A 46-year-old woman underwent laparoscopic supracervical hysterectomy to remove her uterus but preserve her cervix. Postsurgically, she had difficulty breathing deeply and reported abdominal pain. The nurses and on-call physician reassured her that she was experiencing “gas pains” due to insufflation. After same-day discharge, she stayed in a motel room to avoid a second-floor bedroom at home.

She called the gynecologist’s office the following day to report continued pain and severe hot flashes and sweats. The gynecologist instructed his nurse to advise the patient to stop taking her birth control pill (ethinyl estradiol/norethindrone, Microgestin) and “to ride out” the hot flashes.

The woman was found dead in her motel room the next morning. An autopsy revealed a perforated small intestine with leakage into the abdominal cavity causing sepsis, multi-organ failure, and death.

ESTATE’S CLAIM The gynecologist reviewed the medical records and found an error in the operative report, but he made no addendum or late entry to correct the operative report. His defense counsel instructed him to draft a letter clarifying the surgery; this clarification was given to defense experts. The description of the procedure in the clarification was different from what was described in the medical records. For example, the clarification reported making 4 incisions for 4 trocars; the operative report indicated using 3 trocars. The pathologist and 2 nurses who treated the patient after surgery confirmed that there were 3 trocar incisions. The pathologist found no tissue necrosis at or around the perforation site, indicating that the perforation likely occurred during surgery.

PHYSICIAN’S DEFENSE Bowel perforation is a known complication of the procedure. The perforation was not present at the time of surgery because leakage of bowel content would have been obvious.

VERDICT A $1.5 million Virginia settlement was reached.

Retained products of conception after D&C
When sonography indicated that a 30-year-old woman was pregnant, she decided to abort the pregnancy and was given mifepristone.

Another sonogram 5 weeks later showed retained products of conception within the uterus. An ObGyn performed dilation and curettage (D&C) at an outpatient clinic. Because he believed the cannula did not remove everything, he used a curette to scrape the uterus. After the patient was dizzy, hypotensive, and in pain for 4 hours, an ambulance transported her to a hospital. Perforations of the uterus and sigmoid colon were discovered and repaired during emergency surgery. The patient has a large scar on her abdomen.

PATIENT'S CLAIM The ObGyn did not perform the D&C properly and perforated the uterus and colon. An earlier response to symptoms could have prevented repair surgery. Damage to the uterus may now preclude her from having a successful pregnancy.

DEFENDANTS’ DEFENSE The ObGyn argued that the aborted pregnancy was ectopic; spontaneous rupture caused the perforations.

VERDICT A $340,000 New York settlement was reached with the ObGyn. By the time of trial, the clinic had closed. 

Wrong-site biopsy; records altered

A 40-year-old woman underwent excisional breast biopsy. The wrong lump was removed and the woman had to have another procedure.

PATIENT'S CLAIM The hospital’s nursing staff failed to properly mark the operative site. The breast surgeon did not confirm that the markings were correct. The surgeon altered the written operative report after the surgery to conceal negligence.

DEFENDANTS’ DEFENSE The nurses properly marked the biopsy site, but the surgeon chose another route. The surgeon edited the original report to reflect events that occurred during surgery that had not been included in the original dictation. The added material gave justification for performing the procedure at a different site than originally intended.  

VERDICT A $15,500 Connecticut verdict was returned.    

Second twin has CP and brain damage: $10M settlement
A woman gave birth to twins at an Army hospital. The first twin was delivered without complications. The second twin developed a prolapsed cord during delivery of the first twin. A resident and the attending physician allowed the mother to continue with vaginal delivery. The heart-rate monitor showed fetal distress, but the medical staff did not respond. After an hour, another physician was consulted, and he ordered immediate delivery. The attending physician decided to continue with vaginal delivery using forceps, but it took 15 minutes to locate forceps in the hospital. The infant suffered severe brain damage and cerebral palsy. She will require 24-hour nursing care for life, including treatment of a tracheostomy.

PARENTS' CLAIM The physicians were negligent for not reacting to non-reassuring monitor strips and for allowing the vaginal delivery to continue. An emergency cesarean delivery should have been performed.

DEFENDANTS’ DEFENSE The case was settled before trial. 

VERDICT A $10 million North Carolina settlement was reached for past medical bills and future care.

Faulty biopsies: breast cancer diagnosis missed
In September 2006, a 40-year-old woman underwent breast sonography. A radiologist, Dr. A, reported finding a mass and a smaller nodule in the right breast, and recommended a biopsy of each area. Two weeks later, a second radiologist, Dr. B, biopsied the larger of the two areas and diagnosed a hyalinized fibroadenoma. He did not biopsy the smaller growth, but reported it as a benign nodule. He recommended more frequent screenings. The patient was referred to a surgeon, who determined that she should be seen in 6 months.

In June 2007, the patient underwent right-breast sonography that revealed cysts and three nodules. The surgeon recommended a biopsy, but the biopsy was performed on only two of three nodules. A third radiologist, Dr. C, determined that the nodules were all benign.

In November 2007, when the patient reported a painful lump in her right breast, her gynecologist ordered mammography, which revealed lesions. A biopsy revealed that one lesion was stage III invasive ductal carcinoma. The patient underwent extensive treatment, including a mastectomy, lymphadenectomy, chemotherapy, and radiation therapy, and prophylactic surgical reduction of the left breast.

PATIENT'S CLAIM The cancer should have been diagnosed in September 2006. Prompt treatment would have decreased the progression of the disease. The September 2006 biopsy should have included both lumps, as recommended by Dr. A.

DEFENDANTS’ DEFENSE There was no indication of cancer in September 2006. Reasonable follow-up care was given. 

VERDICT A New York defense verdict was returned.

Tumor not found during surgery; BSO performed
A 41-year-old woman underwent surgery to remove a pelvic tumor in November 2004. The gynecologist was unable to locate the tumor during surgery. He performed bilateral salpingo-oophorectomy (BSO) because of a visual diagnosis of endometriosis. In August 2005, the patient underwent surgical removal of the tumor by another surgeon. She was hospitalized for several weeks and suffered a large scar that required additional surgery.

PATIENT'S CLAIM BSO was unnecessary, and caused early menopause, with vaginal atrophy and dryness, depression, fatigue, insomnia, loss of hair, and other symptoms.

The patient claimed lack of informed consent. From Ecuador, the patient’s command of English was not sufficient for her to completely understand the consent form; an interpreter should have been provided.

DEFENDANTS’ DEFENSE BSO did not cause a significant acceleration of the onset of menopause. It was necessary to treat the endometriosis.

The patient signed a consent form that included BSO. The patient did not indicate that she did not understand the language on the form; had she asked, an interpreter would have been provided.

VERDICT A $750,000 New York settlement was reached with the gynecologist and medical center.

These cases were selected by the editors of OBG Management from Medical Malpractice Verdicts, Settlements & Experts, with permission of the editor, Lewis Laska (www.verdictslaska.com). The information available to the editors about the cases presented here is sometimes incomplete. Moreover, the cases may or may not have merit. Nevertheless, these cases represent the types of clinical situations that typically result in litigation and are meant to illustrate nationwide variation in jury verdicts and awards.

A 46-year-old woman underwent laparoscopic supracervical hysterectomy to remove her uterus but preserve her cervix. Postsurgically, she had difficulty breathing deeply and reported abdominal pain. The nurses and on-call physician reassured her that she was experiencing “gas pains” due to insufflation. After same-day discharge, she stayed in a motel room to avoid a second-floor bedroom at home.

She called the gynecologist’s office the following day to report continued pain and severe hot flashes and sweats. The gynecologist instructed his nurse to advise the patient to stop taking her birth control pill (ethinyl estradiol/norethindrone, Microgestin) and “to ride out” the hot flashes.

The woman was found dead in her motel room the next morning. An autopsy revealed a perforated small intestine with leakage into the abdominal cavity causing sepsis, multi-organ failure, and death.

ESTATE’S CLAIM The gynecologist reviewed the medical records and found an error in the operative report, but he made no addendum or late entry to correct the operative report. His defense counsel instructed him to draft a letter clarifying the surgery; this clarification was given to defense experts. The description of the procedure in the clarification was different from what was described in the medical records. For example, the clarification reported making 4 incisions for 4 trocars; the operative report indicated using 3 trocars. The pathologist and 2 nurses who treated the patient after surgery confirmed that there were 3 trocar incisions. The pathologist found no tissue necrosis at or around the perforation site, indicating that the perforation likely occurred during surgery.

PHYSICIAN’S DEFENSE Bowel perforation is a known complication of the procedure. The perforation was not present at the time of surgery because leakage of bowel content would have been obvious.

VERDICT A $1.5 million Virginia settlement was reached.

Retained products of conception after D&C
When sonography indicated that a 30-year-old woman was pregnant, she decided to abort the pregnancy and was given mifepristone.

Another sonogram 5 weeks later showed retained products of conception within the uterus. An ObGyn performed dilation and curettage (D&C) at an outpatient clinic. Because he believed the cannula did not remove everything, he used a curette to scrape the uterus. After the patient was dizzy, hypotensive, and in pain for 4 hours, an ambulance transported her to a hospital. Perforations of the uterus and sigmoid colon were discovered and repaired during emergency surgery. The patient has a large scar on her abdomen.

PATIENT'S CLAIM The ObGyn did not perform the D&C properly and perforated the uterus and colon. An earlier response to symptoms could have prevented repair surgery. Damage to the uterus may now preclude her from having a successful pregnancy.

DEFENDANTS’ DEFENSE The ObGyn argued that the aborted pregnancy was ectopic; spontaneous rupture caused the perforations.

VERDICT A $340,000 New York settlement was reached with the ObGyn. By the time of trial, the clinic had closed. 

Wrong-site biopsy; records altered

A 40-year-old woman underwent excisional breast biopsy. The wrong lump was removed and the woman had to have another procedure.

PATIENT'S CLAIM The hospital’s nursing staff failed to properly mark the operative site. The breast surgeon did not confirm that the markings were correct. The surgeon altered the written operative report after the surgery to conceal negligence.

DEFENDANTS’ DEFENSE The nurses properly marked the biopsy site, but the surgeon chose another route. The surgeon edited the original report to reflect events that occurred during surgery that had not been included in the original dictation. The added material gave justification for performing the procedure at a different site than originally intended.  

VERDICT A $15,500 Connecticut verdict was returned.    

Second twin has CP and brain damage: $10M settlement
A woman gave birth to twins at an Army hospital. The first twin was delivered without complications. The second twin developed a prolapsed cord during delivery of the first twin. A resident and the attending physician allowed the mother to continue with vaginal delivery. The heart-rate monitor showed fetal distress, but the medical staff did not respond. After an hour, another physician was consulted, and he ordered immediate delivery. The attending physician decided to continue with vaginal delivery using forceps, but it took 15 minutes to locate forceps in the hospital. The infant suffered severe brain damage and cerebral palsy. She will require 24-hour nursing care for life, including treatment of a tracheostomy.

PARENTS' CLAIM The physicians were negligent for not reacting to non-reassuring monitor strips and for allowing the vaginal delivery to continue. An emergency cesarean delivery should have been performed.

DEFENDANTS’ DEFENSE The case was settled before trial. 

VERDICT A $10 million North Carolina settlement was reached for past medical bills and future care.

Faulty biopsies: breast cancer diagnosis missed
In September 2006, a 40-year-old woman underwent breast sonography. A radiologist, Dr. A, reported finding a mass and a smaller nodule in the right breast, and recommended a biopsy of each area. Two weeks later, a second radiologist, Dr. B, biopsied the larger of the two areas and diagnosed a hyalinized fibroadenoma. He did not biopsy the smaller growth, but reported it as a benign nodule. He recommended more frequent screenings. The patient was referred to a surgeon, who determined that she should be seen in 6 months.

In June 2007, the patient underwent right-breast sonography that revealed cysts and three nodules. The surgeon recommended a biopsy, but the biopsy was performed on only two of three nodules. A third radiologist, Dr. C, determined that the nodules were all benign.

In November 2007, when the patient reported a painful lump in her right breast, her gynecologist ordered mammography, which revealed lesions. A biopsy revealed that one lesion was stage III invasive ductal carcinoma. The patient underwent extensive treatment, including a mastectomy, lymphadenectomy, chemotherapy, and radiation therapy, and prophylactic surgical reduction of the left breast.

PATIENT'S CLAIM The cancer should have been diagnosed in September 2006. Prompt treatment would have decreased the progression of the disease. The September 2006 biopsy should have included both lumps, as recommended by Dr. A.

DEFENDANTS’ DEFENSE There was no indication of cancer in September 2006. Reasonable follow-up care was given. 

VERDICT A New York defense verdict was returned.

Tumor not found during surgery; BSO performed
A 41-year-old woman underwent surgery to remove a pelvic tumor in November 2004. The gynecologist was unable to locate the tumor during surgery. He performed bilateral salpingo-oophorectomy (BSO) because of a visual diagnosis of endometriosis. In August 2005, the patient underwent surgical removal of the tumor by another surgeon. She was hospitalized for several weeks and suffered a large scar that required additional surgery.

PATIENT'S CLAIM BSO was unnecessary, and caused early menopause, with vaginal atrophy and dryness, depression, fatigue, insomnia, loss of hair, and other symptoms.

The patient claimed lack of informed consent. From Ecuador, the patient’s command of English was not sufficient for her to completely understand the consent form; an interpreter should have been provided.

DEFENDANTS’ DEFENSE BSO did not cause a significant acceleration of the onset of menopause. It was necessary to treat the endometriosis.

The patient signed a consent form that included BSO. The patient did not indicate that she did not understand the language on the form; had she asked, an interpreter would have been provided.

VERDICT A $750,000 New York settlement was reached with the gynecologist and medical center.

These cases were selected by the editors of OBG Management from Medical Malpractice Verdicts, Settlements & Experts, with permission of the editor, Lewis Laska (www.verdictslaska.com). The information available to the editors about the cases presented here is sometimes incomplete. Moreover, the cases may or may not have merit. Nevertheless, these cases represent the types of clinical situations that typically result in litigation and are meant to illustrate nationwide variation in jury verdicts and awards.

A 46-year-old woman underwent laparoscopic supracervical hysterectomy to remove her uterus but preserve her cervix. Postsurgically, she had difficulty breathing deeply and reported abdominal pain. The nurses and on-call physician reassured her that she was experiencing “gas pains” due to insufflation. After same-day discharge, she stayed in a motel room to avoid a second-floor bedroom at home.

She called the gynecologist’s office the following day to report continued pain and severe hot flashes and sweats. The gynecologist instructed his nurse to advise the patient to stop taking her birth control pill (ethinyl estradiol/norethindrone, Microgestin) and “to ride out” the hot flashes.

The woman was found dead in her motel room the next morning. An autopsy revealed a perforated small intestine with leakage into the abdominal cavity causing sepsis, multi-organ failure, and death.

ESTATE’S CLAIM The gynecologist reviewed the medical records and found an error in the operative report, but he made no addendum or late entry to correct the operative report. His defense counsel instructed him to draft a letter clarifying the surgery; this clarification was given to defense experts. The description of the procedure in the clarification was different from what was described in the medical records. For example, the clarification reported making 4 incisions for 4 trocars; the operative report indicated using 3 trocars. The pathologist and 2 nurses who treated the patient after surgery confirmed that there were 3 trocar incisions. The pathologist found no tissue necrosis at or around the perforation site, indicating that the perforation likely occurred during surgery.

PHYSICIAN’S DEFENSE Bowel perforation is a known complication of the procedure. The perforation was not present at the time of surgery because leakage of bowel content would have been obvious.

VERDICT A $1.5 million Virginia settlement was reached.

Retained products of conception after D&C
When sonography indicated that a 30-year-old woman was pregnant, she decided to abort the pregnancy and was given mifepristone.

Another sonogram 5 weeks later showed retained products of conception within the uterus. An ObGyn performed dilation and curettage (D&C) at an outpatient clinic. Because he believed the cannula did not remove everything, he used a curette to scrape the uterus. After the patient was dizzy, hypotensive, and in pain for 4 hours, an ambulance transported her to a hospital. Perforations of the uterus and sigmoid colon were discovered and repaired during emergency surgery. The patient has a large scar on her abdomen.

PATIENT'S CLAIM The ObGyn did not perform the D&C properly and perforated the uterus and colon. An earlier response to symptoms could have prevented repair surgery. Damage to the uterus may now preclude her from having a successful pregnancy.

DEFENDANTS’ DEFENSE The ObGyn argued that the aborted pregnancy was ectopic; spontaneous rupture caused the perforations.

VERDICT A $340,000 New York settlement was reached with the ObGyn. By the time of trial, the clinic had closed. 

Wrong-site biopsy; records altered

A 40-year-old woman underwent excisional breast biopsy. The wrong lump was removed and the woman had to have another procedure.

PATIENT'S CLAIM The hospital’s nursing staff failed to properly mark the operative site. The breast surgeon did not confirm that the markings were correct. The surgeon altered the written operative report after the surgery to conceal negligence.

DEFENDANTS’ DEFENSE The nurses properly marked the biopsy site, but the surgeon chose another route. The surgeon edited the original report to reflect events that occurred during surgery that had not been included in the original dictation. The added material gave justification for performing the procedure at a different site than originally intended.  

VERDICT A $15,500 Connecticut verdict was returned.    

Second twin has CP and brain damage: $10M settlement
A woman gave birth to twins at an Army hospital. The first twin was delivered without complications. The second twin developed a prolapsed cord during delivery of the first twin. A resident and the attending physician allowed the mother to continue with vaginal delivery. The heart-rate monitor showed fetal distress, but the medical staff did not respond. After an hour, another physician was consulted, and he ordered immediate delivery. The attending physician decided to continue with vaginal delivery using forceps, but it took 15 minutes to locate forceps in the hospital. The infant suffered severe brain damage and cerebral palsy. She will require 24-hour nursing care for life, including treatment of a tracheostomy.

PARENTS' CLAIM The physicians were negligent for not reacting to non-reassuring monitor strips and for allowing the vaginal delivery to continue. An emergency cesarean delivery should have been performed.

DEFENDANTS’ DEFENSE The case was settled before trial. 

VERDICT A $10 million North Carolina settlement was reached for past medical bills and future care.

Faulty biopsies: breast cancer diagnosis missed
In September 2006, a 40-year-old woman underwent breast sonography. A radiologist, Dr. A, reported finding a mass and a smaller nodule in the right breast, and recommended a biopsy of each area. Two weeks later, a second radiologist, Dr. B, biopsied the larger of the two areas and diagnosed a hyalinized fibroadenoma. He did not biopsy the smaller growth, but reported it as a benign nodule. He recommended more frequent screenings. The patient was referred to a surgeon, who determined that she should be seen in 6 months.

In June 2007, the patient underwent right-breast sonography that revealed cysts and three nodules. The surgeon recommended a biopsy, but the biopsy was performed on only two of three nodules. A third radiologist, Dr. C, determined that the nodules were all benign.

In November 2007, when the patient reported a painful lump in her right breast, her gynecologist ordered mammography, which revealed lesions. A biopsy revealed that one lesion was stage III invasive ductal carcinoma. The patient underwent extensive treatment, including a mastectomy, lymphadenectomy, chemotherapy, and radiation therapy, and prophylactic surgical reduction of the left breast.

PATIENT'S CLAIM The cancer should have been diagnosed in September 2006. Prompt treatment would have decreased the progression of the disease. The September 2006 biopsy should have included both lumps, as recommended by Dr. A.

DEFENDANTS’ DEFENSE There was no indication of cancer in September 2006. Reasonable follow-up care was given. 

VERDICT A New York defense verdict was returned.

Tumor not found during surgery; BSO performed
A 41-year-old woman underwent surgery to remove a pelvic tumor in November 2004. The gynecologist was unable to locate the tumor during surgery. He performed bilateral salpingo-oophorectomy (BSO) because of a visual diagnosis of endometriosis. In August 2005, the patient underwent surgical removal of the tumor by another surgeon. She was hospitalized for several weeks and suffered a large scar that required additional surgery.

PATIENT'S CLAIM BSO was unnecessary, and caused early menopause, with vaginal atrophy and dryness, depression, fatigue, insomnia, loss of hair, and other symptoms.

The patient claimed lack of informed consent. From Ecuador, the patient’s command of English was not sufficient for her to completely understand the consent form; an interpreter should have been provided.

DEFENDANTS’ DEFENSE BSO did not cause a significant acceleration of the onset of menopause. It was necessary to treat the endometriosis.

The patient signed a consent form that included BSO. The patient did not indicate that she did not understand the language on the form; had she asked, an interpreter would have been provided.

VERDICT A $750,000 New York settlement was reached with the gynecologist and medical center.

These cases were selected by the editors of OBG Management from Medical Malpractice Verdicts, Settlements & Experts, with permission of the editor, Lewis Laska (www.verdictslaska.com). The information available to the editors about the cases presented here is sometimes incomplete. Moreover, the cases may or may not have merit. Nevertheless, these cases represent the types of clinical situations that typically result in litigation and are meant to illustrate nationwide variation in jury verdicts and awards.

Urinary incontinence affects more than one-third of women aged 70 years or older. As the authors of this study point out, urinary incontinence is comparable to other chronic conditions, such as hypertension and diabetes mellitus, in its impact on quality of life, and is a common reason for institutionalization.

The risk of urinary incontinence increases with age. In elderly women, it is often mixed (ie, having both urge- and stress-related components) and associated with functional impairments, including reduced mobility. It is thought that the association between incontinence and functional impairment is related primarily to the urge component:

• Women with impaired mobility take longer to reach the toilet, increasing the risk of leakage when the urge to urinate is strong
• Women with urge incontinence may be more likely to limit their activities so that they are always near a toilet.

Cognitive impairment may also play a role, affecting motor skills and bladder control.

To better understand why advancing age is linked with urinary incontinence, Fritel and colleagues studied a population of 1,942 urban-dwelling French women aged 75 to 85 years (mean age, 79.3 years; mean body mass index, 25.9 kg/m²).

Details of the study
Investigators assessed the frequency and quantity of urine leaks, the impact of urinary incontinence on daily life, and the participants’ mobility and balance. Data on urinary incontinence were collected via a self-administered questionnaire (the International Consultation on Incontinence Questionnaire–Short Form). Motor-related physical function was assessed using standardized balance and gait tests.

