The Journal of Family Practice

Managing diabetes is a complex undertaking, with an extensive regimen of self-care—including regular exercise, meal planning, blood glucose monitoring, medication scheduling, and multiple visits—that is critically linked to glycemic control and the prevention of complications. Incorporating all of these elements into daily life can be daunting.^1-3

In fact, nearly half of US adults with diabetes fail to meet the recommended targets.⁴ This leads to frustration, which often manifests in psychosocial problems that further hamper efforts to manage the disease.^5-10 The most notable is a psychosocial disorder known as diabetes distress, which affects close to 45% of those with diabetes.^11,12

It is important to note that diabetes distress is not a psychiatric disorder;¹³ rather, it is a broad affective reaction to the stress of living with this chronic and complex disease.^14,15 By negatively affecting adherence to a self-care regimen, diabetes distress contributes to worsening glycemic control and increasing morbidity.^16-18

Recognizing that about 80% of those with diabetes are treated in primary care settings,¹⁹ we wrote this review to call your attention to diabetes distress, alert you to brief screening tools that can easily be incorporated into clinic visits, and offer guidance in matching proposed interventions to the aspects of diabetes self-management that cause patients the greatest distress.

Diabetes distress: What it is, what it’s not

For patients with type 2 diabetes, diabetes distress centers around 4 main issues:

• frustration with the demands of self-care;
• apprehension about the future and the possibility of developing serious complications;
• concern about both the quality and the cost of required medical care; and
• perceived lack of support from family and/or friends.^11,12,20

As mentioned earlier, diabetes distress is not a psychiatric condition and should not be confused with major depressive disorder (MDD). Here’s help in telling the difference.

Unlike major depressive disorder, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.

For starters, a diagnosis of depression is symptom-based.¹³ MDD requires the presence of at least 5 of the 9 symptoms defined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth ed. (DSM-5)—eg, persistent feelings of worthlessness or guilt, sleep disturbances, lack of interest in normal activities—for at least 2 weeks.²¹ What’s more, the diagnostic criteria for MDD do not specify a cause or disease process. Nor do they distinguish between a pathological response and an expected reaction to a stressful life event.²² Further, depression measures reflect symptoms (eg, hyperglycemia), as well as stressful experiences resulting from diabetes self-care, which may contribute to the high rate of false positives or incorrect diagnoses of MDD and missed diagnoses of diabetes distress.²³

Unlike MDD, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.¹³ And, because the source of the problem is identified, diabetes distress can be treated with specific interventions targeting the areas causing the highest levels of stress.

When a psychiatric condition and diabetes distress overlap

MDD, anxiety disorders, and diabetes distress are all common in patients with diabetes,²⁴ and the co-occurrence of a psychiatric disorder and diabetes distress is high.²⁵ Thus, it is important not only to identify cases of diabetes distress but also to consider comorbid depression and/or anxiety in patients with diabetes distress.

More often, though, it is the other way around, according to the Distress and Depression in Diabetes (3D) study. The researchers recently found that 84% of patients with moderate or high diabetes distress did not fulfill the criteria for MDD, but that 67% of diabetes patients with MDD also had moderate or high diabetes distress.^13,15,17,25

The data highlight the importance of screening patients with a dual diagnosis of diabetes and MDD for diabetes distress. Keep in mind that individuals diagnosed with both diabetes distress and a comorbid psychiatric condition may require more complex and intensive treatment than those with either diabetes distress or MDD alone.²⁵

Screening for diabetes distress

Diabetes distress can be easily assessed using one of several patient-reported outcome measures. Six validated measures, ranging in length from one to 28 questions, are designed for use in primary care (TABLE).^26-30 Some of the measures are easily accessible online; others require subscription to MEDLINE.

Problem Areas in Diabetes (PAID): There are 3 versions of PAID—a 20-item screen assessing a broad range of feelings related to living with diabetes and its treatment, a 5-item version (PAID-5) with high rates of sensitivity (95%) and specificity (89%), and a single-item test (PAID-1) that is highly correlated with the longer version.^26,27

Diabetes Distress Scale (DDS): This tool is available in a 17-item measure assessing diabetes distress as it relates to the emotional burden, physician-related distress, regimen-related distress, and interpersonal distress.²⁸ DDS is also available in a short form (DDS-2) with 2 items²⁹ and a 28-item scale specifically for patients with type 1 diabetes.³⁰ T1-DDS, the only diabetes distress measure focused on this particular patient population, assesses the 7 sources of distress found to be common among adults with type 1 diabetes: powerlessness, negative social perceptions, physician distress, friend/family distress, hypoglycemia distress, management distress, and eating distress.

Studies have shown that not only do those with type 1 diabetes experience different stressors compared with their type 2 counterparts, but that they tend to experience distress differently. For patients with type 1 diabetes, for example, powerlessness ranked as the highest source of distress, followed by eating distress and hypoglycemia distress. These sources of distress differ from the regimen distress, emotional burden, interpersonal distress, and physician distress identified by those with type 2 diabetes.³⁰

How to respond to diabetes distress

Diabetes distress is easier to identify than to successfully treat. Few validated treatments for diabetes distress exist and, to our knowledge, only 2 studies have assessed interventions aimed at reduction of such distress.^31,32

The REDEEM trial³¹ recruited adults with type 2 diabetes and diabetes distress to participate in a 12-month randomized controlled trial (RCT). The trial had 3 arms, comparing the effectiveness of a computer-assisted self-management (CASM) program alone, a CASM program plus in-person diabetes distress-specific problem-solving therapy, and a computer-assisted minimally supportive intervention. The main outcomes included diabetes distress (using the DDS scale and subscales), along with self-management behaviors and HbA1c.

Participants in all 3 arms showed significant reductions in total diabetes distress and improvements in self-management behaviors, with no significant differences among the groups. No differences in HbA1c were found. However, those in the CASM program plus distress-specific therapy arm showed a larger reduction in regimen distress compared with the other 2 groups.³¹

The DIAMOS trial³² recruited adults who had type 1 or type 2 diabetes, diabetes distress, and subclinical depressive symptoms for a 2-arm RCT. One group underwent cognitive behavioral interventions, while the controls had standard group-based diabetes education. The main outcomes included diabetes distress (measured via the PAID scale), depressive symptoms, well-being, diabetes self-care, diabetes acceptance, satisfaction with diabetes treatment, HbA1c, and subclinical inflammation.

Major depressive disorder, anxiety disorders, and diabetes distress are all common in patients with diabetes.

The intervention group showed greater improvement in diabetes distress and depressive symptoms compared with the control group, but no differences in well-being, self-care, treatment satisfaction, HbA1c, or subclinical inflammation were observed.³²

Both studies support the use of problem-solving therapy and cognitive behavioral interventions for patients with diabetes distress. Future research should evaluate the effectiveness of these interventions in the primary care setting.

What else to offer when challenges mount?

Diabetes is a progressive disease, and most patients experience multiple challenges over time. These typically include complications and comorbidities, physical limitations, polypharmacy, hypoglycemia, and cognitive impairment, as well as changes in everything from medication and lifestyle to insurance coverage and social support.^33,34 All increase the risk for diabetes distress, as well as related psychiatric conditions.

Eighty-four percent of patients with moderate or high diabetes distress didn’t fulfill the criteria for MDD, but 67% of diabetes patients with MDD also had diabetes distress.

Aging and diabetes are independent risk factors for cognitive impairment, for example, and the presence of both increases this risk.³⁵ What’s more, diabetes alone is associated with poorer executive function,^36-38 the higher-level cognitive processes that allow individuals to engage in independent, purposeful, and flexible goal-related behaviors. Both poor cognitive function and impaired executive function interfere with the ability to perform self-care behaviors such as adjusting insulin doses, drawing insulin into a syringe, or dialing an insulin dose with an insulin pen.³⁹ This in turn can lead to frustration and increase the likelihood of moderate to high diabetes distress.

Assessing diabetes distress in patients with cognitive impairment, poor executive functioning, or other psychological limitations is particularly difficult, however, as no diabetes distress measures take such deficits into account. Thus, primary care physicians without expertise in neuropsychology should consider referring patients with such problems to specialists for assessment.

Be alert to socioeconomic changes—in employment, insurance coverage, and living situations—that are not addressed in the screening tools.

The progressive nature of diabetes also highlights the need for primary care physicians to periodically screen for diabetes distress and engage in ongoing discussions about what type of care is best for individual patients, and why. When developing or updating treatment plans and making recommendations, it is crucial to consider the impact the treatment would likely have on the patient’s physical and mental health and to explicitly inquire about and acknowledge his or her values and preferences for care.^40-44

It is also important to remain aware of socioeconomic changes—in employment, insurance coverage, and living situations, for example—which are not addressed in the screening tools.

Moderate to high diabetes distress scores, as well as individual items patients identify as “very serious” problems, represent clinical red flags that should be the focus of careful discussion during a medical visit. Patients with moderate to high distress should be referred to a therapist trained in cognitive behavioral therapy or problem-solving therapy. Physicians who lack access to such resources can incorporate cognitive behavioral and problem-solving techniques into patient discussion. (See “Directing help where it’s most needed.”) All patients should be referred to a certified diabetes educator—a key component of diabetes care.^45,46

CORRESPONDENCE
Elizabeth A. Beverly, PhD, Department of Family Medicine, Ohio University Heritage College of Osteopathic Medicine, 35 W. Green Drive, Athens, OH 45701; [email protected].

Managing diabetes is a complex undertaking, with an extensive regimen of self-care—including regular exercise, meal planning, blood glucose monitoring, medication scheduling, and multiple visits—that is critically linked to glycemic control and the prevention of complications. Incorporating all of these elements into daily life can be daunting.^1-3

In fact, nearly half of US adults with diabetes fail to meet the recommended targets.⁴ This leads to frustration, which often manifests in psychosocial problems that further hamper efforts to manage the disease.^5-10 The most notable is a psychosocial disorder known as diabetes distress, which affects close to 45% of those with diabetes.^11,12

It is important to note that diabetes distress is not a psychiatric disorder;¹³ rather, it is a broad affective reaction to the stress of living with this chronic and complex disease.^14,15 By negatively affecting adherence to a self-care regimen, diabetes distress contributes to worsening glycemic control and increasing morbidity.^16-18

Recognizing that about 80% of those with diabetes are treated in primary care settings,¹⁹ we wrote this review to call your attention to diabetes distress, alert you to brief screening tools that can easily be incorporated into clinic visits, and offer guidance in matching proposed interventions to the aspects of diabetes self-management that cause patients the greatest distress.

Diabetes distress: What it is, what it’s not

For patients with type 2 diabetes, diabetes distress centers around 4 main issues:

• frustration with the demands of self-care;
• apprehension about the future and the possibility of developing serious complications;
• concern about both the quality and the cost of required medical care; and
• perceived lack of support from family and/or friends.^11,12,20

As mentioned earlier, diabetes distress is not a psychiatric condition and should not be confused with major depressive disorder (MDD). Here’s help in telling the difference.

Unlike major depressive disorder, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.

For starters, a diagnosis of depression is symptom-based.¹³ MDD requires the presence of at least 5 of the 9 symptoms defined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth ed. (DSM-5)—eg, persistent feelings of worthlessness or guilt, sleep disturbances, lack of interest in normal activities—for at least 2 weeks.²¹ What’s more, the diagnostic criteria for MDD do not specify a cause or disease process. Nor do they distinguish between a pathological response and an expected reaction to a stressful life event.²² Further, depression measures reflect symptoms (eg, hyperglycemia), as well as stressful experiences resulting from diabetes self-care, which may contribute to the high rate of false positives or incorrect diagnoses of MDD and missed diagnoses of diabetes distress.²³

Unlike MDD, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.¹³ And, because the source of the problem is identified, diabetes distress can be treated with specific interventions targeting the areas causing the highest levels of stress.

