterial counts on surgeons’ hands or gloves after perforation, precautions are still recommended. Al-Maiyah and colleagues⁸¹ went as far as to recommend glove changes at 20-minute intervals and before cementation.

Surgical instruments can be sources of contamination. Some authors change the suction tip every hour to minimize the risk of deep wound infection.^82-85 Others change it before femoral canal preparation and prosthesis insertion during THA.⁸⁶ The splash basin is frequently contaminated, and instruments placed in it should not be returned to the operative field.⁸⁷ Hargrove and colleagues⁸⁸ suggested pulsatile lavage decreases PJI more than bulb syringe irrigation does, whereas others argued that high-pressure lavage allows bacteria to penetrate more deeply, which could lead to retention of more bacteria.⁸⁹ Minimizing operating room time was found by Kurtz and colleagues⁹⁰ and Peersman and colleagues⁹¹ to decrease PJI incidence. Carroll and colleagues⁷¹ correlated longer tourniquet use with a higher rate of infection after TKA; proposed mechanisms include local tissue hypoxia and lowered concentrations of prophylactic antibiotics.

Similarly, minimizing blood loss and transfusion needs is another strategy for preventing infection. Allogenic transfusion may increase the risk of PJI 2 times.^23,71,92 The mechanism seems to be immune system modulation by allogenic blood, which impairs microcirculation and oxygen delivery at the surgical site.^23,75 Transfusions should be approached with caution, and consideration given to preoperative optimization and autologous blood donation. Cherian and colleagues⁹³ reviewed different blood management strategies and found preoperative iron therapy, intravenous erythropoietin, and autologous blood donation to be equally effective in reducing the need for allogenic transfusions. Numerous studies of tranexamic acid, thrombin-based hemostatic matrix (Floseal; Baxter Inc), and bipolar sealer with radiofrequency ablation (Aquamantys; Medtronic Inc) have found no alterations in infection rates, but most have used calculated blood loss, not PJI, as the primary endpoint.^94-105 Antibiotic cement also can be used to block infection.^63,106-110 Although liquid gentamicin may weaken bone cement,¹¹¹ most antibiotics, including powdered tobramycin and vancomycin, do not weaken its fatigue strength.^111-114 A recent meta-analysis by Parvizi and colleagues¹¹⁵ revealed that deep infection rates dropped from 2.3% to 1.2% with use of antibiotic cement for primary THAs. Cummins and colleagues,¹¹⁶ however, reported the limited cost-effectiveness of antibiotic cement in primary arthroplasty. Performing povidone-iodine lavage at the end of the case may be a more inexpensive alternative. Brown and colleagues¹¹⁷ found that rinsing with dilute povidone-iodine (.35%) for 3 minutes significantly decreased the incidence of PJI.

Closure techniques and sutures have been a focus of much of the recent literature. Winiarsky and colleagues³⁴ advocated using a longer incision for obese patients and augmenting closure in fattier areas with vertical mattress retention sutures, which are removed after 5 days. A barbed monofilament suture (Quill; Angiotech Inc) is gaining in popularity. Laboratory research has shown that bacteria adhere less to barbed monofilament sutures than to braided sutures.¹¹⁸ Smith and colleagues¹¹⁹ found a statistically nonsignificant higher rate of wound complications with barbed monofilament sutures, whereas Ting and colleagues120 found no difference in complications. These studies were powered to detect differences in time and cost, not postoperative complications. Skin adhesive (Dermabond; Ethicon Inc), also used in closure, may be superior to staples in avoiding superficial skin abscesses.¹²¹ Although expensive, silver-impregnated dressing has antimicrobial activity that reduces PJI incidence by up to 74%.¹²² One brand of this dressing (Aquacel; ConvaTec Inc) has a polyurethane waterproof barrier that allows it to be worn for 7 days.

Three factors commonly mentioned in PJI prevention show little supporting evidence. Drains, which are often used, may create a passage for postoperative infection and are associated with increased transfusion needs.^123,124 Adding antibiotics to irrigation solution¹²⁵ and routinely changing scalpel blades^126-129 also have little supporting evidence. In 2014, the utility of changing scalpel blades after incision was studied by Lee and colleagues,¹³⁰ who reported persistence of Propionibacterium acnes in the dermal layer after skin preparation. Their study, however, was isolated to the upper back region, not the hip or knee.

Postoperative Risk Prevention

Most arthroplasty patients receive anticoagulation after surgery, but it must be used with caution. Large hematomas can predispose to wound complications. Parvizi and colleagues¹³¹ associated wound drainage, hematoma, and subsequent PJI with an INR above 1.5 in the early postoperative period. Therefore, balanced anticoagulation is crucial. Postoperative glucose control is also essential, particularly for patients with diabetes. Although preoperative blood glucose levels may or may not affect PJI risk,^15,17,132 postoperative blood glucose levels of 126 mg/dL or higher are strongly associated with joint infections.¹³³ Even nondiabetic patients with postoperative morning levels over 140 mg/dL are 3 times more likely to develop an infection.¹⁷

Efforts should be made to discharge patients as soon as it is safe to do so. With longer hospital stays, patients are more exposed to nosocomial organisms and increased antibiotic resistance.^5,23,134 Outpatient antibiotics should be considered for dental, gastrointestinal, and genitourinary procedures. Oral antibiotic prophylaxis is controversial, as there is some evidence that dental procedures increase the risk of PJI only minimally.^10,135-138

Conclusion

PJI is a potentially devastating complication of TJA. For this reason, much research has been devoted to proper diagnosis and treatment. Although the literature on PJI prophylaxis is abundant, there is relatively little consensus on appropriate PJI precautions. Preoperative considerations should include medical comorbidities, use of immunocompromising medications, obesity, nutritional factors, smoking, alcohol use, and MRSA colonization. Surgeons must have a consistent intraoperative method of antibiotic administration, skin preparation, scrubbing, draping, gloving, instrument exchange, blood loss management, cementing, and closure. In addition, monitoring of postoperative anticoagulation and blood glucose management is important. Having a thorough understanding of PJI risk factors may help reduce the incidence of this devastating complication.

Nearly 2% of patients who undergo total knee arthroplasty (TKA) or total hip arthroplasty (THA) develop a periprosthetic joint infection (PJI) within 20 years of surgery, and 41% of these infections occur within the first 2 years.¹ PJI is the most common cause of TKA failure and the third leading complication of THA.² The estimated total hospital cost of treating PJI increased from $320 million in 2001 to $566 million in 2009, which can be extrapolated to $1.62 billion in 2020.³ By 2030, the projected increase in demand for TKA and THA will be 673% and 174% of what it was in 2005, respectively.⁴ Treatment of PJI of the knee is estimated to cost 3 to 4 times more than a primary TKA, and the cost of revision THA for PJI is almost $6000 more than that of revision TKA for PJI.³

In this article, we review the numerous preoperative, intraoperative, and postoperative methods of decreasing PJI incidence after total joint arthroplasty (TJA).

Preoperative Risk Prevention

Medical Comorbidities

Preoperative medical optimization is a key element in PJI prevention (Table 1). An American Society of Anesthesiologists classification score of 3 or more has been associated with doubled risk for surgical site infections (SSIs) after THA.5 Autoimmune conditions confer a particularly higher risk. In a retrospective double-cohort study of 924 subjects, Bongartz and colleagues⁶ found that, compared with osteoarthritis, rheumatoid arthritis tripled the risk of PJI. Small case series originally suggested a higher risk of PJI in patients with psoriasis,^7,8 but more recent studies have contradicted that finding.^9,10 Nevertheless, psoriatic plaques have elevated bacterial counts,¹¹ and planned incisions should circumvent these areas.

Diabetes mellitus is a clear risk factor for PJI.^12-16 Regarding whether preoperative glucose control affects risk, findings have been mixed. Mraovic and colleagues¹⁷ showed preoperative hyperglycemia to be an independent risk factor; Jämsen and colleagues,¹⁵ in a single-center analysis of more than 7000 TJAs, suggested preoperative blood glucose levels were not independently associated with PJI; and Iorio and colleagues¹⁶ found no association between surgical infections and hemoglobin A1c levels.

TJA incidence is higher in patients with chronic kidney disease (CKD) than in the general population.¹⁸ Dialysis users have a post-THA PJI rate as high as 13% to 19%.^19,20 Early clinical data suggested that outcomes are improved in dialysis users who undergo renal transplant, but this finding recently has been questioned.^19,21 Deegan and colleagues²² found an increased PJA rate of 3.5% even in low-level CKD (stage 1, 2, or 3), but this may be confounded by the increased association of CKD with other PJI-predisposing comorbidities.

Given a higher incidence of urinary tract infections (UTIs) among patients with PJI, some surgeons think UTIs predispose to PJIs by hematogenous seeding.^12,23,24 Symptomatic UTIs should be cleared before surgery and confirmed on urinalysis. Obstructive symptoms should prompt urologic evaluation. As asymptomatic pyuria and bacteriuria (colony counts, >1 × 105/mL) do not predispose to PJI, patients without symptoms do not require intervention.^25,26 Past history of malignancy may also have a role in PJI. In a case-control study of the Mayo Clinic arthroplasty experience from 1969 to 1991, Berbari and colleagues¹ found an association between malignancy and PJI (odds ratio, 2.4). They theorized the immunosuppressive effects of cancer treatment might be responsible for this increased risk.

Immunocompromising Medications

Immunocompromising medications are modifiable and should be adjusted before surgery. Stopping any disease-modifying antirheumatic drug (DMARD) more than 4 weeks before surgery is not recommended.²⁷

Corticosteroid use can lead to immunosuppression and increased protein catabolism, which impairs soft-tissue healing. To avoid flares or adrenal insufficiency, however, chronic corticosteroid users should continue their regular doses perioperatively.²⁸ On the day of surgery, they should also receive a stress dose of hydrocortisone 50 to 75 mg (for primary arthroplasty) or 100 to 150 mg (for revision arthroplasty), followed by expeditious tapering over 1 to 2 days.²⁹ DMARDs are increasingly used by rheumatologists. One of the most effective DMARDs is methotrexate. Despite its immunocompromising activity, methotrexate should be continued perioperatively, as stopping for even 2 days may increase flare-related complications.³⁰ Hydroxychloroquine can be continued perioperatively and has even been shown, by Johnson and Charnley,³¹ to prevent deep vein thromboses. Sulfasalazine can also be continued perioperatively—but with caution, as it may elevate international normalized ratio (INR) levels in patients receiving warfarin.²⁹ Most other DMARDs should be temporarily discontinued. Leflunomide and interleukin 1 antagonists, such as anakinra, should be stopped 1 to 2 days before surgery and restarted 10 to 14 days after surgery.²⁹ Rituximab should be stopped 1 week before surgery and restarted 10 to 14 days after surgery. Tumor necrosis factor α inhibitors should be discontinued for 2 half-lives before and after surgery.³² Etanercept has a half-life of 3 to 5 days; infliximab, 8 to 10 days; and adalimumab, 10 to 13 days. Most surgeons schedule surgery for the end of a dosing cycle and discontinue these biologic agents for another 10 to 14 days after surgery.

Metabolic Factors

Obese patients are susceptible to longer surgeries, more extensive dissection, poorly vascularized subcutaneous tissue, and higher requirements of weight-adjusted antibiotic dosing.¹³ Body mass index (BMI) of 40 kg/m² or more (morbid obesity) and BMI over 50 kg/m² have been associated with 9 times and 21.3 times increased risk of PJI, respectively.^13,14 Delaying surgery with dietary consultation has been suggested,^33,34 and bariatric surgery before TKA may decrease infection rates by 3.5 times.³⁵

Nutritional markers are considered before arthroplasty. According to most laboratories, a serum transferrin level under 200 mg/dL, albumin level under 3.5 g/dL, and total lymphocyte count under 1500 cells/mm³ indicate malnourishment, which can increase the incidence of wound complications by 5 to 7 times.³⁶ Patients should also have sufficient protein, vitamin, and mineral supplementation, particularly vitamins A and C, zinc, and copper.³⁷Smokers who cease smoking at least 4 to 6 weeks before surgery lower their wound complication rate by up to 26%.^38,39 When nicotine leaves the bloodstream, vasodilation occurs, oxygenation improves, and the immune system recovers.³⁹ Studies have found more SSIs in patients who abuse alcohol,40 and numerous authors have confirmed this finding in the arthroplasty population.^24,41,42 Alcohol inhibits platelet function and may predispose to a postoperative hematoma. In contrast to smoking cessation evidence, evidence regarding alcohol interventions in preventing postoperative infections is less conclusive.^43,44

MRSA Colonization

Methicillin-resistant Staphylococcus aureus (MRSA) is a particularly difficult bacterium to eradicate in PJI. As the mean cost of treating a single case of MRSA-related prosthetic infection is $107,264 vs $68,053 for susceptible strains,45,46 many infection-containment strategies focus on addressing benign MRSA colonization before surgery.

MRSA is present in the nares of 25 million people in the United States. Nasal colonization increases the risk of bacteremia 4-fold⁴⁷ and SSI 2- to 9-fold.^48,49 Nasal swabs are analyzed with either a rapid polymerase chain reaction (PCR) test, which provides results in 2 hours, or a bacterial culture, which provides results in 1 to 4 days. The PCR test is more expensive.

Eradication of MRSA colonization is increasingly prevalent. Several Scandinavian countries have instituted strict practices by which patients are denied elective surgery until negative nasal swabs are obtained.⁴⁹ Nasal decontamination is one method of colonization reduction. Topical mupirocin, which yields eradication in 91% of nasal carriers immediately after treatment and in 87% after 4 weeks,⁵⁰ is effective in reducing SSI rates only when used in conjunction with a body wash, which is used to clean the axilla and groin.⁵¹ There is no consensus on optimal timing, but Bode and colleagues⁵² found a significant decrease in deep SSIs when decontamination occurred just 24 hours before surgery.

Povidone-iodine showers went out of favor with the realization that chlorhexidine gluconate acts longer on the skin surface.^53,54 Preoperative showers involve rinsing with liquid chlorhexidine soap 24 to 48 hours before surgery. However, chlorhexidine binds preferentially to the cotton in washcloths instead of the skin. Edmiston and colleagues^54,55 found that 4% chlorhexidine liquid soaps achieve much lower skin chlorhexidine concentrations than 2% polyester cloths do. Use of these “chlorhexidine wipes” the night before and the day of surgery has decreased PJI after TKA from 2.2% to 0.6%.^56,57

Intraoperative Risk Prevention

Preparation

Which preoperative antibiotic to use is one of the first operative considerations in PJI prophylaxis (Table 2). Cefazolin is recommended as a first-line agent for its excellent soft-tissue penetration, long half-life, and activity against gram-positive bacteria such as skin flora.⁵⁸ Clindamycin may be considered for patients allergic to β-lactam antibiotics. Vancomycin may be considered for adjunctive use with cephalosporins in cases of known MRSA colonization. Vancomycin infusion should be started earlier than infusion with other antibiotics, as vancomycin must be infused slowly and takes longer to become therapeutic.

Antibiotic dosing should be based on local antibiograms, adjusted dosing weight, or BMI.⁵⁹ For revision arthroplasty, preoperative prophylaxis should not be stopped out of fear of affecting operative cultures.⁶⁰ Some surgeons pause antibiotic use if a preoperative joint aspirate has not been obtained. Infusion within 1 hour of incision is part of the pay-for-performance guidelines established by the US Centers for Medicare & Medicaid Services.⁶¹ An antibiotic should be redosed if the operation will take longer than 2 half-lives of the drug.⁵⁹ Surgeons should consider administering a dose every 4 hours or whenever blood loss exceeds 1000 mL.⁶² Engesæter and colleagues⁶³ found that antibiotic prophylaxis was most effective given 4 times perioperatively (1 time before surgery, 3 times after surgery). Postoperative antibiotics should not be administered longer than 24 hours, as prolonged dosing confers no benefit.⁵⁸ Operating room conditions must be optimized for prophylaxis. More people and operating room traffic in nonsterile corridors increase contamination of instruments open to air.⁶⁴ Laminar airflow systems are commonly used. Although there is little dispute that laminar flow decreases the bacterial load of air, there are mixed results regarding its benefit in preventing PJI.^65-68 Skin preparation may address patient risk factors. Hair clipping is preferred to shaving, which may cause microabrasions and increased susceptibility to skin flora.69 Patients should be prepared with antiseptic solution. One randomized controlled trial found that 2% chlorhexidine gluconate mixed with 70% isopropyl alcohol was superior to 10% povidone-iodine in preventing SSIs.⁷⁰ However, a recent cohort study showed a lower rate of superficial wound infections when 1% povidone-iodine (vs 0.5% chlorhexidine) was used with alcohol.⁷¹ This finding may indicate the need for alcohol preparation, higher concentrations of chlorhexidine, or both.