Urinary incontinence was reported by 42% of participants. Of these women, 57% reported daily urine leakage, with mixed incontinence found to be more prevalent than urge incontinence, which was more prevalent than stress incontinence. Overall, women with urinary incontinence reported that its impact on daily life was mild. Among those with mixed or urge incontinence, limitations in mobility and balance were correlated significantly with the severity of incontinence.

What this evidence means for practice
These findings support earlier data suggesting that impaired mobility can promote urge and mixed urinary incontinence. Because there also is a possibility that cerebral deterioration in aging women causes gait and balance problems that increase the likelihood of urge incontinence, the authors advise against the use of anticholinergic agents in elderly women, as these drugs can impair cognitive function. Another take-home message from this French report is that we should counsel elderly patients about the importance of maintaining balance and mobility through exercise, physical therapy, or other strategies.
Andrew M. Kaunitz, MD

Urinary incontinence affects more than one-third of women aged 70 years or older. As the authors of this study point out, urinary incontinence is comparable to other chronic conditions, such as hypertension and diabetes mellitus, in its impact on quality of life, and is a common reason for institutionalization.

The risk of urinary incontinence increases with age. In elderly women, it is often mixed (ie, having both urge- and stress-related components) and associated with functional impairments, including reduced mobility. It is thought that the association between incontinence and functional impairment is related primarily to the urge component:

• Women with impaired mobility take longer to reach the toilet, increasing the risk of leakage when the urge to urinate is strong
• Women with urge incontinence may be more likely to limit their activities so that they are always near a toilet.

Cognitive impairment may also play a role, affecting motor skills and bladder control.

To better understand why advancing age is linked with urinary incontinence, Fritel and colleagues studied a population of 1,942 urban-dwelling French women aged 75 to 85 years (mean age, 79.3 years; mean body mass index, 25.9 kg/m²).

Details of the study
Investigators assessed the frequency and quantity of urine leaks, the impact of urinary incontinence on daily life, and the participants’ mobility and balance. Data on urinary incontinence were collected via a self-administered questionnaire (the International Consultation on Incontinence Questionnaire–Short Form). Motor-related physical function was assessed using standardized balance and gait tests.

Urinary incontinence was reported by 42% of participants. Of these women, 57% reported daily urine leakage, with mixed incontinence found to be more prevalent than urge incontinence, which was more prevalent than stress incontinence. Overall, women with urinary incontinence reported that its impact on daily life was mild. Among those with mixed or urge incontinence, limitations in mobility and balance were correlated significantly with the severity of incontinence.

What this evidence means for practice
These findings support earlier data suggesting that impaired mobility can promote urge and mixed urinary incontinence. Because there also is a possibility that cerebral deterioration in aging women causes gait and balance problems that increase the likelihood of urge incontinence, the authors advise against the use of anticholinergic agents in elderly women, as these drugs can impair cognitive function. Another take-home message from this French report is that we should counsel elderly patients about the importance of maintaining balance and mobility through exercise, physical therapy, or other strategies.
Andrew M. Kaunitz, MD

Urinary incontinence affects more than one-third of women aged 70 years or older. As the authors of this study point out, urinary incontinence is comparable to other chronic conditions, such as hypertension and diabetes mellitus, in its impact on quality of life, and is a common reason for institutionalization.

The risk of urinary incontinence increases with age. In elderly women, it is often mixed (ie, having both urge- and stress-related components) and associated with functional impairments, including reduced mobility. It is thought that the association between incontinence and functional impairment is related primarily to the urge component:

• Women with impaired mobility take longer to reach the toilet, increasing the risk of leakage when the urge to urinate is strong
• Women with urge incontinence may be more likely to limit their activities so that they are always near a toilet.

Cognitive impairment may also play a role, affecting motor skills and bladder control.

To better understand why advancing age is linked with urinary incontinence, Fritel and colleagues studied a population of 1,942 urban-dwelling French women aged 75 to 85 years (mean age, 79.3 years; mean body mass index, 25.9 kg/m²).

Details of the study
Investigators assessed the frequency and quantity of urine leaks, the impact of urinary incontinence on daily life, and the participants’ mobility and balance. Data on urinary incontinence were collected via a self-administered questionnaire (the International Consultation on Incontinence Questionnaire–Short Form). Motor-related physical function was assessed using standardized balance and gait tests.

Urinary incontinence was reported by 42% of participants. Of these women, 57% reported daily urine leakage, with mixed incontinence found to be more prevalent than urge incontinence, which was more prevalent than stress incontinence. Overall, women with urinary incontinence reported that its impact on daily life was mild. Among those with mixed or urge incontinence, limitations in mobility and balance were correlated significantly with the severity of incontinence.

What this evidence means for practice
These findings support earlier data suggesting that impaired mobility can promote urge and mixed urinary incontinence. Because there also is a possibility that cerebral deterioration in aging women causes gait and balance problems that increase the likelihood of urge incontinence, the authors advise against the use of anticholinergic agents in elderly women, as these drugs can impair cognitive function. Another take-home message from this French report is that we should counsel elderly patients about the importance of maintaining balance and mobility through exercise, physical therapy, or other strategies.
Andrew M. Kaunitz, MD

Delirium is a common condition in hospitalized older patients. Often, a report of a “change in mental status” is the reason geriatric patients are sent to the emergency room for evaluation, although delirium also can develop after admission.

Delirium is a marker of underlying medical illness that needs careful workup and treatment. The condition can be iatrogenic, resulting from prescribed medication or a surgical procedure; most often, it is the consequence of multiple factors. Delirium can be expensive, because it increases hospital length of stay and overall costs—particularly if the patient is discharged to a nursing facility, not to home. Patients with delirium are at higher risk of death.

Delirium often goes unrecognized by physicians and nursing staff, and is not documented in medical records. Educating the medical staff on the identification and management of delirium is a key role for consulting psychiatrists.

CASE: Confused and agitated

Ms. T, a 93-year-old resident of an assisted living facility with a history of three
cerebral vascular accidents, atrial fibrillation, hypertension, multiple deep venous thromboses, blindness in her right eye, and deafness in her right ear without a hearing aid, is brought to the hospital after a syncopal episode lasting 10 minutes that was followed by slurred speech, confusion, and transient hypotension. Her dentist recently started her on azithromycin.

In the emergency room, Ms. T’s elevated blood pressure is managed with hydralazine and diltiazem. A CT scan of the head rules out hemorrhagic stroke. Complete blood count and tests of electrolytes, vitamin B12, and thyroid-stimulating hormone are within normal limits; urinalysis is negative for urinary tract infection.

Ms. T is noted to be in and out of sleep, with some confusion. She is maintained without oral food or fluids because of concerns about her ability to swallow. After 5 or 6 hours in the ER, Ms. T is transferred to a medical unit, where she becomes agitated and paranoid, with the delusion that her daughter is an impostor. She yells, is combative, and refuses medication.

Her confusion and behaviors become worse at night: She pulls out her IV line and telemetry leads. Blood pressure remains elevated, for which she receives additional doses of hydralazine.

For behavioral management, the medical team orders a one-time IM dose of haloperidol and starts her on risperidone, 0.5 mg every 4 hours as needed, which Ms. T refuses to take. She is incontinent and has foul-smelling urine.

Ms. T’s family is shocked at her condition; nursing staff is frustrated. With her worsening paranoia, delusions, and combative behaviors towards the nursing staff, psychiatry is consulted.

How to recognize and diagnose

The Box lists DSM-5 criteria for delirium.¹ The key feature is a disturbance in attention—what was referred to in DSM-IV-TR as “disturbance in consciousness.” That finding contrasts with what is seen in dementia, with its hallmark memory impairment and chronic deterioration.

In a hospital setting, the question is often asked: Does this patient have dementia or delirium? In many cases, the answer is both, because preexisting cognitive impairment is an important risk factor for delirium.

In addition to the standard clinical interview, several screening instruments or delirium rating scales have been developed. The most commonly used (Table 1) is the Confusion Assessment Method developed by Inouye and colleagues.²

Subtypes of delirium have been described, largely based on motor activity. Patients can present as hyperactive, hypoactive, mixed, or neither.³ Psychiatrists are more likely to be consulted regarding patients with hyperactive delirium, because they are the ones who scream, pull out their IV line, hallucinate, and are delusional, insisting they “have to go home”—such as the patient described in the case above.

Patients with hypoactive delirium often, on the other hand, are difficult to recognize; they present with lethargy, drowsiness, apathy, and confusion. They become withdrawn and answer slowly⁴; often, psychiatry is consulted to assess them for depression.

Delirium can be difficult to diagnose in patients with underlying dementia, who are not able to provide information. In such cases, obtaining collateral information from a family member or primary caretaker is crucial. Knowing the patient’s baseline helps to determine whether there has been an acute change in mental status.

CASE CONTINUED: Acute mental status changes

Ms. T’s daughter reports that her mother has not been in this condition before. At baseline, Ms. T has had memory problems but no paranoia, delusions, or agitated behaviors. Her daughter also reports that Ms. T has visual and hearing impairments and is not wearing her hearing aid.

The acute change in mental status and the perceptual disturbances indicate that Ms. T has delirium, not dementia.

Who is likely to develop delirium?

Risk factors for delirium (Table 2) include preexisting cognitive impairment, older age, vision and hearing impairment, use of psychoactive drugs, severe illness, azotemia and dehydration, a metabolic abnormality, and infection. Male sex also seems to be a risk factor, perhaps because men are more likely to abuse alcohol before admission.

Many patients become delirious after starting a new medication. An experienced geriatrician teaches that the main causes of delirium are “drugs, drugs, drugs, infections, and everything else” (Kenneth Rockwood, MD, personal communication, 2012). At admission, urinary tract infection and pneumonia are common causes of delirium, especially in geriatric patients.

What is the clinical course?

The clinical course varies widely. Delirium often is the reason that a patient is brought to the hospital, presenting with the condition at admission or early in hospitalization. The highest incidence among surgical patients appears to be on the third postoperative day—in some cases because of alcohol or drug withdrawal.

As noted in the DSM-5 criteria, delirium often comes on acutely, over hours or days. Symptoms can persist for weeks after initial onset of episodes of delirium.⁵ Symptoms fluctuate over the course of the day; at times, they can be missed if a provider sees the patient only while she (he) is clearer and doesn’t review nursing notes from other shifts.

How does delirium affect outcome?

Delirium has been shown to be associated with prolonged hospital stay (21 days, compared with 11 days in the absence of delirium), functional decline during hospitalization, and increased admission to long-term care (36% compared with 13%).⁶ In a study by O’Keefe and Lavan,⁶ delirious patients were more likely to sustain falls and to develop urinary incontinence, pressure sores, and other complications during hospitalization.

Older patients with delirium superimposed on dementia had a more than twofold increased risk of mortality compared with patients with dementia alone or with neither dementia nor delirium.⁷ Rockwood found that an episode of delirium was associated with a much higher rate of subsequent dementia.⁸

Think of an acute medical illness as a “stress test” for the brain, such that, if the patient develops delirium, it suggests an underlying brain disease that was not evident before the acute episode. After hip fracture, for example, delirium was independently associated with poor functional recovery at 1 month⁹ and at 2 years.¹⁰

Older patients admitted to a skilled nursing facility with delirium are more likely to experience one or more complications (73% compared with 41%).¹¹ In the study by Marcantonio and colleagues, patients with delirium were more than twice as likely to be hospitalized again within 30 days (30% and 13%), and less than half as likely to be discharged to the community (30% and 73%). Table 3 summarizes the impact of delirium on outcomes.

Appropriate management steps

Identifying and treating underlying medical illness is the definitive treatment for delirium; in a geriatric patient with multiple medical comorbidities the pathogenesis often is multifactorial or a definitive precipitant cannot always be identified.¹²

Managing a patient with delirium includes both non-pharmacotherapeutic interventions, which should be considered first-line, and pharmacotherapeutic interventions. Non-pharmacotherapeutic interventions include, but are not limited to:

•  support and close observation by nursing staff

•  placing a clock or calendar in the room

•  frequent reorientation and reminders

•  placing familiar possessions in the room

•  putting the patient in an isolated room with a window

•  regulating the sleep-wake cycle.⁴

Pharmacotherapeutic intervention in delirium should be used for behavioral symptoms, but only for the minimum duration necessary⁴ and preferably oral or IV. No drugs are FDA-approved for delirium, which means that use of any agent is off-label.¹³

Antipsychotics are the mainstay of pharmacotherapy for delirium in most settings. The use of antipsychotics relates to the dopamine excess-acetylcholine deficiency hypothesis of delirium pathophysiology.¹² Haloperidol remains the first-line agent because it is available in multiple dosages and can be given by various routes. IV haloperidol appears to carry less risk of extrapyramidal symptoms than oral haloperidol does but, as with all antipsychotics, its use warrants monitoring for QTc prolongation.¹²

Studies have not shown that atypical antipsychotics are superior to typical antipsychotics for delirium. Multiple studies have shown that atypicals are as efficacious as haloperidol.

Benzodiazepines are the treatment of choice for delirium caused by alcohol withdrawal. A Cochrane review found no evidence that benzodiazepines were helpful in treating delirium unrelated to alcohol withdrawal.¹⁴ In some studies, benzodiazepines were associated with an increased risk of delirium, especially in patients in the intensive care unit.¹⁵

More recently, cholinesterase inhibitors have been used to treat delirium. The reasoning behind their use is the hypothesis of a central cholinergic deficiency in delirium.¹² Regrettably, there have been few well-conducted studies of these agents in delirium, and a Cochrane review found no significant benefit for cholinesterase inhibitors.¹⁶ With the same hypothesis in mind, anticholinergic medications in patients with delirium should be avoided because these agents could exacerbate delirium by further decreasing the acetylcholine level.

Because delirium is common in the hospitalized population (especially older patients), a number of studies have examined strategies to prevent or reduce its development. Inouye and colleagues conducted a controlled clinical trial, in which they intervened to reduce six risk factors for delirium: cognitive impairment, sleep deprivation, immobility, visual and hearing impairment, and dehydration in hospitalized geriatric patients. The number and duration of events of delirium were significantly lower in the intervention group.¹⁷

Brummel et al reported that reducing modifiable risk factors in intensive care unit patients—including sedation management, minimizing deliriogenic medications (anticholinergics, antihistamines), minimizing sleep disruption, and encouraging early mobility—could prevent or reduce the incidence of delirium.¹⁵

CASE CONCLUDED: Return to baseline

Ms. T’s medications are minimized or discontinued, including azithromycin, based on case reports in the literature. She is stabilized hemodynamically.

Clinicians educate Ms. T’s family about delirium. To address Ms. T’s aggressive and paranoid behaviors, clinicians request that a family member is present to reassure Ms. T. She is continued on low-dose haloperidol. The family also is asked to bring Ms. T’s hearing aid and eyeglasses.

MRI is performed after Ms. T’s behavior is under control. The scan is negative for a new stroke.

Repeat blood tests the following day show an elevated white blood cell count; urinalysis is positive for a urinary tract infection. Ms. T is started on antibiotics. Subsequent urine culture shows no bacterial growth; the antibiotics are stopped after 3 days.

Ms. T slowly improves. According to her family, she is back at baseline in 3 or 4 days.

This case illustrates the complexity of trying to identify the precise cause of delirium among the many that could be involved. Often, no single cause can be found.¹⁸

Bottom Line
Delirium is a common and potentially life-threatening condition in hospitalized geriatric patients. General hospital psychiatrists should know how to recognize and treat the condition in collaboration with their medical colleagues.

Related Resources

• Treating delirium: a quick reference guide. Arlington, VA: American Psychiatric Association. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1662986.
• Cook IA. Guideline watch: practice guidelines for the treatment of patients with delirium. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1681952.
• Fearing MA, Inouye SK. Delirium. In: Blazer DG, Steffens D, eds. The American Psychiatric Publishing textbook of geriatric psychiatry. 4th ed. Arlington, VA: American Psychiatric Publishing, Inc.; 2009:229-241.
• Ghandour A, Saab R, Mehr D. Detecting and treating delirium—key interventions you may be missing. J Fam Pract. 2011;60(12):726-734.
• Leentjens AF, Rundell J, Rummans T, et al. Delirium: an evidence-based medicine (EBM) monograph for psychosomatic medicine practice. J Psychosom Res. 2012;73:149-152.
• Liptzin B, Jacobson SA. Delirium. In: Sadock BJ, Sadock VA, Ruiz P, eds. Comprehensive textbook of psychiatry. 9th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2009:4066-4073.

Drug Brand Names

Azithromycin • Zithromax                 Hydralazine • Apresoline

Diltiazem • Cardizem                        Risperidone • Risperdal

Haloperidol • Haldo

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.


Featured Audio
Benjamin Liptzin, MD, describes the distinction between dementia and delirium. Dr. Liptzin is Chair of Psychiatry, Baystate Medical Center, Springfield, Massachusetts, and Professor and Deputy Chair, Department of Psychiatry, Tufts University School of Medicine, Boston, Massachusetts.

Delirium is a common condition in hospitalized older patients. Often, a report of a “change in mental status” is the reason geriatric patients are sent to the emergency room for evaluation, although delirium also can develop after admission.

Delirium is a marker of underlying medical illness that needs careful workup and treatment. The condition can be iatrogenic, resulting from prescribed medication or a surgical procedure; most often, it is the consequence of multiple factors. Delirium can be expensive, because it increases hospital length of stay and overall costs—particularly if the patient is discharged to a nursing facility, not to home. Patients with delirium are at higher risk of death.

Delirium often goes unrecognized by physicians and nursing staff, and is not documented in medical records. Educating the medical staff on the identification and management of delirium is a key role for consulting psychiatrists.

CASE: Confused and agitated

Ms. T, a 93-year-old resident of an assisted living facility with a history of three
cerebral vascular accidents, atrial fibrillation, hypertension, multiple deep venous thromboses, blindness in her right eye, and deafness in her right ear without a hearing aid, is brought to the hospital after a syncopal episode lasting 10 minutes that was followed by slurred speech, confusion, and transient hypotension. Her dentist recently started her on azithromycin.

In the emergency room, Ms. T’s elevated blood pressure is managed with hydralazine and diltiazem. A CT scan of the head rules out hemorrhagic stroke. Complete blood count and tests of electrolytes, vitamin B12, and thyroid-stimulating hormone are within normal limits; urinalysis is negative for urinary tract infection.

Ms. T is noted to be in and out of sleep, with some confusion. She is maintained without oral food or fluids because of concerns about her ability to swallow. After 5 or 6 hours in the ER, Ms. T is transferred to a medical unit, where she becomes agitated and paranoid, with the delusion that her daughter is an impostor. She yells, is combative, and refuses medication.

Her confusion and behaviors become worse at night: She pulls out her IV line and telemetry leads. Blood pressure remains elevated, for which she receives additional doses of hydralazine.

For behavioral management, the medical team orders a one-time IM dose of haloperidol and starts her on risperidone, 0.5 mg every 4 hours as needed, which Ms. T refuses to take. She is incontinent and has foul-smelling urine.

Ms. T’s family is shocked at her condition; nursing staff is frustrated. With her worsening paranoia, delusions, and combative behaviors towards the nursing staff, psychiatry is consulted.

How to recognize and diagnose

The Box lists DSM-5 criteria for delirium.¹ The key feature is a disturbance in attention—what was referred to in DSM-IV-TR as “disturbance in consciousness.” That finding contrasts with what is seen in dementia, with its hallmark memory impairment and chronic deterioration.

In a hospital setting, the question is often asked: Does this patient have dementia or delirium? In many cases, the answer is both, because preexisting cognitive impairment is an important risk factor for delirium.

In addition to the standard clinical interview, several screening instruments or delirium rating scales have been developed. The most commonly used (Table 1) is the Confusion Assessment Method developed by Inouye and colleagues.²

Subtypes of delirium have been described, largely based on motor activity. Patients can present as hyperactive, hypoactive, mixed, or neither.³ Psychiatrists are more likely to be consulted regarding patients with hyperactive delirium, because they are the ones who scream, pull out their IV line, hallucinate, and are delusional, insisting they “have to go home”—such as the patient described in the case above.

Patients with hypoactive delirium often, on the other hand, are difficult to recognize; they present with lethargy, drowsiness, apathy, and confusion. They become withdrawn and answer slowly⁴; often, psychiatry is consulted to assess them for depression.

Delirium can be difficult to diagnose in patients with underlying dementia, who are not able to provide information. In such cases, obtaining collateral information from a family member or primary caretaker is crucial. Knowing the patient’s baseline helps to determine whether there has been an acute change in mental status.

CASE CONTINUED: Acute mental status changes

Ms. T’s daughter reports that her mother has not been in this condition before. At baseline, Ms. T has had memory problems but no paranoia, delusions, or agitated behaviors. Her daughter also reports that Ms. T has visual and hearing impairments and is not wearing her hearing aid.

The acute change in mental status and the perceptual disturbances indicate that Ms. T has delirium, not dementia.

Who is likely to develop delirium?

Risk factors for delirium (Table 2) include preexisting cognitive impairment, older age, vision and hearing impairment, use of psychoactive drugs, severe illness, azotemia and dehydration, a metabolic abnormality, and infection. Male sex also seems to be a risk factor, perhaps because men are more likely to abuse alcohol before admission.

Many patients become delirious after starting a new medication. An experienced geriatrician teaches that the main causes of delirium are “drugs, drugs, drugs, infections, and everything else” (Kenneth Rockwood, MD, personal communication, 2012). At admission, urinary tract infection and pneumonia are common causes of delirium, especially in geriatric patients.

What is the clinical course?

The clinical course varies widely. Delirium often is the reason that a patient is brought to the hospital, presenting with the condition at admission or early in hospitalization. The highest incidence among surgical patients appears to be on the third postoperative day—in some cases because of alcohol or drug withdrawal.

As noted in the DSM-5 criteria, delirium often comes on acutely, over hours or days. Symptoms can persist for weeks after initial onset of episodes of delirium.⁵ Symptoms fluctuate over the course of the day; at times, they can be missed if a provider sees the patient only while she (he) is clearer and doesn’t review nursing notes from other shifts.

How does delirium affect outcome?

Delirium has been shown to be associated with prolonged hospital stay (21 days, compared with 11 days in the absence of delirium), functional decline during hospitalization, and increased admission to long-term care (36% compared with 13%).⁶ In a study by O’Keefe and Lavan,⁶ delirious patients were more likely to sustain falls and to develop urinary incontinence, pressure sores, and other complications during hospitalization.