When a psychiatric condition and diabetes distress overlap

MDD, anxiety disorders, and diabetes distress are all common in patients with diabetes,²⁴ and the co-occurrence of a psychiatric disorder and diabetes distress is high.²⁵ Thus, it is important not only to identify cases of diabetes distress but also to consider comorbid depression and/or anxiety in patients with diabetes distress.

More often, though, it is the other way around, according to the Distress and Depression in Diabetes (3D) study. The researchers recently found that 84% of patients with moderate or high diabetes distress did not fulfill the criteria for MDD, but that 67% of diabetes patients with MDD also had moderate or high diabetes distress.^13,15,17,25

The data highlight the importance of screening patients with a dual diagnosis of diabetes and MDD for diabetes distress. Keep in mind that individuals diagnosed with both diabetes distress and a comorbid psychiatric condition may require more complex and intensive treatment than those with either diabetes distress or MDD alone.²⁵

Screening for diabetes distress

Diabetes distress can be easily assessed using one of several patient-reported outcome measures. Six validated measures, ranging in length from one to 28 questions, are designed for use in primary care (TABLE).^26-30 Some of the measures are easily accessible online; others require subscription to MEDLINE.

Problem Areas in Diabetes (PAID): There are 3 versions of PAID—a 20-item screen assessing a broad range of feelings related to living with diabetes and its treatment, a 5-item version (PAID-5) with high rates of sensitivity (95%) and specificity (89%), and a single-item test (PAID-1) that is highly correlated with the longer version.^26,27

Diabetes Distress Scale (DDS): This tool is available in a 17-item measure assessing diabetes distress as it relates to the emotional burden, physician-related distress, regimen-related distress, and interpersonal distress.²⁸ DDS is also available in a short form (DDS-2) with 2 items²⁹ and a 28-item scale specifically for patients with type 1 diabetes.³⁰ T1-DDS, the only diabetes distress measure focused on this particular patient population, assesses the 7 sources of distress found to be common among adults with type 1 diabetes: powerlessness, negative social perceptions, physician distress, friend/family distress, hypoglycemia distress, management distress, and eating distress.

Studies have shown that not only do those with type 1 diabetes experience different stressors compared with their type 2 counterparts, but that they tend to experience distress differently. For patients with type 1 diabetes, for example, powerlessness ranked as the highest source of distress, followed by eating distress and hypoglycemia distress. These sources of distress differ from the regimen distress, emotional burden, interpersonal distress, and physician distress identified by those with type 2 diabetes.³⁰

How to respond to diabetes distress

Diabetes distress is easier to identify than to successfully treat. Few validated treatments for diabetes distress exist and, to our knowledge, only 2 studies have assessed interventions aimed at reduction of such distress.^31,32

The REDEEM trial³¹ recruited adults with type 2 diabetes and diabetes distress to participate in a 12-month randomized controlled trial (RCT). The trial had 3 arms, comparing the effectiveness of a computer-assisted self-management (CASM) program alone, a CASM program plus in-person diabetes distress-specific problem-solving therapy, and a computer-assisted minimally supportive intervention. The main outcomes included diabetes distress (using the DDS scale and subscales), along with self-management behaviors and HbA1c.

Participants in all 3 arms showed significant reductions in total diabetes distress and improvements in self-management behaviors, with no significant differences among the groups. No differences in HbA1c were found. However, those in the CASM program plus distress-specific therapy arm showed a larger reduction in regimen distress compared with the other 2 groups.³¹

The DIAMOS trial³² recruited adults who had type 1 or type 2 diabetes, diabetes distress, and subclinical depressive symptoms for a 2-arm RCT. One group underwent cognitive behavioral interventions, while the controls had standard group-based diabetes education. The main outcomes included diabetes distress (measured via the PAID scale), depressive symptoms, well-being, diabetes self-care, diabetes acceptance, satisfaction with diabetes treatment, HbA1c, and subclinical inflammation.

Major depressive disorder, anxiety disorders, and diabetes distress are all common in patients with diabetes.

The intervention group showed greater improvement in diabetes distress and depressive symptoms compared with the control group, but no differences in well-being, self-care, treatment satisfaction, HbA1c, or subclinical inflammation were observed.³²

Both studies support the use of problem-solving therapy and cognitive behavioral interventions for patients with diabetes distress. Future research should evaluate the effectiveness of these interventions in the primary care setting.

What else to offer when challenges mount?

Diabetes is a progressive disease, and most patients experience multiple challenges over time. These typically include complications and comorbidities, physical limitations, polypharmacy, hypoglycemia, and cognitive impairment, as well as changes in everything from medication and lifestyle to insurance coverage and social support.^33,34 All increase the risk for diabetes distress, as well as related psychiatric conditions.

Eighty-four percent of patients with moderate or high diabetes distress didn’t fulfill the criteria for MDD, but 67% of diabetes patients with MDD also had diabetes distress.

Aging and diabetes are independent risk factors for cognitive impairment, for example, and the presence of both increases this risk.³⁵ What’s more, diabetes alone is associated with poorer executive function,^36-38 the higher-level cognitive processes that allow individuals to engage in independent, purposeful, and flexible goal-related behaviors. Both poor cognitive function and impaired executive function interfere with the ability to perform self-care behaviors such as adjusting insulin doses, drawing insulin into a syringe, or dialing an insulin dose with an insulin pen.³⁹ This in turn can lead to frustration and increase the likelihood of moderate to high diabetes distress.

Assessing diabetes distress in patients with cognitive impairment, poor executive functioning, or other psychological limitations is particularly difficult, however, as no diabetes distress measures take such deficits into account. Thus, primary care physicians without expertise in neuropsychology should consider referring patients with such problems to specialists for assessment.

Be alert to socioeconomic changes—in employment, insurance coverage, and living situations—that are not addressed in the screening tools.

The progressive nature of diabetes also highlights the need for primary care physicians to periodically screen for diabetes distress and engage in ongoing discussions about what type of care is best for individual patients, and why. When developing or updating treatment plans and making recommendations, it is crucial to consider the impact the treatment would likely have on the patient’s physical and mental health and to explicitly inquire about and acknowledge his or her values and preferences for care.^40-44

It is also important to remain aware of socioeconomic changes—in employment, insurance coverage, and living situations, for example—which are not addressed in the screening tools.

Moderate to high diabetes distress scores, as well as individual items patients identify as “very serious” problems, represent clinical red flags that should be the focus of careful discussion during a medical visit. Patients with moderate to high distress should be referred to a therapist trained in cognitive behavioral therapy or problem-solving therapy. Physicians who lack access to such resources can incorporate cognitive behavioral and problem-solving techniques into patient discussion. (See “Directing help where it’s most needed.”) All patients should be referred to a certified diabetes educator—a key component of diabetes care.^45,46

CORRESPONDENCE
Elizabeth A. Beverly, PhD, Department of Family Medicine, Ohio University Heritage College of Osteopathic Medicine, 35 W. Green Drive, Athens, OH 45701; [email protected].

Managing diabetes is a complex undertaking, with an extensive regimen of self-care—including regular exercise, meal planning, blood glucose monitoring, medication scheduling, and multiple visits—that is critically linked to glycemic control and the prevention of complications. Incorporating all of these elements into daily life can be daunting.^1-3

In fact, nearly half of US adults with diabetes fail to meet the recommended targets.⁴ This leads to frustration, which often manifests in psychosocial problems that further hamper efforts to manage the disease.^5-10 The most notable is a psychosocial disorder known as diabetes distress, which affects close to 45% of those with diabetes.^11,12

It is important to note that diabetes distress is not a psychiatric disorder;¹³ rather, it is a broad affective reaction to the stress of living with this chronic and complex disease.^14,15 By negatively affecting adherence to a self-care regimen, diabetes distress contributes to worsening glycemic control and increasing morbidity.^16-18

Recognizing that about 80% of those with diabetes are treated in primary care settings,¹⁹ we wrote this review to call your attention to diabetes distress, alert you to brief screening tools that can easily be incorporated into clinic visits, and offer guidance in matching proposed interventions to the aspects of diabetes self-management that cause patients the greatest distress.

Diabetes distress: What it is, what it’s not

For patients with type 2 diabetes, diabetes distress centers around 4 main issues:

• frustration with the demands of self-care;
• apprehension about the future and the possibility of developing serious complications;
• concern about both the quality and the cost of required medical care; and
• perceived lack of support from family and/or friends.^11,12,20

As mentioned earlier, diabetes distress is not a psychiatric condition and should not be confused with major depressive disorder (MDD). Here’s help in telling the difference.

Unlike major depressive disorder, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.

For starters, a diagnosis of depression is symptom-based.¹³ MDD requires the presence of at least 5 of the 9 symptoms defined by the Diagnostic and Statistical Manual of Mental Disorders, Fifth ed. (DSM-5)—eg, persistent feelings of worthlessness or guilt, sleep disturbances, lack of interest in normal activities—for at least 2 weeks.²¹ What’s more, the diagnostic criteria for MDD do not specify a cause or disease process. Nor do they distinguish between a pathological response and an expected reaction to a stressful life event.²² Further, depression measures reflect symptoms (eg, hyperglycemia), as well as stressful experiences resulting from diabetes self-care, which may contribute to the high rate of false positives or incorrect diagnoses of MDD and missed diagnoses of diabetes distress.²³

Unlike MDD, diabetes distress has a specific cause—diabetes—and can best be understood as an emotional response to a demanding health condition.¹³ And, because the source of the problem is identified, diabetes distress can be treated with specific interventions targeting the areas causing the highest levels of stress.

When a psychiatric condition and diabetes distress overlap

MDD, anxiety disorders, and diabetes distress are all common in patients with diabetes,²⁴ and the co-occurrence of a psychiatric disorder and diabetes distress is high.²⁵ Thus, it is important not only to identify cases of diabetes distress but also to consider comorbid depression and/or anxiety in patients with diabetes distress.

More often, though, it is the other way around, according to the Distress and Depression in Diabetes (3D) study. The researchers recently found that 84% of patients with moderate or high diabetes distress did not fulfill the criteria for MDD, but that 67% of diabetes patients with MDD also had moderate or high diabetes distress.^13,15,17,25

The data highlight the importance of screening patients with a dual diagnosis of diabetes and MDD for diabetes distress. Keep in mind that individuals diagnosed with both diabetes distress and a comorbid psychiatric condition may require more complex and intensive treatment than those with either diabetes distress or MDD alone.²⁵

Screening for diabetes distress

Diabetes distress can be easily assessed using one of several patient-reported outcome measures. Six validated measures, ranging in length from one to 28 questions, are designed for use in primary care (TABLE).^26-30 Some of the measures are easily accessible online; others require subscription to MEDLINE.

Problem Areas in Diabetes (PAID): There are 3 versions of PAID—a 20-item screen assessing a broad range of feelings related to living with diabetes and its treatment, a 5-item version (PAID-5) with high rates of sensitivity (95%) and specificity (89%), and a single-item test (PAID-1) that is highly correlated with the longer version.^26,27

Diabetes Distress Scale (DDS): This tool is available in a 17-item measure assessing diabetes distress as it relates to the emotional burden, physician-related distress, regimen-related distress, and interpersonal distress.²⁸ DDS is also available in a short form (DDS-2) with 2 items²⁹ and a 28-item scale specifically for patients with type 1 diabetes.³⁰ T1-DDS, the only diabetes distress measure focused on this particular patient population, assesses the 7 sources of distress found to be common among adults with type 1 diabetes: powerlessness, negative social perceptions, physician distress, friend/family distress, hypoglycemia distress, management distress, and eating distress.