Proper scrubbing and protective gear are needed to reduce surgeon risk factors. Hand washing is a routine part of any surgery. Alcohol-based hand scrubs are as effective as hand scrubbing.65 They reduce local skin flora by 95% immediately and by 99% with repeated applications.⁷² Lidwell and colleagues⁷³ found a 75% reduction in infection when body exhaust suits were used in combination with laminar flow in a multicenter randomized controlled trial of 8052 patients. Sterile draping with impermeable drapes should be done over properly prepared skin. Ioban drapes (3M) are often used as a protective barrier. Interestingly, a Cochrane review found no benefit in using plastic adhesives impregnated with iodine over sterilely prepared skin.⁷⁴

Operative Considerations

Surgical gloves become contaminated in almost one third of cases, half the time during draping.⁷⁵ For this reason, many surgeons change gloves after draping. In addition, double gloving prevents a breech of aseptic technique should the outer glove become perforated.⁷⁶ Demircay and colleagues⁷⁷ assessed double latex gloving in arthroplasty and found the outer and inner gloves perforated in 18.4% and 8.4% of cases, respectively. Punctures are most common along the nondominant index finger, and then the dominant thumb.^77,78 Perforation is more common when 2 latex gloves are worn—vs 1 latex glove plus an outer cloth glove—and the chance of perforation increases with surgery duration. The inner glove may become punctured in up to 100% of operations that last over 3 hours.⁷⁹ Although Dodds and colleagues⁸⁰ found no change in bacterial counts on surgeons’ hands or gloves after perforation, precautions are still recommended. Al-Maiyah and colleagues⁸¹ went as far as to recommend glove changes at 20-minute intervals and before cementation.

Surgical instruments can be sources of contamination. Some authors change the suction tip every hour to minimize the risk of deep wound infection.^82-85 Others change it before femoral canal preparation and prosthesis insertion during THA.⁸⁶ The splash basin is frequently contaminated, and instruments placed in it should not be returned to the operative field.⁸⁷ Hargrove and colleagues⁸⁸ suggested pulsatile lavage decreases PJI more than bulb syringe irrigation does, whereas others argued that high-pressure lavage allows bacteria to penetrate more deeply, which could lead to retention of more bacteria.⁸⁹ Minimizing operating room time was found by Kurtz and colleagues⁹⁰ and Peersman and colleagues⁹¹ to decrease PJI incidence. Carroll and colleagues⁷¹ correlated longer tourniquet use with a higher rate of infection after TKA; proposed mechanisms include local tissue hypoxia and lowered concentrations of prophylactic antibiotics.

Similarly, minimizing blood loss and transfusion needs is another strategy for preventing infection. Allogenic transfusion may increase the risk of PJI 2 times.^23,71,92 The mechanism seems to be immune system modulation by allogenic blood, which impairs microcirculation and oxygen delivery at the surgical site.^23,75 Transfusions should be approached with caution, and consideration given to preoperative optimization and autologous blood donation. Cherian and colleagues⁹³ reviewed different blood management strategies and found preoperative iron therapy, intravenous erythropoietin, and autologous blood donation to be equally effective in reducing the need for allogenic transfusions. Numerous studies of tranexamic acid, thrombin-based hemostatic matrix (Floseal; Baxter Inc), and bipolar sealer with radiofrequency ablation (Aquamantys; Medtronic Inc) have found no alterations in infection rates, but most have used calculated blood loss, not PJI, as the primary endpoint.^94-105 Antibiotic cement also can be used to block infection.^63,106-110 Although liquid gentamicin may weaken bone cement,¹¹¹ most antibiotics, including powdered tobramycin and vancomycin, do not weaken its fatigue strength.^111-114 A recent meta-analysis by Parvizi and colleagues¹¹⁵ revealed that deep infection rates dropped from 2.3% to 1.2% with use of antibiotic cement for primary THAs. Cummins and colleagues,¹¹⁶ however, reported the limited cost-effectiveness of antibiotic cement in primary arthroplasty. Performing povidone-iodine lavage at the end of the case may be a more inexpensive alternative. Brown and colleagues¹¹⁷ found that rinsing with dilute povidone-iodine (.35%) for 3 minutes significantly decreased the incidence of PJI.

Closure techniques and sutures have been a focus of much of the recent literature. Winiarsky and colleagues³⁴ advocated using a longer incision for obese patients and augmenting closure in fattier areas with vertical mattress retention sutures, which are removed after 5 days. A barbed monofilament suture (Quill; Angiotech Inc) is gaining in popularity. Laboratory research has shown that bacteria adhere less to barbed monofilament sutures than to braided sutures.¹¹⁸ Smith and colleagues¹¹⁹ found a statistically nonsignificant higher rate of wound complications with barbed monofilament sutures, whereas Ting and colleagues120 found no difference in complications. These studies were powered to detect differences in time and cost, not postoperative complications. Skin adhesive (Dermabond; Ethicon Inc), also used in closure, may be superior to staples in avoiding superficial skin abscesses.¹²¹ Although expensive, silver-impregnated dressing has antimicrobial activity that reduces PJI incidence by up to 74%.¹²² One brand of this dressing (Aquacel; ConvaTec Inc) has a polyurethane waterproof barrier that allows it to be worn for 7 days.

Three factors commonly mentioned in PJI prevention show little supporting evidence. Drains, which are often used, may create a passage for postoperative infection and are associated with increased transfusion needs.^123,124 Adding antibiotics to irrigation solution¹²⁵ and routinely changing scalpel blades^126-129 also have little supporting evidence. In 2014, the utility of changing scalpel blades after incision was studied by Lee and colleagues,¹³⁰ who reported persistence of Propionibacterium acnes in the dermal layer after skin preparation. Their study, however, was isolated to the upper back region, not the hip or knee.

Postoperative Risk Prevention

Most arthroplasty patients receive anticoagulation after surgery, but it must be used with caution. Large hematomas can predispose to wound complications. Parvizi and colleagues¹³¹ associated wound drainage, hematoma, and subsequent PJI with an INR above 1.5 in the early postoperative period. Therefore, balanced anticoagulation is crucial. Postoperative glucose control is also essential, particularly for patients with diabetes. Although preoperative blood glucose levels may or may not affect PJI risk,^15,17,132 postoperative blood glucose levels of 126 mg/dL or higher are strongly associated with joint infections.¹³³ Even nondiabetic patients with postoperative morning levels over 140 mg/dL are 3 times more likely to develop an infection.¹⁷

Efforts should be made to discharge patients as soon as it is safe to do so. With longer hospital stays, patients are more exposed to nosocomial organisms and increased antibiotic resistance.^5,23,134 Outpatient antibiotics should be considered for dental, gastrointestinal, and genitourinary procedures. Oral antibiotic prophylaxis is controversial, as there is some evidence that dental procedures increase the risk of PJI only minimally.^10,135-138

Conclusion

PJI is a potentially devastating complication of TJA. For this reason, much research has been devoted to proper diagnosis and treatment. Although the literature on PJI prophylaxis is abundant, there is relatively little consensus on appropriate PJI precautions. Preoperative considerations should include medical comorbidities, use of immunocompromising medications, obesity, nutritional factors, smoking, alcohol use, and MRSA colonization. Surgeons must have a consistent intraoperative method of antibiotic administration, skin preparation, scrubbing, draping, gloving, instrument exchange, blood loss management, cementing, and closure. In addition, monitoring of postoperative anticoagulation and blood glucose management is important. Having a thorough understanding of PJI risk factors may help reduce the incidence of this devastating complication.

Although elbow arthroscopy was first described in the 1930s, it has become increasingly popular in the last 30 years.¹ While initially considered as a tool for diagnosis and loose body removal, indications have expanded to include treatment of osteochondritis dissecans (OCD), treatment of lateral epicondylitis, fixation of fractures, and others.^2-5 Miyake and colleagues⁶ found a significant improvement in range of motion, both flexion and extension, and outcome scores when elbow arthroscopy was used to remove impinging osteophytes. Babaqi and colleagues⁷ found significant improvement in pain, satisfaction, and outcome scores in 31 patients who underwent elbow arthroscopy for lateral epicondylitis refractory to nonsurgical management. The technical difficulty of the procedure, lower frequency of pathology amenable to arthroscopic intervention, and potential neurovascular complications make the elbow less frequently evaluated with the arthroscope vs other joints, such as the knee and shoulder.^2,8,9

Geographic distribution of subjects undergoing elbow arthroscopy, the indications used, surgical techniques being performed, and their associated clinical outcomes have received little to no recognition in the peer-reviewed literature.10 Differences in the elbow arthroscopy literature include characteristics related to the patient (age, gender, hand dominance, duration of symptoms), study (level of evidence, number of subjects, number of participating centers, design), indication (lateral epicondylitis, loose bodies, olecranon osteophytes, OCD), surgical technique, and outcome. Evidence-based medicine and clinical practice guidelines direct surgeons in clinical decision-making. Payers investigate the cost of surgical interventions and the value that surgery may provide, while following trends in different surgical techniques. Regulatory agencies and associations emphasize subjective patient-reported outcomes as the primary outcome measured in high-quality trials. Thus, in discussion of complex surgical interventions such as elbow arthroscopy, it is important to characterize the studies, subjects, and surgeries across the world to understand the geographic similarities and differences to optimize care in this clinical situation.

The goal of this study was to perform a systematic review and meta-analysis of elbow arthroscopy literature to identify and compare the characteristics of the studies published, the subjects analyzed, and surgical techniques performed across continents and countries to answer these questions: “Across the world, what demographic of patients are undergoing elbow arthroscopy, what are the most common indications for elbow arthroscopy, and how good is the evidence?” The authors hypothesized that patients who undergo elbow arthroscopy will be largely age <40 years, the most common indication for elbow arthroscopy will be a release/débridement, and the evidence for elbow arthroscopy will be poor. Also, no significant differences will exist in elbow arthroscopy publications, subjects, outcomes, and techniques based on continent/country of publication.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹¹ Systematic review registration was performed using the International Prospective Register of Ongoing Systematic Reviews (PROSPERO; registration number, CRD42014010580; registration date, July 15, 2014).¹² Two study authors independently conducted the search on June 23, 2014 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm used was: (elbow) AND arthroscopy) NOT shoulder) NOT knee) NOT ankle) NOT wrist) NOT hip) NOT dog) NOT cadaver). English language Level I-IV evidence (2012 update by the Oxford Centre for Evidence-Based Medicine¹³) clinical studies were eligible for inclusion into this study. Abstracts were ineligible for inclusion. All references in selected studies were cross-referenced for inclusion if they were missed during the initial search. Duplicate subject publications within separate unique studies were not reported twice. The study with longer duration follow-up, higher level of evidence, greater number of subjects, or more detailed subject, surgical technique, or outcome reporting was retained for inclusion. Level V evidence reviews, expert opinion articles, letters to the editor, basic science, biomechanical studies, open elbow surgery, imaging, surgical technique, and classification studies were excluded.

All included patients underwent elbow arthroscopy for either intra- or extra-articular elbow pathology (ulnotrochlear osteoarthritis, lateral epicondylitis, rheumatoid arthritis, post-traumatic contracture, osteonecrosis of the capitellum or radial head, osteoid osteoma, and others). There was no minimum follow-up duration or rehabilitation requirement. The study and subject demographic parameters that we analyzed included year of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and elbows, elbow dominance, gender, age, body mass index, diagnoses treated, type of anesthesia (block or general), and surgical positioning. Postoperative splint application and pain management, and whether a continuous passive motion machine was used and whether a drain was placed were recorded. Clinical outcome scores were DASH (Disability of the Arm, Shoulder, and Hand), Morrey score, MEPS (Mayo Elbow Performance Score), Andrews-Carson score, Timmerman-Andrews score, LES (Liverpool Elbow Score), Tegner score, HSS (Hospital for Special Surgery Score), VAS (Visual Analog Scale), EFA (Elbow Functional Assessment), Short Form-12 (SF-12), Short Form-36 (SF-36), Kerlan-Jobe Orthopaedic Clinic (KJOC) Shoulder and Elbow Questionnaire, and MAESS (Modified Andrews Elbow Scoring System). Radiographs, computed tomography (CT), computed tomography arthrography (CTA), magnetic resonance imaging (MRI), and magnetic resonance arthrography (MRA) data were extracted when available. Range of motion (flexion, extension, supination, and pronation) and grip strength data, both preoperative and postoperative, were extracted when available. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁴

Statistical Analysis

Study descriptive statistics were calculated. Continuous variable data were reported as weighted means ± weighted standard deviations. Categorical variable data were reported as frequencies with percentages. For all statistical analysis either measured and calculated from study data extraction or directly reported from the individual studies, P < .05 was considered statistically significant. Study, subject, and surgical outcomes data were compared using 1-way analysis of variance (ANOVA) tests. Where applicable, study, subject, and surgical outcomes data were also compared using 2-sample and 2-proportion Z-test calculators with α .05 because of the difference in sample sizes between compared groups. To examine trends over time, Pearson’s correlation coefficients were calculated. For the purposes of analysis, the indications of “osteoarthritis,” “arthrofibrosis,” “loose body removal,” “ulnotrochlear osteoarthritis causing stiffness,” “post-traumatic contracture/stiffness,” and “post-operative elbow contracture” were combined into the indication “release and débridement.” For the 3 most common indications for arthroscopy (OCD, lateral epicondylitis, and release and débridement) data were combined into 5-year increments to overcome the smaller sample size within each of these categories, and Pearson’s correlation coefficients were calculated to determine if number of reported cases covaried with year period. Within these 3 diagnoses, ANOVA analyses were performed to determine whether the number of cases differed between continents and countries.

Results

A total of 353 studies were located, and, after implementation of the exclusion criteria, 112 studies were included in the final analysis (Figure 1; 3093 subjects; 3168 elbows; 64% male; mean age, 34.9 ± 14.68 years). There was a mean of 33.4 ± 26.02 months of follow-up, and 75% of surgeries involved the dominant elbow (Table 1). Most studies were level IV evidence (94.6%), had a low MCMS (mean 28.1 ± 8.06; poor rating), and were single-center investigations (94.6%). Most studies did not report financial conflicts of interest (56.3%) (Tables 1 and 2). From 1985 through 2014, the number of publications significantly increased with time (P = .004) among all continents. The MCMS was unchanged over time (P = .247) (Figure 2A), as was the level of evidence (P = .094) (Figure 2B). Conflicts of interest significantly increased with time (P = .025) (Figure 3).

Among continents, North America published the largest number of studies (54), and had the largest number of patients (1395) and elbow surgeries (1425) (Table 1). The United States published the largest number of studies (43%). There were no significant differences between age (P = .331), length of follow-up (P = .403), MCMS (P = .123), and level of evidence (P = .288) between continents. Of the 32 studies that reported the use of preoperative MRI, studies from Asia reported significantly more MRI scans than those from other continents (P = .040); there were no other significant differences between continents in reference to preoperative imaging studies or other demographic information.

The most common surgical indications were OCD (Figure 4), lateral epicondylitis (Figure 5), and release and débridement (Figure 6, Table 3; all studies listed indications). The number of reported cases for these 3 indications significantly increased over time (OCD P = .005, lateral epicondylitis P = .044, release and débridement P = .042) but did not significantly differ between regions (P > .05 in all cases).

Thirty-two (28.6%) studies reported the use of outcome measures (16 different outcome scores were used by the included studies). Asia reported outcome measures in 9 of 23 studies (39%), Europe in 12 of 35 studies (34%) and North America in 11 of 54 (20%) of studies. The MEPS was the most frequently used outcome score in 9.8% of studies, followed by VAS for pain in 5.3% of cases. North American studies reported a significantly higher increase in extension after elbow arthroscopy than Asia (P = .0432) (Figure 7), with no differences in flexion (P = .699), pronation (P = .376), or supination (P = .408). No significant differences were observed between continents in the type of anesthesia chosen (general anesthesia [P = .94] or regional anesthesia [P = .85]). Asia and Europe performed elbow arthroscopy most frequently in the lateral decubitus position, while North American studies most often used the supine position (Table 4).