Older patients with delirium superimposed on dementia had a more than twofold increased risk of mortality compared with patients with dementia alone or with neither dementia nor delirium.⁷ Rockwood found that an episode of delirium was associated with a much higher rate of subsequent dementia.⁸

Think of an acute medical illness as a “stress test” for the brain, such that, if the patient develops delirium, it suggests an underlying brain disease that was not evident before the acute episode. After hip fracture, for example, delirium was independently associated with poor functional recovery at 1 month⁹ and at 2 years.¹⁰

Older patients admitted to a skilled nursing facility with delirium are more likely to experience one or more complications (73% compared with 41%).¹¹ In the study by Marcantonio and colleagues, patients with delirium were more than twice as likely to be hospitalized again within 30 days (30% and 13%), and less than half as likely to be discharged to the community (30% and 73%). Table 3 summarizes the impact of delirium on outcomes.

Appropriate management steps

Identifying and treating underlying medical illness is the definitive treatment for delirium; in a geriatric patient with multiple medical comorbidities the pathogenesis often is multifactorial or a definitive precipitant cannot always be identified.¹²

Managing a patient with delirium includes both non-pharmacotherapeutic interventions, which should be considered first-line, and pharmacotherapeutic interventions. Non-pharmacotherapeutic interventions include, but are not limited to:

•  support and close observation by nursing staff

•  placing a clock or calendar in the room

•  frequent reorientation and reminders

•  placing familiar possessions in the room

•  putting the patient in an isolated room with a window

•  regulating the sleep-wake cycle.⁴

Pharmacotherapeutic intervention in delirium should be used for behavioral symptoms, but only for the minimum duration necessary⁴ and preferably oral or IV. No drugs are FDA-approved for delirium, which means that use of any agent is off-label.¹³

Antipsychotics are the mainstay of pharmacotherapy for delirium in most settings. The use of antipsychotics relates to the dopamine excess-acetylcholine deficiency hypothesis of delirium pathophysiology.¹² Haloperidol remains the first-line agent because it is available in multiple dosages and can be given by various routes. IV haloperidol appears to carry less risk of extrapyramidal symptoms than oral haloperidol does but, as with all antipsychotics, its use warrants monitoring for QTc prolongation.¹²

Studies have not shown that atypical antipsychotics are superior to typical antipsychotics for delirium. Multiple studies have shown that atypicals are as efficacious as haloperidol.

Benzodiazepines are the treatment of choice for delirium caused by alcohol withdrawal. A Cochrane review found no evidence that benzodiazepines were helpful in treating delirium unrelated to alcohol withdrawal.¹⁴ In some studies, benzodiazepines were associated with an increased risk of delirium, especially in patients in the intensive care unit.¹⁵

More recently, cholinesterase inhibitors have been used to treat delirium. The reasoning behind their use is the hypothesis of a central cholinergic deficiency in delirium.¹² Regrettably, there have been few well-conducted studies of these agents in delirium, and a Cochrane review found no significant benefit for cholinesterase inhibitors.¹⁶ With the same hypothesis in mind, anticholinergic medications in patients with delirium should be avoided because these agents could exacerbate delirium by further decreasing the acetylcholine level.

Because delirium is common in the hospitalized population (especially older patients), a number of studies have examined strategies to prevent or reduce its development. Inouye and colleagues conducted a controlled clinical trial, in which they intervened to reduce six risk factors for delirium: cognitive impairment, sleep deprivation, immobility, visual and hearing impairment, and dehydration in hospitalized geriatric patients. The number and duration of events of delirium were significantly lower in the intervention group.¹⁷

Brummel et al reported that reducing modifiable risk factors in intensive care unit patients—including sedation management, minimizing deliriogenic medications (anticholinergics, antihistamines), minimizing sleep disruption, and encouraging early mobility—could prevent or reduce the incidence of delirium.¹⁵

CASE CONCLUDED: Return to baseline

Ms. T’s medications are minimized or discontinued, including azithromycin, based on case reports in the literature. She is stabilized hemodynamically.

Clinicians educate Ms. T’s family about delirium. To address Ms. T’s aggressive and paranoid behaviors, clinicians request that a family member is present to reassure Ms. T. She is continued on low-dose haloperidol. The family also is asked to bring Ms. T’s hearing aid and eyeglasses.

MRI is performed after Ms. T’s behavior is under control. The scan is negative for a new stroke.

Repeat blood tests the following day show an elevated white blood cell count; urinalysis is positive for a urinary tract infection. Ms. T is started on antibiotics. Subsequent urine culture shows no bacterial growth; the antibiotics are stopped after 3 days.

Ms. T slowly improves. According to her family, she is back at baseline in 3 or 4 days.

This case illustrates the complexity of trying to identify the precise cause of delirium among the many that could be involved. Often, no single cause can be found.¹⁸

Bottom Line
Delirium is a common and potentially life-threatening condition in hospitalized geriatric patients. General hospital psychiatrists should know how to recognize and treat the condition in collaboration with their medical colleagues.

Related Resources

• Treating delirium: a quick reference guide. Arlington, VA: American Psychiatric Association. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1662986.
• Cook IA. Guideline watch: practice guidelines for the treatment of patients with delirium. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1681952.
• Fearing MA, Inouye SK. Delirium. In: Blazer DG, Steffens D, eds. The American Psychiatric Publishing textbook of geriatric psychiatry. 4th ed. Arlington, VA: American Psychiatric Publishing, Inc.; 2009:229-241.
• Ghandour A, Saab R, Mehr D. Detecting and treating delirium—key interventions you may be missing. J Fam Pract. 2011;60(12):726-734.
• Leentjens AF, Rundell J, Rummans T, et al. Delirium: an evidence-based medicine (EBM) monograph for psychosomatic medicine practice. J Psychosom Res. 2012;73:149-152.
• Liptzin B, Jacobson SA. Delirium. In: Sadock BJ, Sadock VA, Ruiz P, eds. Comprehensive textbook of psychiatry. 9th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2009:4066-4073.

Drug Brand Names

Azithromycin • Zithromax                 Hydralazine • Apresoline

Diltiazem • Cardizem                        Risperidone • Risperdal

Haloperidol • Haldo

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.


Featured Audio
Benjamin Liptzin, MD, describes the distinction between dementia and delirium. Dr. Liptzin is Chair of Psychiatry, Baystate Medical Center, Springfield, Massachusetts, and Professor and Deputy Chair, Department of Psychiatry, Tufts University School of Medicine, Boston, Massachusetts.

Delirium is a common condition in hospitalized older patients. Often, a report of a “change in mental status” is the reason geriatric patients are sent to the emergency room for evaluation, although delirium also can develop after admission.

Delirium is a marker of underlying medical illness that needs careful workup and treatment. The condition can be iatrogenic, resulting from prescribed medication or a surgical procedure; most often, it is the consequence of multiple factors. Delirium can be expensive, because it increases hospital length of stay and overall costs—particularly if the patient is discharged to a nursing facility, not to home. Patients with delirium are at higher risk of death.

Delirium often goes unrecognized by physicians and nursing staff, and is not documented in medical records. Educating the medical staff on the identification and management of delirium is a key role for consulting psychiatrists.

CASE: Confused and agitated

Ms. T, a 93-year-old resident of an assisted living facility with a history of three
cerebral vascular accidents, atrial fibrillation, hypertension, multiple deep venous thromboses, blindness in her right eye, and deafness in her right ear without a hearing aid, is brought to the hospital after a syncopal episode lasting 10 minutes that was followed by slurred speech, confusion, and transient hypotension. Her dentist recently started her on azithromycin.

In the emergency room, Ms. T’s elevated blood pressure is managed with hydralazine and diltiazem. A CT scan of the head rules out hemorrhagic stroke. Complete blood count and tests of electrolytes, vitamin B12, and thyroid-stimulating hormone are within normal limits; urinalysis is negative for urinary tract infection.

Ms. T is noted to be in and out of sleep, with some confusion. She is maintained without oral food or fluids because of concerns about her ability to swallow. After 5 or 6 hours in the ER, Ms. T is transferred to a medical unit, where she becomes agitated and paranoid, with the delusion that her daughter is an impostor. She yells, is combative, and refuses medication.

Her confusion and behaviors become worse at night: She pulls out her IV line and telemetry leads. Blood pressure remains elevated, for which she receives additional doses of hydralazine.

For behavioral management, the medical team orders a one-time IM dose of haloperidol and starts her on risperidone, 0.5 mg every 4 hours as needed, which Ms. T refuses to take. She is incontinent and has foul-smelling urine.

Ms. T’s family is shocked at her condition; nursing staff is frustrated. With her worsening paranoia, delusions, and combative behaviors towards the nursing staff, psychiatry is consulted.

How to recognize and diagnose

The Box lists DSM-5 criteria for delirium.¹ The key feature is a disturbance in attention—what was referred to in DSM-IV-TR as “disturbance in consciousness.” That finding contrasts with what is seen in dementia, with its hallmark memory impairment and chronic deterioration.

In a hospital setting, the question is often asked: Does this patient have dementia or delirium? In many cases, the answer is both, because preexisting cognitive impairment is an important risk factor for delirium.

In addition to the standard clinical interview, several screening instruments or delirium rating scales have been developed. The most commonly used (Table 1) is the Confusion Assessment Method developed by Inouye and colleagues.²

Subtypes of delirium have been described, largely based on motor activity. Patients can present as hyperactive, hypoactive, mixed, or neither.³ Psychiatrists are more likely to be consulted regarding patients with hyperactive delirium, because they are the ones who scream, pull out their IV line, hallucinate, and are delusional, insisting they “have to go home”—such as the patient described in the case above.

Patients with hypoactive delirium often, on the other hand, are difficult to recognize; they present with lethargy, drowsiness, apathy, and confusion. They become withdrawn and answer slowly⁴; often, psychiatry is consulted to assess them for depression.

Delirium can be difficult to diagnose in patients with underlying dementia, who are not able to provide information. In such cases, obtaining collateral information from a family member or primary caretaker is crucial. Knowing the patient’s baseline helps to determine whether there has been an acute change in mental status.

CASE CONTINUED: Acute mental status changes

Ms. T’s daughter reports that her mother has not been in this condition before. At baseline, Ms. T has had memory problems but no paranoia, delusions, or agitated behaviors. Her daughter also reports that Ms. T has visual and hearing impairments and is not wearing her hearing aid.

The acute change in mental status and the perceptual disturbances indicate that Ms. T has delirium, not dementia.

Who is likely to develop delirium?

Risk factors for delirium (Table 2) include preexisting cognitive impairment, older age, vision and hearing impairment, use of psychoactive drugs, severe illness, azotemia and dehydration, a metabolic abnormality, and infection. Male sex also seems to be a risk factor, perhaps because men are more likely to abuse alcohol before admission.

Many patients become delirious after starting a new medication. An experienced geriatrician teaches that the main causes of delirium are “drugs, drugs, drugs, infections, and everything else” (Kenneth Rockwood, MD, personal communication, 2012). At admission, urinary tract infection and pneumonia are common causes of delirium, especially in geriatric patients.

What is the clinical course?

The clinical course varies widely. Delirium often is the reason that a patient is brought to the hospital, presenting with the condition at admission or early in hospitalization. The highest incidence among surgical patients appears to be on the third postoperative day—in some cases because of alcohol or drug withdrawal.

As noted in the DSM-5 criteria, delirium often comes on acutely, over hours or days. Symptoms can persist for weeks after initial onset of episodes of delirium.⁵ Symptoms fluctuate over the course of the day; at times, they can be missed if a provider sees the patient only while she (he) is clearer and doesn’t review nursing notes from other shifts.

How does delirium affect outcome?

Delirium has been shown to be associated with prolonged hospital stay (21 days, compared with 11 days in the absence of delirium), functional decline during hospitalization, and increased admission to long-term care (36% compared with 13%).⁶ In a study by O’Keefe and Lavan,⁶ delirious patients were more likely to sustain falls and to develop urinary incontinence, pressure sores, and other complications during hospitalization.

Older patients with delirium superimposed on dementia had a more than twofold increased risk of mortality compared with patients with dementia alone or with neither dementia nor delirium.⁷ Rockwood found that an episode of delirium was associated with a much higher rate of subsequent dementia.⁸

Think of an acute medical illness as a “stress test” for the brain, such that, if the patient develops delirium, it suggests an underlying brain disease that was not evident before the acute episode. After hip fracture, for example, delirium was independently associated with poor functional recovery at 1 month⁹ and at 2 years.¹⁰

Older patients admitted to a skilled nursing facility with delirium are more likely to experience one or more complications (73% compared with 41%).¹¹ In the study by Marcantonio and colleagues, patients with delirium were more than twice as likely to be hospitalized again within 30 days (30% and 13%), and less than half as likely to be discharged to the community (30% and 73%). Table 3 summarizes the impact of delirium on outcomes.

Appropriate management steps

Identifying and treating underlying medical illness is the definitive treatment for delirium; in a geriatric patient with multiple medical comorbidities the pathogenesis often is multifactorial or a definitive precipitant cannot always be identified.¹²

Managing a patient with delirium includes both non-pharmacotherapeutic interventions, which should be considered first-line, and pharmacotherapeutic interventions. Non-pharmacotherapeutic interventions include, but are not limited to:

•  support and close observation by nursing staff

•  placing a clock or calendar in the room

•  frequent reorientation and reminders

•  placing familiar possessions in the room

•  putting the patient in an isolated room with a window

•  regulating the sleep-wake cycle.⁴

Pharmacotherapeutic intervention in delirium should be used for behavioral symptoms, but only for the minimum duration necessary⁴ and preferably oral or IV. No drugs are FDA-approved for delirium, which means that use of any agent is off-label.¹³

Antipsychotics are the mainstay of pharmacotherapy for delirium in most settings. The use of antipsychotics relates to the dopamine excess-acetylcholine deficiency hypothesis of delirium pathophysiology.¹² Haloperidol remains the first-line agent because it is available in multiple dosages and can be given by various routes. IV haloperidol appears to carry less risk of extrapyramidal symptoms than oral haloperidol does but, as with all antipsychotics, its use warrants monitoring for QTc prolongation.¹²

Studies have not shown that atypical antipsychotics are superior to typical antipsychotics for delirium. Multiple studies have shown that atypicals are as efficacious as haloperidol.

Benzodiazepines are the treatment of choice for delirium caused by alcohol withdrawal. A Cochrane review found no evidence that benzodiazepines were helpful in treating delirium unrelated to alcohol withdrawal.¹⁴ In some studies, benzodiazepines were associated with an increased risk of delirium, especially in patients in the intensive care unit.¹⁵

More recently, cholinesterase inhibitors have been used to treat delirium. The reasoning behind their use is the hypothesis of a central cholinergic deficiency in delirium.¹² Regrettably, there have been few well-conducted studies of these agents in delirium, and a Cochrane review found no significant benefit for cholinesterase inhibitors.¹⁶ With the same hypothesis in mind, anticholinergic medications in patients with delirium should be avoided because these agents could exacerbate delirium by further decreasing the acetylcholine level.

Because delirium is common in the hospitalized population (especially older patients), a number of studies have examined strategies to prevent or reduce its development. Inouye and colleagues conducted a controlled clinical trial, in which they intervened to reduce six risk factors for delirium: cognitive impairment, sleep deprivation, immobility, visual and hearing impairment, and dehydration in hospitalized geriatric patients. The number and duration of events of delirium were significantly lower in the intervention group.¹⁷

Brummel et al reported that reducing modifiable risk factors in intensive care unit patients—including sedation management, minimizing deliriogenic medications (anticholinergics, antihistamines), minimizing sleep disruption, and encouraging early mobility—could prevent or reduce the incidence of delirium.¹⁵

CASE CONCLUDED: Return to baseline

Ms. T’s medications are minimized or discontinued, including azithromycin, based on case reports in the literature. She is stabilized hemodynamically.

Clinicians educate Ms. T’s family about delirium. To address Ms. T’s aggressive and paranoid behaviors, clinicians request that a family member is present to reassure Ms. T. She is continued on low-dose haloperidol. The family also is asked to bring Ms. T’s hearing aid and eyeglasses.

MRI is performed after Ms. T’s behavior is under control. The scan is negative for a new stroke.

Repeat blood tests the following day show an elevated white blood cell count; urinalysis is positive for a urinary tract infection. Ms. T is started on antibiotics. Subsequent urine culture shows no bacterial growth; the antibiotics are stopped after 3 days.

Ms. T slowly improves. According to her family, she is back at baseline in 3 or 4 days.

This case illustrates the complexity of trying to identify the precise cause of delirium among the many that could be involved. Often, no single cause can be found.¹⁸

Bottom Line
Delirium is a common and potentially life-threatening condition in hospitalized geriatric patients. General hospital psychiatrists should know how to recognize and treat the condition in collaboration with their medical colleagues.

Related Resources

• Treating delirium: a quick reference guide. Arlington, VA: American Psychiatric Association. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1662986.
• Cook IA. Guideline watch: practice guidelines for the treatment of patients with delirium. http://psychiatryonline.org/content.aspx?bookid=28&sectionid=1681952.
• Fearing MA, Inouye SK. Delirium. In: Blazer DG, Steffens D, eds. The American Psychiatric Publishing textbook of geriatric psychiatry. 4th ed. Arlington, VA: American Psychiatric Publishing, Inc.; 2009:229-241.
• Ghandour A, Saab R, Mehr D. Detecting and treating delirium—key interventions you may be missing. J Fam Pract. 2011;60(12):726-734.
• Leentjens AF, Rundell J, Rummans T, et al. Delirium: an evidence-based medicine (EBM) monograph for psychosomatic medicine practice. J Psychosom Res. 2012;73:149-152.
• Liptzin B, Jacobson SA. Delirium. In: Sadock BJ, Sadock VA, Ruiz P, eds. Comprehensive textbook of psychiatry. 9th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2009:4066-4073.

Drug Brand Names

Azithromycin • Zithromax                 Hydralazine • Apresoline

Diltiazem • Cardizem                        Risperidone • Risperdal

Haloperidol • Haldo

Disclosures
The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.


Featured Audio
Benjamin Liptzin, MD, describes the distinction between dementia and delirium. Dr. Liptzin is Chair of Psychiatry, Baystate Medical Center, Springfield, Massachusetts, and Professor and Deputy Chair, Department of Psychiatry, Tufts University School of Medicine, Boston, Massachusetts.

Ovarian cysts, sometimes reported as ovarian masses or adnexal masses, are frequently found incidentally in women who have no symptoms. These cysts can be physiologic (having to do with ovulation) or neoplastic—either benign, borderline (having low malignant potential), or frankly malignant. Thus, these incidental lesions pose many diagnostic challenges to the clinician.

The vast majority of cysts are benign, but a few are malignant, and ovarian malignancies have a notoriously poor survival rate. The diagnosis can only be obtained surgically, as aspiration and biopsy are not definitive and may be harmful. Therefore, the clinician must try to balance the risks of surgery for what may be a benign lesion with the risk of delaying diagnosis of a malignancy.

In this article we provide an approach to evaluating these cysts, with guidance on when the patient can be reassured and when referral is needed.

THE DILEMMA OF OVARIAN CYSTS

Ovarian cysts are common

Premenopausal women can be expected to make at least a small cyst or follicle almost every month. The point prevalence for significant cysts has been reported to be almost 8% in premenopausal women.¹

Surprisingly, the prevalence in postmenopausal women is as high as 14% to 18%, with a yearly incidence of 8%. From 30% to 54% of postmenopausal ovarian cysts persist for years.^2,3

Little is known about the cause of most cysts

Little is known about the cause of most ovarian cysts. Functional or physiologic cysts are thought to be variations in the ovulatory process. They do not seem to be precursors to ovarian cancer.

Most benign neoplastic cysts are also not thought to be precancerous, with the possible exception of the mucinous kind.⁴ Ovarian cysts do not increase the risk of ovarian cancer later in life,^3,9 and removing benign cysts has not been shown to decrease the risk of death from ovarian cancer.¹⁰

Most ovarian cysts and masses are benign

Simple ovarian cysts are much more likely to be benign than malignant. Complex and solid ovarian masses are also more likely to be benign, regardless of menopausal status, but more malignancies are found in this group.

With any kind of mass, the chances of malignancy increase with age. Children and adolescents are not discussed in this article; they should be referred to a specialist.

Ovarian cancer often has a poor prognosis

This “silent” cancer is most often discovered and treated when it has already spread, contributing to a reported 5-year survival rate of only 33% to 46%.^11–13 Ideally, ovarian cancer would be found and removed while still confined to the ovary, when the 5-year survival rate is greater than 90%.

Unfortunately, there does not seem to be a precursor lesion for most ovarian cancers, and there is no good way of finding it in the stage 1 phase, so detecting this cancer before it spreads remains an elusive goal.^11,14

Surgery is required to diagnose difficult cases

There is no perfect test for the preoperative assessment of a cystic ovarian mass. Every method has drawbacks (Table 1).^15–18 Therefore, the National Institutes of Health estimates that 5% to 10% of women in the United States will undergo surgical exploration for an ovarian cyst in their lifetime. Only 13% to 21% of these cysts will be malignant.⁵

ASSESSING AN INCIDENTALLY DISCOVERED OVARIAN MASS

Certain factors in the history, physical examination, and blood work may suggest the cyst is either benign or malignant and may influence the subsequent assessment. However, in most cases, the best next step is to perform transvaginal ultrasonography, which we will discuss later in this paper.

History

Age is a major risk factor for ovarian cancer; the median age at diagnosis is 63 years.⁹ In the reproductive-age group, ovarian cysts are much more likely to be functional than neoplastic. Epithelial cancers are rare before the age of 40, but other cancer types such as borderline, germ cell, and sex cord stromal tumors may occur.¹⁹

In every age group a cyst is more likely to be benign than malignant, although, as noted above, the probability of malignancy increases with age.

Symptoms. Most ovarian cysts, benign or malignant, are asymptomatic and are found only incidentally.

The most commonly reported symptoms are pelvic or lower-abdominal pressure or pain. Acutely painful conditions include ovarian torsion, hemorrhage into the cyst, cyst rupture with or without intra-abdominal hemorrhage, ectopic pregnancy, and pelvic inflammatory disease with tubo-ovarian abscess.