Studies have shown that not only do those with type 1 diabetes experience different stressors compared with their type 2 counterparts, but that they tend to experience distress differently. For patients with type 1 diabetes, for example, powerlessness ranked as the highest source of distress, followed by eating distress and hypoglycemia distress. These sources of distress differ from the regimen distress, emotional burden, interpersonal distress, and physician distress identified by those with type 2 diabetes.³⁰

How to respond to diabetes distress

Diabetes distress is easier to identify than to successfully treat. Few validated treatments for diabetes distress exist and, to our knowledge, only 2 studies have assessed interventions aimed at reduction of such distress.^31,32

The REDEEM trial³¹ recruited adults with type 2 diabetes and diabetes distress to participate in a 12-month randomized controlled trial (RCT). The trial had 3 arms, comparing the effectiveness of a computer-assisted self-management (CASM) program alone, a CASM program plus in-person diabetes distress-specific problem-solving therapy, and a computer-assisted minimally supportive intervention. The main outcomes included diabetes distress (using the DDS scale and subscales), along with self-management behaviors and HbA1c.

Participants in all 3 arms showed significant reductions in total diabetes distress and improvements in self-management behaviors, with no significant differences among the groups. No differences in HbA1c were found. However, those in the CASM program plus distress-specific therapy arm showed a larger reduction in regimen distress compared with the other 2 groups.³¹

The DIAMOS trial³² recruited adults who had type 1 or type 2 diabetes, diabetes distress, and subclinical depressive symptoms for a 2-arm RCT. One group underwent cognitive behavioral interventions, while the controls had standard group-based diabetes education. The main outcomes included diabetes distress (measured via the PAID scale), depressive symptoms, well-being, diabetes self-care, diabetes acceptance, satisfaction with diabetes treatment, HbA1c, and subclinical inflammation.

Major depressive disorder, anxiety disorders, and diabetes distress are all common in patients with diabetes.

The intervention group showed greater improvement in diabetes distress and depressive symptoms compared with the control group, but no differences in well-being, self-care, treatment satisfaction, HbA1c, or subclinical inflammation were observed.³²

Both studies support the use of problem-solving therapy and cognitive behavioral interventions for patients with diabetes distress. Future research should evaluate the effectiveness of these interventions in the primary care setting.

What else to offer when challenges mount?

Diabetes is a progressive disease, and most patients experience multiple challenges over time. These typically include complications and comorbidities, physical limitations, polypharmacy, hypoglycemia, and cognitive impairment, as well as changes in everything from medication and lifestyle to insurance coverage and social support.^33,34 All increase the risk for diabetes distress, as well as related psychiatric conditions.

Eighty-four percent of patients with moderate or high diabetes distress didn’t fulfill the criteria for MDD, but 67% of diabetes patients with MDD also had diabetes distress.

Aging and diabetes are independent risk factors for cognitive impairment, for example, and the presence of both increases this risk.³⁵ What’s more, diabetes alone is associated with poorer executive function,^36-38 the higher-level cognitive processes that allow individuals to engage in independent, purposeful, and flexible goal-related behaviors. Both poor cognitive function and impaired executive function interfere with the ability to perform self-care behaviors such as adjusting insulin doses, drawing insulin into a syringe, or dialing an insulin dose with an insulin pen.³⁹ This in turn can lead to frustration and increase the likelihood of moderate to high diabetes distress.

Assessing diabetes distress in patients with cognitive impairment, poor executive functioning, or other psychological limitations is particularly difficult, however, as no diabetes distress measures take such deficits into account. Thus, primary care physicians without expertise in neuropsychology should consider referring patients with such problems to specialists for assessment.

Be alert to socioeconomic changes—in employment, insurance coverage, and living situations—that are not addressed in the screening tools.

The progressive nature of diabetes also highlights the need for primary care physicians to periodically screen for diabetes distress and engage in ongoing discussions about what type of care is best for individual patients, and why. When developing or updating treatment plans and making recommendations, it is crucial to consider the impact the treatment would likely have on the patient’s physical and mental health and to explicitly inquire about and acknowledge his or her values and preferences for care.^40-44

It is also important to remain aware of socioeconomic changes—in employment, insurance coverage, and living situations, for example—which are not addressed in the screening tools.

Moderate to high diabetes distress scores, as well as individual items patients identify as “very serious” problems, represent clinical red flags that should be the focus of careful discussion during a medical visit. Patients with moderate to high distress should be referred to a therapist trained in cognitive behavioral therapy or problem-solving therapy. Physicians who lack access to such resources can incorporate cognitive behavioral and problem-solving techniques into patient discussion. (See “Directing help where it’s most needed.”) All patients should be referred to a certified diabetes educator—a key component of diabetes care.^45,46

CORRESPONDENCE
Elizabeth A. Beverly, PhD, Department of Family Medicine, Ohio University Heritage College of Osteopathic Medicine, 35 W. Green Drive, Athens, OH 45701; [email protected].

Atrial fibrillation (AF)—the most common supraventricular tachycardia—affects as many as 6.1 million adults in the United States.¹ It is associated with a 5-fold increased risk of stroke,² a 3-fold increased risk of heart failure (HF),³ and about a 2-fold increased risk of dementia⁴ and mortality.² The prevalence of AF increases with maturity, from 2% in people <65 years of age to 9% in those ≥65 years,⁵ and that prevalence is expected to double over the next 25 years as the population ages.¹

The primary goals of treatment are to alleviate symptoms and prevent thromboembolism. Strokes related to AF are more likely to result in severe disability or death when compared with those unrelated to AF.⁶ And yet anticoagulation remains underutilized.⁷

The net clinical benefit of oral anticoagulation appears to be greatest in patients with the highest risk of bleeding, since these patients are also at the highest risk for stroke.⁸ Patients at increased risk of stroke are more likely to receive oral anticoagulation; however, for unknown reasons, more than half of people with the highest risk of stroke are not prescribed these important anti-blood-clotting medications.⁷ One theory is that physicians may be relying on their gut rather than objective risk scores, and underuse of validated schemata leads to poor estimation of risk.

IMAGE: © ALICIA BUELO

For example, results from the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) trial, which involved over 10,000 people with AF, found that although 72% (n=7251) had high-risk CHADS₂ scores (≥2), only 16% were assessed as having a high risk of stroke by physicians.⁹ Along the same lines, a recent study of Canadian primary care physicians showed that stroke risk and bleeding risk were not evaluated with validated tools in 58% and 81% of patients, respectively, leading to both significant underestimation and overestimation of risk.¹⁰

This review provides the tools to identify when anticoagulation is indicated, reports the advantages and disadvantages of the currently available anticoagulants, and discusses the selection and implementation of rate- vs rhythm-control strategies. But first, a word about the etiology, classification, and diagnosis of AF.

AF: The result of any number of cardiac and non-cardiac causes

AF is characterized by uncoordinated activation of the atria, which results in ineffective atrial contractions and an irregular, often rapid, ventricular response. It is the ultimate clinical manifestation of multiple diseases that alter atrial tissue through inflammation, fibrosis, or hypertrophy.⁵ The most common causes are hypertension, coronary artery disease, HF, cardiomyopathies, and valvular heart disease, all of which stimulate the renin-angiotensin-aldosterone system, leading to increased susceptibility to arrhythmia.⁵ Atrial ectopic tachycardia, Wolff-Parkinson-White (WPW) syndrome, and atrioventricular (AV) nodal reentrant tachycardia also may precipitate AF.⁵ In these cases, AF usually resolves after catheter ablation (CA) of the primary arrhythmia.¹¹ Unrecognized AF may trigger atrial flutter, and more than 80% of patients who undergo radiofrequency ablation for atrial flutter experience AF at some point in the subsequent 5 years.¹²

Strokes related to atrial fibrillation are more likely to result in severe disability or death when compared with those unrelated to AF. And yet anticoagulation remains underutilized.

Non-cardiac causes of AF include sleep apnea, obesity, hyperthyroidism, drugs, electrocution, pneumonia, and pulmonary embolism.⁵ An association between binge drinking and AF (“holiday heart syndrome”) has long been recognized. The evidence now suggests that alcohol increases the risk of AF in a dose-dependent manner with intakes of ≥1 drink per day (12 g per drink).¹³

Classification schema no longer includes “lone AF”

AF is classified in terms of the duration of episodes:⁵

• Paroxysmal AF is characterized by brief episodes that terminate spontaneously or with intervention within 7 days of onset. These episodes recur with variable frequency.
• Persistent AF refers to AF that is continuously sustained for more than 7 days.
• Longstanding persistent AF refers to continuous AF that lasts longer than 12 months.
• Permanent AF is not an inherent pathophysiologic attribute of AF, but rather an acceptance of AF where the patient and physician abandon further efforts to restore and/or maintain sinus rhythm.
• Nonvalvular AF occurs in the absence of a valve replacement (mechanical or bioprosthetic), rheumatic mitral stenosis, or mitral valve repair.

Although paroxysmal and persistent AF may occur in the same individual, the distinction is still clinically relevant, as outcomes of certain therapies, such as CA, are superior in patients with paroxysmal AF.¹⁴ With a more complete understanding of AF pathophysiology, guidelines now discourage use of the potentially confusing term “lone AF,” which has historically been applied to younger patients with no known clinical risk factors or echocardiographic abnormalities. As a result, therapeutic decisions are no longer based on this nomenclature, according to the 2014 AF practice guideline from the American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS).⁵

Patient complaints—or incidental findings—can prompt a Dx

Fatigue is the most common symptom of AF. Other signs and symptoms include palpitations, dyspnea, HF, hypotension, syncope, chest pain, and stroke. Some patients are asymptomatic, and AF is an incidental finding when an irregular pulse is discovered during a physical examination. The diagnosis is confirmed by electrocardiogram (EKG), telemetry, Holter monitor, event recorder, or an implanted electrocardiographic recording device. A chest x-ray, serum electrolyte levels, a complete blood count, thyroid testing, and renal and hepatic function testing are recommended. Transthoracic echocardiography to measure cardiac function, detect underlying structural heart disease, and evaluate atrial size is essential.⁵

Warfarin remains the only recommended anticoagulation strategy for patients with severe renal impairment or valvular atrial fibrillation.

An electrophysiologic (EP) study may be needed for diagnosis or treatment if another arrhythmia is present. Aberrant conduction may cause AF to present as a wide complex tachycardia and be mislabeled as ventricular tachycardia. The presence of delta waves is an indication for an EP study targeting the WPW accessory pathway. Transesophageal echocardiography (TEE) is the most sensitive and specific test for left atrial thrombi. If you are considering a TEE for a patient with AF of unknown, or >48 hours’, duration who has not been anticoagulated in the preceding 3 weeks, obtain it before performing cardioversion because of the risk of embolism.⁵

Stroke prevention

The ACC/AHA/HRS AF guideline recommends basing anticoagulation decisions on thromboembolic risk, regardless of AF pattern (paroxysmal, persistent, or permanent) (Class I recommendation).⁵ For patients with nonvalvular AF and atrial flutter, the guideline recommends using the Birmingham 2009 schema (CHA₂DS₂-VASc score) (TABLE 1^15-18) to estimate thromboembolic risk.^5,15 CHA₂DS₂-VASc improves on the older CHADS₂ score by significantly reducing the number of patients categorized as having intermediate risk and better identifying truly low-risk patients who are unlikely to benefit from anticoagulation.^16,17,19

Men with a CHA₂DS₂-VASc score of zero and women with a score of one do not need anticoagulation.^5,20 Discuss the risks and benefits of oral anticoagulation with men who have a score of one. In these intermediate-risk men, antiplatelet therapy with aspirin and/or clopidogrel may be reasonable, especially if there is an indication other than stroke prevention (eg, post-myocardial infarction). Oral anticoagulation is strongly recommended for all patients with a CHA₂DS₂-VASc score of 2 or higher.^5,18,21,22

Anticoagulant considerations: Warfarin vs DOACs

Warfarin was the gold standard for stroke prevention in nonvalvular AF until the direct oral anticoagulants (DOACs) became available in 2010. Guidelines in the United States and the United Kingdom recommend shared decision-making to help patients with AF who do not have a specific indication for warfarin choose between warfarin and the DOACs.^5,21 Canadian and European guidelines recommend DOACs as the first-line option for anticoagulation and reserve warfarin for patients who have contraindications to, or are unable to afford, DOACs.^18,22 All current guidelines recommend continuing warfarin in patients who are stable, well controlled, and satisfied with warfarin therapy and the monitoring and dietary restrictions it entails.