Twenty (17.9%) studies reported the use of a postoperative splint, 12 (10.7%) studies reported use of a drain, 2 (1.79%) studies reported use of a hinged elbow brace, 9 (8.03%) studies reported use of a continuous passive motion machine postoperatively, and 3 (2.68%) studies reported use of an indwelling axillary catheter for postoperative pain management. Of 130 reported surgical complications (4.1%), the most frequent complication was transient sensory ulnar nerve palsy (1.5%), followed by persistent wound drainage (.76%), and transient sensory radial nerve palsy (.38%). Other reported complications included infection (.22%), transient sensory palsy of the median nerve (.19%), heterotopic ossification (.13%), complete transection of the ulnar nerve (.10%), loose body formation (.06%), hematoma formation (.06%), transient sensory palsy of the posterior interosseous (.06%), or anterior interosseous nerve (.03%), and complete transection of the radial (.03%), or median nerve (.03%).

Discussion

Elbow arthroscopy is an evolving surgical procedure that is used to treat intra- and extra-articular pathologies of the elbow. Outcomes of elbow arthroscopy for certain conditions have generally been reported as good, with improvements seen in pain, functional scores, and range of motion.^6,15-17 The authors’ hypotheses were mostly confirmed in that the average age of patients undergoing elbow arthroscopy was <40 years, release/débridement was one of the most common indications (along with lateral epicondylitis and OCD), and the general evidence for elbow arthroscopy was poor. Also, there were almost no differences between continents/countries related to patient indications, preoperative imaging, anesthesia choice, indications, postoperative protocols, and outcomes (although the number of studies that reported outcomes was low and could have skewed the results), with the exception of a higher number of preoperative MRI scans in Asia. Some of the notable findings of this study included: 1) the number of studies published on elbow arthroscopy is significantly increasing with time, despite a lack of improvement in the level of evidence; 2) the majority of studies on elbow arthroscopy do not report a surgical outcome score; and 3) the number of reported cases for the 3 most common indications significantly increased over time (OCD, P = .005; lateral epicondylitis, P = .044; release and débridement, P = .042) but did not differ between regions (P > .05 in all cases).

The indications for elbow arthroscopy have grown dramatically in the past 2 decades to include both intra- and extra-articular pathologies.¹⁸ Despite this increase in the number of indications for elbow arthroscopy, the study did not find a significant difference between countries/continents in the indications each used for elbow arthroscopy patients. There was a trend towards an increase in OCD cases in all continents, especially Asia (Figure 4), with time. Interestingly, while not statistically significant, there was variation among countries for surgical indications. In North America, removal of loose bodies accounts for 18% of patients, while in Europe this accounted for only 9% and in Asia for 1%. Post-traumatic stiffness was the indication for elbow arthroscopy in Europe in 19% of patients vs 7% in North America and 10% in Asia. In Asia, OCD accounts for 40% of arthroscopies, 7% in Europe, and 14% in North America (Figure 4) (Table 3).

This study demonstrated that the mean increase in elbow extension gained after surgery in North America was significantly greater when compared with studies from Asia, but the gain in flexion, pronation, and supination was similar across continents. The underlying cause of this difference in improvement in elbow extension between nations is unclear, although differences in diagnosis could account for some variation. This study did not examine differences in rehabilitation protocols, and certainly, it is plausible that protocol variations by country could account for some discrepancy. Furthermore, differences in functional needs may vary by continent and could have driven this result.

This study found no routine reporting of outcome scores by elbow arthroscopy studies from any continent, and that when outcome scores are reported, there is substantial inconsistency with regard to the actual scoring system used. No continent reported outcome scores in more than 40% of the studies published from that area, and the variation of outcome scores used, even from a single region, was large. This makes comparing clinical outcomes between studies difficult, even when performing identical procedures for identical indications, because there is no standardized method of reporting outcomes. To allow comparison of studies and generalizability of the results to different populations, a more standardized approach to outcome reporting needs to be instituted in the elbow arthroscopy literature. To date, there is no standardized score that has been validated for reporting clinical outcomes after elbow arthroscopy.19 Hence, it is not surprising that there were 16 different outcome scores reported throughout the 112 studies analyzed in this review, with the most frequent score, the MEPS, reported in a total of 10 studies. As medicine moves towards pay scales that are based on patient outcomes, it will become more important to define a clear outcome score that can be used to assess these patients, and reliably report scores. This will allow comparison of patients across nations to determine the best surgical treatment for different clinical problems. A validation study comparing these outcome scores to determine which score best summarizes the patient’s level of pain and function after surgery would be beneficial, because this could identify 1 score that could be standardized to allow comparison among reported outcomes.

Limitations

This study had several limitations. Despite having 2 authors search independently for studies, some studies could have been missed during the search process, introducing possible selection bias. Including only published studies could have introduced publication bias. Numerous studies did not report all the variables the authors examined. This could have skewed some results, and had additional variables been reported, could have altered the data to show significant differences in some measured variables. Because this study did not compare outcome measures for varying pathologies, conclusions cannot be drawn on the best treatment options for different indications. Case reports could have lowered the MCMS score and the average in studies reporting outcomes. Furthermore, the poor quality of the underlying data used in this study could limit the validity/generalizability of the results because this is a systematic review, and its level of evidence is only as high as the studies it includes. Because the primary aim was to report on demographics, this study did not examine concomitant pathology at the time of surgery or rehabilitation protocols.

Conclusion

The quantity, but not the quality, of arthroscopic elbow publications has significantly increased over time. Most patients undergo elbow arthroscopy for lateral epicondylitis, OCD, and release and débridement. Pathology and indications do not appear to differ geographically with more men undergoing elbow arthroscopy than women.

Although elbow arthroscopy was first described in the 1930s, it has become increasingly popular in the last 30 years.¹ While initially considered as a tool for diagnosis and loose body removal, indications have expanded to include treatment of osteochondritis dissecans (OCD), treatment of lateral epicondylitis, fixation of fractures, and others.^2-5 Miyake and colleagues⁶ found a significant improvement in range of motion, both flexion and extension, and outcome scores when elbow arthroscopy was used to remove impinging osteophytes. Babaqi and colleagues⁷ found significant improvement in pain, satisfaction, and outcome scores in 31 patients who underwent elbow arthroscopy for lateral epicondylitis refractory to nonsurgical management. The technical difficulty of the procedure, lower frequency of pathology amenable to arthroscopic intervention, and potential neurovascular complications make the elbow less frequently evaluated with the arthroscope vs other joints, such as the knee and shoulder.^2,8,9

Geographic distribution of subjects undergoing elbow arthroscopy, the indications used, surgical techniques being performed, and their associated clinical outcomes have received little to no recognition in the peer-reviewed literature.10 Differences in the elbow arthroscopy literature include characteristics related to the patient (age, gender, hand dominance, duration of symptoms), study (level of evidence, number of subjects, number of participating centers, design), indication (lateral epicondylitis, loose bodies, olecranon osteophytes, OCD), surgical technique, and outcome. Evidence-based medicine and clinical practice guidelines direct surgeons in clinical decision-making. Payers investigate the cost of surgical interventions and the value that surgery may provide, while following trends in different surgical techniques. Regulatory agencies and associations emphasize subjective patient-reported outcomes as the primary outcome measured in high-quality trials. Thus, in discussion of complex surgical interventions such as elbow arthroscopy, it is important to characterize the studies, subjects, and surgeries across the world to understand the geographic similarities and differences to optimize care in this clinical situation.

The goal of this study was to perform a systematic review and meta-analysis of elbow arthroscopy literature to identify and compare the characteristics of the studies published, the subjects analyzed, and surgical techniques performed across continents and countries to answer these questions: “Across the world, what demographic of patients are undergoing elbow arthroscopy, what are the most common indications for elbow arthroscopy, and how good is the evidence?” The authors hypothesized that patients who undergo elbow arthroscopy will be largely age <40 years, the most common indication for elbow arthroscopy will be a release/débridement, and the evidence for elbow arthroscopy will be poor. Also, no significant differences will exist in elbow arthroscopy publications, subjects, outcomes, and techniques based on continent/country of publication.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹¹ Systematic review registration was performed using the International Prospective Register of Ongoing Systematic Reviews (PROSPERO; registration number, CRD42014010580; registration date, July 15, 2014).¹² Two study authors independently conducted the search on June 23, 2014 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm used was: (elbow) AND arthroscopy) NOT shoulder) NOT knee) NOT ankle) NOT wrist) NOT hip) NOT dog) NOT cadaver). English language Level I-IV evidence (2012 update by the Oxford Centre for Evidence-Based Medicine¹³) clinical studies were eligible for inclusion into this study. Abstracts were ineligible for inclusion. All references in selected studies were cross-referenced for inclusion if they were missed during the initial search. Duplicate subject publications within separate unique studies were not reported twice. The study with longer duration follow-up, higher level of evidence, greater number of subjects, or more detailed subject, surgical technique, or outcome reporting was retained for inclusion. Level V evidence reviews, expert opinion articles, letters to the editor, basic science, biomechanical studies, open elbow surgery, imaging, surgical technique, and classification studies were excluded.

All included patients underwent elbow arthroscopy for either intra- or extra-articular elbow pathology (ulnotrochlear osteoarthritis, lateral epicondylitis, rheumatoid arthritis, post-traumatic contracture, osteonecrosis of the capitellum or radial head, osteoid osteoma, and others). There was no minimum follow-up duration or rehabilitation requirement. The study and subject demographic parameters that we analyzed included year of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and elbows, elbow dominance, gender, age, body mass index, diagnoses treated, type of anesthesia (block or general), and surgical positioning. Postoperative splint application and pain management, and whether a continuous passive motion machine was used and whether a drain was placed were recorded. Clinical outcome scores were DASH (Disability of the Arm, Shoulder, and Hand), Morrey score, MEPS (Mayo Elbow Performance Score), Andrews-Carson score, Timmerman-Andrews score, LES (Liverpool Elbow Score), Tegner score, HSS (Hospital for Special Surgery Score), VAS (Visual Analog Scale), EFA (Elbow Functional Assessment), Short Form-12 (SF-12), Short Form-36 (SF-36), Kerlan-Jobe Orthopaedic Clinic (KJOC) Shoulder and Elbow Questionnaire, and MAESS (Modified Andrews Elbow Scoring System). Radiographs, computed tomography (CT), computed tomography arthrography (CTA), magnetic resonance imaging (MRI), and magnetic resonance arthrography (MRA) data were extracted when available. Range of motion (flexion, extension, supination, and pronation) and grip strength data, both preoperative and postoperative, were extracted when available. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁴

Statistical Analysis

Study descriptive statistics were calculated. Continuous variable data were reported as weighted means ± weighted standard deviations. Categorical variable data were reported as frequencies with percentages. For all statistical analysis either measured and calculated from study data extraction or directly reported from the individual studies, P < .05 was considered statistically significant. Study, subject, and surgical outcomes data were compared using 1-way analysis of variance (ANOVA) tests. Where applicable, study, subject, and surgical outcomes data were also compared using 2-sample and 2-proportion Z-test calculators with α .05 because of the difference in sample sizes between compared groups. To examine trends over time, Pearson’s correlation coefficients were calculated. For the purposes of analysis, the indications of “osteoarthritis,” “arthrofibrosis,” “loose body removal,” “ulnotrochlear osteoarthritis causing stiffness,” “post-traumatic contracture/stiffness,” and “post-operative elbow contracture” were combined into the indication “release and débridement.” For the 3 most common indications for arthroscopy (OCD, lateral epicondylitis, and release and débridement) data were combined into 5-year increments to overcome the smaller sample size within each of these categories, and Pearson’s correlation coefficients were calculated to determine if number of reported cases covaried with year period. Within these 3 diagnoses, ANOVA analyses were performed to determine whether the number of cases differed between continents and countries.

Results

A total of 353 studies were located, and, after implementation of the exclusion criteria, 112 studies were included in the final analysis (Figure 1; 3093 subjects; 3168 elbows; 64% male; mean age, 34.9 ± 14.68 years). There was a mean of 33.4 ± 26.02 months of follow-up, and 75% of surgeries involved the dominant elbow (Table 1). Most studies were level IV evidence (94.6%), had a low MCMS (mean 28.1 ± 8.06; poor rating), and were single-center investigations (94.6%). Most studies did not report financial conflicts of interest (56.3%) (Tables 1 and 2). From 1985 through 2014, the number of publications significantly increased with time (P = .004) among all continents. The MCMS was unchanged over time (P = .247) (Figure 2A), as was the level of evidence (P = .094) (Figure 2B). Conflicts of interest significantly increased with time (P = .025) (Figure 3).

Among continents, North America published the largest number of studies (54), and had the largest number of patients (1395) and elbow surgeries (1425) (Table 1). The United States published the largest number of studies (43%). There were no significant differences between age (P = .331), length of follow-up (P = .403), MCMS (P = .123), and level of evidence (P = .288) between continents. Of the 32 studies that reported the use of preoperative MRI, studies from Asia reported significantly more MRI scans than those from other continents (P = .040); there were no other significant differences between continents in reference to preoperative imaging studies or other demographic information.

The most common surgical indications were OCD (Figure 4), lateral epicondylitis (Figure 5), and release and débridement (Figure 6, Table 3; all studies listed indications). The number of reported cases for these 3 indications significantly increased over time (OCD P = .005, lateral epicondylitis P = .044, release and débridement P = .042) but did not significantly differ between regions (P > .05 in all cases).

Thirty-two (28.6%) studies reported the use of outcome measures (16 different outcome scores were used by the included studies). Asia reported outcome measures in 9 of 23 studies (39%), Europe in 12 of 35 studies (34%) and North America in 11 of 54 (20%) of studies. The MEPS was the most frequently used outcome score in 9.8% of studies, followed by VAS for pain in 5.3% of cases. North American studies reported a significantly higher increase in extension after elbow arthroscopy than Asia (P = .0432) (Figure 7), with no differences in flexion (P = .699), pronation (P = .376), or supination (P = .408). No significant differences were observed between continents in the type of anesthesia chosen (general anesthesia [P = .94] or regional anesthesia [P = .85]). Asia and Europe performed elbow arthroscopy most frequently in the lateral decubitus position, while North American studies most often used the supine position (Table 4).

Twenty (17.9%) studies reported the use of a postoperative splint, 12 (10.7%) studies reported use of a drain, 2 (1.79%) studies reported use of a hinged elbow brace, 9 (8.03%) studies reported use of a continuous passive motion machine postoperatively, and 3 (2.68%) studies reported use of an indwelling axillary catheter for postoperative pain management. Of 130 reported surgical complications (4.1%), the most frequent complication was transient sensory ulnar nerve palsy (1.5%), followed by persistent wound drainage (.76%), and transient sensory radial nerve palsy (.38%). Other reported complications included infection (.22%), transient sensory palsy of the median nerve (.19%), heterotopic ossification (.13%), complete transection of the ulnar nerve (.10%), loose body formation (.06%), hematoma formation (.06%), transient sensory palsy of the posterior interosseous (.06%), or anterior interosseous nerve (.03%), and complete transection of the radial (.03%), or median nerve (.03%).

Discussion

Elbow arthroscopy is an evolving surgical procedure that is used to treat intra- and extra-articular pathologies of the elbow. Outcomes of elbow arthroscopy for certain conditions have generally been reported as good, with improvements seen in pain, functional scores, and range of motion.^6,15-17 The authors’ hypotheses were mostly confirmed in that the average age of patients undergoing elbow arthroscopy was <40 years, release/débridement was one of the most common indications (along with lateral epicondylitis and OCD), and the general evidence for elbow arthroscopy was poor. Also, there were almost no differences between continents/countries related to patient indications, preoperative imaging, anesthesia choice, indications, postoperative protocols, and outcomes (although the number of studies that reported outcomes was low and could have skewed the results), with the exception of a higher number of preoperative MRI scans in Asia. Some of the notable findings of this study included: 1) the number of studies published on elbow arthroscopy is significantly increasing with time, despite a lack of improvement in the level of evidence; 2) the majority of studies on elbow arthroscopy do not report a surgical outcome score; and 3) the number of reported cases for the 3 most common indications significantly increased over time (OCD, P = .005; lateral epicondylitis, P = .044; release and débridement, P = .042) but did not differ between regions (P > .05 in all cases).

The indications for elbow arthroscopy have grown dramatically in the past 2 decades to include both intra- and extra-articular pathologies.¹⁸ Despite this increase in the number of indications for elbow arthroscopy, the study did not find a significant difference between countries/continents in the indications each used for elbow arthroscopy patients. There was a trend towards an increase in OCD cases in all continents, especially Asia (Figure 4), with time. Interestingly, while not statistically significant, there was variation among countries for surgical indications. In North America, removal of loose bodies accounts for 18% of patients, while in Europe this accounted for only 9% and in Asia for 1%. Post-traumatic stiffness was the indication for elbow arthroscopy in Europe in 19% of patients vs 7% in North America and 10% in Asia. In Asia, OCD accounts for 40% of arthroscopies, 7% in Europe, and 14% in North America (Figure 4) (Table 3).