Some patients who have ovarian cancer report vague symptoms such as urinary urgency or frequency, abdominal distention or bloating, and difficulty eating or early satiety.²⁰ Although the positive predictive value of this symptom constellation is only about 1%, its usefulness increases if these symptoms arose recently (within the past year) and occur than 12 days a month.²¹

Family history of ovarian, breast, endometrial, or colon cancer is of particular interest. The greater the number of affected relatives and the closer the degree of relation, the greater the risk; in some cases the relative risk is 40 times greater.²² Breast-ovarian cancer syndromes, hereditary nonpolyposis colorectal cancer syndrome, and family cancer syndrome, as well as extremely high-risk pedigrees such as BRCA1, BRCA2, and Lynch syndrome, all place women at significantly higher risk. Daughters tend to develop cancer at a younger age than their affected mothers.

However, only 10% of ovarian cancers occur in patients who have a family history of it, leaving 90% as sporadic occurrences.

Other history. Factors protective against ovarian cancer include use of oral contraceptives at any time, tubal ligation, hysterectomy, having had children, breastfeeding, a low-fat diet, and possibly use of aspirin and acetaminophen.^23,24

Risk factors for malignancy include advanced age; nulliparity; family history of ovarian or breast cancer; personal history of breast cancer; talc use; asbestos exposure; white ethnicity; pelvic irradiation; smoking; alcohol use; possibly the previous use of fertility drugs, estrogen, or androgen; history of mumps; urban location; early menarche; and late menopause.²⁴

Vital signs. Fever can indicate an infectious process or torsion of the ovary. A sudden onset of low blood pressure or rapid pulse can indicate a hemorrhagic condition such as ectopic pregnancy or ruptured hemorrhagic cyst.

Bimanual pelvic examination is notoriously inaccurate for detecting and characterizing ovarian cysts. In one prospective study, examiners blinded to the reason for surgery evaluated women under anesthesia. The authors concluded that bimanual examination was of limited value even under the best circumstances.¹⁵ Pelvic examination can be even more difficult in patients who are obese, are virginal, have vaginal atrophy, or are in pain.

Useful information that can be obtained through the bimanual examination includes the exact location of pelvic tenderness, the relative firmness of an identified mass, and the existence of nodularity in the posterior cul-de-sac, suggesting advanced ovarian cancer.

Tumor markers

Cancer antigen 125 (CA125) is the most studied and widely used of the ovarian cancer tumor markers. When advanced epithelial ovarian cancer is associated with a markedly elevated level, the value correlates with tumor burden.²⁵

Unfortunately, only about half of early-stage ovarian cancers and 75% to 80% of advanced ovarian cancers express this marker.²⁶ Especially in premenopausal women, there are many pelvic conditions that can falsely elevate CA125. Therefore, its sensitivity and specificity for predicting ovarian cancer are suboptimal. Nevertheless, CA125 is often used to help stratify risk when assessing known ovarian cysts and masses.

The value considered abnormal in postmenopausal women is 35 U/mL or greater, while in premenopausal women the cutoff is less well defined. The lower the cutoff level is set, the more sensitive the test. Recent recommendations advise 50 U/mL or 67 U/mL, rather than the 200 U/mL recommended in the 2002 joint guidelines of the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology.^27,28

However, specificity is likely to be lower with these lower cutoff values. Conditions that can elevate CA125 levels include almost anything that irritates the peritoneum, including pregnancy, menstruation, fibroids, endometriosis, infection, and ovarian hyperstimulation, as well as medical conditions such as liver or renal disease, colitis, diverticulitis, congestive heart failure, diabetes, autoimmune diseases, and ascites.

Following serial CA125 levels may be more sensitive than trying to establish a single cutoff value.²⁹ CA125 should not be used as a screening tool in average-risk women.²⁶

OVA1. Several biomarker panels have been developed and evaluated for risk assessment in women with pelvic masses. OVA1, a proprietary panel of tests (Vermillion; Austin, TX) received US Food and Drug Administration approval in 2009. It includes CA125 and four other proteins, from which it calculates a probability score (high or low) using a proprietary formula.

In prospective studies, OVA1 was more sensitive than clinical assessment or CA125 alone.³⁰ The higher sensitivity and negative predictive value were counterbalanced by a lower specificity and positive predictive value.³¹ Its cost ($650) is not always covered by insurance. OVA1 is not a screening tool.

EVALUATION WITH ULTRASONOGRAPHY

Ultrasonography is the imaging test of choice in assessing adnexal cysts and masses, and therefore it is the best next step after taking a history, performing a physical examination, and obtaining blood work.³² In cases in which an incidental ovarian mass is discovered on computed tomography (CT), further characterization by ultrasonography will likely yield helpful information.

Pelvic ultrasonography can be performed transabdominally or transvaginally. Vaginal ultrasonography gives the clearest images in most patients. Abdominal scanning is indicated for large masses, when vaginal access is difficult (as in virginal patients or those with vaginal atrophy) or when the mass is out of the focal length of the vaginal probe. A full bladder is usually required for the best transabdominal images.

The value of the images obtained depends on the experience of the ultrasonographer and reader and on the equipment. Also, there is currently no widely used standard for reporting the findings³³—descriptions are individualized, leading some authors to recommend that the clinician personally review the films to get the most accurate picture.¹⁹

Size

Size alone cannot be used to distinguish between benign and malignant lesions. Simple cysts up to 10 cm are most likely benign regardless of menopausal status.^2,34 However, in a complex or solid mass, size correlates somewhat with the chance of malignancy, with notable exceptions, such as the famously large sizes of some solid fibromas or mucinous cystadenomas. Also, size may correlate with risk of other complications such as torsion or symptomatic rupture.

Complexity

Simple cysts have clear fluid, thin smooth walls, no loculations or septae, and enhanced through-transmission of echo waves.^32,33

Complexity is described in terms of septations, wall thickness, internal echoes, and solid nodules. Increasing complexity does correlate with increased risk of malignancy.

Worrisome findings

The most worrisome findings are:

• Solid areas that are not hyperechoic, especially when there is blood flow to them
• Thick septations, more than 2 or 3 mm wide, especially if there is blood flow within them
• Excrescences on the inner or outer aspect of a cystic area
• Ascites
• Other pelvic or omental masses.

Benign conditions

Several benign conditions have characteristic complex findings on ultrasonography (Table 2), whereas other findings can be indeterminate (Table 3) or worrisome for malignancy (Table 4).

Hemorrhagic corpus luteum cysts can be complex with an internal reticular pattern due to organizing clot and fibrin strands. A “ring of fire” vascular pattern is often seen around the cyst bed.

Dermoids (mature cystic teratomas) may have hyperechoic elements with acoustic shadowing and no internal Doppler flow. They can have a complex appearance due to fat, hair, and sebum within the cyst. Dermoid cysts have a pathognomonic appearance on CT with a clear fat-fluid level.

Endometriomas classically have a homogeneous “ground-glass” appearance or low-level echoes, without internal color Doppler flow, wall nodules, or other malignant features.

Fibroids may be pedunculated and may appear to be complex or solid adnexal masses.

Hydrosalpinges may present as tortuous tubular-shaped cystic masses. There may be incomplete septations or indentations seen on opposite sides (the “waist” sign).

Paratubal and paraovarian cysts are usually simple round cysts that can be demonstrated as separate from the ovary. Sometimes these appear complex as well.

Peritoneal inclusion cysts, also known as pseudocysts, are seen in patients with intra-abdominal adhesions. Often multiple septations are seen through clear fluid, with the cyst conforming to the shape of other pelvic structures.

Torsion of the ovary may occur with either benign or malignant masses. Torsion can be diagnosed when venous flow is absent on Doppler. The presence of flow, however, doesn’t rule out torsion, as torsion is often intermittent. The twisted ovary is most often enlarged and can have an edematous appearance. Although typically benign, these should be referred for urgent surgical treatment.

Vascularity

Doppler imaging is being extensively studied. The general principle is that malignant masses will be more vascular, with a high-volume, low-resistance pattern of flow. This can result in a pulsatility index of less than 1 or a resistive index of less than 0.4. In practice, however, there is significant overlap between high and low pulsatility indices and resistive indices in benign and malignant cysts. Low resistance can also be found in endometriomas, corpus luteum cysts, inflammatory masses, and vascular benign neoplasms. A normal (high) resistive index does not rule out malignancy.^32,33

One Doppler finding that does seem to correlate with malignancy is the presence of any flow within a solid nodule or wall excrescence.

3D ultrasonography

As the use of 3D ultrasonography increases, studies are yielding different results as to its utility in describing ovarian masses. 3D ultrasonography may be useful in finding centrally located vessels so that Doppler can be applied.³²

OTHER IMAGING

Although ultrasonography is the initial imaging study of choice in the evaluation of adnexal masses owing to its high sensitivity, availability, and low cost, studies have shown that up to 20% of adnexal masses can be reported as indeterminate by ultrasonography (Table 1).

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is emerging as a very valuable tool when ultrasonography is inconclusive or limited.³⁵ Although MRI is very accurate (Table 1), it is not considered a first-line imaging test because it is more expensive, less available, and more inconvenient for the patient than ultrasonography.

MRI provides additional information on the composition of soft-tissue tumors. Usually, MRI is ordered with contrast, unless there are contraindications to it. The radiologist will evaluate morphologic features, signal intensity, and enhancement of solid areas. Techniques such as dynamic contrast-enhanced MRI (following the distribution of contrast material over time), in- and out-of-phase T1 imaging (looking for fat, such as in dermoids), and the newer diffusion-weighted imaging may further improve characterization.

In one study of MRI as second-line imaging, contrast-enhanced MRI contributed to a greater change in the probability of ovarian cancer than did CT, Doppler ultrasonography, or MRI without contrast.³⁶ This may result in a reduction in unnecessary surgeries and in an increase in proper referrals in cases of suspected malignancy.

Computed tomography

Disadvantages of CT include radiation exposure and poor discrimination of soft tissue. It can, however, differentiate fat or calcifications that may be found in dermoids. While CT is not often used to describe an ovarian lesion, it may be used preoperatively to stage an ovarian cancer or to look for a primary intra-abdominal cancer when an ovarian mass may represent metastasis.³²

MANAGING AN INCIDENTAL OVARIAN CYST OR CYSTIC MASS

Combining information from the history, physical examination, imaging, and blood work to assign a level of risk of malignancy is not straightforward. The clinician must weigh several imperfect tests, each with its own sensitivity and specificity, against the background of the individual patient’s likelihood of malignancy. Whereas a 4-cm simple cyst in a premenopausal woman can be assigned to a low-risk category and a complex mass with flow to a solid component in a postmenopausal woman can be assigned to a high-risk category, many lesions are more difficult to assess.

Several systems have been proposed for analyzing data and standardizing risk assessment. There are a number of scoring systems based on ultrasonographic morphology and several mathematical regression models that include menopausal status and tumor markers. But each has drawbacks, and none is definitively superior to expert opinion.^16,17,37,38

A 2012 systematic review and meta-analysis³⁹ calculated sensitivity and specificity for several imaging tests, scoring systems, and blood tumor markers. Some results are presented in Table 1.

The management of an ovarian cyst depends on symptoms, likelihood of torsion or rupture, and the level of concern for malignancy. At the lower-risk end of the spectrum, reassurance or observation over time may be appropriate. A general gynecologist can evaluate indeterminate or symptomatic ovarian cysts. Patients with masses frankly suspicious for malignancy are best referred to a gynecologic oncologist.

Expectant management for low-risk lesions

Low-risk lesions such as simple cysts, endometriomas, and dermoids have a less than 1% chance of malignancy. Most patients who have them require only reassurance or follow-up with serial ultrasonography. Oral contraceptives may prevent new cysts from forming. Aspiration is not recommended.

In 2010, the Society of Radiologists in Ultrasound issued a consensus statement regarding re-imaging of simple ovarian cysts.³³

In premenopausal women, they recommend no further testing for cysts 5 cm or smaller, yearly follow-up for cysts larger than 5 cm and up to and including 7 cm, and MRI or surgical evaluation for cysts larger than 7 cm, as it is difficult to completely image a large cyst with ultrasonography.

In postmenopausal women, if the cyst is 1 cm in diameter or smaller, no further studies need to be done. For simple cysts larger than 1 cm and up to and including 7 cm, yearly re-imaging is recommended. And for cysts larger than 7 cm, MRI or surgery is indicated. The American College of Radiology recommends repeat ultrasonography and CA125 testing for cysts 3 cm and larger but doesn’t specify an interval.³²

A cyst that is otherwise simple but has a single thin septation (< 3 mm) or a small calcification in the wall is almost always benign. Such a cyst should be followed as if it were a simple cyst, as indicated by patient age and cyst size.

There are no official guidelines as to when to stop serial imaging,^22,32 but a recent paper suggested one or two ultrasonographic examinations to confirm size and morphologic stability.¹⁹ Once a lesion has resolved, there is no need for further imaging (Figures 1–3).

Birth control pills for suppression of new cysts. Oral contraceptives do not hasten the resolution of ovarian cysts, according to a 2011 Cochrane review.40 Some practitioners will, nevertheless, prescribe them in an attempt to prevent new cysts from confusing the picture.

Aspiration is not recommended for either diagnosis or treatment. It can only be considered in patients at high risk who are not surgical candidates. Results of cytologic study of specimens obtained by fine-needle aspiration cannot reliably determine the presence or absence of malignancy.⁴¹ There is also a theoretical risk of spreading cancer from an early-stage lesion. A retrospective study has suggested that spillage of cyst contents during surgery in early ovarian cancer is associated with a worse prognosis.⁴²

From a therapeutic point of view, studies have shown the same resolution rate at 6 months for aspirated cysts vs those followed expectantly.⁴³ Another study found a recurrence rate of 25% within 1 year of aspiration.⁴⁴

Referral for medium-risk or indeterminant-risk ovarian masses

Patients who have medium- or indeterminaterisk ovarian masses (Table 3) should be referred to a gynecologist. Further testing will help stratify the risk of malignancy. This can include tumor marker blood tests, MRI, or CT, the addition of Doppler or 3D ultrasonography, serial ultrasonography, or surgical exploration.

If repeat ultrasonography is chosen, the interval will likely be 6 to 12 weeks. Surgery may consist of removing only the cyst itself, or the whole ovary with or without the tube, or sometimes both ovaries. Purely diagnostic laparoscopy is rarely performed, as direct visualization of a lesion is rarely helpful. Frozen section should be employed, and the operating gynecologist should have oncologic backup, since the surgery is performed to rule out malignancy.

In the case of a benign-appearing cyst larger than 6 cm, thought must be given as to whether it is likely to rupture or twist. Rupture of a large cyst can lead to pain and in some cases to hemorrhage. Contents of a ruptured dermoid cyst can cause chemical peritonitis. Torsion of an ovary can result in loss of the ovary through compromised perfusion. A general gynecologist can decide with the patient whether preemptive surgery is indicated.

Operative evaluation for high-risk masses

Patients with high-risk ovarian masses (Table 4) are best referred to a gynecologic oncologist for operative evaluation. If features are seen that indicate malignancy, such as thick septations, solid areas with blood flow, ascites, or other pelvic masses, surgery is indicated. The surgical approach may be through laparoscopy or laparotomy.⁴⁵ It should be noted that even in the face of worrisome features on ultrasonography, many masses turn out to be benign.

In 2011, the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology issued new guidelines recommending oncologic referral of patients with high-risk masses. Elevated CA125, ascites, a nodular or fixed pelvic mass, or evidence of metastasis in postmenopausal women requires oncologic evaluation.²⁶ For premenopausal women, a very elevated CA125, ascites, or metastasis requires referral (Table 4).²⁶

Direct referral to a gynecologic oncologist is underutilized. A recent study found that fewer than half of primary care physicians said that they would refer a classic suspicious case directly to a subspecialist.⁴⁶ It is estimated that only 33% of ovarian cancers are first operated on by a gynecologic oncologist.⁴⁷

A 2011 Cochrane review confirmed a survival benefit for women with cancer who are operated on by gynecologic oncologists primarily, rather than by a general gynecologist and then referred.⁴⁸ A gynecologic oncologist is most likely to perform proper staging and debulking at the time of initial diagnosis.⁴⁹

Special situations require consultation

Ovarian cysts in pregnancy are most often benign,⁵⁰ but malignancy is always a possibility. Functional cysts and dermoids are the most common. These may remain asymptomatic or may rupture or twist or cause difficulty with delivering the baby. Surgical intervention, if needed, should be performed in the second trimester if possible. A multidisciplinary approach and referral to a perinatologist and gynecologic oncologist are advised.

Symptomatic ovarian cysts that may need surgical intervention are the purview of the general gynecologist. If the risk of a surgical emergency is judged to be low, a symptomatic patient may be supported with pain medication and may be managed on an outpatient basis. Immediate surgical consultation is appropriate if the patient appears toxic or in shock. Depending on the clinical picture, there may be a ruptured tubo-ovarian abscess, ruptured ectopic pregnancy, ruptured hemorrhagic cyst, or ovarian torsion, any of which may need immediate surgical intervention.

If a symptomatic mass is highly suspicious for cancer, a gynecologic oncologist should be consulted directly.

WHEN TO REASSURE, REASSESS, REFER

Ovarian masses often pose diagnostic and management dilemmas. Reassurance can be offered to women with small simple cysts. Interval follow-up with ultrasonography is appropriate for cysts that are most likely to be benign. If malignancy is suspected based on ultrasonography, other imaging, blood testing, or expert opinion, referral to a surgical gynecologist or gynecologic oncologist is recommended. If malignancy is strongly suspected, direct referral to a gynecologic oncologist offers the best chance of survival if cancer is actually present.

Reassure

• When simple cysts are less than 1 cm in postmenopausal women
• When simple cysts are less than 5 cm in premenopausal patients.

Reassess

• With yearly ultrasonography in cases of very low risk
• With repeat ultrasonography in 6 to 12 weeks when the diagnosis is not clear but the cyst is likely benign.

Refer

• To a gynecologist for symptomatic cysts, cysts larger than 6 cm, and cysts that require ancillary testing
• To a gynecologic oncologist for findings worrisome for cancer, such as thick septations, solid areas with flow, ascites, evidence of metastasis, or high cancer antigen 125 levels.

Ovarian cysts, sometimes reported as ovarian masses or adnexal masses, are frequently found incidentally in women who have no symptoms. These cysts can be physiologic (having to do with ovulation) or neoplastic—either benign, borderline (having low malignant potential), or frankly malignant. Thus, these incidental lesions pose many diagnostic challenges to the clinician.

The vast majority of cysts are benign, but a few are malignant, and ovarian malignancies have a notoriously poor survival rate. The diagnosis can only be obtained surgically, as aspiration and biopsy are not definitive and may be harmful. Therefore, the clinician must try to balance the risks of surgery for what may be a benign lesion with the risk of delaying diagnosis of a malignancy.

In this article we provide an approach to evaluating these cysts, with guidance on when the patient can be reassured and when referral is needed.

THE DILEMMA OF OVARIAN CYSTS

Ovarian cysts are common

Premenopausal women can be expected to make at least a small cyst or follicle almost every month. The point prevalence for significant cysts has been reported to be almost 8% in premenopausal women.¹

Surprisingly, the prevalence in postmenopausal women is as high as 14% to 18%, with a yearly incidence of 8%. From 30% to 54% of postmenopausal ovarian cysts persist for years.^2,3

Little is known about the cause of most cysts

Little is known about the cause of most ovarian cysts. Functional or physiologic cysts are thought to be variations in the ovulatory process. They do not seem to be precursors to ovarian cancer.

Most benign neoplastic cysts are also not thought to be precancerous, with the possible exception of the mucinous kind.⁴ Ovarian cysts do not increase the risk of ovarian cancer later in life,^3,9 and removing benign cysts has not been shown to decrease the risk of death from ovarian cancer.¹⁰

Most ovarian cysts and masses are benign

Simple ovarian cysts are much more likely to be benign than malignant. Complex and solid ovarian masses are also more likely to be benign, regardless of menopausal status, but more malignancies are found in this group.

With any kind of mass, the chances of malignancy increase with age. Children and adolescents are not discussed in this article; they should be referred to a specialist.

Ovarian cancer often has a poor prognosis

This “silent” cancer is most often discovered and treated when it has already spread, contributing to a reported 5-year survival rate of only 33% to 46%.^11–13 Ideally, ovarian cancer would be found and removed while still confined to the ovary, when the 5-year survival rate is greater than 90%.

Unfortunately, there does not seem to be a precursor lesion for most ovarian cancers, and there is no good way of finding it in the stage 1 phase, so detecting this cancer before it spreads remains an elusive goal.^11,14

Surgery is required to diagnose difficult cases

There is no perfect test for the preoperative assessment of a cystic ovarian mass. Every method has drawbacks (Table 1).^15–18 Therefore, the National Institutes of Health estimates that 5% to 10% of women in the United States will undergo surgical exploration for an ovarian cyst in their lifetime. Only 13% to 21% of these cysts will be malignant.⁵

ASSESSING AN INCIDENTALLY DISCOVERED OVARIAN MASS

Certain factors in the history, physical examination, and blood work may suggest the cyst is either benign or malignant and may influence the subsequent assessment. However, in most cases, the best next step is to perform transvaginal ultrasonography, which we will discuss later in this paper.

History

Age is a major risk factor for ovarian cancer; the median age at diagnosis is 63 years.⁹ In the reproductive-age group, ovarian cysts are much more likely to be functional than neoplastic. Epithelial cancers are rare before the age of 40, but other cancer types such as borderline, germ cell, and sex cord stromal tumors may occur.¹⁹

In every age group a cyst is more likely to be benign than malignant, although, as noted above, the probability of malignancy increases with age.

Symptoms. Most ovarian cysts, benign or malignant, are asymptomatic and are found only incidentally.

The most commonly reported symptoms are pelvic or lower-abdominal pressure or pain. Acutely painful conditions include ovarian torsion, hemorrhage into the cyst, cyst rupture with or without intra-abdominal hemorrhage, ectopic pregnancy, and pelvic inflammatory disease with tubo-ovarian abscess.

Some patients who have ovarian cancer report vague symptoms such as urinary urgency or frequency, abdominal distention or bloating, and difficulty eating or early satiety.²⁰ Although the positive predictive value of this symptom constellation is only about 1%, its usefulness increases if these symptoms arose recently (within the past year) and occur than 12 days a month.²¹

Family history of ovarian, breast, endometrial, or colon cancer is of particular interest. The greater the number of affected relatives and the closer the degree of relation, the greater the risk; in some cases the relative risk is 40 times greater.²² Breast-ovarian cancer syndromes, hereditary nonpolyposis colorectal cancer syndrome, and family cancer syndrome, as well as extremely high-risk pedigrees such as BRCA1, BRCA2, and Lynch syndrome, all place women at significantly higher risk. Daughters tend to develop cancer at a younger age than their affected mothers.