DOACs are as effective as warfarin. All of the DOACs are approved for stroke prevention based on individual phase III non-inferiority trials in which they were compared to warfarin.^23-26 In addition, a meta-analysis of these 4 trials involving a total of 71,683 patients (mean age 70-73 years; median follow-up, 1.8-2.8 years) evaluated the benefits and risks of the 4 DOACs against the former gold standard.²⁷

Higher doses of the DOACs (dabigatran 150 mg BID, rivaroxaban 20 mg/d, edoxaban 60 mg/d, and apixaban 5 mg BID) reduced the rates of stroke or systemic embolism (relative risk [RR]=0.81; 95% confidence interval [CI], 0.73-0.91; P<.0001; number needed to treat [NNT]=147), hemorrhagic stroke (RR=0.49; 95% CI, 0.38-0.64; P<.0001; NNT=219), and all-cause mortality (RR=0.90; 95% CI, 0.85-0.95; P=.0003; NNT=128), compared with warfarin.²⁷ It is important to note that while lower doses of some DOACs (dabigatran 110 mg BID and edoxaban 30 mg/d) were not as effective at preventing ischemic stroke when compared with warfarin (RR=1.3; 95% CI, 1-1.6; P=.045), they still significantly reduced hemorrhagic stroke (RR=0.33; 95% CI, 0.23-0.46; P<.0001) and all-cause mortality (RR=0.89; 95% CI, 0.83-0.96; P=.003).

Of course, the biggest concern is bleeding. In that same meta-analysis, the difference in major bleeding events with DOACs vs warfarin was not statistically significant (RR=0.86; 95% CI, 0.73-1; P=.06). While DOACs likely lower rates of intracranial hemorrhage (RR=0.48; 95% CI, 0.39-0.59; P<.0001; NNT=132), they seem to increase the risk of gastrointestinal (GI) bleeding (RR=1.3; 95% CI, 1-1.6; P=.043; number needed to harm [NNH]=185).²⁷

Without head-to-head trials, it is impossible to know if one direct oral anticoagulant is superior to another.

There was significant heterogeneity in the GI bleeding outcome, however. When compared with warfarin, GI bleeding was increased by dabigatran 150 mg BID (RR=1.5; 95% CI, 1.2-1.9; P<.001) and edoxaban 60 mg/d (HR=1.2; 95% CI, 1.02-1.5; P=.03), but there were no significant differences for dabigatran 110 mg BID or apixaban 5 mg BID.^23,25,26

On the other hand, edoxaban 30 mg/d had a lower risk of GI bleeding when compared with warfarin (HR=0.67; 95% CI, 0.53-0.83; P<.001).²⁵ Without head-to-head trials, it is impossible to know if one DOAC is superior to another. Apixaban 5 mg BID appears to offer the best overall balance between efficacy and safety. Other DOACs may be better options for patients who have specific concerns regarding efficacy or safety.^28,29

Convenience, interactions, and cost may be the deciding factors. Since all DOACs are fairly comparable in efficacy and safety, other factors such as convenience, interactions with other medications, and cost should be considered when deciding on a medication for an individual patient (TABLE 2^30,31). The DOACs require no lab monitoring or dose titration, and all 4 have fewer potential drug interactions than warfarin.³⁰ Due to their relatively short half-lives, strict adherence is critical; DOACs are not suitable for patients who frequently miss doses.⁵ (For more information on starting or switching to DOACs, see, “Is a novel anticoagulant right for your patient?” J Fam Pract. 2014;63:22-28.)

A word about DOACs and renal impairment. Another concern with DOACs is their reliance on renal metabolism and excretion. A meta-analysis of the 4 phase III trials of the DOACs, this time involving 58,338 patients, evaluated DOAC efficacy and safety compared to warfarin in the presence of kidney dysfunction.³² Renal function was categorized as normal (estimated glomerular filtration rate [eGFR] >80 mL/min/1.73 m²), mildly impaired (eGFR 50-80 mL/min/1.73 m²), or moderately impaired (eGFR <50 mL/min/1.73m²). Compared with warfarin, DOACs lowered stroke risk in patients with mild (RR=0.71; 95% CI, 0.62-0.81) or moderate (RR=0.79; 95% CI, 0.66-0.94) renal impairment. DOACs also reduced major bleeding compared to warfarin in patients with mild (RR=0.88; 95% CI, 0.80-0.97) or moderate (RR=0.80; 95% CI, 0.66-0.94) renal impairment. How the DOACs fare in patients with severe renal dysfunction could not be determined because such patients were excluded from the trials.

Keep in mind that the DOACs require dose adjustment at different levels of renal impairment (TABLE 2^30,31), and warfarin remains the only recommended treatment for patients with severe renal impairment, according to both AHA/ACC/HRS and European Society of Cardiology guidelines.^5,18

Tools to help assess patients’ bleeding risk

Of the available scoring mechanisms to identify risk factors for bleeding, 3 have been specifically validated in AF populations (ie, ATRIA,³³ HEMORR₂HAGES,³⁴ and HAS-BLED³⁵). Of the 3, HAS-BLED is superior,³⁶ the most practical, and recommended by expert guidelines.^18,21,22 Additionally, HAS-BLED has good correlation with intracranial hemorrhage risk. The HAS-BLED score ranges from 0 to 9 points with one point assigned for each of the following:³⁵

• Hypertension–uncontrolled with systolic BP >160 mm Hg
• Abnormal liver function–cirrhosis, bilirubin >2× normal, or liver enzymes >3× normal
• Abnormal renal function–dialysis, transplant, or serum creatinine >2.26 mg/dL
• Stroke history–including lacunar infarcts
• Bleeding predisposition–history of major bleeding due to any cause
• Labile international normalized ratio (INR)–time in therapeutic range <60%
• Elderly–age >65 years
• Drug–antiplatelet agents, including nonsteroidal anti-inflammatory drugs
• Alcohol usage–>8 drinks per week.

Patients with a HAS-BLED score ≥3 warrant additional monitoring and attempts to reduce bleeding risk by addressing modifiable risk factors. Bleeding risk scores should not be used to exclude patients from anticoagulation therapy.⁵ In fact, the British National Institute for Health and Clinical Excellence (NICE) guidelines state that anticoagulation should not be withheld solely due to fall risk.²¹

Also, anticoagulation with warfarin should not be permanently discontinued because of a single GI bleed, since restarting warfarin is associated with decreased risks of thromboembolism and mortality and a statistically insignificant increase in recurrent GI bleeding.³⁷ Restarting DOAC therapy following a GI bleed has not been evaluated in clinical trials; however, it may be reasonable to use one of the DOAC doses with a lower risk of GI bleeding (dabigatran 110 mg BID, apixaban 5 mg BID, or edoxaban 30 mg/d) in patients who have experienced a GI bleed on warfarin or another DOAC.^18,22

An online calculator is available that uses CHA₂DS₂-VASc and HAS-BLED scores to determine an individual’s risk/benefit profile with the various anticoagulation strategies available (http://www.sparctool.com). Consider percutaneous left atrial appendage occlusion if the risks of anticoagulation truly exceed the benefits.³⁸

Rate control vs rhythm control

Most patients who present with AF require immediate ventricular rate control to reduce symptoms. In the acute setting, this can be accomplished with intravenous (IV) beta-blockers or IV calcium channel antagonists.^5,39 If the patient is hemodynamically unstable, urgent direct-current cardioversion is the preferred treatment strategy and should not be delayed pending anticoagulation. IV amiodarone can be used in the ICU patient who does not require cardioversion, but is unable to tolerate beta-blockers or calcium channel antagonists.⁴⁰ Once the patient is stable, long-term treatment focuses on ventricular rate control or restoration and maintenance of sinus rhythm.

Direct oral anticoagulants are not suitable for patients who frequently miss doses.

The AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) trial enrolled 4060 patients (mean age 70 years, mean follow-up 3.5 years) with paroxysmal and persistent AF and randomized them to either pharmacologic rate control or rhythm control.⁴¹ No significant differences were found in all-cause mortality or in the composite secondary endpoint of death, ischemic stroke, anoxic encephalopathy, major bleeding, or cardiac arrest. In addition, no significant differences emerged in quality of life or global functional status. The number of patients requiring hospitalization during follow-up was significantly lower in the rate-control group vs the rhythm-control group (73% vs 80%; P<.001). Anticoagulation was encouraged but not mandated in the rhythm-control group after 4 weeks in sinus rhythm, and there was a trend toward higher mortality in the rhythm-control group (27% vs 26%; P=.08).

Patients <65 years were excluded from the AFFIRM trial. When younger patients experience significant symptoms, early referral to Cardiology should be considered to discuss the long-term benefits and risks of a rhythm-control strategy. Regardless of age, when patients remain symptomatic despite rate- or rhythm-control management, the strategy should be changed.⁵

Rate-control targets and options

Target heart rates should be individualized. The 2014 ACC/AHA/HRS guideline recommends a resting target heart rate <80 beats per minute (bpm) in symptomatic patients.⁵ In patients with permanent AF who remain asymptomatic at higher resting heart rates, a more lenient rate-control strategy (resting heart rate <110 bpm) has demonstrated outcomes equivalent to those of a more strict approach (resting heart rate <80 bpm and heart rate during moderate exercise <110 bpm).⁴² Pharmacologic rate-control options include beta-blockers, non-dihydropyridine calcium channel antagonists, and digoxin (TABLE 3⁵). Digoxin is associated with increased all-cause mortality in patients with AF regardless of HF status (HR=1.4; 95% CI, 1.2-1.6, P=.0001).⁴³ Digoxin should be reserved for patients who are sedentary or have inadequate control with first-line medications.⁵

Indications for rhythm control

The NICE guidelines, which are consistent with the ACC/AHA/HRS guidelines, recommend rate control as the first-line strategy for AF management, except in people:²¹

• whose AF has a reversible cause
• who have HF believed to be primarily caused by AF
• with new-onset AF
• with atrial flutter that is considered suitable for an ablation strategy to restore sinus rhythm
• for whom a rhythm-control strategy would be more suitable based on clinical judgment.

In addition, patients who continue to experience symptomatic AF despite an adequate trial of rate control should be offered rhythm control.⁵

Pharmacologic rhythm-control strategies. Antiarrhythmic drugs can be used for chemical cardioversion, reduction of paroxysms, and long-term maintenance of sinus rhythm. The most commonly used antiarrhythmic drugs are Class IC and Class III agents (TABLE 3).⁵ Tailored drug selection for each patient is key. Patients with left atrial diameters >4.5 cm are less likely to remain in sinus rhythm, and patients with left ventricular hypertrophy are at increased risk for proarrhythmic adverse effects.⁴⁴ Patients with paroxysmal AF may be candidates for a “pill-in-the-pocket” strategy using propafenone or flecainide.⁵

AF frequently progresses from paroxysmal to persistent and can subsequently result in electrical and structural remodeling that becomes irreversible over time.⁴⁵ The patient with uncontrolled symptoms despite attempts at rate control and rhythm control should be promptly referred to an electrophysiologist.

Surgical interventions for rate or rhythm control

Electrophysiology interventions include AV nodal ablation with pacemaker placement for rate control, or catheter-directed ablation (radiofrequency or cryotherapy) for rhythm control. CA appears to be more effective than pharmacologic rhythm control.^46,47 Treatment with CA is indicated for symptomatic paroxysmal AF when a rhythm-control strategy is desired and the AF is refractory to, or the patient is intolerant of, at least one class I or III antiarrhythmic medication.⁵ With these same caveats, CA is a reasonable strategy for symptomatic persistent AF.