This study demonstrated that the mean increase in elbow extension gained after surgery in North America was significantly greater when compared with studies from Asia, but the gain in flexion, pronation, and supination was similar across continents. The underlying cause of this difference in improvement in elbow extension between nations is unclear, although differences in diagnosis could account for some variation. This study did not examine differences in rehabilitation protocols, and certainly, it is plausible that protocol variations by country could account for some discrepancy. Furthermore, differences in functional needs may vary by continent and could have driven this result.

This study found no routine reporting of outcome scores by elbow arthroscopy studies from any continent, and that when outcome scores are reported, there is substantial inconsistency with regard to the actual scoring system used. No continent reported outcome scores in more than 40% of the studies published from that area, and the variation of outcome scores used, even from a single region, was large. This makes comparing clinical outcomes between studies difficult, even when performing identical procedures for identical indications, because there is no standardized method of reporting outcomes. To allow comparison of studies and generalizability of the results to different populations, a more standardized approach to outcome reporting needs to be instituted in the elbow arthroscopy literature. To date, there is no standardized score that has been validated for reporting clinical outcomes after elbow arthroscopy.19 Hence, it is not surprising that there were 16 different outcome scores reported throughout the 112 studies analyzed in this review, with the most frequent score, the MEPS, reported in a total of 10 studies. As medicine moves towards pay scales that are based on patient outcomes, it will become more important to define a clear outcome score that can be used to assess these patients, and reliably report scores. This will allow comparison of patients across nations to determine the best surgical treatment for different clinical problems. A validation study comparing these outcome scores to determine which score best summarizes the patient’s level of pain and function after surgery would be beneficial, because this could identify 1 score that could be standardized to allow comparison among reported outcomes.

Limitations

This study had several limitations. Despite having 2 authors search independently for studies, some studies could have been missed during the search process, introducing possible selection bias. Including only published studies could have introduced publication bias. Numerous studies did not report all the variables the authors examined. This could have skewed some results, and had additional variables been reported, could have altered the data to show significant differences in some measured variables. Because this study did not compare outcome measures for varying pathologies, conclusions cannot be drawn on the best treatment options for different indications. Case reports could have lowered the MCMS score and the average in studies reporting outcomes. Furthermore, the poor quality of the underlying data used in this study could limit the validity/generalizability of the results because this is a systematic review, and its level of evidence is only as high as the studies it includes. Because the primary aim was to report on demographics, this study did not examine concomitant pathology at the time of surgery or rehabilitation protocols.

Conclusion

The quantity, but not the quality, of arthroscopic elbow publications has significantly increased over time. Most patients undergo elbow arthroscopy for lateral epicondylitis, OCD, and release and débridement. Pathology and indications do not appear to differ geographically with more men undergoing elbow arthroscopy than women.

Although elbow arthroscopy was first described in the 1930s, it has become increasingly popular in the last 30 years.¹ While initially considered as a tool for diagnosis and loose body removal, indications have expanded to include treatment of osteochondritis dissecans (OCD), treatment of lateral epicondylitis, fixation of fractures, and others.^2-5 Miyake and colleagues⁶ found a significant improvement in range of motion, both flexion and extension, and outcome scores when elbow arthroscopy was used to remove impinging osteophytes. Babaqi and colleagues⁷ found significant improvement in pain, satisfaction, and outcome scores in 31 patients who underwent elbow arthroscopy for lateral epicondylitis refractory to nonsurgical management. The technical difficulty of the procedure, lower frequency of pathology amenable to arthroscopic intervention, and potential neurovascular complications make the elbow less frequently evaluated with the arthroscope vs other joints, such as the knee and shoulder.^2,8,9

Geographic distribution of subjects undergoing elbow arthroscopy, the indications used, surgical techniques being performed, and their associated clinical outcomes have received little to no recognition in the peer-reviewed literature.10 Differences in the elbow arthroscopy literature include characteristics related to the patient (age, gender, hand dominance, duration of symptoms), study (level of evidence, number of subjects, number of participating centers, design), indication (lateral epicondylitis, loose bodies, olecranon osteophytes, OCD), surgical technique, and outcome. Evidence-based medicine and clinical practice guidelines direct surgeons in clinical decision-making. Payers investigate the cost of surgical interventions and the value that surgery may provide, while following trends in different surgical techniques. Regulatory agencies and associations emphasize subjective patient-reported outcomes as the primary outcome measured in high-quality trials. Thus, in discussion of complex surgical interventions such as elbow arthroscopy, it is important to characterize the studies, subjects, and surgeries across the world to understand the geographic similarities and differences to optimize care in this clinical situation.

The goal of this study was to perform a systematic review and meta-analysis of elbow arthroscopy literature to identify and compare the characteristics of the studies published, the subjects analyzed, and surgical techniques performed across continents and countries to answer these questions: “Across the world, what demographic of patients are undergoing elbow arthroscopy, what are the most common indications for elbow arthroscopy, and how good is the evidence?” The authors hypothesized that patients who undergo elbow arthroscopy will be largely age <40 years, the most common indication for elbow arthroscopy will be a release/débridement, and the evidence for elbow arthroscopy will be poor. Also, no significant differences will exist in elbow arthroscopy publications, subjects, outcomes, and techniques based on continent/country of publication.

Methods

A systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines using a PRISMA checklist.¹¹ Systematic review registration was performed using the International Prospective Register of Ongoing Systematic Reviews (PROSPERO; registration number, CRD42014010580; registration date, July 15, 2014).¹² Two study authors independently conducted the search on June 23, 2014 using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm used was: (elbow) AND arthroscopy) NOT shoulder) NOT knee) NOT ankle) NOT wrist) NOT hip) NOT dog) NOT cadaver). English language Level I-IV evidence (2012 update by the Oxford Centre for Evidence-Based Medicine¹³) clinical studies were eligible for inclusion into this study. Abstracts were ineligible for inclusion. All references in selected studies were cross-referenced for inclusion if they were missed during the initial search. Duplicate subject publications within separate unique studies were not reported twice. The study with longer duration follow-up, higher level of evidence, greater number of subjects, or more detailed subject, surgical technique, or outcome reporting was retained for inclusion. Level V evidence reviews, expert opinion articles, letters to the editor, basic science, biomechanical studies, open elbow surgery, imaging, surgical technique, and classification studies were excluded.

All included patients underwent elbow arthroscopy for either intra- or extra-articular elbow pathology (ulnotrochlear osteoarthritis, lateral epicondylitis, rheumatoid arthritis, post-traumatic contracture, osteonecrosis of the capitellum or radial head, osteoid osteoma, and others). There was no minimum follow-up duration or rehabilitation requirement. The study and subject demographic parameters that we analyzed included year of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and elbows, elbow dominance, gender, age, body mass index, diagnoses treated, type of anesthesia (block or general), and surgical positioning. Postoperative splint application and pain management, and whether a continuous passive motion machine was used and whether a drain was placed were recorded. Clinical outcome scores were DASH (Disability of the Arm, Shoulder, and Hand), Morrey score, MEPS (Mayo Elbow Performance Score), Andrews-Carson score, Timmerman-Andrews score, LES (Liverpool Elbow Score), Tegner score, HSS (Hospital for Special Surgery Score), VAS (Visual Analog Scale), EFA (Elbow Functional Assessment), Short Form-12 (SF-12), Short Form-36 (SF-36), Kerlan-Jobe Orthopaedic Clinic (KJOC) Shoulder and Elbow Questionnaire, and MAESS (Modified Andrews Elbow Scoring System). Radiographs, computed tomography (CT), computed tomography arthrography (CTA), magnetic resonance imaging (MRI), and magnetic resonance arthrography (MRA) data were extracted when available. Range of motion (flexion, extension, supination, and pronation) and grip strength data, both preoperative and postoperative, were extracted when available. Study methodological quality was evaluated using the Modified Coleman Methodology Score (MCMS).¹⁴

Statistical Analysis

Study descriptive statistics were calculated. Continuous variable data were reported as weighted means ± weighted standard deviations. Categorical variable data were reported as frequencies with percentages. For all statistical analysis either measured and calculated from study data extraction or directly reported from the individual studies, P < .05 was considered statistically significant. Study, subject, and surgical outcomes data were compared using 1-way analysis of variance (ANOVA) tests. Where applicable, study, subject, and surgical outcomes data were also compared using 2-sample and 2-proportion Z-test calculators with α .05 because of the difference in sample sizes between compared groups. To examine trends over time, Pearson’s correlation coefficients were calculated. For the purposes of analysis, the indications of “osteoarthritis,” “arthrofibrosis,” “loose body removal,” “ulnotrochlear osteoarthritis causing stiffness,” “post-traumatic contracture/stiffness,” and “post-operative elbow contracture” were combined into the indication “release and débridement.” For the 3 most common indications for arthroscopy (OCD, lateral epicondylitis, and release and débridement) data were combined into 5-year increments to overcome the smaller sample size within each of these categories, and Pearson’s correlation coefficients were calculated to determine if number of reported cases covaried with year period. Within these 3 diagnoses, ANOVA analyses were performed to determine whether the number of cases differed between continents and countries.

Results

A total of 353 studies were located, and, after implementation of the exclusion criteria, 112 studies were included in the final analysis (Figure 1; 3093 subjects; 3168 elbows; 64% male; mean age, 34.9 ± 14.68 years). There was a mean of 33.4 ± 26.02 months of follow-up, and 75% of surgeries involved the dominant elbow (Table 1). Most studies were level IV evidence (94.6%), had a low MCMS (mean 28.1 ± 8.06; poor rating), and were single-center investigations (94.6%). Most studies did not report financial conflicts of interest (56.3%) (Tables 1 and 2). From 1985 through 2014, the number of publications significantly increased with time (P = .004) among all continents. The MCMS was unchanged over time (P = .247) (Figure 2A), as was the level of evidence (P = .094) (Figure 2B). Conflicts of interest significantly increased with time (P = .025) (Figure 3).

Among continents, North America published the largest number of studies (54), and had the largest number of patients (1395) and elbow surgeries (1425) (Table 1). The United States published the largest number of studies (43%). There were no significant differences between age (P = .331), length of follow-up (P = .403), MCMS (P = .123), and level of evidence (P = .288) between continents. Of the 32 studies that reported the use of preoperative MRI, studies from Asia reported significantly more MRI scans than those from other continents (P = .040); there were no other significant differences between continents in reference to preoperative imaging studies or other demographic information.

The most common surgical indications were OCD (Figure 4), lateral epicondylitis (Figure 5), and release and débridement (Figure 6, Table 3; all studies listed indications). The number of reported cases for these 3 indications significantly increased over time (OCD P = .005, lateral epicondylitis P = .044, release and débridement P = .042) but did not significantly differ between regions (P > .05 in all cases).

Thirty-two (28.6%) studies reported the use of outcome measures (16 different outcome scores were used by the included studies). Asia reported outcome measures in 9 of 23 studies (39%), Europe in 12 of 35 studies (34%) and North America in 11 of 54 (20%) of studies. The MEPS was the most frequently used outcome score in 9.8% of studies, followed by VAS for pain in 5.3% of cases. North American studies reported a significantly higher increase in extension after elbow arthroscopy than Asia (P = .0432) (Figure 7), with no differences in flexion (P = .699), pronation (P = .376), or supination (P = .408). No significant differences were observed between continents in the type of anesthesia chosen (general anesthesia [P = .94] or regional anesthesia [P = .85]). Asia and Europe performed elbow arthroscopy most frequently in the lateral decubitus position, while North American studies most often used the supine position (Table 4).

Twenty (17.9%) studies reported the use of a postoperative splint, 12 (10.7%) studies reported use of a drain, 2 (1.79%) studies reported use of a hinged elbow brace, 9 (8.03%) studies reported use of a continuous passive motion machine postoperatively, and 3 (2.68%) studies reported use of an indwelling axillary catheter for postoperative pain management. Of 130 reported surgical complications (4.1%), the most frequent complication was transient sensory ulnar nerve palsy (1.5%), followed by persistent wound drainage (.76%), and transient sensory radial nerve palsy (.38%). Other reported complications included infection (.22%), transient sensory palsy of the median nerve (.19%), heterotopic ossification (.13%), complete transection of the ulnar nerve (.10%), loose body formation (.06%), hematoma formation (.06%), transient sensory palsy of the posterior interosseous (.06%), or anterior interosseous nerve (.03%), and complete transection of the radial (.03%), or median nerve (.03%).

Discussion

Elbow arthroscopy is an evolving surgical procedure that is used to treat intra- and extra-articular pathologies of the elbow. Outcomes of elbow arthroscopy for certain conditions have generally been reported as good, with improvements seen in pain, functional scores, and range of motion.^6,15-17 The authors’ hypotheses were mostly confirmed in that the average age of patients undergoing elbow arthroscopy was <40 years, release/débridement was one of the most common indications (along with lateral epicondylitis and OCD), and the general evidence for elbow arthroscopy was poor. Also, there were almost no differences between continents/countries related to patient indications, preoperative imaging, anesthesia choice, indications, postoperative protocols, and outcomes (although the number of studies that reported outcomes was low and could have skewed the results), with the exception of a higher number of preoperative MRI scans in Asia. Some of the notable findings of this study included: 1) the number of studies published on elbow arthroscopy is significantly increasing with time, despite a lack of improvement in the level of evidence; 2) the majority of studies on elbow arthroscopy do not report a surgical outcome score; and 3) the number of reported cases for the 3 most common indications significantly increased over time (OCD, P = .005; lateral epicondylitis, P = .044; release and débridement, P = .042) but did not differ between regions (P > .05 in all cases).

The indications for elbow arthroscopy have grown dramatically in the past 2 decades to include both intra- and extra-articular pathologies.¹⁸ Despite this increase in the number of indications for elbow arthroscopy, the study did not find a significant difference between countries/continents in the indications each used for elbow arthroscopy patients. There was a trend towards an increase in OCD cases in all continents, especially Asia (Figure 4), with time. Interestingly, while not statistically significant, there was variation among countries for surgical indications. In North America, removal of loose bodies accounts for 18% of patients, while in Europe this accounted for only 9% and in Asia for 1%. Post-traumatic stiffness was the indication for elbow arthroscopy in Europe in 19% of patients vs 7% in North America and 10% in Asia. In Asia, OCD accounts for 40% of arthroscopies, 7% in Europe, and 14% in North America (Figure 4) (Table 3).

This study demonstrated that the mean increase in elbow extension gained after surgery in North America was significantly greater when compared with studies from Asia, but the gain in flexion, pronation, and supination was similar across continents. The underlying cause of this difference in improvement in elbow extension between nations is unclear, although differences in diagnosis could account for some variation. This study did not examine differences in rehabilitation protocols, and certainly, it is plausible that protocol variations by country could account for some discrepancy. Furthermore, differences in functional needs may vary by continent and could have driven this result.

This study found no routine reporting of outcome scores by elbow arthroscopy studies from any continent, and that when outcome scores are reported, there is substantial inconsistency with regard to the actual scoring system used. No continent reported outcome scores in more than 40% of the studies published from that area, and the variation of outcome scores used, even from a single region, was large. This makes comparing clinical outcomes between studies difficult, even when performing identical procedures for identical indications, because there is no standardized method of reporting outcomes. To allow comparison of studies and generalizability of the results to different populations, a more standardized approach to outcome reporting needs to be instituted in the elbow arthroscopy literature. To date, there is no standardized score that has been validated for reporting clinical outcomes after elbow arthroscopy.19 Hence, it is not surprising that there were 16 different outcome scores reported throughout the 112 studies analyzed in this review, with the most frequent score, the MEPS, reported in a total of 10 studies. As medicine moves towards pay scales that are based on patient outcomes, it will become more important to define a clear outcome score that can be used to assess these patients, and reliably report scores. This will allow comparison of patients across nations to determine the best surgical treatment for different clinical problems. A validation study comparing these outcome scores to determine which score best summarizes the patient’s level of pain and function after surgery would be beneficial, because this could identify 1 score that could be standardized to allow comparison among reported outcomes.