However, only 10% of ovarian cancers occur in patients who have a family history of it, leaving 90% as sporadic occurrences.

Other history. Factors protective against ovarian cancer include use of oral contraceptives at any time, tubal ligation, hysterectomy, having had children, breastfeeding, a low-fat diet, and possibly use of aspirin and acetaminophen.^23,24

Risk factors for malignancy include advanced age; nulliparity; family history of ovarian or breast cancer; personal history of breast cancer; talc use; asbestos exposure; white ethnicity; pelvic irradiation; smoking; alcohol use; possibly the previous use of fertility drugs, estrogen, or androgen; history of mumps; urban location; early menarche; and late menopause.²⁴

Vital signs. Fever can indicate an infectious process or torsion of the ovary. A sudden onset of low blood pressure or rapid pulse can indicate a hemorrhagic condition such as ectopic pregnancy or ruptured hemorrhagic cyst.

Bimanual pelvic examination is notoriously inaccurate for detecting and characterizing ovarian cysts. In one prospective study, examiners blinded to the reason for surgery evaluated women under anesthesia. The authors concluded that bimanual examination was of limited value even under the best circumstances.¹⁵ Pelvic examination can be even more difficult in patients who are obese, are virginal, have vaginal atrophy, or are in pain.

Useful information that can be obtained through the bimanual examination includes the exact location of pelvic tenderness, the relative firmness of an identified mass, and the existence of nodularity in the posterior cul-de-sac, suggesting advanced ovarian cancer.

Tumor markers

Cancer antigen 125 (CA125) is the most studied and widely used of the ovarian cancer tumor markers. When advanced epithelial ovarian cancer is associated with a markedly elevated level, the value correlates with tumor burden.²⁵

Unfortunately, only about half of early-stage ovarian cancers and 75% to 80% of advanced ovarian cancers express this marker.²⁶ Especially in premenopausal women, there are many pelvic conditions that can falsely elevate CA125. Therefore, its sensitivity and specificity for predicting ovarian cancer are suboptimal. Nevertheless, CA125 is often used to help stratify risk when assessing known ovarian cysts and masses.

The value considered abnormal in postmenopausal women is 35 U/mL or greater, while in premenopausal women the cutoff is less well defined. The lower the cutoff level is set, the more sensitive the test. Recent recommendations advise 50 U/mL or 67 U/mL, rather than the 200 U/mL recommended in the 2002 joint guidelines of the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology.^27,28

However, specificity is likely to be lower with these lower cutoff values. Conditions that can elevate CA125 levels include almost anything that irritates the peritoneum, including pregnancy, menstruation, fibroids, endometriosis, infection, and ovarian hyperstimulation, as well as medical conditions such as liver or renal disease, colitis, diverticulitis, congestive heart failure, diabetes, autoimmune diseases, and ascites.

Following serial CA125 levels may be more sensitive than trying to establish a single cutoff value.²⁹ CA125 should not be used as a screening tool in average-risk women.²⁶

OVA1. Several biomarker panels have been developed and evaluated for risk assessment in women with pelvic masses. OVA1, a proprietary panel of tests (Vermillion; Austin, TX) received US Food and Drug Administration approval in 2009. It includes CA125 and four other proteins, from which it calculates a probability score (high or low) using a proprietary formula.

In prospective studies, OVA1 was more sensitive than clinical assessment or CA125 alone.³⁰ The higher sensitivity and negative predictive value were counterbalanced by a lower specificity and positive predictive value.³¹ Its cost ($650) is not always covered by insurance. OVA1 is not a screening tool.

EVALUATION WITH ULTRASONOGRAPHY

Ultrasonography is the imaging test of choice in assessing adnexal cysts and masses, and therefore it is the best next step after taking a history, performing a physical examination, and obtaining blood work.³² In cases in which an incidental ovarian mass is discovered on computed tomography (CT), further characterization by ultrasonography will likely yield helpful information.

Pelvic ultrasonography can be performed transabdominally or transvaginally. Vaginal ultrasonography gives the clearest images in most patients. Abdominal scanning is indicated for large masses, when vaginal access is difficult (as in virginal patients or those with vaginal atrophy) or when the mass is out of the focal length of the vaginal probe. A full bladder is usually required for the best transabdominal images.

The value of the images obtained depends on the experience of the ultrasonographer and reader and on the equipment. Also, there is currently no widely used standard for reporting the findings³³—descriptions are individualized, leading some authors to recommend that the clinician personally review the films to get the most accurate picture.¹⁹

Size

Size alone cannot be used to distinguish between benign and malignant lesions. Simple cysts up to 10 cm are most likely benign regardless of menopausal status.^2,34 However, in a complex or solid mass, size correlates somewhat with the chance of malignancy, with notable exceptions, such as the famously large sizes of some solid fibromas or mucinous cystadenomas. Also, size may correlate with risk of other complications such as torsion or symptomatic rupture.

Complexity

Simple cysts have clear fluid, thin smooth walls, no loculations or septae, and enhanced through-transmission of echo waves.^32,33

Complexity is described in terms of septations, wall thickness, internal echoes, and solid nodules. Increasing complexity does correlate with increased risk of malignancy.

Worrisome findings

The most worrisome findings are:

• Solid areas that are not hyperechoic, especially when there is blood flow to them
• Thick septations, more than 2 or 3 mm wide, especially if there is blood flow within them
• Excrescences on the inner or outer aspect of a cystic area
• Ascites
• Other pelvic or omental masses.

Benign conditions

Several benign conditions have characteristic complex findings on ultrasonography (Table 2), whereas other findings can be indeterminate (Table 3) or worrisome for malignancy (Table 4).

Hemorrhagic corpus luteum cysts can be complex with an internal reticular pattern due to organizing clot and fibrin strands. A “ring of fire” vascular pattern is often seen around the cyst bed.

Dermoids (mature cystic teratomas) may have hyperechoic elements with acoustic shadowing and no internal Doppler flow. They can have a complex appearance due to fat, hair, and sebum within the cyst. Dermoid cysts have a pathognomonic appearance on CT with a clear fat-fluid level.

Endometriomas classically have a homogeneous “ground-glass” appearance or low-level echoes, without internal color Doppler flow, wall nodules, or other malignant features.

Fibroids may be pedunculated and may appear to be complex or solid adnexal masses.

Hydrosalpinges may present as tortuous tubular-shaped cystic masses. There may be incomplete septations or indentations seen on opposite sides (the “waist” sign).

Paratubal and paraovarian cysts are usually simple round cysts that can be demonstrated as separate from the ovary. Sometimes these appear complex as well.

Peritoneal inclusion cysts, also known as pseudocysts, are seen in patients with intra-abdominal adhesions. Often multiple septations are seen through clear fluid, with the cyst conforming to the shape of other pelvic structures.

Torsion of the ovary may occur with either benign or malignant masses. Torsion can be diagnosed when venous flow is absent on Doppler. The presence of flow, however, doesn’t rule out torsion, as torsion is often intermittent. The twisted ovary is most often enlarged and can have an edematous appearance. Although typically benign, these should be referred for urgent surgical treatment.

Vascularity

Doppler imaging is being extensively studied. The general principle is that malignant masses will be more vascular, with a high-volume, low-resistance pattern of flow. This can result in a pulsatility index of less than 1 or a resistive index of less than 0.4. In practice, however, there is significant overlap between high and low pulsatility indices and resistive indices in benign and malignant cysts. Low resistance can also be found in endometriomas, corpus luteum cysts, inflammatory masses, and vascular benign neoplasms. A normal (high) resistive index does not rule out malignancy.^32,33

One Doppler finding that does seem to correlate with malignancy is the presence of any flow within a solid nodule or wall excrescence.

3D ultrasonography

As the use of 3D ultrasonography increases, studies are yielding different results as to its utility in describing ovarian masses. 3D ultrasonography may be useful in finding centrally located vessels so that Doppler can be applied.³²

OTHER IMAGING

Although ultrasonography is the initial imaging study of choice in the evaluation of adnexal masses owing to its high sensitivity, availability, and low cost, studies have shown that up to 20% of adnexal masses can be reported as indeterminate by ultrasonography (Table 1).

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is emerging as a very valuable tool when ultrasonography is inconclusive or limited.³⁵ Although MRI is very accurate (Table 1), it is not considered a first-line imaging test because it is more expensive, less available, and more inconvenient for the patient than ultrasonography.

MRI provides additional information on the composition of soft-tissue tumors. Usually, MRI is ordered with contrast, unless there are contraindications to it. The radiologist will evaluate morphologic features, signal intensity, and enhancement of solid areas. Techniques such as dynamic contrast-enhanced MRI (following the distribution of contrast material over time), in- and out-of-phase T1 imaging (looking for fat, such as in dermoids), and the newer diffusion-weighted imaging may further improve characterization.

In one study of MRI as second-line imaging, contrast-enhanced MRI contributed to a greater change in the probability of ovarian cancer than did CT, Doppler ultrasonography, or MRI without contrast.³⁶ This may result in a reduction in unnecessary surgeries and in an increase in proper referrals in cases of suspected malignancy.

Computed tomography

Disadvantages of CT include radiation exposure and poor discrimination of soft tissue. It can, however, differentiate fat or calcifications that may be found in dermoids. While CT is not often used to describe an ovarian lesion, it may be used preoperatively to stage an ovarian cancer or to look for a primary intra-abdominal cancer when an ovarian mass may represent metastasis.³²

MANAGING AN INCIDENTAL OVARIAN CYST OR CYSTIC MASS

Combining information from the history, physical examination, imaging, and blood work to assign a level of risk of malignancy is not straightforward. The clinician must weigh several imperfect tests, each with its own sensitivity and specificity, against the background of the individual patient’s likelihood of malignancy. Whereas a 4-cm simple cyst in a premenopausal woman can be assigned to a low-risk category and a complex mass with flow to a solid component in a postmenopausal woman can be assigned to a high-risk category, many lesions are more difficult to assess.

Several systems have been proposed for analyzing data and standardizing risk assessment. There are a number of scoring systems based on ultrasonographic morphology and several mathematical regression models that include menopausal status and tumor markers. But each has drawbacks, and none is definitively superior to expert opinion.^16,17,37,38

A 2012 systematic review and meta-analysis³⁹ calculated sensitivity and specificity for several imaging tests, scoring systems, and blood tumor markers. Some results are presented in Table 1.

The management of an ovarian cyst depends on symptoms, likelihood of torsion or rupture, and the level of concern for malignancy. At the lower-risk end of the spectrum, reassurance or observation over time may be appropriate. A general gynecologist can evaluate indeterminate or symptomatic ovarian cysts. Patients with masses frankly suspicious for malignancy are best referred to a gynecologic oncologist.

Expectant management for low-risk lesions

Low-risk lesions such as simple cysts, endometriomas, and dermoids have a less than 1% chance of malignancy. Most patients who have them require only reassurance or follow-up with serial ultrasonography. Oral contraceptives may prevent new cysts from forming. Aspiration is not recommended.

In 2010, the Society of Radiologists in Ultrasound issued a consensus statement regarding re-imaging of simple ovarian cysts.³³

In premenopausal women, they recommend no further testing for cysts 5 cm or smaller, yearly follow-up for cysts larger than 5 cm and up to and including 7 cm, and MRI or surgical evaluation for cysts larger than 7 cm, as it is difficult to completely image a large cyst with ultrasonography.

In postmenopausal women, if the cyst is 1 cm in diameter or smaller, no further studies need to be done. For simple cysts larger than 1 cm and up to and including 7 cm, yearly re-imaging is recommended. And for cysts larger than 7 cm, MRI or surgery is indicated. The American College of Radiology recommends repeat ultrasonography and CA125 testing for cysts 3 cm and larger but doesn’t specify an interval.³²

A cyst that is otherwise simple but has a single thin septation (< 3 mm) or a small calcification in the wall is almost always benign. Such a cyst should be followed as if it were a simple cyst, as indicated by patient age and cyst size.

There are no official guidelines as to when to stop serial imaging,^22,32 but a recent paper suggested one or two ultrasonographic examinations to confirm size and morphologic stability.¹⁹ Once a lesion has resolved, there is no need for further imaging (Figures 1–3).

Birth control pills for suppression of new cysts. Oral contraceptives do not hasten the resolution of ovarian cysts, according to a 2011 Cochrane review.40 Some practitioners will, nevertheless, prescribe them in an attempt to prevent new cysts from confusing the picture.

Aspiration is not recommended for either diagnosis or treatment. It can only be considered in patients at high risk who are not surgical candidates. Results of cytologic study of specimens obtained by fine-needle aspiration cannot reliably determine the presence or absence of malignancy.⁴¹ There is also a theoretical risk of spreading cancer from an early-stage lesion. A retrospective study has suggested that spillage of cyst contents during surgery in early ovarian cancer is associated with a worse prognosis.⁴²

From a therapeutic point of view, studies have shown the same resolution rate at 6 months for aspirated cysts vs those followed expectantly.⁴³ Another study found a recurrence rate of 25% within 1 year of aspiration.⁴⁴

Referral for medium-risk or indeterminant-risk ovarian masses

Patients who have medium- or indeterminaterisk ovarian masses (Table 3) should be referred to a gynecologist. Further testing will help stratify the risk of malignancy. This can include tumor marker blood tests, MRI, or CT, the addition of Doppler or 3D ultrasonography, serial ultrasonography, or surgical exploration.

If repeat ultrasonography is chosen, the interval will likely be 6 to 12 weeks. Surgery may consist of removing only the cyst itself, or the whole ovary with or without the tube, or sometimes both ovaries. Purely diagnostic laparoscopy is rarely performed, as direct visualization of a lesion is rarely helpful. Frozen section should be employed, and the operating gynecologist should have oncologic backup, since the surgery is performed to rule out malignancy.

In the case of a benign-appearing cyst larger than 6 cm, thought must be given as to whether it is likely to rupture or twist. Rupture of a large cyst can lead to pain and in some cases to hemorrhage. Contents of a ruptured dermoid cyst can cause chemical peritonitis. Torsion of an ovary can result in loss of the ovary through compromised perfusion. A general gynecologist can decide with the patient whether preemptive surgery is indicated.

Operative evaluation for high-risk masses

Patients with high-risk ovarian masses (Table 4) are best referred to a gynecologic oncologist for operative evaluation. If features are seen that indicate malignancy, such as thick septations, solid areas with blood flow, ascites, or other pelvic masses, surgery is indicated. The surgical approach may be through laparoscopy or laparotomy.⁴⁵ It should be noted that even in the face of worrisome features on ultrasonography, many masses turn out to be benign.

In 2011, the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology issued new guidelines recommending oncologic referral of patients with high-risk masses. Elevated CA125, ascites, a nodular or fixed pelvic mass, or evidence of metastasis in postmenopausal women requires oncologic evaluation.²⁶ For premenopausal women, a very elevated CA125, ascites, or metastasis requires referral (Table 4).²⁶

Direct referral to a gynecologic oncologist is underutilized. A recent study found that fewer than half of primary care physicians said that they would refer a classic suspicious case directly to a subspecialist.⁴⁶ It is estimated that only 33% of ovarian cancers are first operated on by a gynecologic oncologist.⁴⁷

A 2011 Cochrane review confirmed a survival benefit for women with cancer who are operated on by gynecologic oncologists primarily, rather than by a general gynecologist and then referred.⁴⁸ A gynecologic oncologist is most likely to perform proper staging and debulking at the time of initial diagnosis.⁴⁹

Special situations require consultation

Ovarian cysts in pregnancy are most often benign,⁵⁰ but malignancy is always a possibility. Functional cysts and dermoids are the most common. These may remain asymptomatic or may rupture or twist or cause difficulty with delivering the baby. Surgical intervention, if needed, should be performed in the second trimester if possible. A multidisciplinary approach and referral to a perinatologist and gynecologic oncologist are advised.

Symptomatic ovarian cysts that may need surgical intervention are the purview of the general gynecologist. If the risk of a surgical emergency is judged to be low, a symptomatic patient may be supported with pain medication and may be managed on an outpatient basis. Immediate surgical consultation is appropriate if the patient appears toxic or in shock. Depending on the clinical picture, there may be a ruptured tubo-ovarian abscess, ruptured ectopic pregnancy, ruptured hemorrhagic cyst, or ovarian torsion, any of which may need immediate surgical intervention.

If a symptomatic mass is highly suspicious for cancer, a gynecologic oncologist should be consulted directly.

WHEN TO REASSURE, REASSESS, REFER

Ovarian masses often pose diagnostic and management dilemmas. Reassurance can be offered to women with small simple cysts. Interval follow-up with ultrasonography is appropriate for cysts that are most likely to be benign. If malignancy is suspected based on ultrasonography, other imaging, blood testing, or expert opinion, referral to a surgical gynecologist or gynecologic oncologist is recommended. If malignancy is strongly suspected, direct referral to a gynecologic oncologist offers the best chance of survival if cancer is actually present.

Reassure

• When simple cysts are less than 1 cm in postmenopausal women
• When simple cysts are less than 5 cm in premenopausal patients.

Reassess

• With yearly ultrasonography in cases of very low risk
• With repeat ultrasonography in 6 to 12 weeks when the diagnosis is not clear but the cyst is likely benign.

Refer

• To a gynecologist for symptomatic cysts, cysts larger than 6 cm, and cysts that require ancillary testing
• To a gynecologic oncologist for findings worrisome for cancer, such as thick septations, solid areas with flow, ascites, evidence of metastasis, or high cancer antigen 125 levels.

Ovarian cysts, sometimes reported as ovarian masses or adnexal masses, are frequently found incidentally in women who have no symptoms. These cysts can be physiologic (having to do with ovulation) or neoplastic—either benign, borderline (having low malignant potential), or frankly malignant. Thus, these incidental lesions pose many diagnostic challenges to the clinician.

The vast majority of cysts are benign, but a few are malignant, and ovarian malignancies have a notoriously poor survival rate. The diagnosis can only be obtained surgically, as aspiration and biopsy are not definitive and may be harmful. Therefore, the clinician must try to balance the risks of surgery for what may be a benign lesion with the risk of delaying diagnosis of a malignancy.

In this article we provide an approach to evaluating these cysts, with guidance on when the patient can be reassured and when referral is needed.

THE DILEMMA OF OVARIAN CYSTS

Ovarian cysts are common

Premenopausal women can be expected to make at least a small cyst or follicle almost every month. The point prevalence for significant cysts has been reported to be almost 8% in premenopausal women.¹

Surprisingly, the prevalence in postmenopausal women is as high as 14% to 18%, with a yearly incidence of 8%. From 30% to 54% of postmenopausal ovarian cysts persist for years.^2,3

Little is known about the cause of most cysts

Little is known about the cause of most ovarian cysts. Functional or physiologic cysts are thought to be variations in the ovulatory process. They do not seem to be precursors to ovarian cancer.

Most benign neoplastic cysts are also not thought to be precancerous, with the possible exception of the mucinous kind.⁴ Ovarian cysts do not increase the risk of ovarian cancer later in life,^3,9 and removing benign cysts has not been shown to decrease the risk of death from ovarian cancer.¹⁰

Most ovarian cysts and masses are benign

Simple ovarian cysts are much more likely to be benign than malignant. Complex and solid ovarian masses are also more likely to be benign, regardless of menopausal status, but more malignancies are found in this group.

With any kind of mass, the chances of malignancy increase with age. Children and adolescents are not discussed in this article; they should be referred to a specialist.

Ovarian cancer often has a poor prognosis

This “silent” cancer is most often discovered and treated when it has already spread, contributing to a reported 5-year survival rate of only 33% to 46%.^11–13 Ideally, ovarian cancer would be found and removed while still confined to the ovary, when the 5-year survival rate is greater than 90%.

Unfortunately, there does not seem to be a precursor lesion for most ovarian cancers, and there is no good way of finding it in the stage 1 phase, so detecting this cancer before it spreads remains an elusive goal.^11,14

Surgery is required to diagnose difficult cases

There is no perfect test for the preoperative assessment of a cystic ovarian mass. Every method has drawbacks (Table 1).^15–18 Therefore, the National Institutes of Health estimates that 5% to 10% of women in the United States will undergo surgical exploration for an ovarian cyst in their lifetime. Only 13% to 21% of these cysts will be malignant.⁵

ASSESSING AN INCIDENTALLY DISCOVERED OVARIAN MASS

Certain factors in the history, physical examination, and blood work may suggest the cyst is either benign or malignant and may influence the subsequent assessment. However, in most cases, the best next step is to perform transvaginal ultrasonography, which we will discuss later in this paper.

History

Age is a major risk factor for ovarian cancer; the median age at diagnosis is 63 years.⁹ In the reproductive-age group, ovarian cysts are much more likely to be functional than neoplastic. Epithelial cancers are rare before the age of 40, but other cancer types such as borderline, germ cell, and sex cord stromal tumors may occur.¹⁹

In every age group a cyst is more likely to be benign than malignant, although, as noted above, the probability of malignancy increases with age.

Symptoms. Most ovarian cysts, benign or malignant, are asymptomatic and are found only incidentally.

The most commonly reported symptoms are pelvic or lower-abdominal pressure or pain. Acutely painful conditions include ovarian torsion, hemorrhage into the cyst, cyst rupture with or without intra-abdominal hemorrhage, ectopic pregnancy, and pelvic inflammatory disease with tubo-ovarian abscess.

Some patients who have ovarian cancer report vague symptoms such as urinary urgency or frequency, abdominal distention or bloating, and difficulty eating or early satiety.²⁰ Although the positive predictive value of this symptom constellation is only about 1%, its usefulness increases if these symptoms arose recently (within the past year) and occur than 12 days a month.²¹

Family history of ovarian, breast, endometrial, or colon cancer is of particular interest. The greater the number of affected relatives and the closer the degree of relation, the greater the risk; in some cases the relative risk is 40 times greater.²² Breast-ovarian cancer syndromes, hereditary nonpolyposis colorectal cancer syndrome, and family cancer syndrome, as well as extremely high-risk pedigrees such as BRCA1, BRCA2, and Lynch syndrome, all place women at significantly higher risk. Daughters tend to develop cancer at a younger age than their affected mothers.

However, only 10% of ovarian cancers occur in patients who have a family history of it, leaving 90% as sporadic occurrences.