Consider more invasive interventions, such as an atrial maze procedure, when patients require cardiac surgery for another indication. Patients with an increased risk of thromboembolism (based on CHA₂DS₂-VASc) remain at high risk even after successful ablation.⁴⁸ As a result, some guidelines recommend continued long-term anticoagulation following CA.^18,22

CORRESPONDENCE
Philip Dooley, MD, University of Kansas School of Medicine–Wichita Family Medicine Residency at Via Christi, 707 North Emporia, Wichita, KS 67207; [email protected].

ACKNOWLEDGMENTS
We thank Professor Anne Walling, MB, ChB, FFPHM, Department of Family and Community Medicine, University of Kansas School of Medicine–Wichita for her suggestions and critical review of an earlier version of this manuscript.

Atrial fibrillation (AF)—the most common supraventricular tachycardia—affects as many as 6.1 million adults in the United States.¹ It is associated with a 5-fold increased risk of stroke,² a 3-fold increased risk of heart failure (HF),³ and about a 2-fold increased risk of dementia⁴ and mortality.² The prevalence of AF increases with maturity, from 2% in people <65 years of age to 9% in those ≥65 years,⁵ and that prevalence is expected to double over the next 25 years as the population ages.¹

The primary goals of treatment are to alleviate symptoms and prevent thromboembolism. Strokes related to AF are more likely to result in severe disability or death when compared with those unrelated to AF.⁶ And yet anticoagulation remains underutilized.⁷

The net clinical benefit of oral anticoagulation appears to be greatest in patients with the highest risk of bleeding, since these patients are also at the highest risk for stroke.⁸ Patients at increased risk of stroke are more likely to receive oral anticoagulation; however, for unknown reasons, more than half of people with the highest risk of stroke are not prescribed these important anti-blood-clotting medications.⁷ One theory is that physicians may be relying on their gut rather than objective risk scores, and underuse of validated schemata leads to poor estimation of risk.

IMAGE: © ALICIA BUELO

For example, results from the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) trial, which involved over 10,000 people with AF, found that although 72% (n=7251) had high-risk CHADS₂ scores (≥2), only 16% were assessed as having a high risk of stroke by physicians.⁹ Along the same lines, a recent study of Canadian primary care physicians showed that stroke risk and bleeding risk were not evaluated with validated tools in 58% and 81% of patients, respectively, leading to both significant underestimation and overestimation of risk.¹⁰

This review provides the tools to identify when anticoagulation is indicated, reports the advantages and disadvantages of the currently available anticoagulants, and discusses the selection and implementation of rate- vs rhythm-control strategies. But first, a word about the etiology, classification, and diagnosis of AF.

AF: The result of any number of cardiac and non-cardiac causes

AF is characterized by uncoordinated activation of the atria, which results in ineffective atrial contractions and an irregular, often rapid, ventricular response. It is the ultimate clinical manifestation of multiple diseases that alter atrial tissue through inflammation, fibrosis, or hypertrophy.⁵ The most common causes are hypertension, coronary artery disease, HF, cardiomyopathies, and valvular heart disease, all of which stimulate the renin-angiotensin-aldosterone system, leading to increased susceptibility to arrhythmia.⁵ Atrial ectopic tachycardia, Wolff-Parkinson-White (WPW) syndrome, and atrioventricular (AV) nodal reentrant tachycardia also may precipitate AF.⁵ In these cases, AF usually resolves after catheter ablation (CA) of the primary arrhythmia.¹¹ Unrecognized AF may trigger atrial flutter, and more than 80% of patients who undergo radiofrequency ablation for atrial flutter experience AF at some point in the subsequent 5 years.¹²

Strokes related to atrial fibrillation are more likely to result in severe disability or death when compared with those unrelated to AF. And yet anticoagulation remains underutilized.

Non-cardiac causes of AF include sleep apnea, obesity, hyperthyroidism, drugs, electrocution, pneumonia, and pulmonary embolism.⁵ An association between binge drinking and AF (“holiday heart syndrome”) has long been recognized. The evidence now suggests that alcohol increases the risk of AF in a dose-dependent manner with intakes of ≥1 drink per day (12 g per drink).¹³

Classification schema no longer includes “lone AF”

AF is classified in terms of the duration of episodes:⁵

• Paroxysmal AF is characterized by brief episodes that terminate spontaneously or with intervention within 7 days of onset. These episodes recur with variable frequency.
• Persistent AF refers to AF that is continuously sustained for more than 7 days.
• Longstanding persistent AF refers to continuous AF that lasts longer than 12 months.
• Permanent AF is not an inherent pathophysiologic attribute of AF, but rather an acceptance of AF where the patient and physician abandon further efforts to restore and/or maintain sinus rhythm.
• Nonvalvular AF occurs in the absence of a valve replacement (mechanical or bioprosthetic), rheumatic mitral stenosis, or mitral valve repair.

Although paroxysmal and persistent AF may occur in the same individual, the distinction is still clinically relevant, as outcomes of certain therapies, such as CA, are superior in patients with paroxysmal AF.¹⁴ With a more complete understanding of AF pathophysiology, guidelines now discourage use of the potentially confusing term “lone AF,” which has historically been applied to younger patients with no known clinical risk factors or echocardiographic abnormalities. As a result, therapeutic decisions are no longer based on this nomenclature, according to the 2014 AF practice guideline from the American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS).⁵

Patient complaints—or incidental findings—can prompt a Dx

Fatigue is the most common symptom of AF. Other signs and symptoms include palpitations, dyspnea, HF, hypotension, syncope, chest pain, and stroke. Some patients are asymptomatic, and AF is an incidental finding when an irregular pulse is discovered during a physical examination. The diagnosis is confirmed by electrocardiogram (EKG), telemetry, Holter monitor, event recorder, or an implanted electrocardiographic recording device. A chest x-ray, serum electrolyte levels, a complete blood count, thyroid testing, and renal and hepatic function testing are recommended. Transthoracic echocardiography to measure cardiac function, detect underlying structural heart disease, and evaluate atrial size is essential.⁵

Warfarin remains the only recommended anticoagulation strategy for patients with severe renal impairment or valvular atrial fibrillation.

An electrophysiologic (EP) study may be needed for diagnosis or treatment if another arrhythmia is present. Aberrant conduction may cause AF to present as a wide complex tachycardia and be mislabeled as ventricular tachycardia. The presence of delta waves is an indication for an EP study targeting the WPW accessory pathway. Transesophageal echocardiography (TEE) is the most sensitive and specific test for left atrial thrombi. If you are considering a TEE for a patient with AF of unknown, or >48 hours’, duration who has not been anticoagulated in the preceding 3 weeks, obtain it before performing cardioversion because of the risk of embolism.⁵

Stroke prevention

The ACC/AHA/HRS AF guideline recommends basing anticoagulation decisions on thromboembolic risk, regardless of AF pattern (paroxysmal, persistent, or permanent) (Class I recommendation).⁵ For patients with nonvalvular AF and atrial flutter, the guideline recommends using the Birmingham 2009 schema (CHA₂DS₂-VASc score) (TABLE 1^15-18) to estimate thromboembolic risk.^5,15 CHA₂DS₂-VASc improves on the older CHADS₂ score by significantly reducing the number of patients categorized as having intermediate risk and better identifying truly low-risk patients who are unlikely to benefit from anticoagulation.^16,17,19

Men with a CHA₂DS₂-VASc score of zero and women with a score of one do not need anticoagulation.^5,20 Discuss the risks and benefits of oral anticoagulation with men who have a score of one. In these intermediate-risk men, antiplatelet therapy with aspirin and/or clopidogrel may be reasonable, especially if there is an indication other than stroke prevention (eg, post-myocardial infarction). Oral anticoagulation is strongly recommended for all patients with a CHA₂DS₂-VASc score of 2 or higher.^5,18,21,22

Anticoagulant considerations: Warfarin vs DOACs

Warfarin was the gold standard for stroke prevention in nonvalvular AF until the direct oral anticoagulants (DOACs) became available in 2010. Guidelines in the United States and the United Kingdom recommend shared decision-making to help patients with AF who do not have a specific indication for warfarin choose between warfarin and the DOACs.^5,21 Canadian and European guidelines recommend DOACs as the first-line option for anticoagulation and reserve warfarin for patients who have contraindications to, or are unable to afford, DOACs.^18,22 All current guidelines recommend continuing warfarin in patients who are stable, well controlled, and satisfied with warfarin therapy and the monitoring and dietary restrictions it entails.

DOACs are as effective as warfarin. All of the DOACs are approved for stroke prevention based on individual phase III non-inferiority trials in which they were compared to warfarin.^23-26 In addition, a meta-analysis of these 4 trials involving a total of 71,683 patients (mean age 70-73 years; median follow-up, 1.8-2.8 years) evaluated the benefits and risks of the 4 DOACs against the former gold standard.²⁷

Higher doses of the DOACs (dabigatran 150 mg BID, rivaroxaban 20 mg/d, edoxaban 60 mg/d, and apixaban 5 mg BID) reduced the rates of stroke or systemic embolism (relative risk [RR]=0.81; 95% confidence interval [CI], 0.73-0.91; P<.0001; number needed to treat [NNT]=147), hemorrhagic stroke (RR=0.49; 95% CI, 0.38-0.64; P<.0001; NNT=219), and all-cause mortality (RR=0.90; 95% CI, 0.85-0.95; P=.0003; NNT=128), compared with warfarin.²⁷ It is important to note that while lower doses of some DOACs (dabigatran 110 mg BID and edoxaban 30 mg/d) were not as effective at preventing ischemic stroke when compared with warfarin (RR=1.3; 95% CI, 1-1.6; P=.045), they still significantly reduced hemorrhagic stroke (RR=0.33; 95% CI, 0.23-0.46; P<.0001) and all-cause mortality (RR=0.89; 95% CI, 0.83-0.96; P=.003).

Of course, the biggest concern is bleeding. In that same meta-analysis, the difference in major bleeding events with DOACs vs warfarin was not statistically significant (RR=0.86; 95% CI, 0.73-1; P=.06). While DOACs likely lower rates of intracranial hemorrhage (RR=0.48; 95% CI, 0.39-0.59; P<.0001; NNT=132), they seem to increase the risk of gastrointestinal (GI) bleeding (RR=1.3; 95% CI, 1-1.6; P=.043; number needed to harm [NNH]=185).²⁷

Without head-to-head trials, it is impossible to know if one direct oral anticoagulant is superior to another.

There was significant heterogeneity in the GI bleeding outcome, however. When compared with warfarin, GI bleeding was increased by dabigatran 150 mg BID (RR=1.5; 95% CI, 1.2-1.9; P<.001) and edoxaban 60 mg/d (HR=1.2; 95% CI, 1.02-1.5; P=.03), but there were no significant differences for dabigatran 110 mg BID or apixaban 5 mg BID.^23,25,26

On the other hand, edoxaban 30 mg/d had a lower risk of GI bleeding when compared with warfarin (HR=0.67; 95% CI, 0.53-0.83; P<.001).²⁵ Without head-to-head trials, it is impossible to know if one DOAC is superior to another. Apixaban 5 mg BID appears to offer the best overall balance between efficacy and safety. Other DOACs may be better options for patients who have specific concerns regarding efficacy or safety.^28,29

Convenience, interactions, and cost may be the deciding factors. Since all DOACs are fairly comparable in efficacy and safety, other factors such as convenience, interactions with other medications, and cost should be considered when deciding on a medication for an individual patient (TABLE 2^30,31). The DOACs require no lab monitoring or dose titration, and all 4 have fewer potential drug interactions than warfarin.³⁰ Due to their relatively short half-lives, strict adherence is critical; DOACs are not suitable for patients who frequently miss doses.⁵ (For more information on starting or switching to DOACs, see, “Is a novel anticoagulant right for your patient?” J Fam Pract. 2014;63:22-28.)