Limitations

This study had several limitations. Despite having 2 authors search independently for studies, some studies could have been missed during the search process, introducing possible selection bias. Including only published studies could have introduced publication bias. Numerous studies did not report all the variables the authors examined. This could have skewed some results, and had additional variables been reported, could have altered the data to show significant differences in some measured variables. Because this study did not compare outcome measures for varying pathologies, conclusions cannot be drawn on the best treatment options for different indications. Case reports could have lowered the MCMS score and the average in studies reporting outcomes. Furthermore, the poor quality of the underlying data used in this study could limit the validity/generalizability of the results because this is a systematic review, and its level of evidence is only as high as the studies it includes. Because the primary aim was to report on demographics, this study did not examine concomitant pathology at the time of surgery or rehabilitation protocols.

Conclusion

The quantity, but not the quality, of arthroscopic elbow publications has significantly increased over time. Most patients undergo elbow arthroscopy for lateral epicondylitis, OCD, and release and débridement. Pathology and indications do not appear to differ geographically with more men undergoing elbow arthroscopy than women.

From the Division of Endocrinology, Diabetes, and Metabolism, University of Alabama at Birmingham, Birmingham, Alabama.

Abstract

• Objective: To review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.
• Methods: Review of the literature.
• Results: With the growing aging population, there is an increased number of people at risk of osteoporosis and fracture. Several medications are available that reduce the risk of fracture. However, adherence to osteoporosis medications is suboptimal. Factors related to nonadherence include dosing frequency, real side effects, and concern about potential side effects. Interventions that may improve adherence include clinician and patient education, less frequent and less complex dosing regimens, medication reminders, and adherence counseling.
• Conclusions: Improving adherence to osteoporosis medications is a complex and challenging issue. Considering and implementing strategies to improve adherence tailored to patient preferences may enhance long-term outcomes for patients with osteoporosis.

Osteoporosis is a chronic but asymptomatic disease that is characterized by an increased fragility of bones and increased risk of fractures. Hip and vertebral fractures are associated with the greatest morbidity and mortality. The prevalence of osteoporosis is estimated to be 10.3% in the US, with approximately 10.2 million adults over the age of 50 having osteoporosis based on 2010 census data and results from the National Health and Nutrition Examination Survey (NHANES) [1].

Several drugs are currently available for the treatment of osteoporosis, but adherence to treatment is low. Understanding the factors associated with low adherence and actions that can be taken to improve adherence to treatment is important given the large number of individuals with osteoporosis and the need to reduce the burden caused by fragility fracture. In this article, we review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.

Nonprescription Medications

Calcium

There have been several published studies over the last decade evaluating calcium supplementation and its efficacy in reducing fractures. Although these studies showed that calcium reduces bone turnover by 20% and slowed postmenopausal bone loss by one third [2,3], none of these studies or a recent systematic review [4] showed any degree of fracture risk reduction with calcium supplements alone.

Although some calcium intake may be good, too much calcium has the potential to cause harm, including an increased risk of nephrolithiasis and constipation/bloating. An analysis of the Women’s Health Initiative (WHI) study reported a 17% increase in renal calculi in women who received calcium and vitamin D supplements [5]. Another recently published meta-analysis showed a 43% increase in gastrointestinal complaints in patients who were taking calcium supplements [6]. The potential for increased cardiovascular risk with calcium supplements is controversial [7]. The WHI study did not show an increased occurrence of cardiovascular events among those taking calcium supplements [8]. In a different population, men who consumed more than 1000 mg per day of supplemental calcium had higher all-cause and cardiovascular disease-specific mortality [9]. Large, well-conducted randomized controlled trials will be needed to further elucidate the question of calcium supplementation and risk of cardiovascular disease.

Vitamin D

Deficiency of vitamin D is common with one study finding more than 90% of older adults deficient in vitamin D [10]. Vitamin D is essential for proper calcium metabolism and deficiency is known to induce secondary hyperparathyroidism. Studies in mouse models have also shown that normal vitamin D receptors in enterocytes are essential for normal bone mineralization [11,12]. A systematic Cochrane database review showed that vitamin D3 supplementation decreased mortality in elderly people living independently or in institutional care [13]. Vitamin D was administered for a weighted mean of 4.4 years. Vitamin D2, alfacalcidol, and calcitriol had no statistically significant beneficial effects on mortality. Vitamin D3 combined with calcium was associated with an increased risk of nephrolithiasis during a follow-up period of 1.25 to 7 years (relative risk [RR] 1.17, 95% confidence interval (CI) 1.02–1.34) [13]. The inconsistencies of published reports looking at benefits of vitamin D supplementation may be due in part to variability in compliance with taking the supplements and baseline vitamin D levels.

Two randomized controlled trials have shown that low vitamin D appears to be an independent predictor of fall risk, and vitamin D supplementation has been found to reduce this risk of falls, through improved musculo-skeletal function [14–16]. Thus, vitamin D may play a role in fracture risk reduction beyond direct bone effects.

Prescription Osteoporosis Treatments

Bisphosphonates

Bisphosphonates are the most commonly prescribed medication for osteoporosis. The efficacy of bisphosphonates to reduce fractures is well established. There are oral bisphosphonates, which can be dosed daily, weekly, or monthly, and intravenous bisphosphonates, which can be given every 3 months or annually. Side effects with this class of medications include gastrointestinal effects with the oral options in up to 20% to 30% of users [17]. With intravenous bisphosphonates, the greatest risk is an acute phase response, which can occur in up to 42% of patients [18]. The risk of an acute phase reaction is much lower with doses beyond the first dose and lower if patients have ever previously taken an oral bisphosphonate and/or receive acetaminophen prior to the infusion. Other potential side effects with all bisphosphonates include osteonecrosis of jaw (ONJ) and atypical subtrochanteric fractures. Post marketing studies have indicated that the incidence of ONJ is less than 2 per 100,000 patient-years among those taking bisphosphonates [19,20]. A number of database analyses have shown that ONJ-like lesions can also occur in older individuals with osteoporosis who have never been exposed to bisphosphonates [21]. A case series of an osteoporotic population showed that ONJ-like lesions are lower grade than those typically seen in cancer patients who usually are exposed to higher doses of bisphosphonates [22]. A study of Swedish older men and women reported that long-term use of bisphosphonates (4 years or more) was associated with an increased incidence of atypical fractures. The RR for women was 126.0 (95% CI, 55.1–288.1) after 4 years of bisphosphonates [23]. A U.S. health care database analysis reported that 90% of those with atypical fractures were bisphosphonate users, almost half were Asian (49%), and use beyond 6 years showed the greatest risk [24].

Non-Bisphosphonate Medications

Other osteoporosis medications include denosumab, raloxifene, estrogen, and teriparatide (calcitonin will not be discussed here). Newer options currently under study, including cathepsin K inhibitors and anti-sclerostin therapies, are not available in the United States.

Denosumab is a monoclonal antibody that interferes with the receptor activator of nuclear kappa B ligand (RANK-L), which is the principal stimulus for osteoclastogenesis. Denosumab is administered once every 6 months subcutaneously. Phase III trials of denosumab demonstrated a 68% reduction in vertebral fractures and 40% and 20% reduction in fractures at hip and non-vertebral sites, respectively [25]. Similar to bisphosphonates, other risks include atypical femoral fractures and ONJ. In addition, hypocalcemia, including severe, symptomatic hypocalcemia, has been reported at rates higher than initially reported in the original clinical trials [26]. Hypocalcemia can be severe, especially in patients who are deficient in vitamin D [10,27].

Estrogen is effective in reducing the risk of vertebral fractures. Selective estrogen receptor modulators (SERMs) have both estrogen agonist and antagonist effects. The SERM, raloxifene, has been used in osteoporosis for its antiresorptive effects through the estrogen receptor [28,29]. A newer SERM, bazedoxifene, has been studied in combination with conjugated estrogen and has been reported to improve bone mineral density and other symptoms of menopause, like vasomotor symptoms and vulvo-vaginal atrophy, but its efficacy in reducing fracture risk has not been demonstrated [30,31].

Teriparatide is an anabolic agent that works by stimulating osteoblastic bone formation which results in an increase in bone density and reduction in both vertebral and non-vertebral fracture risk. In women with postmenopausal osteoporosis, it is typically reserved for those with very low bone mineral density (BMD) or those who continue to have fractures despite a bisphosphonate [32]. Barriers to use of teriparatide include high cost, the need for daily injections, and approved use for a total of two years in a lifetime. There is also a theoretical risk of osteosarcoma shown in animal studies but human cases have not been reported when used for postmenopausal osteoporosis. Published studies have shown that combination zolendronate and teriparatide have additive benefits to spine and hip BMD [33]. Another study reported that the combination of denosumab and teriparatide resulted in additive effects, ie, an increase in lumbar, hip, and femoral neck BMD [34]. These combinations have not been studied in populations large enough or for long enough duration to evaluate fracture risk reduction.

Adherence to Osteoporosis Medications

Treatment of osteoporosis reduces risk of fracture, but the benefit of osteoporosis medications is dependent on adherence. Adherence is associated with improved clinical outcomes [35,36] as well as reduced costs and utilization [37,38]; however, adherence to osteoporosis medications is poor. In a meta-analysis of 24 observational studies conducted in large populations, overall adherence for all osteoporosis therapies ranged from approximately 40% to 70% [39]. A recent retrospective claims database analysis in the U.S. reported a 60% noncompliance rate among the 57,913 postmenopausal women prescribed bisphosphonates over 1 year [40]. Another administrative database analysis from a managed care population compared the 3 oral bisphosphonates (risedronate, ibandronate, and alendronate) and found a mean medication possession ratio (MPR) between 0.57–0.58 at 12 months, which dropped to 0.47–0.50 after 24 months and 0.44–0.47 after 36 months [41]. In an observational study of 3200 older women in the U.K. low adherence was self-reported in 8.5%, and 21.6% self-discontinued treatment within 2 years [42]. In a study of Medicare Advantage prescription drug plan members, a small but significant increase in adherence was seen after osteo-porosis treatment change but overall adherence remained low (51% MPR in the change cohort and vs. 44% in the no-change cohort at 24 months, P < 0.01) [43].

Some patients restart osteoporosis therapy after a prolonged lapse in medication use. In one study, re-initiation rates for bisphosphonate therapy among persons who discontinued were as high as 30% within 6 months and 50% within 2 years [44]. Predictors of treatment re-initiation included younger age, female sex, history of fracture, recent hip fracture, nursing home discharge, and BMD testing [44].

Factors that Impact Adherence

Understanding which patients are most likely to be compliant with medications can aid physicians when monitoring osteoporosis treatment responses. In a retro-spective claims analysis, older age was found to be a predictor of compliance: women 65 years and older were more likely to be compliant than younger patients (P = 0.012) [45]. Among women receiving denosumab, improved adherence was found among women with a family history of a parent with a hip fracture, and lower adherence was seen in those with higher age, decreased mobility, and further distance from the clinic where the medication was provided [46].

Major reasons for nonadherence include a fear of potential side effects, occurrence of real side effects, the complicated dosing regimens, and perceived lack of benefit from the medications due to the asymptomatic nature of osteoporosis. In the above noted observational study from the U.K., more than half of the nonadherent patients attributed their nonadherence to side effects (53.9%), with a smaller proportion reporting fear of potential side effects (20.5%) or trouble with the dosing regimen (8.0%)[42].

Patients may also be unwilling to continue to take an osteoporosis medication if a fracture develops while on it and if they are not otherwise provided evidence that the medication is working. In a study by Costa Paiva et al, an understanding and knowledge to osteoporosis was a prerequisite to adherence and the strongest predictor of knowledge was higher education level [47]. Factors that impaired adherence were lower socioeconomic status and presence of comorbidities [47]. In a phenomenological qualitative study, trust in a health care provider was the most common reason for patients’ decision to accept an osteoporosis medication, emphasizing the importance of physician-patient communication [48].

Interventions to Enhance Adherence

Current methods of improving adherence for chronic health problems are mostly complex and not very effective [49]. In a systematic review of interventions to improve medication adherence, only 37 out of 81 studies reported improved adherence in the treatment of chronic diseases, and multifaceted treatments were more likely to succeed [49]. Improving adherence to osteoporosis medications is a complex issue, and a number of interventions evaluated in systematic reviews have shown limited efficacy [50,51]. Simplification of dosing regimens have been found to have a significant impact in chronic disease management [52,53] as well as in some studies of osteoporosis medications.

Simplification of Dosing

Among women prescribed daily vs. weekly bisphosphonates, those on the weekly regimen had significantly higher compliance [54]. However, rates were suboptimal in both groups and more than 50% of women discontinued at 1 year [54]. In addition, in a meta-analysis of osteoporosis medication adherence, a nearly two-fold higher odds of discontinuation with daily vs. weekly bisphosphonates was seen (odds ratio 1.90, 95% CI 1.81–2.00) [55]. Likewise, in a retrospective study in Spain, nearly 85% of those started on a daily bisphosphonate stopped within a year [56], while discontinuation was significantly lower in those prescribed a weekly or monthly bisphosphonate or daily teriparatide; however, discontinuation was still nearly 50% in these groups [56].

Once monthly dosing may be preferred by some patients as there is less time involved in thinking about the disease being treated and a perception of lower likelihood of side effects. In one study, postmenopausal women who had previously stopped oral bisphosphonates due to GI side effects had high adherence rates after self-selecting either monthly oral or quarterly intravenous ibandronate therapy [57]. However, not all studies show significant differences in adherence between weekly and monthly preparations [58–60].

The newer parenteral treatment options that can be given every 6 months or once yearly have the potential to significantly improve adherence. Once a year parenteral administration of a bisphosphonate was preferred over once-weekly oral administration, according to a 1-year study in patients with low bone density previously treated with alendronate [61]. A recent study that looked at persistence with an infusion of zolendronic acid in Taiwanese patients for 48 months found that 85% of patients received at least 2 infusions [62]. In patients treated with denosumab in 4 European countries, adherence and persistence at 12 months were consistently > 80% [46]. Persistence in this study was defined as receiving the subsequent injection within 6 months ± 8 weeks of the previous injection; adherence was defined as receiving 2 consecutive injections within 6 months ± 4 weeks of each other [46].

In a study by Cramer et al, increased adherence and persistence was seen with weekly alendronate compared daily alendronate at the end of 12 months [54]. Similar results were seen in a large longitudinal cohort study of weekly vs. daily bisphosphonates but less than 50% of patients were adherent with the weekly regimen [63]. When once monthly preparations of bisphosphonates became available, studies continued to support a patient preference for less frequently dosed bisphosphonates, with the majority of patients preferring monthly over weekly dosed medications [64–66].

The availability of quarterly ibandronate and yearly zoledronic acid infusions have further simplified dosing. In large, randomized, multicenter studies, patients consistently expressed a preference for yearly infusions over a weekly oral medication [61,67]. Adherence and persistence to osteoporosis medications was also greater in women receiving intravenous ibandronate compared to those receiving oral alendronate [68,69]. However, a study by Curtis et al showed low persistence with intravenous bisphosphonates in a Medicare population [70]. A possible reason for the lower adherence in this population was postulated to include the provision of the infusions at an outpatient center rather than a physician office. Automated nursing reminders with either phone calls or emails have the potential to mitigate the problem of persistence with this less frequent regimen [71,72]. In a review of patient preferences, less frequent dosing of medications was a common desire, but further generalizability were limited, emphasizing the need to individualize treatment [73].

Patient-Provider Communication

Individualizing treatment with better patient-provider communication and identification of potential barriers may increase compliance [74]. In one study, increasing patient participation in determining the treatment option was associated with improved patient adherence [57]. A systematic review of literature on interventions to improve adherence found that periodic follow-up interaction between patients and their health professionals also improved adherence [50]. Positive reinforcement via physician-patient discussion of either bone turnover markers or bone mineral density test results has also been found to improve long-term adherence with osteoporosis medications [71,75].

Better perceived physician knowledge may help with patient adherence. A study by Pickney et al reported that the patient confidence in their health care providers has influence on improved adherence, and patients were more likely to comply when the medications were prescribed by a specialist rather than a general practitioner [76].

Education, Reminders, Phone-Based

Improving patient knowledge of osteoporosis, especially with education using visual aids, may help with improving adherence [47]. In a randomized controlled trial at a single health management organization, an interactive voice response phone call plus a letter 1 week later increased the rate of obtaining a prescribed oral bisphosphonate in the intervention group (48.8% vs. 30.5% control; OR 2.3, 95% CI 1.34–3.94) when adjusted for age, sex, prior BMD, and fracture [77]. Use of an encounter decision aid also improved knowledge of osteoporosis medication options and led to a doubling of medication prescription attainment. However, adherence at 6 months was not improved [78].