Other history. Factors protective against ovarian cancer include use of oral contraceptives at any time, tubal ligation, hysterectomy, having had children, breastfeeding, a low-fat diet, and possibly use of aspirin and acetaminophen.^23,24

Risk factors for malignancy include advanced age; nulliparity; family history of ovarian or breast cancer; personal history of breast cancer; talc use; asbestos exposure; white ethnicity; pelvic irradiation; smoking; alcohol use; possibly the previous use of fertility drugs, estrogen, or androgen; history of mumps; urban location; early menarche; and late menopause.²⁴

Vital signs. Fever can indicate an infectious process or torsion of the ovary. A sudden onset of low blood pressure or rapid pulse can indicate a hemorrhagic condition such as ectopic pregnancy or ruptured hemorrhagic cyst.

Bimanual pelvic examination is notoriously inaccurate for detecting and characterizing ovarian cysts. In one prospective study, examiners blinded to the reason for surgery evaluated women under anesthesia. The authors concluded that bimanual examination was of limited value even under the best circumstances.¹⁵ Pelvic examination can be even more difficult in patients who are obese, are virginal, have vaginal atrophy, or are in pain.

Useful information that can be obtained through the bimanual examination includes the exact location of pelvic tenderness, the relative firmness of an identified mass, and the existence of nodularity in the posterior cul-de-sac, suggesting advanced ovarian cancer.

Tumor markers

Cancer antigen 125 (CA125) is the most studied and widely used of the ovarian cancer tumor markers. When advanced epithelial ovarian cancer is associated with a markedly elevated level, the value correlates with tumor burden.²⁵

Unfortunately, only about half of early-stage ovarian cancers and 75% to 80% of advanced ovarian cancers express this marker.²⁶ Especially in premenopausal women, there are many pelvic conditions that can falsely elevate CA125. Therefore, its sensitivity and specificity for predicting ovarian cancer are suboptimal. Nevertheless, CA125 is often used to help stratify risk when assessing known ovarian cysts and masses.

The value considered abnormal in postmenopausal women is 35 U/mL or greater, while in premenopausal women the cutoff is less well defined. The lower the cutoff level is set, the more sensitive the test. Recent recommendations advise 50 U/mL or 67 U/mL, rather than the 200 U/mL recommended in the 2002 joint guidelines of the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology.^27,28

However, specificity is likely to be lower with these lower cutoff values. Conditions that can elevate CA125 levels include almost anything that irritates the peritoneum, including pregnancy, menstruation, fibroids, endometriosis, infection, and ovarian hyperstimulation, as well as medical conditions such as liver or renal disease, colitis, diverticulitis, congestive heart failure, diabetes, autoimmune diseases, and ascites.

Following serial CA125 levels may be more sensitive than trying to establish a single cutoff value.²⁹ CA125 should not be used as a screening tool in average-risk women.²⁶

OVA1. Several biomarker panels have been developed and evaluated for risk assessment in women with pelvic masses. OVA1, a proprietary panel of tests (Vermillion; Austin, TX) received US Food and Drug Administration approval in 2009. It includes CA125 and four other proteins, from which it calculates a probability score (high or low) using a proprietary formula.

In prospective studies, OVA1 was more sensitive than clinical assessment or CA125 alone.³⁰ The higher sensitivity and negative predictive value were counterbalanced by a lower specificity and positive predictive value.³¹ Its cost ($650) is not always covered by insurance. OVA1 is not a screening tool.

EVALUATION WITH ULTRASONOGRAPHY

Ultrasonography is the imaging test of choice in assessing adnexal cysts and masses, and therefore it is the best next step after taking a history, performing a physical examination, and obtaining blood work.³² In cases in which an incidental ovarian mass is discovered on computed tomography (CT), further characterization by ultrasonography will likely yield helpful information.

Pelvic ultrasonography can be performed transabdominally or transvaginally. Vaginal ultrasonography gives the clearest images in most patients. Abdominal scanning is indicated for large masses, when vaginal access is difficult (as in virginal patients or those with vaginal atrophy) or when the mass is out of the focal length of the vaginal probe. A full bladder is usually required for the best transabdominal images.

The value of the images obtained depends on the experience of the ultrasonographer and reader and on the equipment. Also, there is currently no widely used standard for reporting the findings³³—descriptions are individualized, leading some authors to recommend that the clinician personally review the films to get the most accurate picture.¹⁹

Size

Size alone cannot be used to distinguish between benign and malignant lesions. Simple cysts up to 10 cm are most likely benign regardless of menopausal status.^2,34 However, in a complex or solid mass, size correlates somewhat with the chance of malignancy, with notable exceptions, such as the famously large sizes of some solid fibromas or mucinous cystadenomas. Also, size may correlate with risk of other complications such as torsion or symptomatic rupture.

Complexity

Simple cysts have clear fluid, thin smooth walls, no loculations or septae, and enhanced through-transmission of echo waves.^32,33

Complexity is described in terms of septations, wall thickness, internal echoes, and solid nodules. Increasing complexity does correlate with increased risk of malignancy.

Worrisome findings

The most worrisome findings are:

• Solid areas that are not hyperechoic, especially when there is blood flow to them
• Thick septations, more than 2 or 3 mm wide, especially if there is blood flow within them
• Excrescences on the inner or outer aspect of a cystic area
• Ascites
• Other pelvic or omental masses.

Benign conditions

Several benign conditions have characteristic complex findings on ultrasonography (Table 2), whereas other findings can be indeterminate (Table 3) or worrisome for malignancy (Table 4).

Hemorrhagic corpus luteum cysts can be complex with an internal reticular pattern due to organizing clot and fibrin strands. A “ring of fire” vascular pattern is often seen around the cyst bed.

Dermoids (mature cystic teratomas) may have hyperechoic elements with acoustic shadowing and no internal Doppler flow. They can have a complex appearance due to fat, hair, and sebum within the cyst. Dermoid cysts have a pathognomonic appearance on CT with a clear fat-fluid level.

Endometriomas classically have a homogeneous “ground-glass” appearance or low-level echoes, without internal color Doppler flow, wall nodules, or other malignant features.

Fibroids may be pedunculated and may appear to be complex or solid adnexal masses.

Hydrosalpinges may present as tortuous tubular-shaped cystic masses. There may be incomplete septations or indentations seen on opposite sides (the “waist” sign).

Paratubal and paraovarian cysts are usually simple round cysts that can be demonstrated as separate from the ovary. Sometimes these appear complex as well.

Peritoneal inclusion cysts, also known as pseudocysts, are seen in patients with intra-abdominal adhesions. Often multiple septations are seen through clear fluid, with the cyst conforming to the shape of other pelvic structures.

Torsion of the ovary may occur with either benign or malignant masses. Torsion can be diagnosed when venous flow is absent on Doppler. The presence of flow, however, doesn’t rule out torsion, as torsion is often intermittent. The twisted ovary is most often enlarged and can have an edematous appearance. Although typically benign, these should be referred for urgent surgical treatment.

Vascularity

Doppler imaging is being extensively studied. The general principle is that malignant masses will be more vascular, with a high-volume, low-resistance pattern of flow. This can result in a pulsatility index of less than 1 or a resistive index of less than 0.4. In practice, however, there is significant overlap between high and low pulsatility indices and resistive indices in benign and malignant cysts. Low resistance can also be found in endometriomas, corpus luteum cysts, inflammatory masses, and vascular benign neoplasms. A normal (high) resistive index does not rule out malignancy.^32,33

One Doppler finding that does seem to correlate with malignancy is the presence of any flow within a solid nodule or wall excrescence.

3D ultrasonography

As the use of 3D ultrasonography increases, studies are yielding different results as to its utility in describing ovarian masses. 3D ultrasonography may be useful in finding centrally located vessels so that Doppler can be applied.³²

OTHER IMAGING

Although ultrasonography is the initial imaging study of choice in the evaluation of adnexal masses owing to its high sensitivity, availability, and low cost, studies have shown that up to 20% of adnexal masses can be reported as indeterminate by ultrasonography (Table 1).

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is emerging as a very valuable tool when ultrasonography is inconclusive or limited.³⁵ Although MRI is very accurate (Table 1), it is not considered a first-line imaging test because it is more expensive, less available, and more inconvenient for the patient than ultrasonography.

MRI provides additional information on the composition of soft-tissue tumors. Usually, MRI is ordered with contrast, unless there are contraindications to it. The radiologist will evaluate morphologic features, signal intensity, and enhancement of solid areas. Techniques such as dynamic contrast-enhanced MRI (following the distribution of contrast material over time), in- and out-of-phase T1 imaging (looking for fat, such as in dermoids), and the newer diffusion-weighted imaging may further improve characterization.

In one study of MRI as second-line imaging, contrast-enhanced MRI contributed to a greater change in the probability of ovarian cancer than did CT, Doppler ultrasonography, or MRI without contrast.³⁶ This may result in a reduction in unnecessary surgeries and in an increase in proper referrals in cases of suspected malignancy.

Computed tomography

Disadvantages of CT include radiation exposure and poor discrimination of soft tissue. It can, however, differentiate fat or calcifications that may be found in dermoids. While CT is not often used to describe an ovarian lesion, it may be used preoperatively to stage an ovarian cancer or to look for a primary intra-abdominal cancer when an ovarian mass may represent metastasis.³²

MANAGING AN INCIDENTAL OVARIAN CYST OR CYSTIC MASS

Combining information from the history, physical examination, imaging, and blood work to assign a level of risk of malignancy is not straightforward. The clinician must weigh several imperfect tests, each with its own sensitivity and specificity, against the background of the individual patient’s likelihood of malignancy. Whereas a 4-cm simple cyst in a premenopausal woman can be assigned to a low-risk category and a complex mass with flow to a solid component in a postmenopausal woman can be assigned to a high-risk category, many lesions are more difficult to assess.

Several systems have been proposed for analyzing data and standardizing risk assessment. There are a number of scoring systems based on ultrasonographic morphology and several mathematical regression models that include menopausal status and tumor markers. But each has drawbacks, and none is definitively superior to expert opinion.^16,17,37,38

A 2012 systematic review and meta-analysis³⁹ calculated sensitivity and specificity for several imaging tests, scoring systems, and blood tumor markers. Some results are presented in Table 1.

The management of an ovarian cyst depends on symptoms, likelihood of torsion or rupture, and the level of concern for malignancy. At the lower-risk end of the spectrum, reassurance or observation over time may be appropriate. A general gynecologist can evaluate indeterminate or symptomatic ovarian cysts. Patients with masses frankly suspicious for malignancy are best referred to a gynecologic oncologist.

Expectant management for low-risk lesions

Low-risk lesions such as simple cysts, endometriomas, and dermoids have a less than 1% chance of malignancy. Most patients who have them require only reassurance or follow-up with serial ultrasonography. Oral contraceptives may prevent new cysts from forming. Aspiration is not recommended.

In 2010, the Society of Radiologists in Ultrasound issued a consensus statement regarding re-imaging of simple ovarian cysts.³³

In premenopausal women, they recommend no further testing for cysts 5 cm or smaller, yearly follow-up for cysts larger than 5 cm and up to and including 7 cm, and MRI or surgical evaluation for cysts larger than 7 cm, as it is difficult to completely image a large cyst with ultrasonography.

In postmenopausal women, if the cyst is 1 cm in diameter or smaller, no further studies need to be done. For simple cysts larger than 1 cm and up to and including 7 cm, yearly re-imaging is recommended. And for cysts larger than 7 cm, MRI or surgery is indicated. The American College of Radiology recommends repeat ultrasonography and CA125 testing for cysts 3 cm and larger but doesn’t specify an interval.³²

A cyst that is otherwise simple but has a single thin septation (< 3 mm) or a small calcification in the wall is almost always benign. Such a cyst should be followed as if it were a simple cyst, as indicated by patient age and cyst size.

There are no official guidelines as to when to stop serial imaging,^22,32 but a recent paper suggested one or two ultrasonographic examinations to confirm size and morphologic stability.¹⁹ Once a lesion has resolved, there is no need for further imaging (Figures 1–3).

Birth control pills for suppression of new cysts. Oral contraceptives do not hasten the resolution of ovarian cysts, according to a 2011 Cochrane review.40 Some practitioners will, nevertheless, prescribe them in an attempt to prevent new cysts from confusing the picture.

Aspiration is not recommended for either diagnosis or treatment. It can only be considered in patients at high risk who are not surgical candidates. Results of cytologic study of specimens obtained by fine-needle aspiration cannot reliably determine the presence or absence of malignancy.⁴¹ There is also a theoretical risk of spreading cancer from an early-stage lesion. A retrospective study has suggested that spillage of cyst contents during surgery in early ovarian cancer is associated with a worse prognosis.⁴²

From a therapeutic point of view, studies have shown the same resolution rate at 6 months for aspirated cysts vs those followed expectantly.⁴³ Another study found a recurrence rate of 25% within 1 year of aspiration.⁴⁴

Referral for medium-risk or indeterminant-risk ovarian masses

Patients who have medium- or indeterminaterisk ovarian masses (Table 3) should be referred to a gynecologist. Further testing will help stratify the risk of malignancy. This can include tumor marker blood tests, MRI, or CT, the addition of Doppler or 3D ultrasonography, serial ultrasonography, or surgical exploration.

If repeat ultrasonography is chosen, the interval will likely be 6 to 12 weeks. Surgery may consist of removing only the cyst itself, or the whole ovary with or without the tube, or sometimes both ovaries. Purely diagnostic laparoscopy is rarely performed, as direct visualization of a lesion is rarely helpful. Frozen section should be employed, and the operating gynecologist should have oncologic backup, since the surgery is performed to rule out malignancy.

In the case of a benign-appearing cyst larger than 6 cm, thought must be given as to whether it is likely to rupture or twist. Rupture of a large cyst can lead to pain and in some cases to hemorrhage. Contents of a ruptured dermoid cyst can cause chemical peritonitis. Torsion of an ovary can result in loss of the ovary through compromised perfusion. A general gynecologist can decide with the patient whether preemptive surgery is indicated.

Operative evaluation for high-risk masses

Patients with high-risk ovarian masses (Table 4) are best referred to a gynecologic oncologist for operative evaluation. If features are seen that indicate malignancy, such as thick septations, solid areas with blood flow, ascites, or other pelvic masses, surgery is indicated. The surgical approach may be through laparoscopy or laparotomy.⁴⁵ It should be noted that even in the face of worrisome features on ultrasonography, many masses turn out to be benign.

In 2011, the American Congress of Obstetricians and Gynecologists and the Society of Gynecologic Oncology issued new guidelines recommending oncologic referral of patients with high-risk masses. Elevated CA125, ascites, a nodular or fixed pelvic mass, or evidence of metastasis in postmenopausal women requires oncologic evaluation.²⁶ For premenopausal women, a very elevated CA125, ascites, or metastasis requires referral (Table 4).²⁶

Direct referral to a gynecologic oncologist is underutilized. A recent study found that fewer than half of primary care physicians said that they would refer a classic suspicious case directly to a subspecialist.⁴⁶ It is estimated that only 33% of ovarian cancers are first operated on by a gynecologic oncologist.⁴⁷

A 2011 Cochrane review confirmed a survival benefit for women with cancer who are operated on by gynecologic oncologists primarily, rather than by a general gynecologist and then referred.⁴⁸ A gynecologic oncologist is most likely to perform proper staging and debulking at the time of initial diagnosis.⁴⁹

Special situations require consultation

Ovarian cysts in pregnancy are most often benign,⁵⁰ but malignancy is always a possibility. Functional cysts and dermoids are the most common. These may remain asymptomatic or may rupture or twist or cause difficulty with delivering the baby. Surgical intervention, if needed, should be performed in the second trimester if possible. A multidisciplinary approach and referral to a perinatologist and gynecologic oncologist are advised.

Symptomatic ovarian cysts that may need surgical intervention are the purview of the general gynecologist. If the risk of a surgical emergency is judged to be low, a symptomatic patient may be supported with pain medication and may be managed on an outpatient basis. Immediate surgical consultation is appropriate if the patient appears toxic or in shock. Depending on the clinical picture, there may be a ruptured tubo-ovarian abscess, ruptured ectopic pregnancy, ruptured hemorrhagic cyst, or ovarian torsion, any of which may need immediate surgical intervention.

If a symptomatic mass is highly suspicious for cancer, a gynecologic oncologist should be consulted directly.

WHEN TO REASSURE, REASSESS, REFER

Ovarian masses often pose diagnostic and management dilemmas. Reassurance can be offered to women with small simple cysts. Interval follow-up with ultrasonography is appropriate for cysts that are most likely to be benign. If malignancy is suspected based on ultrasonography, other imaging, blood testing, or expert opinion, referral to a surgical gynecologist or gynecologic oncologist is recommended. If malignancy is strongly suspected, direct referral to a gynecologic oncologist offers the best chance of survival if cancer is actually present.

Reassure

• When simple cysts are less than 1 cm in postmenopausal women
• When simple cysts are less than 5 cm in premenopausal patients.

Reassess

• With yearly ultrasonography in cases of very low risk
• With repeat ultrasonography in 6 to 12 weeks when the diagnosis is not clear but the cyst is likely benign.

Refer

• To a gynecologist for symptomatic cysts, cysts larger than 6 cm, and cysts that require ancillary testing
• To a gynecologic oncologist for findings worrisome for cancer, such as thick septations, solid areas with flow, ascites, evidence of metastasis, or high cancer antigen 125 levels.

A 49-year-old man was referred for evaluation of an abnormal chest radiograph. A 25-pack-year smoker, he had a history of chronic shortness of breath on exertion with occasional coughing and whitish sputum production. He also had a history of hypertension. He had not had hemoptysis, fever, chills, weight loss, or other symptoms, and he had not traveled recently.

On examination, he appeared comfortable. His breath sounds were decreased bilaterally; the rest of his physical examination was normal. His medical history, social history, and review of systems were otherwise unremarkable.

His white blood cell count was 9.4 × 10⁹/L (reference range 4.5–11.0), with a normal differential. His hemoglobin concentration was 166 g/L (140–175).

Pulmonary function testing demonstrated moderate obstruction, with the following values:

• Forced expiratory volume in the first second of expiration/ forced vital capacity 0.65
• Forced expiratory volume in the first second of expiration 2.40 L (72% of predicted)
• Total lung capacity 7.11 L (92% of predicted)
• Diffusing capacity of lung for carbon monoxide 58% of predicted.

He underwent radiography (Figure 1) and computed tomography of the chest (Figure 2).

DIAGNOSIS: INFECTED EMPHYSEMATOUS BULLAE

The patient had infected emphysematous bullae.

The diagnosis can typically be made by the new development of an air-fluid level in a patient known to have preexisting emphysematous bullae.¹ If previous images are not available, the presence of other bullae in a patient with established chronic obstructive pulmonary disease, a thin-walled cavity, and a disproportionate presentation with impressive radiographic findings along with a subtle clinical picture can support the diagnosis.² In most reported cases, patients are not significantly symptomatic or ill.³ The differential diagnosis includes loculated parapneumonic pleural effusion,⁴ lung abscess,⁵ tuberculosis,⁶ and infected pneumatocele.

Since percutaneous aspiration of the bullae has been discouraged,² the causative organism is often not identified. Also, the role of bronchoscopy in the diagnostic evaluation and treatment of infected emphysematous bullae appears to be limited.⁷

Our patient had minimal symptoms and did not appear ill; he had a relatively unremarkable physical examination, no leukocytosis, and negative blood and sputum cultures, suggesting a benign presentation. In addition, chest radiography a few months before this presentation showed multiple large emphysematous bullae (Figure 3). The current chest radiograph suggested multiple thin-walled cavitary lesions with an air-fluid level, which was confirmed on computed tomography.

TREATMENT OF INFECTED EMPHYSEMATOUS BULLAE

Currently, there is no established therapy for infected emphysematous bullae. Because the presentation is usually relatively benign in most case series, conservative treatment with a prolonged course of antibiotics alone seems to be the most appropriate initial course of action. A follow-up evaluation with chest imaging is recommended. On the other hand, in patients with worse symptoms, percutaneous aspiration of the bullae should be considered, as it may guide antibiotic therapy.⁸

We started our patient on clindamycin and scheduled him for follow-up chest imaging in 6 weeks.

A 49-year-old man was referred for evaluation of an abnormal chest radiograph. A 25-pack-year smoker, he had a history of chronic shortness of breath on exertion with occasional coughing and whitish sputum production. He also had a history of hypertension. He had not had hemoptysis, fever, chills, weight loss, or other symptoms, and he had not traveled recently.

On examination, he appeared comfortable. His breath sounds were decreased bilaterally; the rest of his physical examination was normal. His medical history, social history, and review of systems were otherwise unremarkable.

His white blood cell count was 9.4 × 10⁹/L (reference range 4.5–11.0), with a normal differential. His hemoglobin concentration was 166 g/L (140–175).

Pulmonary function testing demonstrated moderate obstruction, with the following values:

• Forced expiratory volume in the first second of expiration/ forced vital capacity 0.65
• Forced expiratory volume in the first second of expiration 2.40 L (72% of predicted)
• Total lung capacity 7.11 L (92% of predicted)
• Diffusing capacity of lung for carbon monoxide 58% of predicted.

He underwent radiography (Figure 1) and computed tomography of the chest (Figure 2).

DIAGNOSIS: INFECTED EMPHYSEMATOUS BULLAE

The patient had infected emphysematous bullae.

The diagnosis can typically be made by the new development of an air-fluid level in a patient known to have preexisting emphysematous bullae.¹ If previous images are not available, the presence of other bullae in a patient with established chronic obstructive pulmonary disease, a thin-walled cavity, and a disproportionate presentation with impressive radiographic findings along with a subtle clinical picture can support the diagnosis.² In most reported cases, patients are not significantly symptomatic or ill.³ The differential diagnosis includes loculated parapneumonic pleural effusion,⁴ lung abscess,⁵ tuberculosis,⁶ and infected pneumatocele.

Since percutaneous aspiration of the bullae has been discouraged,² the causative organism is often not identified. Also, the role of bronchoscopy in the diagnostic evaluation and treatment of infected emphysematous bullae appears to be limited.⁷

Our patient had minimal symptoms and did not appear ill; he had a relatively unremarkable physical examination, no leukocytosis, and negative blood and sputum cultures, suggesting a benign presentation. In addition, chest radiography a few months before this presentation showed multiple large emphysematous bullae (Figure 3). The current chest radiograph suggested multiple thin-walled cavitary lesions with an air-fluid level, which was confirmed on computed tomography.