A word about DOACs and renal impairment. Another concern with DOACs is their reliance on renal metabolism and excretion. A meta-analysis of the 4 phase III trials of the DOACs, this time involving 58,338 patients, evaluated DOAC efficacy and safety compared to warfarin in the presence of kidney dysfunction.³² Renal function was categorized as normal (estimated glomerular filtration rate [eGFR] >80 mL/min/1.73 m²), mildly impaired (eGFR 50-80 mL/min/1.73 m²), or moderately impaired (eGFR <50 mL/min/1.73m²). Compared with warfarin, DOACs lowered stroke risk in patients with mild (RR=0.71; 95% CI, 0.62-0.81) or moderate (RR=0.79; 95% CI, 0.66-0.94) renal impairment. DOACs also reduced major bleeding compared to warfarin in patients with mild (RR=0.88; 95% CI, 0.80-0.97) or moderate (RR=0.80; 95% CI, 0.66-0.94) renal impairment. How the DOACs fare in patients with severe renal dysfunction could not be determined because such patients were excluded from the trials.

Keep in mind that the DOACs require dose adjustment at different levels of renal impairment (TABLE 2^30,31), and warfarin remains the only recommended treatment for patients with severe renal impairment, according to both AHA/ACC/HRS and European Society of Cardiology guidelines.^5,18

Tools to help assess patients’ bleeding risk

Of the available scoring mechanisms to identify risk factors for bleeding, 3 have been specifically validated in AF populations (ie, ATRIA,³³ HEMORR₂HAGES,³⁴ and HAS-BLED³⁵). Of the 3, HAS-BLED is superior,³⁶ the most practical, and recommended by expert guidelines.^18,21,22 Additionally, HAS-BLED has good correlation with intracranial hemorrhage risk. The HAS-BLED score ranges from 0 to 9 points with one point assigned for each of the following:³⁵

• Hypertension–uncontrolled with systolic BP >160 mm Hg
• Abnormal liver function–cirrhosis, bilirubin >2× normal, or liver enzymes >3× normal
• Abnormal renal function–dialysis, transplant, or serum creatinine >2.26 mg/dL
• Stroke history–including lacunar infarcts
• Bleeding predisposition–history of major bleeding due to any cause
• Labile international normalized ratio (INR)–time in therapeutic range <60%
• Elderly–age >65 years
• Drug–antiplatelet agents, including nonsteroidal anti-inflammatory drugs
• Alcohol usage–>8 drinks per week.

Patients with a HAS-BLED score ≥3 warrant additional monitoring and attempts to reduce bleeding risk by addressing modifiable risk factors. Bleeding risk scores should not be used to exclude patients from anticoagulation therapy.⁵ In fact, the British National Institute for Health and Clinical Excellence (NICE) guidelines state that anticoagulation should not be withheld solely due to fall risk.²¹

Also, anticoagulation with warfarin should not be permanently discontinued because of a single GI bleed, since restarting warfarin is associated with decreased risks of thromboembolism and mortality and a statistically insignificant increase in recurrent GI bleeding.³⁷ Restarting DOAC therapy following a GI bleed has not been evaluated in clinical trials; however, it may be reasonable to use one of the DOAC doses with a lower risk of GI bleeding (dabigatran 110 mg BID, apixaban 5 mg BID, or edoxaban 30 mg/d) in patients who have experienced a GI bleed on warfarin or another DOAC.^18,22

An online calculator is available that uses CHA₂DS₂-VASc and HAS-BLED scores to determine an individual’s risk/benefit profile with the various anticoagulation strategies available (http://www.sparctool.com). Consider percutaneous left atrial appendage occlusion if the risks of anticoagulation truly exceed the benefits.³⁸

Rate control vs rhythm control

Most patients who present with AF require immediate ventricular rate control to reduce symptoms. In the acute setting, this can be accomplished with intravenous (IV) beta-blockers or IV calcium channel antagonists.^5,39 If the patient is hemodynamically unstable, urgent direct-current cardioversion is the preferred treatment strategy and should not be delayed pending anticoagulation. IV amiodarone can be used in the ICU patient who does not require cardioversion, but is unable to tolerate beta-blockers or calcium channel antagonists.⁴⁰ Once the patient is stable, long-term treatment focuses on ventricular rate control or restoration and maintenance of sinus rhythm.

Direct oral anticoagulants are not suitable for patients who frequently miss doses.

The AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) trial enrolled 4060 patients (mean age 70 years, mean follow-up 3.5 years) with paroxysmal and persistent AF and randomized them to either pharmacologic rate control or rhythm control.⁴¹ No significant differences were found in all-cause mortality or in the composite secondary endpoint of death, ischemic stroke, anoxic encephalopathy, major bleeding, or cardiac arrest. In addition, no significant differences emerged in quality of life or global functional status. The number of patients requiring hospitalization during follow-up was significantly lower in the rate-control group vs the rhythm-control group (73% vs 80%; P<.001). Anticoagulation was encouraged but not mandated in the rhythm-control group after 4 weeks in sinus rhythm, and there was a trend toward higher mortality in the rhythm-control group (27% vs 26%; P=.08).

Patients <65 years were excluded from the AFFIRM trial. When younger patients experience significant symptoms, early referral to Cardiology should be considered to discuss the long-term benefits and risks of a rhythm-control strategy. Regardless of age, when patients remain symptomatic despite rate- or rhythm-control management, the strategy should be changed.⁵

Rate-control targets and options

Target heart rates should be individualized. The 2014 ACC/AHA/HRS guideline recommends a resting target heart rate <80 beats per minute (bpm) in symptomatic patients.⁵ In patients with permanent AF who remain asymptomatic at higher resting heart rates, a more lenient rate-control strategy (resting heart rate <110 bpm) has demonstrated outcomes equivalent to those of a more strict approach (resting heart rate <80 bpm and heart rate during moderate exercise <110 bpm).⁴² Pharmacologic rate-control options include beta-blockers, non-dihydropyridine calcium channel antagonists, and digoxin (TABLE 3⁵). Digoxin is associated with increased all-cause mortality in patients with AF regardless of HF status (HR=1.4; 95% CI, 1.2-1.6, P=.0001).⁴³ Digoxin should be reserved for patients who are sedentary or have inadequate control with first-line medications.⁵

Indications for rhythm control

The NICE guidelines, which are consistent with the ACC/AHA/HRS guidelines, recommend rate control as the first-line strategy for AF management, except in people:²¹

• whose AF has a reversible cause
• who have HF believed to be primarily caused by AF
• with new-onset AF
• with atrial flutter that is considered suitable for an ablation strategy to restore sinus rhythm
• for whom a rhythm-control strategy would be more suitable based on clinical judgment.

In addition, patients who continue to experience symptomatic AF despite an adequate trial of rate control should be offered rhythm control.⁵

Pharmacologic rhythm-control strategies. Antiarrhythmic drugs can be used for chemical cardioversion, reduction of paroxysms, and long-term maintenance of sinus rhythm. The most commonly used antiarrhythmic drugs are Class IC and Class III agents (TABLE 3).⁵ Tailored drug selection for each patient is key. Patients with left atrial diameters >4.5 cm are less likely to remain in sinus rhythm, and patients with left ventricular hypertrophy are at increased risk for proarrhythmic adverse effects.⁴⁴ Patients with paroxysmal AF may be candidates for a “pill-in-the-pocket” strategy using propafenone or flecainide.⁵

AF frequently progresses from paroxysmal to persistent and can subsequently result in electrical and structural remodeling that becomes irreversible over time.⁴⁵ The patient with uncontrolled symptoms despite attempts at rate control and rhythm control should be promptly referred to an electrophysiologist.

Surgical interventions for rate or rhythm control

Electrophysiology interventions include AV nodal ablation with pacemaker placement for rate control, or catheter-directed ablation (radiofrequency or cryotherapy) for rhythm control. CA appears to be more effective than pharmacologic rhythm control.^46,47 Treatment with CA is indicated for symptomatic paroxysmal AF when a rhythm-control strategy is desired and the AF is refractory to, or the patient is intolerant of, at least one class I or III antiarrhythmic medication.⁵ With these same caveats, CA is a reasonable strategy for symptomatic persistent AF.

Consider more invasive interventions, such as an atrial maze procedure, when patients require cardiac surgery for another indication. Patients with an increased risk of thromboembolism (based on CHA₂DS₂-VASc) remain at high risk even after successful ablation.⁴⁸ As a result, some guidelines recommend continued long-term anticoagulation following CA.^18,22

CORRESPONDENCE
Philip Dooley, MD, University of Kansas School of Medicine–Wichita Family Medicine Residency at Via Christi, 707 North Emporia, Wichita, KS 67207; [email protected].

ACKNOWLEDGMENTS
We thank Professor Anne Walling, MB, ChB, FFPHM, Department of Family and Community Medicine, University of Kansas School of Medicine–Wichita for her suggestions and critical review of an earlier version of this manuscript.

Atrial fibrillation (AF)—the most common supraventricular tachycardia—affects as many as 6.1 million adults in the United States.¹ It is associated with a 5-fold increased risk of stroke,² a 3-fold increased risk of heart failure (HF),³ and about a 2-fold increased risk of dementia⁴ and mortality.² The prevalence of AF increases with maturity, from 2% in people <65 years of age to 9% in those ≥65 years,⁵ and that prevalence is expected to double over the next 25 years as the population ages.¹

The primary goals of treatment are to alleviate symptoms and prevent thromboembolism. Strokes related to AF are more likely to result in severe disability or death when compared with those unrelated to AF.⁶ And yet anticoagulation remains underutilized.⁷

The net clinical benefit of oral anticoagulation appears to be greatest in patients with the highest risk of bleeding, since these patients are also at the highest risk for stroke.⁸ Patients at increased risk of stroke are more likely to receive oral anticoagulation; however, for unknown reasons, more than half of people with the highest risk of stroke are not prescribed these important anti-blood-clotting medications.⁷ One theory is that physicians may be relying on their gut rather than objective risk scores, and underuse of validated schemata leads to poor estimation of risk.

IMAGE: © ALICIA BUELO

For example, results from the ORBIT-AF (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation) trial, which involved over 10,000 people with AF, found that although 72% (n=7251) had high-risk CHADS₂ scores (≥2), only 16% were assessed as having a high risk of stroke by physicians.⁹ Along the same lines, a recent study of Canadian primary care physicians showed that stroke risk and bleeding risk were not evaluated with validated tools in 58% and 81% of patients, respectively, leading to both significant underestimation and overestimation of risk.¹⁰

This review provides the tools to identify when anticoagulation is indicated, reports the advantages and disadvantages of the currently available anticoagulants, and discusses the selection and implementation of rate- vs rhythm-control strategies. But first, a word about the etiology, classification, and diagnosis of AF.

AF: The result of any number of cardiac and non-cardiac causes

AF is characterized by uncoordinated activation of the atria, which results in ineffective atrial contractions and an irregular, often rapid, ventricular response. It is the ultimate clinical manifestation of multiple diseases that alter atrial tissue through inflammation, fibrosis, or hypertrophy.⁵ The most common causes are hypertension, coronary artery disease, HF, cardiomyopathies, and valvular heart disease, all of which stimulate the renin-angiotensin-aldosterone system, leading to increased susceptibility to arrhythmia.⁵ Atrial ectopic tachycardia, Wolff-Parkinson-White (WPW) syndrome, and atrioventricular (AV) nodal reentrant tachycardia also may precipitate AF.⁵ In these cases, AF usually resolves after catheter ablation (CA) of the primary arrhythmia.¹¹ Unrecognized AF may trigger atrial flutter, and more than 80% of patients who undergo radiofrequency ablation for atrial flutter experience AF at some point in the subsequent 5 years.¹²

Strokes related to atrial fibrillation are more likely to result in severe disability or death when compared with those unrelated to AF. And yet anticoagulation remains underutilized.