Pill reminders in the form of text messages, paging systems on medication devices, and alarm beeps have been studied in patients with chronic diseases, and these technologies could be utilized for osteoporosis treatment [79–81]. A study of smart phone applications showed that many apps help with adherence, especially in noncompliant patients [82]. The researchers reported that of apps studied, MyMedSchedule, MyMeds, and RxmindMe were among the most highly rated due to their ease of use and enhanced functions. Solomon et all studied the effectiveness of a telephone-based counseling program using motivational interviewing in a large randomized study. They found no significant improvement in adherence to an osteoporosis regimen with the telephonic motivational interview compared to mailed educational materials (control group) (P = 0.07) [83]. In a 12-month multicenter, prospective randomized study, Bianchi et al examined the effectiveness of an intervention of reminders or reminders plus phone calls and meetings at the referral center in postmenopausal women initiating an oral osteoporosis prescription. No significant difference was seen in adherence at 12 months compared to standard care [84]. Adherence among the entire cohort, however, was very high [84]

Pharmacist-Based

The role of pharmacists in the treatment of chronic diseases, including osteoporosis, has been studied and shown to be cost-effective. In a study by van Boven et al, an algorithm was designed to detect patients with nonadherence and then tailor an intervention that consisted of structured counseling and active monitoring by pharmacists in initial and continuous phases [85]. This effort-intensive intervention resulted in reduced discontinuation of bisphosphonates after 12 months (reduction from 31.7% to 16.2% at 12 months) [85]. Despite the effort required, findings from the study support overall cost-effectiveness of this intervention [85]. A randomized controlled study by Lai et al showed that pharmacists can play a role in improving medication adherence through counseling patients on the importance of adherence, side effects, and goals of therapy [86]. The same authors also showed that involvement of a clinical pharmacist in the care of patients helped to further improve patient knowledge of medications and osteoporosis treatments, resolve medication-related concerns, and improve overall quality of life [87]. Such pharmacist-led interventions would require pharmacists to understand their role and the potential for drug holidays in the course of osteoporosis treatments and not mislabel patients as nonadherent when in fact purposefully holding osteoporosis medications [88].

Conclusion

Osteoporosis is a growing problem with increasing numbers of patients at risk for osteoporosis and related fractures. Currently available osteoporosis medications have shown clear benefit in reducing fracture risk; however, adherence to these therapies is required to obtain benefit. Unfortunately, osteoporosis medications have several limitations to full compliance, particularly the oral treatment options, including known possible side effects acutely and chronically, potential/feared side effects, irregular dosing intervals, complicated dosing instructions, and absence of an immediate recognizable benefit/effect. Improving adherence is complex [89] and tailoring to individual patients is of importance. Successful techniques for improving adherence may include a focus on physician-patient communication, use of the less frequently dosed medications, various medication reminders, use of available technology, and use of pharmacists for patient counseling and monitoring. Recognition of this common problem by clinicians is of utmost importance.

Corresponding author: Amy H. Warriner, MD, The University of Alabama at Birmingham, Division of Endocrinology, Diabetes and Metabolism, 702 Faculty Office Tower, 510 20th Street South, Birmingham, AL 35294, [email protected].

Financial disclosures: None.

From the Division of Endocrinology, Diabetes, and Metabolism, University of Alabama at Birmingham, Birmingham, Alabama.

Abstract

• Objective: To review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.
• Methods: Review of the literature.
• Results: With the growing aging population, there is an increased number of people at risk of osteoporosis and fracture. Several medications are available that reduce the risk of fracture. However, adherence to osteoporosis medications is suboptimal. Factors related to nonadherence include dosing frequency, real side effects, and concern about potential side effects. Interventions that may improve adherence include clinician and patient education, less frequent and less complex dosing regimens, medication reminders, and adherence counseling.
• Conclusions: Improving adherence to osteoporosis medications is a complex and challenging issue. Considering and implementing strategies to improve adherence tailored to patient preferences may enhance long-term outcomes for patients with osteoporosis.

Osteoporosis is a chronic but asymptomatic disease that is characterized by an increased fragility of bones and increased risk of fractures. Hip and vertebral fractures are associated with the greatest morbidity and mortality. The prevalence of osteoporosis is estimated to be 10.3% in the US, with approximately 10.2 million adults over the age of 50 having osteoporosis based on 2010 census data and results from the National Health and Nutrition Examination Survey (NHANES) [1].

Several drugs are currently available for the treatment of osteoporosis, but adherence to treatment is low. Understanding the factors associated with low adherence and actions that can be taken to improve adherence to treatment is important given the large number of individuals with osteoporosis and the need to reduce the burden caused by fragility fracture. In this article, we review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.

Nonprescription Medications

Calcium

There have been several published studies over the last decade evaluating calcium supplementation and its efficacy in reducing fractures. Although these studies showed that calcium reduces bone turnover by 20% and slowed postmenopausal bone loss by one third [2,3], none of these studies or a recent systematic review [4] showed any degree of fracture risk reduction with calcium supplements alone.

Although some calcium intake may be good, too much calcium has the potential to cause harm, including an increased risk of nephrolithiasis and constipation/bloating. An analysis of the Women’s Health Initiative (WHI) study reported a 17% increase in renal calculi in women who received calcium and vitamin D supplements [5]. Another recently published meta-analysis showed a 43% increase in gastrointestinal complaints in patients who were taking calcium supplements [6]. The potential for increased cardiovascular risk with calcium supplements is controversial [7]. The WHI study did not show an increased occurrence of cardiovascular events among those taking calcium supplements [8]. In a different population, men who consumed more than 1000 mg per day of supplemental calcium had higher all-cause and cardiovascular disease-specific mortality [9]. Large, well-conducted randomized controlled trials will be needed to further elucidate the question of calcium supplementation and risk of cardiovascular disease.

Vitamin D

Deficiency of vitamin D is common with one study finding more than 90% of older adults deficient in vitamin D [10]. Vitamin D is essential for proper calcium metabolism and deficiency is known to induce secondary hyperparathyroidism. Studies in mouse models have also shown that normal vitamin D receptors in enterocytes are essential for normal bone mineralization [11,12]. A systematic Cochrane database review showed that vitamin D3 supplementation decreased mortality in elderly people living independently or in institutional care [13]. Vitamin D was administered for a weighted mean of 4.4 years. Vitamin D2, alfacalcidol, and calcitriol had no statistically significant beneficial effects on mortality. Vitamin D3 combined with calcium was associated with an increased risk of nephrolithiasis during a follow-up period of 1.25 to 7 years (relative risk [RR] 1.17, 95% confidence interval (CI) 1.02–1.34) [13]. The inconsistencies of published reports looking at benefits of vitamin D supplementation may be due in part to variability in compliance with taking the supplements and baseline vitamin D levels.

Two randomized controlled trials have shown that low vitamin D appears to be an independent predictor of fall risk, and vitamin D supplementation has been found to reduce this risk of falls, through improved musculo-skeletal function [14–16]. Thus, vitamin D may play a role in fracture risk reduction beyond direct bone effects.

Prescription Osteoporosis Treatments

Bisphosphonates

Bisphosphonates are the most commonly prescribed medication for osteoporosis. The efficacy of bisphosphonates to reduce fractures is well established. There are oral bisphosphonates, which can be dosed daily, weekly, or monthly, and intravenous bisphosphonates, which can be given every 3 months or annually. Side effects with this class of medications include gastrointestinal effects with the oral options in up to 20% to 30% of users [17]. With intravenous bisphosphonates, the greatest risk is an acute phase response, which can occur in up to 42% of patients [18]. The risk of an acute phase reaction is much lower with doses beyond the first dose and lower if patients have ever previously taken an oral bisphosphonate and/or receive acetaminophen prior to the infusion. Other potential side effects with all bisphosphonates include osteonecrosis of jaw (ONJ) and atypical subtrochanteric fractures. Post marketing studies have indicated that the incidence of ONJ is less than 2 per 100,000 patient-years among those taking bisphosphonates [19,20]. A number of database analyses have shown that ONJ-like lesions can also occur in older individuals with osteoporosis who have never been exposed to bisphosphonates [21]. A case series of an osteoporotic population showed that ONJ-like lesions are lower grade than those typically seen in cancer patients who usually are exposed to higher doses of bisphosphonates [22]. A study of Swedish older men and women reported that long-term use of bisphosphonates (4 years or more) was associated with an increased incidence of atypical fractures. The RR for women was 126.0 (95% CI, 55.1–288.1) after 4 years of bisphosphonates [23]. A U.S. health care database analysis reported that 90% of those with atypical fractures were bisphosphonate users, almost half were Asian (49%), and use beyond 6 years showed the greatest risk [24].

Non-Bisphosphonate Medications

Other osteoporosis medications include denosumab, raloxifene, estrogen, and teriparatide (calcitonin will not be discussed here). Newer options currently under study, including cathepsin K inhibitors and anti-sclerostin therapies, are not available in the United States.

Denosumab is a monoclonal antibody that interferes with the receptor activator of nuclear kappa B ligand (RANK-L), which is the principal stimulus for osteoclastogenesis. Denosumab is administered once every 6 months subcutaneously. Phase III trials of denosumab demonstrated a 68% reduction in vertebral fractures and 40% and 20% reduction in fractures at hip and non-vertebral sites, respectively [25]. Similar to bisphosphonates, other risks include atypical femoral fractures and ONJ. In addition, hypocalcemia, including severe, symptomatic hypocalcemia, has been reported at rates higher than initially reported in the original clinical trials [26]. Hypocalcemia can be severe, especially in patients who are deficient in vitamin D [10,27].

Estrogen is effective in reducing the risk of vertebral fractures. Selective estrogen receptor modulators (SERMs) have both estrogen agonist and antagonist effects. The SERM, raloxifene, has been used in osteoporosis for its antiresorptive effects through the estrogen receptor [28,29]. A newer SERM, bazedoxifene, has been studied in combination with conjugated estrogen and has been reported to improve bone mineral density and other symptoms of menopause, like vasomotor symptoms and vulvo-vaginal atrophy, but its efficacy in reducing fracture risk has not been demonstrated [30,31].

Teriparatide is an anabolic agent that works by stimulating osteoblastic bone formation which results in an increase in bone density and reduction in both vertebral and non-vertebral fracture risk. In women with postmenopausal osteoporosis, it is typically reserved for those with very low bone mineral density (BMD) or those who continue to have fractures despite a bisphosphonate [32]. Barriers to use of teriparatide include high cost, the need for daily injections, and approved use for a total of two years in a lifetime. There is also a theoretical risk of osteosarcoma shown in animal studies but human cases have not been reported when used for postmenopausal osteoporosis. Published studies have shown that combination zolendronate and teriparatide have additive benefits to spine and hip BMD [33]. Another study reported that the combination of denosumab and teriparatide resulted in additive effects, ie, an increase in lumbar, hip, and femoral neck BMD [34]. These combinations have not been studied in populations large enough or for long enough duration to evaluate fracture risk reduction.

Adherence to Osteoporosis Medications

Treatment of osteoporosis reduces risk of fracture, but the benefit of osteoporosis medications is dependent on adherence. Adherence is associated with improved clinical outcomes [35,36] as well as reduced costs and utilization [37,38]; however, adherence to osteoporosis medications is poor. In a meta-analysis of 24 observational studies conducted in large populations, overall adherence for all osteoporosis therapies ranged from approximately 40% to 70% [39]. A recent retrospective claims database analysis in the U.S. reported a 60% noncompliance rate among the 57,913 postmenopausal women prescribed bisphosphonates over 1 year [40]. Another administrative database analysis from a managed care population compared the 3 oral bisphosphonates (risedronate, ibandronate, and alendronate) and found a mean medication possession ratio (MPR) between 0.57–0.58 at 12 months, which dropped to 0.47–0.50 after 24 months and 0.44–0.47 after 36 months [41]. In an observational study of 3200 older women in the U.K. low adherence was self-reported in 8.5%, and 21.6% self-discontinued treatment within 2 years [42]. In a study of Medicare Advantage prescription drug plan members, a small but significant increase in adherence was seen after osteo-porosis treatment change but overall adherence remained low (51% MPR in the change cohort and vs. 44% in the no-change cohort at 24 months, P < 0.01) [43].

Some patients restart osteoporosis therapy after a prolonged lapse in medication use. In one study, re-initiation rates for bisphosphonate therapy among persons who discontinued were as high as 30% within 6 months and 50% within 2 years [44]. Predictors of treatment re-initiation included younger age, female sex, history of fracture, recent hip fracture, nursing home discharge, and BMD testing [44].

Factors that Impact Adherence

Understanding which patients are most likely to be compliant with medications can aid physicians when monitoring osteoporosis treatment responses. In a retro-spective claims analysis, older age was found to be a predictor of compliance: women 65 years and older were more likely to be compliant than younger patients (P = 0.012) [45]. Among women receiving denosumab, improved adherence was found among women with a family history of a parent with a hip fracture, and lower adherence was seen in those with higher age, decreased mobility, and further distance from the clinic where the medication was provided [46].

Major reasons for nonadherence include a fear of potential side effects, occurrence of real side effects, the complicated dosing regimens, and perceived lack of benefit from the medications due to the asymptomatic nature of osteoporosis. In the above noted observational study from the U.K., more than half of the nonadherent patients attributed their nonadherence to side effects (53.9%), with a smaller proportion reporting fear of potential side effects (20.5%) or trouble with the dosing regimen (8.0%)[42].

Patients may also be unwilling to continue to take an osteoporosis medication if a fracture develops while on it and if they are not otherwise provided evidence that the medication is working. In a study by Costa Paiva et al, an understanding and knowledge to osteoporosis was a prerequisite to adherence and the strongest predictor of knowledge was higher education level [47]. Factors that impaired adherence were lower socioeconomic status and presence of comorbidities [47]. In a phenomenological qualitative study, trust in a health care provider was the most common reason for patients’ decision to accept an osteoporosis medication, emphasizing the importance of physician-patient communication [48].

Interventions to Enhance Adherence

Current methods of improving adherence for chronic health problems are mostly complex and not very effective [49]. In a systematic review of interventions to improve medication adherence, only 37 out of 81 studies reported improved adherence in the treatment of chronic diseases, and multifaceted treatments were more likely to succeed [49]. Improving adherence to osteoporosis medications is a complex issue, and a number of interventions evaluated in systematic reviews have shown limited efficacy [50,51]. Simplification of dosing regimens have been found to have a significant impact in chronic disease management [52,53] as well as in some studies of osteoporosis medications.

Simplification of Dosing

Among women prescribed daily vs. weekly bisphosphonates, those on the weekly regimen had significantly higher compliance [54]. However, rates were suboptimal in both groups and more than 50% of women discontinued at 1 year [54]. In addition, in a meta-analysis of osteoporosis medication adherence, a nearly two-fold higher odds of discontinuation with daily vs. weekly bisphosphonates was seen (odds ratio 1.90, 95% CI 1.81–2.00) [55]. Likewise, in a retrospective study in Spain, nearly 85% of those started on a daily bisphosphonate stopped within a year [56], while discontinuation was significantly lower in those prescribed a weekly or monthly bisphosphonate or daily teriparatide; however, discontinuation was still nearly 50% in these groups [56].

Once monthly dosing may be preferred by some patients as there is less time involved in thinking about the disease being treated and a perception of lower likelihood of side effects. In one study, postmenopausal women who had previously stopped oral bisphosphonates due to GI side effects had high adherence rates after self-selecting either monthly oral or quarterly intravenous ibandronate therapy [57]. However, not all studies show significant differences in adherence between weekly and monthly preparations [58–60].

The newer parenteral treatment options that can be given every 6 months or once yearly have the potential to significantly improve adherence. Once a year parenteral administration of a bisphosphonate was preferred over once-weekly oral administration, according to a 1-year study in patients with low bone density previously treated with alendronate [61]. A recent study that looked at persistence with an infusion of zolendronic acid in Taiwanese patients for 48 months found that 85% of patients received at least 2 infusions [62]. In patients treated with denosumab in 4 European countries, adherence and persistence at 12 months were consistently > 80% [46]. Persistence in this study was defined as receiving the subsequent injection within 6 months ± 8 weeks of the previous injection; adherence was defined as receiving 2 consecutive injections within 6 months ± 4 weeks of each other [46].

In a study by Cramer et al, increased adherence and persistence was seen with weekly alendronate compared daily alendronate at the end of 12 months [54]. Similar results were seen in a large longitudinal cohort study of weekly vs. daily bisphosphonates but less than 50% of patients were adherent with the weekly regimen [63]. When once monthly preparations of bisphosphonates became available, studies continued to support a patient preference for less frequently dosed bisphosphonates, with the majority of patients preferring monthly over weekly dosed medications [64–66].