TREATMENT OF INFECTED EMPHYSEMATOUS BULLAE

Currently, there is no established therapy for infected emphysematous bullae. Because the presentation is usually relatively benign in most case series, conservative treatment with a prolonged course of antibiotics alone seems to be the most appropriate initial course of action. A follow-up evaluation with chest imaging is recommended. On the other hand, in patients with worse symptoms, percutaneous aspiration of the bullae should be considered, as it may guide antibiotic therapy.⁸

We started our patient on clindamycin and scheduled him for follow-up chest imaging in 6 weeks.

A 49-year-old man was referred for evaluation of an abnormal chest radiograph. A 25-pack-year smoker, he had a history of chronic shortness of breath on exertion with occasional coughing and whitish sputum production. He also had a history of hypertension. He had not had hemoptysis, fever, chills, weight loss, or other symptoms, and he had not traveled recently.

On examination, he appeared comfortable. His breath sounds were decreased bilaterally; the rest of his physical examination was normal. His medical history, social history, and review of systems were otherwise unremarkable.

His white blood cell count was 9.4 × 10⁹/L (reference range 4.5–11.0), with a normal differential. His hemoglobin concentration was 166 g/L (140–175).

Pulmonary function testing demonstrated moderate obstruction, with the following values:

• Forced expiratory volume in the first second of expiration/ forced vital capacity 0.65
• Forced expiratory volume in the first second of expiration 2.40 L (72% of predicted)
• Total lung capacity 7.11 L (92% of predicted)
• Diffusing capacity of lung for carbon monoxide 58% of predicted.

He underwent radiography (Figure 1) and computed tomography of the chest (Figure 2).

DIAGNOSIS: INFECTED EMPHYSEMATOUS BULLAE

The patient had infected emphysematous bullae.

The diagnosis can typically be made by the new development of an air-fluid level in a patient known to have preexisting emphysematous bullae.¹ If previous images are not available, the presence of other bullae in a patient with established chronic obstructive pulmonary disease, a thin-walled cavity, and a disproportionate presentation with impressive radiographic findings along with a subtle clinical picture can support the diagnosis.² In most reported cases, patients are not significantly symptomatic or ill.³ The differential diagnosis includes loculated parapneumonic pleural effusion,⁴ lung abscess,⁵ tuberculosis,⁶ and infected pneumatocele.

Since percutaneous aspiration of the bullae has been discouraged,² the causative organism is often not identified. Also, the role of bronchoscopy in the diagnostic evaluation and treatment of infected emphysematous bullae appears to be limited.⁷

Our patient had minimal symptoms and did not appear ill; he had a relatively unremarkable physical examination, no leukocytosis, and negative blood and sputum cultures, suggesting a benign presentation. In addition, chest radiography a few months before this presentation showed multiple large emphysematous bullae (Figure 3). The current chest radiograph suggested multiple thin-walled cavitary lesions with an air-fluid level, which was confirmed on computed tomography.

TREATMENT OF INFECTED EMPHYSEMATOUS BULLAE

Currently, there is no established therapy for infected emphysematous bullae. Because the presentation is usually relatively benign in most case series, conservative treatment with a prolonged course of antibiotics alone seems to be the most appropriate initial course of action. A follow-up evaluation with chest imaging is recommended. On the other hand, in patients with worse symptoms, percutaneous aspiration of the bullae should be considered, as it may guide antibiotic therapy.⁸

We started our patient on clindamycin and scheduled him for follow-up chest imaging in 6 weeks.

A 51-year-old woman presents to the emergency department with dyspnea, which began 4 days ago. She reports no chest pain, palpitations, hemoptysis, fevers, chills, weight loss, recent travel, immobility, or surgery. One week ago she noticed cramping in her right calf, but that has since resolved.

Her history includes hypertension, hypothyroidism, and immune-mediated glomerulonephritis with proteinuria. She is premenopausal. She takes losartan and levothyroxine; she is not taking oral contraceptives or herbal supplements. She is up to date with her cancer screening and has had negative findings on colonoscopy and mammography within the past year.

She has never smoked and she does not drink alcohol or use illicit drugs. Her mother has a history of provoked deep vein thrombosis and colon cancer.

Her temperature is 36.2°C (97.2°F), heart rate 163 beats per minute, blood pressure 158/102 mm Hg, respiratory rate 40 breaths per minute, and oxygen saturation by pulse oximetry 80% while breathing room air, corrected to 94% with oxygen 6 L/min via nasal cannula.

On physical examination, she is sitting upright on a stretcher and appears uncomfortable and anxious. She is awake and able to communicate clearly. Examination of the head, ears, eyes, nose, and throat is unremarkable, with moist mucous membranes. Her lungs are clear to auscultation. Her heart beat is very rapid, with a regular rhythm and an accentuated P2 heart sound. A right parasternal heave can be palpated in addition to a rightwardly displaced point of maximal impulse. The abdomen is normal, with no tenderness or organomegaly. She has no pain, edema, or erythema in the legs or feet, and she has strong, symmetric pulses (2+) in all extremities. The neurologic examination is nonfocal.

Electrocardiography (ECG) done on arrival (Figure 1) reveals supraventricular tachycardia, a normal axis, and nonspecific ST-segment and T-wave abnormalities, findings commonly seen in pulmonary embolism.^1,2 On the other hand, her ECG does not show some of the other signs of right ventricular strain due to pulmonary embolism such as atrial arrhythmias, complete right bundle branch block, or inferior Q-waves.^3,4

In view of her ECG findings and her symptoms of dyspnea, calf pain, tachypnea, tachycardia, and a pronounced P2 heart sound, her physician concludes that she very likely has a pulmonary embolism1 and orders an intravenous infusion of unfractionated heparin to be started immediately.

TESTING FOR PULMONARY EMBOLISM

1. Which of the following would be the best initial diagnostic imaging study to perform in this patient, who has a high pretest probability of pulmonary embolism?

• Multidetector computed tomographic (CT) pulmonary angiography
• Transthoracic echocardiography
• Magnetic resonance imaging
• Lower-extremity duplex ultrasonography
• Pulmonary angiography
• Ventilation-perfusion scintigraphy

Multidetector CT angiography is rapid, noninvasive, and highly sensitive (83%–90%) and specific (96%) for pulmonary embolism.^5,6 In patients such as ours who have a high pretest probability of having the disease, its positive predictive value is 96%.⁵ Therefore, it would be the initial diagnostic study to perform in our patient.

Although transthoracic echocardiography is noninvasive and can detect right ventricular strain in the setting of pulmonary embolism, it may miss half of all pulmonary emboli detected by angiography.^7,8

When technically adequate images are obtained, the combination of magnetic resonance angiography and magnetic resonance venography is very sensitive (92%) and specific (96%) for pulmonary embolism.⁹ However, one-fourth of patients undergoing these studies may have technically inadequate results, so this is not the best choice for diagnosis.⁹

As our patient complained of recent cramping in the right calf, lower-extremity duplex ultrasonography would be a reasonable test to screen for acute deep vein thrombosis as the source of pulmonary embolism. However, given her worrisome vital signs and impending hemodynamic collapse, CT pulmonary angiography would be a better initial test as it may guide more aggressive therapy. Furthermore, even if ultrasonography showed no evidence of deep vein thrombosis, clinical suspicion for pulmonary embolism would remain high enough that therapeutic anticoagulation would be continued until further testing ruled out this diagnosis.

Pulmonary angiography is the gold-standard test for pulmonary embolism. However, it is time-consuming, expensive, and invasive and so is not usually done unless the diagnosis cannot be made with other imaging studies.

Ventilation-perfusion scintigraphy is an established and safe diagnostic test for pulmonary embolism. It is particularly helpful in patients who have renal dysfunction or contrast allergy. The sensitivity of a high-probability scan is 78%, while the specificity of a very-low-probability scan is 97%.10 However, this study is often nondiagnostic (in 26.5% of cases),10 and further imaging may be required.

RESULTS OF CT ANGIOGRAPHY

Our patient undergoes CT angiography, which reveals multiple bilateral pulmonary emboli and right ventricular enlargement (Figure 2). Transthoracic echocardiography shows dilation of the right ventricle, with severely reduced systolic function, an underfilled and hyperdynamic left ventricle (ejection fraction 75%), and moderate tricuspid valve regurgitation. Her right ventricular systolic pressure is estimated to be 47 mm Hg.

Doppler ultrasonography of the legs reveals an occlusive thrombus within the right small saphenous vein that bulges and extends into the right popliteal vein. Also noted is a nonocclusive thrombus in the upper right popliteal vein that likely originated from the thrombus in the small saphenous vein.

Initial laboratory testing (Table 1) shows elevations of the cardiac enzymes troponin T and N-terminal pro-B-type natriuretic peptide (NT-pro-BNP).

ESTIMATING PROGNOSIS IN PULMONARY EMBOLISM

2. Which of the following laboratory results at presentation is independently associated with a worse outcome in patients with pulmonary embolism?

• Elevated NT-pro-BNP
• Hypercalcemia
• Thrombocytosis
• Hypernatremia
• Elevated procalcitonin

The Pulmonary Embolism Severity Index¹¹ and the Simplified Pulmonary Embolism Severity Index¹² (Table 2) are clinical calculators that help predict 30-day risk of death in patients with pulmonary embolism. Our patient’s Pulmonary Embolism Severity Index score is 60, indicating a very low risk, but her simplified severity index score is 2, indicating a high risk.

A shock index score (the heart rate divided by the systolic blood pressure) greater than 1 is also a sensitive measure of risk.¹³ (Our patient’s shock index score is 1.03.) Although the simplified version is more accurate,¹⁴ the shock index is also helpful when deciding whether patients with suspected pulmonary embolism should receive early fibrinolysis.¹⁵

In a large registry of patients with confirmed pulmonary embolism, risk factors for death were age greater than 70, cancer, clinical congestive heart failure, chronic obstructive pulmonary disease, systolic blood pressure lower than 90 mm Hg, respiratory rate less than 20 per minute, and right ventricular hypokinesis.¹⁶ Right ventricular dysfunction progressing to right ventricular failure and cardiogenic shock is the most common cause of death in patients with pulmonary embolism.^16–18

Post hoc analysis has also shown that elevations of the biomarkers BNP, NT-pro-BNP, and cardiac troponins I and T are associated with a prolonged hospital course and a higher risk of death within 30 days.¹⁹ Interestingly, a recent retrospective analysis found hyponatremia to be an independent risk factor for death in the short term.²⁰

Thrombocytopenia, not thrombocytosis, is associated with worse outcomes in patients with pulmonary embolism.¹⁶ Procalcitonin is elevated in bacterial pneumonia but is normal in pulmonary embolism and so may be helpful in differentiating between the two.^21,22 Hypernatremia, hypercalcemia, and elevated procalcitonin have not been shown to be independently associated with worse outcomes in acute pulmonary embolism.

Thus, of the answer choices shown above, elevated NT-pro-BNP is the correct answer.

Pulmonary embolism is often stratified as massive, submassive, or low-risk, reflecting the severity and the degree of cardiovascular collapse. The treatment depends on the classification.

Pulmonary embolism is classified as massive if the patient has a cardiac arrest or a systolic blood pressure lower than 90 mm Hg for more than 15 minutes.²³ Nearly half of patients in this category die.²⁴

Pulmonary embolism is submassive if the patient has systolic pressure greater than 90 mm Hg but has right ventricular dysfunction, as evidenced by physical examination, elevated cardiac biomarkers, electrocardiography, transthoracic echocardiography, or computed tomography. The death rate is as high as 15%.²⁴

Pulmonary embolism in a normotensive patient with no right ventricular dysfunction is defined as low-risk.

Our patient so far

Our patient has bilateral pulmonary emboli, most likely originating from a deep vein thrombosis in her right lower leg. Her pulmonary embolism would be classified as submassive, as her systolic pressure is greater than 90 mm Hg and right ventricular dysfunction—significant right ventricular strain—was noted on both transthoracic echocardiography and computed tomography. Also, cardiac biomarkers are elevated, and the physical examination revealed a prominent P2 sound and right parasternal heave, also suggestive of right ventricular dysfunction.

Now, 6 hours have passed, and even though she has been receiving intravenous heparin during this time, her shock index remains greater than 1, indicating hemodynamic instability. Her pulse rate is still markedly high—over 160 bpm—and she still appears quite anxious and uncomfortable.

HOW SHOULD THIS PATIENT BE TREATED?

3. Which of the following is the most appropriate treatment for this patient?

• Start warfarin immediately while bridging with unfractionated heparin, low-molecular-weight heparin, or fondaparinux
• Start fibrinolysis with alteplase
• Give metoprolol intravenously to control her heart rate
• Start dabigatran immediately while bridging with unfractionated heparin
• Place an inferior vena cava filter
• Consult cardiothoracic surgery for emergency pulmonary embolectomy

All patients with confirmed pulmonary embolism and no contraindications to anticoagulation should begin treatment with low-molecular-weight heparin, unfractionated heparin, or fondaparinux.²³ In addition, this therapy should be started empirically while the patient is still undergoing diagnostic testing if the pretest probability of pulmonary embolism is intermediate or high.²³

Warfarin is indicated for all patients with pulmonary embolism who do not have contraindications to it (Table 3). If unfractionated heparin, low-molecular-weight heparin, or fondaparinux has not already been started, it should be started at the same time as warfarin and should be continued until the international normalized ratio (INR) is within the therapeutic range.

Fibrinolysis. Treatment with a fibrinolytic agent in addition to heparin results in faster improvement of right ventricular function and pulmonary perfusion than with heparin alone.²⁵ It may also decrease the incidence of pulmonary hypertension secondary to chronic thromboembolic disease.²⁶ It should be considered in patients with massive pulmonary embolism.²³

Whether fibrinolysis is appropriate for all patients with submassive pulmonary embolism remains controversial. Currently, it is not recommended for minor right ventricular dysfunction or myocardial necrosis if the patient has no signs of overt clinical decline.²³ The Pulmonary Embolism Thrombolysis trial²⁷ is an ongoing prospective randomized comparison of tenecteplase in a single bolus plus heparin vs heparin alone in normotensive patients with submassive pulmonary embolism, such as our patient. This trial may elucidate the benefit of fibrinolytic therapy in patients with submassive pulmonary embolism.

Patients at low risk are generally treated with heparin and warfarin anticoagulation alone. Fibrinolysis is not recommended in these patients, as the risk of bleeding outweighs the potential benefits.²³

Metoprolol may not be advisable for our patient, as her tachycardia is likely compensatory, and beta-blocker therapy could blunt this compensatory response, leading to inadequate systemic perfusion.

Dabigatran is an oral direct thrombin inhibitor that does not require laboratory monitoring. It is currently approved for the prevention of stroke in patients with atrial fibrillation. It has been shown to be as effective as warfarin in the treatment of acute venous thromboembolism²⁸ and may be a viable option in the future, but as of this writing it has not yet been approved in the United States for this indication. Furthermore, dabigatran inhibits thrombin immediately, so continued heparin bridging would not be necessary.

An inferior vena cava filter may prevent recurrent pulmonary embolism for patients who have absolute contraindications to anticoagulation, most significantly in the short term, ie, in the first few weeks after placement. However, these devices have not yet been shown to improve long-term mortality rates.

Embolectomy, percutaneous or surgical, is also an option. For patients in whom thrombolytic therapy is not effective, “rescue” surgical embolectomy has been associated with better outcomes compared with secondary thrombolysis and so should be considered.²⁹

Back to our patient

An intravenous infusion of alteplase is started, and the patient’s tachycardia improves. Her oxygen requirements normalize, and she is transferred to the general medical floor the next day. She receives subcutaneous dalteparin as a bridge therapy, and warfarin is titrated to a goal INR of 2.0 to 3.0. Because of the acute deep vein thrombosis in her right lower leg, she is instructed to wear knee-high fitted compression hose for primary prevention of postphlebitic syndrome.

HOW LONG TO TREAT? IS GENETIC TESTING INDICATED?

Patients with a first episode of unprovoked venous thromboembolism should receive oral anticoagulants for 6 months, while those with recurrent unprovoked venous thromboembolism require lifelong oral anticoagulation.²³

Whether to test for inherited thrombophilia after a first episode of venous thromboembolism to guide the duration of anticoagulation is controversial.³⁰ Indiscriminate testing has not been recommended in these patients,³¹ but the American College of Medical Genetics recommends genetic screening for factor V Leiden in patients who have an unprovoked incident of venous thromboembolism before age 50.³²

No randomized controlled trial has assessed whether thrombophilia testing decreases the recurrence rate of venous thromboembolism.³³ One uncontrolled study suggested that testing for inherited thrombophilias in patients with a first episode does not affect the risk of recurrence.³⁴ Testing is costly and may cause psychological distress for patients and family members.

Our patient is discharged home on warfarin for 6 months with subsequent follow-up evaluation in the thrombophilia clinic.

WHEN SHOULD WARFARIN BE RESTARTED?

4. If our patient were to discontinue oral anticoagulation in 6 months, which of the following, if present 1 month afterwards, would be a reason to restart oral anticoagulation?

• Elevated serum cotinine
• Positive pregnancy test
• Elevated follicle-stimulating hormone and luteinizing hormone and low estradiol levels
• Elevated D-dimer

Cotinine is a nicotine metabolite, and serum levels are elevated in smokers. Smoking and pregnancy both increase the risk of venous thromboembolism. However, smoking or pregnancy alone would not be a reason to increase the duration of anticoagulation.

Warfarin is contraindicated in pregnancy because of its teratogenic effects, particularly in the first trimester. Follicle-stimulating hormone and luteinizing hormone levels increase in response to decreased estradiol at menopause. Postmenopausal women are not at increased risk of venous thromboembolism unless they are taking oral estrogen hormone replacement therapy.

An elevated D-dimer level 1 month after stopping of oral anticoagulation has been associated with a higher rate of recurrence of venous thromboembolism, which is reduced by a resumption of anticoagulation.³⁵ Therefore, consideration should be given to resuming anticoagulation in patients who have an elevated D-dimer level.

TAKE-HOME POINTS

It is important to distinguish between massive, submassive, and low-risk pulmonary embolism, since each type has a different treatment and prognosis.

Fibrinolytic therapy is indicated in patients with massive pulmonary embolism when no contraindication is present, whereas it is not indicated in those with low-risk pulmonary embolism.

Management of submassive pulmonary embolism continues to be an area of considerable debate. Current American Heart Association guidelines recommend consideration of fibrinolysis initially in patients with submassive acute pulmonary embolism if there is concurrent clinical evidence of adverse prognosis—eg, new hemodynamic instability, worsening respiratory insufficiency, severe right ventricular dysfunction, or major myocardial necrosis.²³ On the other hand, the American College of Chest Physicians recommends against initial systemic fibrinolysis in submassive acute pulmonary embolism based on biomarkers or findings on ECG, transthoracic echocardiography, or CT, recommending it only in therapeutically anticoagulated patients deemed to be at high risk of hypotension.³⁶

Since the optimal treatment of submassive pulmonary embolism is still not known, it is important that clinicians remain up to date on the evidence and guidelines.

A 51-year-old woman presents to the emergency department with dyspnea, which began 4 days ago. She reports no chest pain, palpitations, hemoptysis, fevers, chills, weight loss, recent travel, immobility, or surgery. One week ago she noticed cramping in her right calf, but that has since resolved.

Her history includes hypertension, hypothyroidism, and immune-mediated glomerulonephritis with proteinuria. She is premenopausal. She takes losartan and levothyroxine; she is not taking oral contraceptives or herbal supplements. She is up to date with her cancer screening and has had negative findings on colonoscopy and mammography within the past year.

She has never smoked and she does not drink alcohol or use illicit drugs. Her mother has a history of provoked deep vein thrombosis and colon cancer.

Her temperature is 36.2°C (97.2°F), heart rate 163 beats per minute, blood pressure 158/102 mm Hg, respiratory rate 40 breaths per minute, and oxygen saturation by pulse oximetry 80% while breathing room air, corrected to 94% with oxygen 6 L/min via nasal cannula.

On physical examination, she is sitting upright on a stretcher and appears uncomfortable and anxious. She is awake and able to communicate clearly. Examination of the head, ears, eyes, nose, and throat is unremarkable, with moist mucous membranes. Her lungs are clear to auscultation. Her heart beat is very rapid, with a regular rhythm and an accentuated P2 heart sound. A right parasternal heave can be palpated in addition to a rightwardly displaced point of maximal impulse. The abdomen is normal, with no tenderness or organomegaly. She has no pain, edema, or erythema in the legs or feet, and she has strong, symmetric pulses (2+) in all extremities. The neurologic examination is nonfocal.

Electrocardiography (ECG) done on arrival (Figure 1) reveals supraventricular tachycardia, a normal axis, and nonspecific ST-segment and T-wave abnormalities, findings commonly seen in pulmonary embolism.^1,2 On the other hand, her ECG does not show some of the other signs of right ventricular strain due to pulmonary embolism such as atrial arrhythmias, complete right bundle branch block, or inferior Q-waves.^3,4

In view of her ECG findings and her symptoms of dyspnea, calf pain, tachypnea, tachycardia, and a pronounced P2 heart sound, her physician concludes that she very likely has a pulmonary embolism1 and orders an intravenous infusion of unfractionated heparin to be started immediately.

TESTING FOR PULMONARY EMBOLISM

1. Which of the following would be the best initial diagnostic imaging study to perform in this patient, who has a high pretest probability of pulmonary embolism?

• Multidetector computed tomographic (CT) pulmonary angiography
• Transthoracic echocardiography
• Magnetic resonance imaging
• Lower-extremity duplex ultrasonography
• Pulmonary angiography
• Ventilation-perfusion scintigraphy

Multidetector CT angiography is rapid, noninvasive, and highly sensitive (83%–90%) and specific (96%) for pulmonary embolism.^5,6 In patients such as ours who have a high pretest probability of having the disease, its positive predictive value is 96%.⁵ Therefore, it would be the initial diagnostic study to perform in our patient.