Non-cardiac causes of AF include sleep apnea, obesity, hyperthyroidism, drugs, electrocution, pneumonia, and pulmonary embolism.⁵ An association between binge drinking and AF (“holiday heart syndrome”) has long been recognized. The evidence now suggests that alcohol increases the risk of AF in a dose-dependent manner with intakes of ≥1 drink per day (12 g per drink).¹³

Classification schema no longer includes “lone AF”

AF is classified in terms of the duration of episodes:⁵

• Paroxysmal AF is characterized by brief episodes that terminate spontaneously or with intervention within 7 days of onset. These episodes recur with variable frequency.
• Persistent AF refers to AF that is continuously sustained for more than 7 days.
• Longstanding persistent AF refers to continuous AF that lasts longer than 12 months.
• Permanent AF is not an inherent pathophysiologic attribute of AF, but rather an acceptance of AF where the patient and physician abandon further efforts to restore and/or maintain sinus rhythm.
• Nonvalvular AF occurs in the absence of a valve replacement (mechanical or bioprosthetic), rheumatic mitral stenosis, or mitral valve repair.

Although paroxysmal and persistent AF may occur in the same individual, the distinction is still clinically relevant, as outcomes of certain therapies, such as CA, are superior in patients with paroxysmal AF.¹⁴ With a more complete understanding of AF pathophysiology, guidelines now discourage use of the potentially confusing term “lone AF,” which has historically been applied to younger patients with no known clinical risk factors or echocardiographic abnormalities. As a result, therapeutic decisions are no longer based on this nomenclature, according to the 2014 AF practice guideline from the American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS).⁵

Patient complaints—or incidental findings—can prompt a Dx

Fatigue is the most common symptom of AF. Other signs and symptoms include palpitations, dyspnea, HF, hypotension, syncope, chest pain, and stroke. Some patients are asymptomatic, and AF is an incidental finding when an irregular pulse is discovered during a physical examination. The diagnosis is confirmed by electrocardiogram (EKG), telemetry, Holter monitor, event recorder, or an implanted electrocardiographic recording device. A chest x-ray, serum electrolyte levels, a complete blood count, thyroid testing, and renal and hepatic function testing are recommended. Transthoracic echocardiography to measure cardiac function, detect underlying structural heart disease, and evaluate atrial size is essential.⁵

Warfarin remains the only recommended anticoagulation strategy for patients with severe renal impairment or valvular atrial fibrillation.

An electrophysiologic (EP) study may be needed for diagnosis or treatment if another arrhythmia is present. Aberrant conduction may cause AF to present as a wide complex tachycardia and be mislabeled as ventricular tachycardia. The presence of delta waves is an indication for an EP study targeting the WPW accessory pathway. Transesophageal echocardiography (TEE) is the most sensitive and specific test for left atrial thrombi. If you are considering a TEE for a patient with AF of unknown, or >48 hours’, duration who has not been anticoagulated in the preceding 3 weeks, obtain it before performing cardioversion because of the risk of embolism.⁵

Stroke prevention

The ACC/AHA/HRS AF guideline recommends basing anticoagulation decisions on thromboembolic risk, regardless of AF pattern (paroxysmal, persistent, or permanent) (Class I recommendation).⁵ For patients with nonvalvular AF and atrial flutter, the guideline recommends using the Birmingham 2009 schema (CHA₂DS₂-VASc score) (TABLE 1^15-18) to estimate thromboembolic risk.^5,15 CHA₂DS₂-VASc improves on the older CHADS₂ score by significantly reducing the number of patients categorized as having intermediate risk and better identifying truly low-risk patients who are unlikely to benefit from anticoagulation.^16,17,19

Men with a CHA₂DS₂-VASc score of zero and women with a score of one do not need anticoagulation.^5,20 Discuss the risks and benefits of oral anticoagulation with men who have a score of one. In these intermediate-risk men, antiplatelet therapy with aspirin and/or clopidogrel may be reasonable, especially if there is an indication other than stroke prevention (eg, post-myocardial infarction). Oral anticoagulation is strongly recommended for all patients with a CHA₂DS₂-VASc score of 2 or higher.^5,18,21,22

Anticoagulant considerations: Warfarin vs DOACs

Warfarin was the gold standard for stroke prevention in nonvalvular AF until the direct oral anticoagulants (DOACs) became available in 2010. Guidelines in the United States and the United Kingdom recommend shared decision-making to help patients with AF who do not have a specific indication for warfarin choose between warfarin and the DOACs.^5,21 Canadian and European guidelines recommend DOACs as the first-line option for anticoagulation and reserve warfarin for patients who have contraindications to, or are unable to afford, DOACs.^18,22 All current guidelines recommend continuing warfarin in patients who are stable, well controlled, and satisfied with warfarin therapy and the monitoring and dietary restrictions it entails.

DOACs are as effective as warfarin. All of the DOACs are approved for stroke prevention based on individual phase III non-inferiority trials in which they were compared to warfarin.^23-26 In addition, a meta-analysis of these 4 trials involving a total of 71,683 patients (mean age 70-73 years; median follow-up, 1.8-2.8 years) evaluated the benefits and risks of the 4 DOACs against the former gold standard.²⁷

Higher doses of the DOACs (dabigatran 150 mg BID, rivaroxaban 20 mg/d, edoxaban 60 mg/d, and apixaban 5 mg BID) reduced the rates of stroke or systemic embolism (relative risk [RR]=0.81; 95% confidence interval [CI], 0.73-0.91; P<.0001; number needed to treat [NNT]=147), hemorrhagic stroke (RR=0.49; 95% CI, 0.38-0.64; P<.0001; NNT=219), and all-cause mortality (RR=0.90; 95% CI, 0.85-0.95; P=.0003; NNT=128), compared with warfarin.²⁷ It is important to note that while lower doses of some DOACs (dabigatran 110 mg BID and edoxaban 30 mg/d) were not as effective at preventing ischemic stroke when compared with warfarin (RR=1.3; 95% CI, 1-1.6; P=.045), they still significantly reduced hemorrhagic stroke (RR=0.33; 95% CI, 0.23-0.46; P<.0001) and all-cause mortality (RR=0.89; 95% CI, 0.83-0.96; P=.003).

Of course, the biggest concern is bleeding. In that same meta-analysis, the difference in major bleeding events with DOACs vs warfarin was not statistically significant (RR=0.86; 95% CI, 0.73-1; P=.06). While DOACs likely lower rates of intracranial hemorrhage (RR=0.48; 95% CI, 0.39-0.59; P<.0001; NNT=132), they seem to increase the risk of gastrointestinal (GI) bleeding (RR=1.3; 95% CI, 1-1.6; P=.043; number needed to harm [NNH]=185).²⁷

Without head-to-head trials, it is impossible to know if one direct oral anticoagulant is superior to another.

There was significant heterogeneity in the GI bleeding outcome, however. When compared with warfarin, GI bleeding was increased by dabigatran 150 mg BID (RR=1.5; 95% CI, 1.2-1.9; P<.001) and edoxaban 60 mg/d (HR=1.2; 95% CI, 1.02-1.5; P=.03), but there were no significant differences for dabigatran 110 mg BID or apixaban 5 mg BID.^23,25,26

On the other hand, edoxaban 30 mg/d had a lower risk of GI bleeding when compared with warfarin (HR=0.67; 95% CI, 0.53-0.83; P<.001).²⁵ Without head-to-head trials, it is impossible to know if one DOAC is superior to another. Apixaban 5 mg BID appears to offer the best overall balance between efficacy and safety. Other DOACs may be better options for patients who have specific concerns regarding efficacy or safety.^28,29

Convenience, interactions, and cost may be the deciding factors. Since all DOACs are fairly comparable in efficacy and safety, other factors such as convenience, interactions with other medications, and cost should be considered when deciding on a medication for an individual patient (TABLE 2^30,31). The DOACs require no lab monitoring or dose titration, and all 4 have fewer potential drug interactions than warfarin.³⁰ Due to their relatively short half-lives, strict adherence is critical; DOACs are not suitable for patients who frequently miss doses.⁵ (For more information on starting or switching to DOACs, see, “Is a novel anticoagulant right for your patient?” J Fam Pract. 2014;63:22-28.)

A word about DOACs and renal impairment. Another concern with DOACs is their reliance on renal metabolism and excretion. A meta-analysis of the 4 phase III trials of the DOACs, this time involving 58,338 patients, evaluated DOAC efficacy and safety compared to warfarin in the presence of kidney dysfunction.³² Renal function was categorized as normal (estimated glomerular filtration rate [eGFR] >80 mL/min/1.73 m²), mildly impaired (eGFR 50-80 mL/min/1.73 m²), or moderately impaired (eGFR <50 mL/min/1.73m²). Compared with warfarin, DOACs lowered stroke risk in patients with mild (RR=0.71; 95% CI, 0.62-0.81) or moderate (RR=0.79; 95% CI, 0.66-0.94) renal impairment. DOACs also reduced major bleeding compared to warfarin in patients with mild (RR=0.88; 95% CI, 0.80-0.97) or moderate (RR=0.80; 95% CI, 0.66-0.94) renal impairment. How the DOACs fare in patients with severe renal dysfunction could not be determined because such patients were excluded from the trials.

Keep in mind that the DOACs require dose adjustment at different levels of renal impairment (TABLE 2^30,31), and warfarin remains the only recommended treatment for patients with severe renal impairment, according to both AHA/ACC/HRS and European Society of Cardiology guidelines.^5,18

Tools to help assess patients’ bleeding risk

Of the available scoring mechanisms to identify risk factors for bleeding, 3 have been specifically validated in AF populations (ie, ATRIA,³³ HEMORR₂HAGES,³⁴ and HAS-BLED³⁵). Of the 3, HAS-BLED is superior,³⁶ the most practical, and recommended by expert guidelines.^18,21,22 Additionally, HAS-BLED has good correlation with intracranial hemorrhage risk. The HAS-BLED score ranges from 0 to 9 points with one point assigned for each of the following:³⁵

• Hypertension–uncontrolled with systolic BP >160 mm Hg
• Abnormal liver function–cirrhosis, bilirubin >2× normal, or liver enzymes >3× normal
• Abnormal renal function–dialysis, transplant, or serum creatinine >2.26 mg/dL
• Stroke history–including lacunar infarcts
• Bleeding predisposition–history of major bleeding due to any cause
• Labile international normalized ratio (INR)–time in therapeutic range <60%
• Elderly–age >65 years
• Drug–antiplatelet agents, including nonsteroidal anti-inflammatory drugs
• Alcohol usage–>8 drinks per week.

Patients with a HAS-BLED score ≥3 warrant additional monitoring and attempts to reduce bleeding risk by addressing modifiable risk factors. Bleeding risk scores should not be used to exclude patients from anticoagulation therapy.⁵ In fact, the British National Institute for Health and Clinical Excellence (NICE) guidelines state that anticoagulation should not be withheld solely due to fall risk.²¹

Also, anticoagulation with warfarin should not be permanently discontinued because of a single GI bleed, since restarting warfarin is associated with decreased risks of thromboembolism and mortality and a statistically insignificant increase in recurrent GI bleeding.³⁷ Restarting DOAC therapy following a GI bleed has not been evaluated in clinical trials; however, it may be reasonable to use one of the DOAC doses with a lower risk of GI bleeding (dabigatran 110 mg BID, apixaban 5 mg BID, or edoxaban 30 mg/d) in patients who have experienced a GI bleed on warfarin or another DOAC.^18,22

An online calculator is available that uses CHA₂DS₂-VASc and HAS-BLED scores to determine an individual’s risk/benefit profile with the various anticoagulation strategies available (http://www.sparctool.com). Consider percutaneous left atrial appendage occlusion if the risks of anticoagulation truly exceed the benefits.³⁸

Rate control vs rhythm control

Most patients who present with AF require immediate ventricular rate control to reduce symptoms. In the acute setting, this can be accomplished with intravenous (IV) beta-blockers or IV calcium channel antagonists.^5,39 If the patient is hemodynamically unstable, urgent direct-current cardioversion is the preferred treatment strategy and should not be delayed pending anticoagulation. IV amiodarone can be used in the ICU patient who does not require cardioversion, but is unable to tolerate beta-blockers or calcium channel antagonists.⁴⁰ Once the patient is stable, long-term treatment focuses on ventricular rate control or restoration and maintenance of sinus rhythm.