The availability of quarterly ibandronate and yearly zoledronic acid infusions have further simplified dosing. In large, randomized, multicenter studies, patients consistently expressed a preference for yearly infusions over a weekly oral medication [61,67]. Adherence and persistence to osteoporosis medications was also greater in women receiving intravenous ibandronate compared to those receiving oral alendronate [68,69]. However, a study by Curtis et al showed low persistence with intravenous bisphosphonates in a Medicare population [70]. A possible reason for the lower adherence in this population was postulated to include the provision of the infusions at an outpatient center rather than a physician office. Automated nursing reminders with either phone calls or emails have the potential to mitigate the problem of persistence with this less frequent regimen [71,72]. In a review of patient preferences, less frequent dosing of medications was a common desire, but further generalizability were limited, emphasizing the need to individualize treatment [73].

Patient-Provider Communication

Individualizing treatment with better patient-provider communication and identification of potential barriers may increase compliance [74]. In one study, increasing patient participation in determining the treatment option was associated with improved patient adherence [57]. A systematic review of literature on interventions to improve adherence found that periodic follow-up interaction between patients and their health professionals also improved adherence [50]. Positive reinforcement via physician-patient discussion of either bone turnover markers or bone mineral density test results has also been found to improve long-term adherence with osteoporosis medications [71,75].

Better perceived physician knowledge may help with patient adherence. A study by Pickney et al reported that the patient confidence in their health care providers has influence on improved adherence, and patients were more likely to comply when the medications were prescribed by a specialist rather than a general practitioner [76].

Education, Reminders, Phone-Based

Improving patient knowledge of osteoporosis, especially with education using visual aids, may help with improving adherence [47]. In a randomized controlled trial at a single health management organization, an interactive voice response phone call plus a letter 1 week later increased the rate of obtaining a prescribed oral bisphosphonate in the intervention group (48.8% vs. 30.5% control; OR 2.3, 95% CI 1.34–3.94) when adjusted for age, sex, prior BMD, and fracture [77]. Use of an encounter decision aid also improved knowledge of osteoporosis medication options and led to a doubling of medication prescription attainment. However, adherence at 6 months was not improved [78].

Pill reminders in the form of text messages, paging systems on medication devices, and alarm beeps have been studied in patients with chronic diseases, and these technologies could be utilized for osteoporosis treatment [79–81]. A study of smart phone applications showed that many apps help with adherence, especially in noncompliant patients [82]. The researchers reported that of apps studied, MyMedSchedule, MyMeds, and RxmindMe were among the most highly rated due to their ease of use and enhanced functions. Solomon et all studied the effectiveness of a telephone-based counseling program using motivational interviewing in a large randomized study. They found no significant improvement in adherence to an osteoporosis regimen with the telephonic motivational interview compared to mailed educational materials (control group) (P = 0.07) [83]. In a 12-month multicenter, prospective randomized study, Bianchi et al examined the effectiveness of an intervention of reminders or reminders plus phone calls and meetings at the referral center in postmenopausal women initiating an oral osteoporosis prescription. No significant difference was seen in adherence at 12 months compared to standard care [84]. Adherence among the entire cohort, however, was very high [84]

Pharmacist-Based

The role of pharmacists in the treatment of chronic diseases, including osteoporosis, has been studied and shown to be cost-effective. In a study by van Boven et al, an algorithm was designed to detect patients with nonadherence and then tailor an intervention that consisted of structured counseling and active monitoring by pharmacists in initial and continuous phases [85]. This effort-intensive intervention resulted in reduced discontinuation of bisphosphonates after 12 months (reduction from 31.7% to 16.2% at 12 months) [85]. Despite the effort required, findings from the study support overall cost-effectiveness of this intervention [85]. A randomized controlled study by Lai et al showed that pharmacists can play a role in improving medication adherence through counseling patients on the importance of adherence, side effects, and goals of therapy [86]. The same authors also showed that involvement of a clinical pharmacist in the care of patients helped to further improve patient knowledge of medications and osteoporosis treatments, resolve medication-related concerns, and improve overall quality of life [87]. Such pharmacist-led interventions would require pharmacists to understand their role and the potential for drug holidays in the course of osteoporosis treatments and not mislabel patients as nonadherent when in fact purposefully holding osteoporosis medications [88].

Conclusion

Osteoporosis is a growing problem with increasing numbers of patients at risk for osteoporosis and related fractures. Currently available osteoporosis medications have shown clear benefit in reducing fracture risk; however, adherence to these therapies is required to obtain benefit. Unfortunately, osteoporosis medications have several limitations to full compliance, particularly the oral treatment options, including known possible side effects acutely and chronically, potential/feared side effects, irregular dosing intervals, complicated dosing instructions, and absence of an immediate recognizable benefit/effect. Improving adherence is complex [89] and tailoring to individual patients is of importance. Successful techniques for improving adherence may include a focus on physician-patient communication, use of the less frequently dosed medications, various medication reminders, use of available technology, and use of pharmacists for patient counseling and monitoring. Recognition of this common problem by clinicians is of utmost importance.

Corresponding author: Amy H. Warriner, MD, The University of Alabama at Birmingham, Division of Endocrinology, Diabetes and Metabolism, 702 Faculty Office Tower, 510 20th Street South, Birmingham, AL 35294, [email protected].

Financial disclosures: None.

From the Division of Endocrinology, Diabetes, and Metabolism, University of Alabama at Birmingham, Birmingham, Alabama.

Abstract

• Objective: To review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.
• Methods: Review of the literature.
• Results: With the growing aging population, there is an increased number of people at risk of osteoporosis and fracture. Several medications are available that reduce the risk of fracture. However, adherence to osteoporosis medications is suboptimal. Factors related to nonadherence include dosing frequency, real side effects, and concern about potential side effects. Interventions that may improve adherence include clinician and patient education, less frequent and less complex dosing regimens, medication reminders, and adherence counseling.
• Conclusions: Improving adherence to osteoporosis medications is a complex and challenging issue. Considering and implementing strategies to improve adherence tailored to patient preferences may enhance long-term outcomes for patients with osteoporosis.

Osteoporosis is a chronic but asymptomatic disease that is characterized by an increased fragility of bones and increased risk of fractures. Hip and vertebral fractures are associated with the greatest morbidity and mortality. The prevalence of osteoporosis is estimated to be 10.3% in the US, with approximately 10.2 million adults over the age of 50 having osteoporosis based on 2010 census data and results from the National Health and Nutrition Examination Survey (NHANES) [1].

Several drugs are currently available for the treatment of osteoporosis, but adherence to treatment is low. Understanding the factors associated with low adherence and actions that can be taken to improve adherence to treatment is important given the large number of individuals with osteoporosis and the need to reduce the burden caused by fragility fracture. In this article, we review the treatment of osteoporosis, challenges to treatment adherence, and factors associated with improved adherence.

Nonprescription Medications

Calcium

There have been several published studies over the last decade evaluating calcium supplementation and its efficacy in reducing fractures. Although these studies showed that calcium reduces bone turnover by 20% and slowed postmenopausal bone loss by one third [2,3], none of these studies or a recent systematic review [4] showed any degree of fracture risk reduction with calcium supplements alone.

Although some calcium intake may be good, too much calcium has the potential to cause harm, including an increased risk of nephrolithiasis and constipation/bloating. An analysis of the Women’s Health Initiative (WHI) study reported a 17% increase in renal calculi in women who received calcium and vitamin D supplements [5]. Another recently published meta-analysis showed a 43% increase in gastrointestinal complaints in patients who were taking calcium supplements [6]. The potential for increased cardiovascular risk with calcium supplements is controversial [7]. The WHI study did not show an increased occurrence of cardiovascular events among those taking calcium supplements [8]. In a different population, men who consumed more than 1000 mg per day of supplemental calcium had higher all-cause and cardiovascular disease-specific mortality [9]. Large, well-conducted randomized controlled trials will be needed to further elucidate the question of calcium supplementation and risk of cardiovascular disease.

Vitamin D

Deficiency of vitamin D is common with one study finding more than 90% of older adults deficient in vitamin D [10]. Vitamin D is essential for proper calcium metabolism and deficiency is known to induce secondary hyperparathyroidism. Studies in mouse models have also shown that normal vitamin D receptors in enterocytes are essential for normal bone mineralization [11,12]. A systematic Cochrane database review showed that vitamin D3 supplementation decreased mortality in elderly people living independently or in institutional care [13]. Vitamin D was administered for a weighted mean of 4.4 years. Vitamin D2, alfacalcidol, and calcitriol had no statistically significant beneficial effects on mortality. Vitamin D3 combined with calcium was associated with an increased risk of nephrolithiasis during a follow-up period of 1.25 to 7 years (relative risk [RR] 1.17, 95% confidence interval (CI) 1.02–1.34) [13]. The inconsistencies of published reports looking at benefits of vitamin D supplementation may be due in part to variability in compliance with taking the supplements and baseline vitamin D levels.

Two randomized controlled trials have shown that low vitamin D appears to be an independent predictor of fall risk, and vitamin D supplementation has been found to reduce this risk of falls, through improved musculo-skeletal function [14–16]. Thus, vitamin D may play a role in fracture risk reduction beyond direct bone effects.

Prescription Osteoporosis Treatments

Bisphosphonates

Bisphosphonates are the most commonly prescribed medication for osteoporosis. The efficacy of bisphosphonates to reduce fractures is well established. There are oral bisphosphonates, which can be dosed daily, weekly, or monthly, and intravenous bisphosphonates, which can be given every 3 months or annually. Side effects with this class of medications include gastrointestinal effects with the oral options in up to 20% to 30% of users [17]. With intravenous bisphosphonates, the greatest risk is an acute phase response, which can occur in up to 42% of patients [18]. The risk of an acute phase reaction is much lower with doses beyond the first dose and lower if patients have ever previously taken an oral bisphosphonate and/or receive acetaminophen prior to the infusion. Other potential side effects with all bisphosphonates include osteonecrosis of jaw (ONJ) and atypical subtrochanteric fractures. Post marketing studies have indicated that the incidence of ONJ is less than 2 per 100,000 patient-years among those taking bisphosphonates [19,20]. A number of database analyses have shown that ONJ-like lesions can also occur in older individuals with osteoporosis who have never been exposed to bisphosphonates [21]. A case series of an osteoporotic population showed that ONJ-like lesions are lower grade than those typically seen in cancer patients who usually are exposed to higher doses of bisphosphonates [22]. A study of Swedish older men and women reported that long-term use of bisphosphonates (4 years or more) was associated with an increased incidence of atypical fractures. The RR for women was 126.0 (95% CI, 55.1–288.1) after 4 years of bisphosphonates [23]. A U.S. health care database analysis reported that 90% of those with atypical fractures were bisphosphonate users, almost half were Asian (49%), and use beyond 6 years showed the greatest risk [24].

Non-Bisphosphonate Medications

Other osteoporosis medications include denosumab, raloxifene, estrogen, and teriparatide (calcitonin will not be discussed here). Newer options currently under study, including cathepsin K inhibitors and anti-sclerostin therapies, are not available in the United States.

Denosumab is a monoclonal antibody that interferes with the receptor activator of nuclear kappa B ligand (RANK-L), which is the principal stimulus for osteoclastogenesis. Denosumab is administered once every 6 months subcutaneously. Phase III trials of denosumab demonstrated a 68% reduction in vertebral fractures and 40% and 20% reduction in fractures at hip and non-vertebral sites, respectively [25]. Similar to bisphosphonates, other risks include atypical femoral fractures and ONJ. In addition, hypocalcemia, including severe, symptomatic hypocalcemia, has been reported at rates higher than initially reported in the original clinical trials [26]. Hypocalcemia can be severe, especially in patients who are deficient in vitamin D [10,27].

Estrogen is effective in reducing the risk of vertebral fractures. Selective estrogen receptor modulators (SERMs) have both estrogen agonist and antagonist effects. The SERM, raloxifene, has been used in osteoporosis for its antiresorptive effects through the estrogen receptor [28,29]. A newer SERM, bazedoxifene, has been studied in combination with conjugated estrogen and has been reported to improve bone mineral density and other symptoms of menopause, like vasomotor symptoms and vulvo-vaginal atrophy, but its efficacy in reducing fracture risk has not been demonstrated [30,31].

Teriparatide is an anabolic agent that works by stimulating osteoblastic bone formation which results in an increase in bone density and reduction in both vertebral and non-vertebral fracture risk. In women with postmenopausal osteoporosis, it is typically reserved for those with very low bone mineral density (BMD) or those who continue to have fractures despite a bisphosphonate [32]. Barriers to use of teriparatide include high cost, the need for daily injections, and approved use for a total of two years in a lifetime. There is also a theoretical risk of osteosarcoma shown in animal studies but human cases have not been reported when used for postmenopausal osteoporosis. Published studies have shown that combination zolendronate and teriparatide have additive benefits to spine and hip BMD [33]. Another study reported that the combination of denosumab and teriparatide resulted in additive effects, ie, an increase in lumbar, hip, and femoral neck BMD [34]. These combinations have not been studied in populations large enough or for long enough duration to evaluate fracture risk reduction.

Adherence to Osteoporosis Medications

Treatment of osteoporosis reduces risk of fracture, but the benefit of osteoporosis medications is dependent on adherence. Adherence is associated with improved clinical outcomes [35,36] as well as reduced costs and utilization [37,38]; however, adherence to osteoporosis medications is poor. In a meta-analysis of 24 observational studies conducted in large populations, overall adherence for all osteoporosis therapies ranged from approximately 40% to 70% [39]. A recent retrospective claims database analysis in the U.S. reported a 60% noncompliance rate among the 57,913 postmenopausal women prescribed bisphosphonates over 1 year [40]. Another administrative database analysis from a managed care population compared the 3 oral bisphosphonates (risedronate, ibandronate, and alendronate) and found a mean medication possession ratio (MPR) between 0.57–0.58 at 12 months, which dropped to 0.47–0.50 after 24 months and 0.44–0.47 after 36 months [41]. In an observational study of 3200 older women in the U.K. low adherence was self-reported in 8.5%, and 21.6% self-discontinued treatment within 2 years [42]. In a study of Medicare Advantage prescription drug plan members, a small but significant increase in adherence was seen after osteo-porosis treatment change but overall adherence remained low (51% MPR in the change cohort and vs. 44% in the no-change cohort at 24 months, P < 0.01) [43].

Some patients restart osteoporosis therapy after a prolonged lapse in medication use. In one study, re-initiation rates for bisphosphonate therapy among persons who discontinued were as high as 30% within 6 months and 50% within 2 years [44]. Predictors of treatment re-initiation included younger age, female sex, history of fracture, recent hip fracture, nursing home discharge, and BMD testing [44].

Factors that Impact Adherence

Understanding which patients are most likely to be compliant with medications can aid physicians when monitoring osteoporosis treatment responses. In a retro-spective claims analysis, older age was found to be a predictor of compliance: women 65 years and older were more likely to be compliant than younger patients (P = 0.012) [45]. Among women receiving denosumab, improved adherence was found among women with a family history of a parent with a hip fracture, and lower adherence was seen in those with higher age, decreased mobility, and further distance from the clinic where the medication was provided [46].

Major reasons for nonadherence include a fear of potential side effects, occurrence of real side effects, the complicated dosing regimens, and perceived lack of benefit from the medications due to the asymptomatic nature of osteoporosis. In the above noted observational study from the U.K., more than half of the nonadherent patients attributed their nonadherence to side effects (53.9%), with a smaller proportion reporting fear of potential side effects (20.5%) or trouble with the dosing regimen (8.0%)[42].

Patients may also be unwilling to continue to take an osteoporosis medication if a fracture develops while on it and if they are not otherwise provided evidence that the medication is working. In a study by Costa Paiva et al, an understanding and knowledge to osteoporosis was a prerequisite to adherence and the strongest predictor of knowledge was higher education level [47]. Factors that impaired adherence were lower socioeconomic status and presence of comorbidities [47]. In a phenomenological qualitative study, trust in a health care provider was the most common reason for patients’ decision to accept an osteoporosis medication, emphasizing the importance of physician-patient communication [48].

Interventions to Enhance Adherence

Current methods of improving adherence for chronic health problems are mostly complex and not very effective [49]. In a systematic review of interventions to improve medication adherence, only 37 out of 81 studies reported improved adherence in the treatment of chronic diseases, and multifaceted treatments were more likely to succeed [49]. Improving adherence to osteoporosis medications is a complex issue, and a number of interventions evaluated in systematic reviews have shown limited efficacy [50,51]. Simplification of dosing regimens have been found to have a significant impact in chronic disease management [52,53] as well as in some studies of osteoporosis medications.

Simplification of Dosing

Among women prescribed daily vs. weekly bisphosphonates, those on the weekly regimen had significantly higher compliance [54]. However, rates were suboptimal in both groups and more than 50% of women discontinued at 1 year [54]. In addition, in a meta-analysis of osteoporosis medication adherence, a nearly two-fold higher odds of discontinuation with daily vs. weekly bisphosphonates was seen (odds ratio 1.90, 95% CI 1.81–2.00) [55]. Likewise, in a retrospective study in Spain, nearly 85% of those started on a daily bisphosphonate stopped within a year [56], while discontinuation was significantly lower in those prescribed a weekly or monthly bisphosphonate or daily teriparatide; however, discontinuation was still nearly 50% in these groups [56].

Once monthly dosing may be preferred by some patients as there is less time involved in thinking about the disease being treated and a perception of lower likelihood of side effects. In one study, postmenopausal women who had previously stopped oral bisphosphonates due to GI side effects had high adherence rates after self-selecting either monthly oral or quarterly intravenous ibandronate therapy [57]. However, not all studies show significant differences in adherence between weekly and monthly preparations [58–60].

The newer parenteral treatment options that can be given every 6 months or once yearly have the potential to significantly improve adherence. Once a year parenteral administration of a bisphosphonate was preferred over once-weekly oral administration, according to a 1-year study in patients with low bone density previously treated with alendronate [61]. A recent study that looked at persistence with an infusion of zolendronic acid in Taiwanese patients for 48 months found that 85% of patients received at least 2 infusions [62]. In patients treated with denosumab in 4 European countries, adherence and persistence at 12 months were consistently > 80% [46]. Persistence in this study was defined as receiving the subsequent injection within 6 months ± 8 weeks of the previous injection; adherence was defined as receiving 2 consecutive injections within 6 months ± 4 weeks of each other [46].

In a study by Cramer et al, increased adherence and persistence was seen with weekly alendronate compared daily alendronate at the end of 12 months [54]. Similar results were seen in a large longitudinal cohort study of weekly vs. daily bisphosphonates but less than 50% of patients were adherent with the weekly regimen [63]. When once monthly preparations of bisphosphonates became available, studies continued to support a patient preference for less frequently dosed bisphosphonates, with the majority of patients preferring monthly over weekly dosed medications [64–66].

The availability of quarterly ibandronate and yearly zoledronic acid infusions have further simplified dosing. In large, randomized, multicenter studies, patients consistently expressed a preference for yearly infusions over a weekly oral medication [61,67]. Adherence and persistence to osteoporosis medications was also greater in women receiving intravenous ibandronate compared to those receiving oral alendronate [68,69]. However, a study by Curtis et al showed low persistence with intravenous bisphosphonates in a Medicare population [70]. A possible reason for the lower adherence in this population was postulated to include the provision of the infusions at an outpatient center rather than a physician office. Automated nursing reminders with either phone calls or emails have the potential to mitigate the problem of persistence with this less frequent regimen [71,72]. In a review of patient preferences, less frequent dosing of medications was a common desire, but further generalizability were limited, emphasizing the need to individualize treatment [73].

Patient-Provider Communication

Individualizing treatment with better patient-provider communication and identification of potential barriers may increase compliance [74]. In one study, increasing patient participation in determining the treatment option was associated with improved patient adherence [57]. A systematic review of literature on interventions to improve adherence found that periodic follow-up interaction between patients and their health professionals also improved adherence [50]. Positive reinforcement via physician-patient discussion of either bone turnover markers or bone mineral density test results has also been found to improve long-term adherence with osteoporosis medications [71,75].

Better perceived physician knowledge may help with patient adherence. A study by Pickney et al reported that the patient confidence in their health care providers has influence on improved adherence, and patients were more likely to comply when the medications were prescribed by a specialist rather than a general practitioner [76].

Education, Reminders, Phone-Based

Improving patient knowledge of osteoporosis, especially with education using visual aids, may help with improving adherence [47]. In a randomized controlled trial at a single health management organization, an interactive voice response phone call plus a letter 1 week later increased the rate of obtaining a prescribed oral bisphosphonate in the intervention group (48.8% vs. 30.5% control; OR 2.3, 95% CI 1.34–3.94) when adjusted for age, sex, prior BMD, and fracture [77]. Use of an encounter decision aid also improved knowledge of osteoporosis medication options and led to a doubling of medication prescription attainment. However, adherence at 6 months was not improved [78].

Pill reminders in the form of text messages, paging systems on medication devices, and alarm beeps have been studied in patients with chronic diseases, and these technologies could be utilized for osteoporosis treatment [79–81]. A study of smart phone applications showed that many apps help with adherence, especially in noncompliant patients [82]. The researchers reported that of apps studied, MyMedSchedule, MyMeds, and RxmindMe were among the most highly rated due to their ease of use and enhanced functions. Solomon et all studied the effectiveness of a telephone-based counseling program using motivational interviewing in a large randomized study. They found no significant improvement in adherence to an osteoporosis regimen with the telephonic motivational interview compared to mailed educational materials (control group) (P = 0.07) [83]. In a 12-month multicenter, prospective randomized study, Bianchi et al examined the effectiveness of an intervention of reminders or reminders plus phone calls and meetings at the referral center in postmenopausal women initiating an oral osteoporosis prescription. No significant difference was seen in adherence at 12 months compared to standard care [84]. Adherence among the entire cohort, however, was very high [84]

Pharmacist-Based

The role of pharmacists in the treatment of chronic diseases, including osteoporosis, has been studied and shown to be cost-effective. In a study by van Boven et al, an algorithm was designed to detect patients with nonadherence and then tailor an intervention that consisted of structured counseling and active monitoring by pharmacists in initial and continuous phases [85]. This effort-intensive intervention resulted in reduced discontinuation of bisphosphonates after 12 months (reduction from 31.7% to 16.2% at 12 months) [85]. Despite the effort required, findings from the study support overall cost-effectiveness of this intervention [85]. A randomized controlled study by Lai et al showed that pharmacists can play a role in improving medication adherence through counseling patients on the importance of adherence, side effects, and goals of therapy [86]. The same authors also showed that involvement of a clinical pharmacist in the care of patients helped to further improve patient knowledge of medications and osteoporosis treatments, resolve medication-related concerns, and improve overall quality of life [87]. Such pharmacist-led interventions would require pharmacists to understand their role and the potential for drug holidays in the course of osteoporosis treatments and not mislabel patients as nonadherent when in fact purposefully holding osteoporosis medications [88].

Conclusion

Osteoporosis is a growing problem with increasing numbers of patients at risk for osteoporosis and related fractures. Currently available osteoporosis medications have shown clear benefit in reducing fracture risk; however, adherence to these therapies is required to obtain benefit. Unfortunately, osteoporosis medications have several limitations to full compliance, particularly the oral treatment options, including known possible side effects acutely and chronically, potential/feared side effects, irregular dosing intervals, complicated dosing instructions, and absence of an immediate recognizable benefit/effect. Improving adherence is complex [89] and tailoring to individual patients is of importance. Successful techniques for improving adherence may include a focus on physician-patient communication, use of the less frequently dosed medications, various medication reminders, use of available technology, and use of pharmacists for patient counseling and monitoring. Recognition of this common problem by clinicians is of utmost importance.

Corresponding author: Amy H. Warriner, MD, The University of Alabama at Birmingham, Division of Endocrinology, Diabetes and Metabolism, 702 Faculty Office Tower, 510 20th Street South, Birmingham, AL 35294, [email protected].

Financial disclosures: None.

Case Scenarios

Case 1

While working abroad in a resource-limited environment, a patient was brought in after falling and hitting his head. Initially, the patient was awake and alert, but he gradually became minimally responsive, with a Glasgow Coma Scale score of 9. Your facility did not have computed tomography (CT) or magnetic resonance imaging (MRI), but did have a point-of-care ultrasound (US) machine. You measured the patient’s optic nerve sheath diameter (ONSD) with the US and found a diameter of 4.5 mm in each eye. With this clinical change, you wondered if repeat US scans to detect increasing intracranial pressure (ICP) would represent changes in the patient’s condition.

Case 2

A patient who presented with an intracranial hemorrhage was treated with hypertonic saline and was awaiting neurosurgical placement of an extraventicular drain. During this time, a resident who was on a US rotation asked you if she would be able to detect changes in the patient’s ICP using US rather than placing an invasive device. How do you respond?

In adults, ICP is normally 10 to 15 mm Hg. It may be pathologically increased in several life-threatening conditions, including traumatic brain injury (TBI), subarachnoid hemorrhage, central venous thrombosis, brain tumor, and abscess. It is also increased by nonacute pathology, such as idiopathic intracranial hypertension (IIH), which also is known as pseudotumor cerebri. In patients with acute pathology, ICP above 20 mm Hg is generally considered an indication for treatment.¹ Indications for ICP monitoring in TBI include positive CT findings, patient age greater than 40 years, systemic hypotension, or abnormal flexion/extension in response to pain.² Other reasons to monitor ICP include the management of pseudotumor cerebri or after ventriculoperitoneal shunt surgery.³

Unfortunately, current methods of ICP monitoring have significant drawbacks and limitations. The gold standard of ICP monitoring—measurement using an intraventricular catheter—increases the risks of infection and hemorrhage, requires the skill of a neurosurgeon, and may be contraindicated due to coagulopathy or thrombocytopenia. It also cannot be done in a prehospital setting and only to a limited extent in the ED.⁴

Computed tomography scans and MRI can assess elevated ICP, but these tests are expensive, may increase patient radiation exposure, require patient transport, and may not always detect raised ICP. In the appropriate clinical context, signs present on physical examination, such as decorticate/decerebrate posturing, papilledema, or fixed/dilated pupils, may be highly suggestive of an increased ICP, but sensitivity and specificity are inadequate. Delay in diagnosis is also a drawback of imaging and physical examination, as findings may not present until ICP has been persistently elevated.

Given the disadvantages of current means of assessing elevated ICP, several noninvasive methods of measuring ICP are being investigated. These include such techniques as transcranial Doppler, electroencephalogram, pupillometry, and ONSD measurements.⁵ This article reviews current applications of ultrasonography measurements of the ONSD in assessing elevations in ICP.

ONSD US

Assessment of ICP via measurement of the ONSD has attracted increasing attention, particularly in emergency medicine. Measurements of the ONSD are possible with CT, MRI, and US. Of these modalities, ONSD US has attracted the most interest, due to its low cost, wide availability, and rapidity. It does not require patient transport, and does not expose a patient to additional radiation. In addition, ONSD US has been utilized in low-resource settings, and may be particularly useful in prehospital and mass-casualty situations.⁶

The underlying relationship between ONSD and ICP is a result of the enclosure of the subarachnoid space by the ONS. Increased ICP leads to expansion of the ONS, particularly at 3 mm behind the globe, in the retrobulbar compartment (Figures 1 and 2).⁷

Unfortunately, it is not possible to precisely determine ICP from an ONSD measurement, because baseline ONSD values and elasticity vary significantly within the population.^4,8 As a result, ONSD US has been investigated mostly for its ability to detect qualitative changes—particularly as a screen for elevated ICP. Optic nerve sheath diameter has high discriminative value in its ability to distinguish normal from elevated ICP. In a meta-analysis, Dubourg et al⁹ showed that the technique had an area under the summary receiver-operating curve of 0.94, signifying excellent test accuracy to diagnose elevated ICPs.

Researchers have attempted to determine a threshold value of ONSD that would serve as a clinically useful predictor of elevated ICP. Currently, this value ranges from 4.8 to 5.9 mm, depending on the study⁹; 5 mm is commonly used clinically as a threshold.¹⁰

Using ONSD US to Monitor Rapid Changes in ICP

While the use of the ONSD technique to screen for elevated ICP is relatively well established, the use of ONSD US to track acute changes in ICP is not as well studied. Serial tracking of acute changes could be useful in a patient at risk for intracranial hypertension secondary to trauma, to monitor the results of treating a patient with IIH, or after ventriculoperitoneal shunt placement.³

In Vivo Data

In 1993, Tamburrelli et al¹¹ performed the first ONSD intrathecal infusion study, using A-scan sonography, and concluded that there was a “direct, biphasic, positive relation between diastolic intracranial pressure and optic nerve diameters” and that the data showed “rapid changes of optic nerve diameters in response to variation of intracranial pressure.”

In 1997, Hansen and Helmke¹² recorded ONSD versus ICP data in the first intrathecal infusion test to use B-scan mode sonography. Ultrasonography was performed at 2- to 4-minute intervals. Their data demonstrated a linear relationship between ICP and ONSD over a particular cerebrospinal fluid pressure interval. They noted that “this interval differed between patients: ONS dilation commenced at pressure thresholds between 15 mm Hg and 30 mm Hg and in some patients saturation of the response (constant ONSD) occurred between 30 mm Hg and 40 mm Hg.”

The slope of ONSD versus ICP curve varied considerably by patient, making it impossible to infer an absolute ICP value from an ONSD without prior knowledge of the patient’s ratio. Similar to the data from Tamburrelli et al,¹¹ Hansen and Helmke¹² also found that there was no lag in ONSD response to ICP: “Within this interval, no temporal delay of the ONS response was noted.”

The only study comparing real-time ONSD data to gold-standard measurements of rapidly changing ICP in humans was performed by Maissan et al¹³ in 2015. This study involved a cohort of 18 patients who had suffered TBI and had intraparenchymal probes inserted. Because ICP rises transiently during endotracheal tube suctioning due to irritation of the trachea, the increase and subsequent decrease after suctioning was an ideal time to perform ONSD measurements and compare them to simultaneous gold-standard ICP measurements. The ONSD US measurements were performed 30 to 60 seconds prior to suctioning, during suctioning, and 30 to 60 seconds after suctioning.

Even during this very rapid time course, a strong correlation between ICP and ONSD measurements was demonstrated. The R2 value was 0.80. There was no perceptible “lag” in ONSD change; changes in ICP were immediately reflected in ONSD. Notably, an absolute change of less than 8 to 10 mm Hg in ICP did not affect ONSD, which is consistent with data collected by Hansen and Helmke.¹²

Therapeutic Lumbar Puncture for IIH

There are two case reports of ONSD US measurements being taken pre- and postlumbar puncture (LP) in patients with IIH. In the first, in 1989 Galetta et al¹⁴ used A-scan US to measure pre- and post-LP ONSD in a woman with papilledema secondary to IIH. They found a significant reduction in ONSD bilaterally “within minutes” of performing the LP.¹⁴

The second case report was published in 2015 by Singleton et al.¹⁵ They recorded ONSD measurements 30 minutes pre- and post-LP in a woman who presented to the ED with symptoms from elevated ICP. After reduction of pressure via LP, they recorded a significant reduction in ONSD bilaterally.¹⁵

Cadaver Data

Hansen et al¹⁶ evaluated the distensibility and elasticity of the ONS using postmortem optic nerve preparations. The ONSD was recorded 200 seconds after each pressure increase, which was long enough to achieve stable diameters. They found a linear correlation between pressure increases of 5 to 45 mm Hg and ONSD. This would suggest a potential positively correlated change in ONSD with in vivo changes in ICP. However, this still needs further clinical study to better assess measurable changes in living patients.

Conclusion

Published data have consistently demonstrated that changes in ICP are rapidly transmitted to the optic nerve sheath and that there does not appear to be a temporal lag in the ONSD. Based on in vivo data, the relationship between ICP and ONSD appears to be linear only over a range of moderately elevated ICP. According to Hansen and Helmke,¹² this range starts at approximately 18 to 30 mm Hg, and ends at approximately 40 to 45 mm Hg. Maissan et al¹³ observed similar findings: “At low levels, ICP changes (8-10 mm Hg) do not affect the ONSD.”

There is still need for additional research to validate and refine these findings. Only one study has compared gold-standard ICP measurements with ONSD US measurements in real time,¹³ and the literature on ONSD US in tracking ICP after therapeutic LP in IIH consists of only two case reports.

Thus, with some caveats, ONSD US appears to permit qualitative tracking of ICP in real time. This supports its use in situations where a patient may have rapidly changing ICP, such as close monitoring of patients at risk for elevated ICP in a critical care setting, and response to treatment in patients with IIH.