Direct oral anticoagulants are not suitable for patients who frequently miss doses.

The AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) trial enrolled 4060 patients (mean age 70 years, mean follow-up 3.5 years) with paroxysmal and persistent AF and randomized them to either pharmacologic rate control or rhythm control.⁴¹ No significant differences were found in all-cause mortality or in the composite secondary endpoint of death, ischemic stroke, anoxic encephalopathy, major bleeding, or cardiac arrest. In addition, no significant differences emerged in quality of life or global functional status. The number of patients requiring hospitalization during follow-up was significantly lower in the rate-control group vs the rhythm-control group (73% vs 80%; P<.001). Anticoagulation was encouraged but not mandated in the rhythm-control group after 4 weeks in sinus rhythm, and there was a trend toward higher mortality in the rhythm-control group (27% vs 26%; P=.08).

Patients <65 years were excluded from the AFFIRM trial. When younger patients experience significant symptoms, early referral to Cardiology should be considered to discuss the long-term benefits and risks of a rhythm-control strategy. Regardless of age, when patients remain symptomatic despite rate- or rhythm-control management, the strategy should be changed.⁵

Rate-control targets and options

Target heart rates should be individualized. The 2014 ACC/AHA/HRS guideline recommends a resting target heart rate <80 beats per minute (bpm) in symptomatic patients.⁵ In patients with permanent AF who remain asymptomatic at higher resting heart rates, a more lenient rate-control strategy (resting heart rate <110 bpm) has demonstrated outcomes equivalent to those of a more strict approach (resting heart rate <80 bpm and heart rate during moderate exercise <110 bpm).⁴² Pharmacologic rate-control options include beta-blockers, non-dihydropyridine calcium channel antagonists, and digoxin (TABLE 3⁵). Digoxin is associated with increased all-cause mortality in patients with AF regardless of HF status (HR=1.4; 95% CI, 1.2-1.6, P=.0001).⁴³ Digoxin should be reserved for patients who are sedentary or have inadequate control with first-line medications.⁵

Indications for rhythm control

The NICE guidelines, which are consistent with the ACC/AHA/HRS guidelines, recommend rate control as the first-line strategy for AF management, except in people:²¹

• whose AF has a reversible cause
• who have HF believed to be primarily caused by AF
• with new-onset AF
• with atrial flutter that is considered suitable for an ablation strategy to restore sinus rhythm
• for whom a rhythm-control strategy would be more suitable based on clinical judgment.

In addition, patients who continue to experience symptomatic AF despite an adequate trial of rate control should be offered rhythm control.⁵

Pharmacologic rhythm-control strategies. Antiarrhythmic drugs can be used for chemical cardioversion, reduction of paroxysms, and long-term maintenance of sinus rhythm. The most commonly used antiarrhythmic drugs are Class IC and Class III agents (TABLE 3).⁵ Tailored drug selection for each patient is key. Patients with left atrial diameters >4.5 cm are less likely to remain in sinus rhythm, and patients with left ventricular hypertrophy are at increased risk for proarrhythmic adverse effects.⁴⁴ Patients with paroxysmal AF may be candidates for a “pill-in-the-pocket” strategy using propafenone or flecainide.⁵

AF frequently progresses from paroxysmal to persistent and can subsequently result in electrical and structural remodeling that becomes irreversible over time.⁴⁵ The patient with uncontrolled symptoms despite attempts at rate control and rhythm control should be promptly referred to an electrophysiologist.

Surgical interventions for rate or rhythm control

Electrophysiology interventions include AV nodal ablation with pacemaker placement for rate control, or catheter-directed ablation (radiofrequency or cryotherapy) for rhythm control. CA appears to be more effective than pharmacologic rhythm control.^46,47 Treatment with CA is indicated for symptomatic paroxysmal AF when a rhythm-control strategy is desired and the AF is refractory to, or the patient is intolerant of, at least one class I or III antiarrhythmic medication.⁵ With these same caveats, CA is a reasonable strategy for symptomatic persistent AF.

Consider more invasive interventions, such as an atrial maze procedure, when patients require cardiac surgery for another indication. Patients with an increased risk of thromboembolism (based on CHA₂DS₂-VASc) remain at high risk even after successful ablation.⁴⁸ As a result, some guidelines recommend continued long-term anticoagulation following CA.^18,22

CORRESPONDENCE
Philip Dooley, MD, University of Kansas School of Medicine–Wichita Family Medicine Residency at Via Christi, 707 North Emporia, Wichita, KS 67207; [email protected].

ACKNOWLEDGMENTS
We thank Professor Anne Walling, MB, ChB, FFPHM, Department of Family and Community Medicine, University of Kansas School of Medicine–Wichita for her suggestions and critical review of an earlier version of this manuscript.

The FP suspected that contact dermatitis was to blame for the rash on the dorsum of the patient’s feet. This would have been an unlikely location for tinea pedis and the patient had already tried topical antifungal medications without any benefit. The FP asked the patient if he had purchased any new shoes or boots before the rash started, and the patient indicated that he’d purchased new running shoes about a year earlier.

The FP prescribed 0.1% triamcinolone cream to be applied twice daily. The FP recognized that patch testing might be indicated, but did not perform this in her office. She also recommended that the patient stay away from the running shoes at this time. One month later, the rash was 95% better, except for some postinflammatory hyperpigmentation that was likely to take months to fade. The patient was happy with the results, but wanted to know the cause of his allergy and what shoes would be safe to wear.

The FP offered the patient a referral to a local dermatologist who performed patch testing. The patient went for patch testing and learned he was allergic to chromates found in many types of leather. Statistically, the most likely offending allergens for a rash in this location would be formaldehyde or a chromate. Both are found in leather and are common allergens that cause contact dermatitis on the feet.

Allergic contact dermatitis to chromate in leather shoes may be seasonal as a result of the allergen being leached out by perspiration in warmer months. There are leather shoes that are made free of chromates for people who suffer from allergic contact dermatitis due to this allergen.

Photos and text for Photo Rounds Friday courtesy of Richard P. Usatine, MD. This case was adapted from: Usatine R, Reppa R. Tinea pedis. In: Usatine R, Smith M, Mayeaux EJ, et al, eds. Color Atlas of Family Medicine. 2nd ed. New York, NY: McGraw-Hill; 2013:799-804.

To learn more about the Color Atlas of Family Medicine, see: www.amazon.com/Color-Family-Medicine-Richard-Usatine/dp/0071769641/

You can now get the second edition of the Color Atlas of Family Medicine as an app by clicking on this link: usatinemedia.com

The FP suspected that contact dermatitis was to blame for the rash on the dorsum of the patient’s feet. This would have been an unlikely location for tinea pedis and the patient had already tried topical antifungal medications without any benefit. The FP asked the patient if he had purchased any new shoes or boots before the rash started, and the patient indicated that he’d purchased new running shoes about a year earlier.

The FP prescribed 0.1% triamcinolone cream to be applied twice daily. The FP recognized that patch testing might be indicated, but did not perform this in her office. She also recommended that the patient stay away from the running shoes at this time. One month later, the rash was 95% better, except for some postinflammatory hyperpigmentation that was likely to take months to fade. The patient was happy with the results, but wanted to know the cause of his allergy and what shoes would be safe to wear.

The FP offered the patient a referral to a local dermatologist who performed patch testing. The patient went for patch testing and learned he was allergic to chromates found in many types of leather. Statistically, the most likely offending allergens for a rash in this location would be formaldehyde or a chromate. Both are found in leather and are common allergens that cause contact dermatitis on the feet.

Allergic contact dermatitis to chromate in leather shoes may be seasonal as a result of the allergen being leached out by perspiration in warmer months. There are leather shoes that are made free of chromates for people who suffer from allergic contact dermatitis due to this allergen.

Photos and text for Photo Rounds Friday courtesy of Richard P. Usatine, MD. This case was adapted from: Usatine R, Reppa R. Tinea pedis. In: Usatine R, Smith M, Mayeaux EJ, et al, eds. Color Atlas of Family Medicine. 2nd ed. New York, NY: McGraw-Hill; 2013:799-804.

To learn more about the Color Atlas of Family Medicine, see: www.amazon.com/Color-Family-Medicine-Richard-Usatine/dp/0071769641/

You can now get the second edition of the Color Atlas of Family Medicine as an app by clicking on this link: usatinemedia.com

The FP suspected that contact dermatitis was to blame for the rash on the dorsum of the patient’s feet. This would have been an unlikely location for tinea pedis and the patient had already tried topical antifungal medications without any benefit. The FP asked the patient if he had purchased any new shoes or boots before the rash started, and the patient indicated that he’d purchased new running shoes about a year earlier.

The FP prescribed 0.1% triamcinolone cream to be applied twice daily. The FP recognized that patch testing might be indicated, but did not perform this in her office. She also recommended that the patient stay away from the running shoes at this time. One month later, the rash was 95% better, except for some postinflammatory hyperpigmentation that was likely to take months to fade. The patient was happy with the results, but wanted to know the cause of his allergy and what shoes would be safe to wear.

The FP offered the patient a referral to a local dermatologist who performed patch testing. The patient went for patch testing and learned he was allergic to chromates found in many types of leather. Statistically, the most likely offending allergens for a rash in this location would be formaldehyde or a chromate. Both are found in leather and are common allergens that cause contact dermatitis on the feet.

Allergic contact dermatitis to chromate in leather shoes may be seasonal as a result of the allergen being leached out by perspiration in warmer months. There are leather shoes that are made free of chromates for people who suffer from allergic contact dermatitis due to this allergen.

Photos and text for Photo Rounds Friday courtesy of Richard P. Usatine, MD. This case was adapted from: Usatine R, Reppa R. Tinea pedis. In: Usatine R, Smith M, Mayeaux EJ, et al, eds. Color Atlas of Family Medicine. 2nd ed. New York, NY: McGraw-Hill; 2013:799-804.

To learn more about the Color Atlas of Family Medicine, see: www.amazon.com/Color-Family-Medicine-Richard-Usatine/dp/0071769641/

You can now get the second edition of the Color Atlas of Family Medicine as an app by clicking on this link: usatinemedia.com

This patient was suffering from pitted keratolysis and interdigital tinea pedis. Pitted keratolysis is caused by the bacterium Kytococcus sedentarius. Like tinea pedis, moist and sweaty feet provide a great environment for growth of this organism. The patient admitted to having sweaty feet—especially while playing soccer for hours. He also didn’t use shower shoes in the gym shower.

In pitted keratolysis, the bacteria live on the dead cells of the stratum corneum of the sole and form visible pits. In interdigital tinea pedis, the skin between the toes is white in appearance.

The FP recommended that the patient wear shower shoes and change his socks during the day if they become sweaty. He also prescribed topical 2% erythromycin solution to apply twice daily to the area with visible pits. In addition, the FP recommended that the patient buy over-the-counter topical terbinafine cream and apply it between the toes once or twice daily—especially after drying his feet well after a shower. A month later, both infections were clear.

Photos and text for Photo Rounds Friday courtesy of Richard P. Usatine, MD. This case was adapted from: Usatine R, Reppa R. Tinea pedis. In: Usatine R, Smith M, Mayeaux EJ, et al, eds. Color Atlas of Family Medicine. 2nd ed. New York, NY: McGraw-Hill; 2013:799-804.

To learn more about the Color Atlas of Family Medicine, see: www.amazon.com/Color-Family-Medicine-Richard-Usatine/dp/0071769641/

You can now get the second edition of the Color Atlas of Family Medicine as an app by clicking on this link: usatinemedia.com