Although transthoracic echocardiography is noninvasive and can detect right ventricular strain in the setting of pulmonary embolism, it may miss half of all pulmonary emboli detected by angiography.^7,8

When technically adequate images are obtained, the combination of magnetic resonance angiography and magnetic resonance venography is very sensitive (92%) and specific (96%) for pulmonary embolism.⁹ However, one-fourth of patients undergoing these studies may have technically inadequate results, so this is not the best choice for diagnosis.⁹

As our patient complained of recent cramping in the right calf, lower-extremity duplex ultrasonography would be a reasonable test to screen for acute deep vein thrombosis as the source of pulmonary embolism. However, given her worrisome vital signs and impending hemodynamic collapse, CT pulmonary angiography would be a better initial test as it may guide more aggressive therapy. Furthermore, even if ultrasonography showed no evidence of deep vein thrombosis, clinical suspicion for pulmonary embolism would remain high enough that therapeutic anticoagulation would be continued until further testing ruled out this diagnosis.

Pulmonary angiography is the gold-standard test for pulmonary embolism. However, it is time-consuming, expensive, and invasive and so is not usually done unless the diagnosis cannot be made with other imaging studies.

Ventilation-perfusion scintigraphy is an established and safe diagnostic test for pulmonary embolism. It is particularly helpful in patients who have renal dysfunction or contrast allergy. The sensitivity of a high-probability scan is 78%, while the specificity of a very-low-probability scan is 97%.10 However, this study is often nondiagnostic (in 26.5% of cases),10 and further imaging may be required.

RESULTS OF CT ANGIOGRAPHY

Our patient undergoes CT angiography, which reveals multiple bilateral pulmonary emboli and right ventricular enlargement (Figure 2). Transthoracic echocardiography shows dilation of the right ventricle, with severely reduced systolic function, an underfilled and hyperdynamic left ventricle (ejection fraction 75%), and moderate tricuspid valve regurgitation. Her right ventricular systolic pressure is estimated to be 47 mm Hg.

Doppler ultrasonography of the legs reveals an occlusive thrombus within the right small saphenous vein that bulges and extends into the right popliteal vein. Also noted is a nonocclusive thrombus in the upper right popliteal vein that likely originated from the thrombus in the small saphenous vein.

Initial laboratory testing (Table 1) shows elevations of the cardiac enzymes troponin T and N-terminal pro-B-type natriuretic peptide (NT-pro-BNP).

ESTIMATING PROGNOSIS IN PULMONARY EMBOLISM

2. Which of the following laboratory results at presentation is independently associated with a worse outcome in patients with pulmonary embolism?

• Elevated NT-pro-BNP
• Hypercalcemia
• Thrombocytosis
• Hypernatremia
• Elevated procalcitonin

The Pulmonary Embolism Severity Index¹¹ and the Simplified Pulmonary Embolism Severity Index¹² (Table 2) are clinical calculators that help predict 30-day risk of death in patients with pulmonary embolism. Our patient’s Pulmonary Embolism Severity Index score is 60, indicating a very low risk, but her simplified severity index score is 2, indicating a high risk.

A shock index score (the heart rate divided by the systolic blood pressure) greater than 1 is also a sensitive measure of risk.¹³ (Our patient’s shock index score is 1.03.) Although the simplified version is more accurate,¹⁴ the shock index is also helpful when deciding whether patients with suspected pulmonary embolism should receive early fibrinolysis.¹⁵

In a large registry of patients with confirmed pulmonary embolism, risk factors for death were age greater than 70, cancer, clinical congestive heart failure, chronic obstructive pulmonary disease, systolic blood pressure lower than 90 mm Hg, respiratory rate less than 20 per minute, and right ventricular hypokinesis.¹⁶ Right ventricular dysfunction progressing to right ventricular failure and cardiogenic shock is the most common cause of death in patients with pulmonary embolism.^16–18

Post hoc analysis has also shown that elevations of the biomarkers BNP, NT-pro-BNP, and cardiac troponins I and T are associated with a prolonged hospital course and a higher risk of death within 30 days.¹⁹ Interestingly, a recent retrospective analysis found hyponatremia to be an independent risk factor for death in the short term.²⁰

Thrombocytopenia, not thrombocytosis, is associated with worse outcomes in patients with pulmonary embolism.¹⁶ Procalcitonin is elevated in bacterial pneumonia but is normal in pulmonary embolism and so may be helpful in differentiating between the two.^21,22 Hypernatremia, hypercalcemia, and elevated procalcitonin have not been shown to be independently associated with worse outcomes in acute pulmonary embolism.

Thus, of the answer choices shown above, elevated NT-pro-BNP is the correct answer.

Pulmonary embolism is often stratified as massive, submassive, or low-risk, reflecting the severity and the degree of cardiovascular collapse. The treatment depends on the classification.

Pulmonary embolism is classified as massive if the patient has a cardiac arrest or a systolic blood pressure lower than 90 mm Hg for more than 15 minutes.²³ Nearly half of patients in this category die.²⁴

Pulmonary embolism is submassive if the patient has systolic pressure greater than 90 mm Hg but has right ventricular dysfunction, as evidenced by physical examination, elevated cardiac biomarkers, electrocardiography, transthoracic echocardiography, or computed tomography. The death rate is as high as 15%.²⁴

Pulmonary embolism in a normotensive patient with no right ventricular dysfunction is defined as low-risk.

Our patient so far

Our patient has bilateral pulmonary emboli, most likely originating from a deep vein thrombosis in her right lower leg. Her pulmonary embolism would be classified as submassive, as her systolic pressure is greater than 90 mm Hg and right ventricular dysfunction—significant right ventricular strain—was noted on both transthoracic echocardiography and computed tomography. Also, cardiac biomarkers are elevated, and the physical examination revealed a prominent P2 sound and right parasternal heave, also suggestive of right ventricular dysfunction.

Now, 6 hours have passed, and even though she has been receiving intravenous heparin during this time, her shock index remains greater than 1, indicating hemodynamic instability. Her pulse rate is still markedly high—over 160 bpm—and she still appears quite anxious and uncomfortable.

HOW SHOULD THIS PATIENT BE TREATED?

3. Which of the following is the most appropriate treatment for this patient?

• Start warfarin immediately while bridging with unfractionated heparin, low-molecular-weight heparin, or fondaparinux
• Start fibrinolysis with alteplase
• Give metoprolol intravenously to control her heart rate
• Start dabigatran immediately while bridging with unfractionated heparin
• Place an inferior vena cava filter
• Consult cardiothoracic surgery for emergency pulmonary embolectomy

All patients with confirmed pulmonary embolism and no contraindications to anticoagulation should begin treatment with low-molecular-weight heparin, unfractionated heparin, or fondaparinux.²³ In addition, this therapy should be started empirically while the patient is still undergoing diagnostic testing if the pretest probability of pulmonary embolism is intermediate or high.²³

Warfarin is indicated for all patients with pulmonary embolism who do not have contraindications to it (Table 3). If unfractionated heparin, low-molecular-weight heparin, or fondaparinux has not already been started, it should be started at the same time as warfarin and should be continued until the international normalized ratio (INR) is within the therapeutic range.

Fibrinolysis. Treatment with a fibrinolytic agent in addition to heparin results in faster improvement of right ventricular function and pulmonary perfusion than with heparin alone.²⁵ It may also decrease the incidence of pulmonary hypertension secondary to chronic thromboembolic disease.²⁶ It should be considered in patients with massive pulmonary embolism.²³

Whether fibrinolysis is appropriate for all patients with submassive pulmonary embolism remains controversial. Currently, it is not recommended for minor right ventricular dysfunction or myocardial necrosis if the patient has no signs of overt clinical decline.²³ The Pulmonary Embolism Thrombolysis trial²⁷ is an ongoing prospective randomized comparison of tenecteplase in a single bolus plus heparin vs heparin alone in normotensive patients with submassive pulmonary embolism, such as our patient. This trial may elucidate the benefit of fibrinolytic therapy in patients with submassive pulmonary embolism.

Patients at low risk are generally treated with heparin and warfarin anticoagulation alone. Fibrinolysis is not recommended in these patients, as the risk of bleeding outweighs the potential benefits.²³

Metoprolol may not be advisable for our patient, as her tachycardia is likely compensatory, and beta-blocker therapy could blunt this compensatory response, leading to inadequate systemic perfusion.

Dabigatran is an oral direct thrombin inhibitor that does not require laboratory monitoring. It is currently approved for the prevention of stroke in patients with atrial fibrillation. It has been shown to be as effective as warfarin in the treatment of acute venous thromboembolism²⁸ and may be a viable option in the future, but as of this writing it has not yet been approved in the United States for this indication. Furthermore, dabigatran inhibits thrombin immediately, so continued heparin bridging would not be necessary.

An inferior vena cava filter may prevent recurrent pulmonary embolism for patients who have absolute contraindications to anticoagulation, most significantly in the short term, ie, in the first few weeks after placement. However, these devices have not yet been shown to improve long-term mortality rates.

Embolectomy, percutaneous or surgical, is also an option. For patients in whom thrombolytic therapy is not effective, “rescue” surgical embolectomy has been associated with better outcomes compared with secondary thrombolysis and so should be considered.²⁹

Back to our patient

An intravenous infusion of alteplase is started, and the patient’s tachycardia improves. Her oxygen requirements normalize, and she is transferred to the general medical floor the next day. She receives subcutaneous dalteparin as a bridge therapy, and warfarin is titrated to a goal INR of 2.0 to 3.0. Because of the acute deep vein thrombosis in her right lower leg, she is instructed to wear knee-high fitted compression hose for primary prevention of postphlebitic syndrome.

HOW LONG TO TREAT? IS GENETIC TESTING INDICATED?

Patients with a first episode of unprovoked venous thromboembolism should receive oral anticoagulants for 6 months, while those with recurrent unprovoked venous thromboembolism require lifelong oral anticoagulation.²³

Whether to test for inherited thrombophilia after a first episode of venous thromboembolism to guide the duration of anticoagulation is controversial.³⁰ Indiscriminate testing has not been recommended in these patients,³¹ but the American College of Medical Genetics recommends genetic screening for factor V Leiden in patients who have an unprovoked incident of venous thromboembolism before age 50.³²

No randomized controlled trial has assessed whether thrombophilia testing decreases the recurrence rate of venous thromboembolism.³³ One uncontrolled study suggested that testing for inherited thrombophilias in patients with a first episode does not affect the risk of recurrence.³⁴ Testing is costly and may cause psychological distress for patients and family members.

Our patient is discharged home on warfarin for 6 months with subsequent follow-up evaluation in the thrombophilia clinic.

WHEN SHOULD WARFARIN BE RESTARTED?

4. If our patient were to discontinue oral anticoagulation in 6 months, which of the following, if present 1 month afterwards, would be a reason to restart oral anticoagulation?

• Elevated serum cotinine
• Positive pregnancy test
• Elevated follicle-stimulating hormone and luteinizing hormone and low estradiol levels
• Elevated D-dimer

Cotinine is a nicotine metabolite, and serum levels are elevated in smokers. Smoking and pregnancy both increase the risk of venous thromboembolism. However, smoking or pregnancy alone would not be a reason to increase the duration of anticoagulation.

Warfarin is contraindicated in pregnancy because of its teratogenic effects, particularly in the first trimester. Follicle-stimulating hormone and luteinizing hormone levels increase in response to decreased estradiol at menopause. Postmenopausal women are not at increased risk of venous thromboembolism unless they are taking oral estrogen hormone replacement therapy.

An elevated D-dimer level 1 month after stopping of oral anticoagulation has been associated with a higher rate of recurrence of venous thromboembolism, which is reduced by a resumption of anticoagulation.³⁵ Therefore, consideration should be given to resuming anticoagulation in patients who have an elevated D-dimer level.

TAKE-HOME POINTS

It is important to distinguish between massive, submassive, and low-risk pulmonary embolism, since each type has a different treatment and prognosis.

Fibrinolytic therapy is indicated in patients with massive pulmonary embolism when no contraindication is present, whereas it is not indicated in those with low-risk pulmonary embolism.

Management of submassive pulmonary embolism continues to be an area of considerable debate. Current American Heart Association guidelines recommend consideration of fibrinolysis initially in patients with submassive acute pulmonary embolism if there is concurrent clinical evidence of adverse prognosis—eg, new hemodynamic instability, worsening respiratory insufficiency, severe right ventricular dysfunction, or major myocardial necrosis.²³ On the other hand, the American College of Chest Physicians recommends against initial systemic fibrinolysis in submassive acute pulmonary embolism based on biomarkers or findings on ECG, transthoracic echocardiography, or CT, recommending it only in therapeutically anticoagulated patients deemed to be at high risk of hypotension.³⁶

Since the optimal treatment of submassive pulmonary embolism is still not known, it is important that clinicians remain up to date on the evidence and guidelines.

A 51-year-old woman presents to the emergency department with dyspnea, which began 4 days ago. She reports no chest pain, palpitations, hemoptysis, fevers, chills, weight loss, recent travel, immobility, or surgery. One week ago she noticed cramping in her right calf, but that has since resolved.

Her history includes hypertension, hypothyroidism, and immune-mediated glomerulonephritis with proteinuria. She is premenopausal. She takes losartan and levothyroxine; she is not taking oral contraceptives or herbal supplements. She is up to date with her cancer screening and has had negative findings on colonoscopy and mammography within the past year.

She has never smoked and she does not drink alcohol or use illicit drugs. Her mother has a history of provoked deep vein thrombosis and colon cancer.

Her temperature is 36.2°C (97.2°F), heart rate 163 beats per minute, blood pressure 158/102 mm Hg, respiratory rate 40 breaths per minute, and oxygen saturation by pulse oximetry 80% while breathing room air, corrected to 94% with oxygen 6 L/min via nasal cannula.

On physical examination, she is sitting upright on a stretcher and appears uncomfortable and anxious. She is awake and able to communicate clearly. Examination of the head, ears, eyes, nose, and throat is unremarkable, with moist mucous membranes. Her lungs are clear to auscultation. Her heart beat is very rapid, with a regular rhythm and an accentuated P2 heart sound. A right parasternal heave can be palpated in addition to a rightwardly displaced point of maximal impulse. The abdomen is normal, with no tenderness or organomegaly. She has no pain, edema, or erythema in the legs or feet, and she has strong, symmetric pulses (2+) in all extremities. The neurologic examination is nonfocal.

Electrocardiography (ECG) done on arrival (Figure 1) reveals supraventricular tachycardia, a normal axis, and nonspecific ST-segment and T-wave abnormalities, findings commonly seen in pulmonary embolism.^1,2 On the other hand, her ECG does not show some of the other signs of right ventricular strain due to pulmonary embolism such as atrial arrhythmias, complete right bundle branch block, or inferior Q-waves.^3,4

In view of her ECG findings and her symptoms of dyspnea, calf pain, tachypnea, tachycardia, and a pronounced P2 heart sound, her physician concludes that she very likely has a pulmonary embolism1 and orders an intravenous infusion of unfractionated heparin to be started immediately.

TESTING FOR PULMONARY EMBOLISM

1. Which of the following would be the best initial diagnostic imaging study to perform in this patient, who has a high pretest probability of pulmonary embolism?

• Multidetector computed tomographic (CT) pulmonary angiography
• Transthoracic echocardiography
• Magnetic resonance imaging
• Lower-extremity duplex ultrasonography
• Pulmonary angiography
• Ventilation-perfusion scintigraphy

Multidetector CT angiography is rapid, noninvasive, and highly sensitive (83%–90%) and specific (96%) for pulmonary embolism.^5,6 In patients such as ours who have a high pretest probability of having the disease, its positive predictive value is 96%.⁵ Therefore, it would be the initial diagnostic study to perform in our patient.

Although transthoracic echocardiography is noninvasive and can detect right ventricular strain in the setting of pulmonary embolism, it may miss half of all pulmonary emboli detected by angiography.^7,8

When technically adequate images are obtained, the combination of magnetic resonance angiography and magnetic resonance venography is very sensitive (92%) and specific (96%) for pulmonary embolism.⁹ However, one-fourth of patients undergoing these studies may have technically inadequate results, so this is not the best choice for diagnosis.⁹

As our patient complained of recent cramping in the right calf, lower-extremity duplex ultrasonography would be a reasonable test to screen for acute deep vein thrombosis as the source of pulmonary embolism. However, given her worrisome vital signs and impending hemodynamic collapse, CT pulmonary angiography would be a better initial test as it may guide more aggressive therapy. Furthermore, even if ultrasonography showed no evidence of deep vein thrombosis, clinical suspicion for pulmonary embolism would remain high enough that therapeutic anticoagulation would be continued until further testing ruled out this diagnosis.

Pulmonary angiography is the gold-standard test for pulmonary embolism. However, it is time-consuming, expensive, and invasive and so is not usually done unless the diagnosis cannot be made with other imaging studies.

Ventilation-perfusion scintigraphy is an established and safe diagnostic test for pulmonary embolism. It is particularly helpful in patients who have renal dysfunction or contrast allergy. The sensitivity of a high-probability scan is 78%, while the specificity of a very-low-probability scan is 97%.10 However, this study is often nondiagnostic (in 26.5% of cases),10 and further imaging may be required.

RESULTS OF CT ANGIOGRAPHY

Our patient undergoes CT angiography, which reveals multiple bilateral pulmonary emboli and right ventricular enlargement (Figure 2). Transthoracic echocardiography shows dilation of the right ventricle, with severely reduced systolic function, an underfilled and hyperdynamic left ventricle (ejection fraction 75%), and moderate tricuspid valve regurgitation. Her right ventricular systolic pressure is estimated to be 47 mm Hg.

Doppler ultrasonography of the legs reveals an occlusive thrombus within the right small saphenous vein that bulges and extends into the right popliteal vein. Also noted is a nonocclusive thrombus in the upper right popliteal vein that likely originated from the thrombus in the small saphenous vein.

Initial laboratory testing (Table 1) shows elevations of the cardiac enzymes troponin T and N-terminal pro-B-type natriuretic peptide (NT-pro-BNP).

ESTIMATING PROGNOSIS IN PULMONARY EMBOLISM

2. Which of the following laboratory results at presentation is independently associated with a worse outcome in patients with pulmonary embolism?

• Elevated NT-pro-BNP
• Hypercalcemia
• Thrombocytosis
• Hypernatremia
• Elevated procalcitonin

The Pulmonary Embolism Severity Index¹¹ and the Simplified Pulmonary Embolism Severity Index¹² (Table 2) are clinical calculators that help predict 30-day risk of death in patients with pulmonary embolism. Our patient’s Pulmonary Embolism Severity Index score is 60, indicating a very low risk, but her simplified severity index score is 2, indicating a high risk.

A shock index score (the heart rate divided by the systolic blood pressure) greater than 1 is also a sensitive measure of risk.¹³ (Our patient’s shock index score is 1.03.) Although the simplified version is more accurate,¹⁴ the shock index is also helpful when deciding whether patients with suspected pulmonary embolism should receive early fibrinolysis.¹⁵

In a large registry of patients with confirmed pulmonary embolism, risk factors for death were age greater than 70, cancer, clinical congestive heart failure, chronic obstructive pulmonary disease, systolic blood pressure lower than 90 mm Hg, respiratory rate less than 20 per minute, and right ventricular hypokinesis.¹⁶ Right ventricular dysfunction progressing to right ventricular failure and cardiogenic shock is the most common cause of death in patients with pulmonary embolism.^16–18

Post hoc analysis has also shown that elevations of the biomarkers BNP, NT-pro-BNP, and cardiac troponins I and T are associated with a prolonged hospital course and a higher risk of death within 30 days.¹⁹ Interestingly, a recent retrospective analysis found hyponatremia to be an independent risk factor for death in the short term.²⁰

Thrombocytopenia, not thrombocytosis, is associated with worse outcomes in patients with pulmonary embolism.¹⁶ Procalcitonin is elevated in bacterial pneumonia but is normal in pulmonary embolism and so may be helpful in differentiating between the two.^21,22 Hypernatremia, hypercalcemia, and elevated procalcitonin have not been shown to be independently associated with worse outcomes in acute pulmonary embolism.

Thus, of the answer choices shown above, elevated NT-pro-BNP is the correct answer.

Pulmonary embolism is often stratified as massive, submassive, or low-risk, reflecting the severity and the degree of cardiovascular collapse. The treatment depends on the classification.

Pulmonary embolism is classified as massive if the patient has a cardiac arrest or a systolic blood pressure lower than 90 mm Hg for more than 15 minutes.²³ Nearly half of patients in this category die.²⁴

Pulmonary embolism is submassive if the patient has systolic pressure greater than 90 mm Hg but has right ventricular dysfunction, as evidenced by physical examination, elevated cardiac biomarkers, electrocardiography, transthoracic echocardiography, or computed tomography. The death rate is as high as 15%.²⁴

Pulmonary embolism in a normotensive patient with no right ventricular dysfunction is defined as low-risk.

Our patient so far

Our patient has bilateral pulmonary emboli, most likely originating from a deep vein thrombosis in her right lower leg. Her pulmonary embolism would be classified as submassive, as her systolic pressure is greater than 90 mm Hg and right ventricular dysfunction—significant right ventricular strain—was noted on both transthoracic echocardiography and computed tomography. Also, cardiac biomarkers are elevated, and the physical examination revealed a prominent P2 sound and right parasternal heave, also suggestive of right ventricular dysfunction.

Now, 6 hours have passed, and even though she has been receiving intravenous heparin during this time, her shock index remains greater than 1, indicating hemodynamic instability. Her pulse rate is still markedly high—over 160 bpm—and she still appears quite anxious and uncomfortable.

HOW SHOULD THIS PATIENT BE TREATED?

3. Which of the following is the most appropriate treatment for this patient?

• Start warfarin immediately while bridging with unfractionated heparin, low-molecular-weight heparin, or fondaparinux
• Start fibrinolysis with alteplase
• Give metoprolol intravenously to control her heart rate
• Start dabigatran immediately while bridging with unfractionated heparin
• Place an inferior vena cava filter
• Consult cardiothoracic surgery for emergency pulmonary embolectomy

All patients with confirmed pulmonary embolism and no contraindications to anticoagulation should begin treatment with low-molecular-weight heparin, unfractionated heparin, or fondaparinux.²³ In addition, this therapy should be started empirically while the patient is still undergoing diagnostic testing if the pretest probability of pulmonary embolism is intermediate or high.²³

Warfarin is indicated for all patients with pulmonary embolism who do not have contraindications to it (Table 3). If unfractionated heparin, low-molecular-weight heparin, or fondaparinux has not already been started, it should be started at the same time as warfarin and should be continued until the international normalized ratio (INR) is within the therapeutic range.

Fibrinolysis. Treatment with a fibrinolytic agent in addition to heparin results in faster improvement of right ventricular function and pulmonary perfusion than with heparin alone.²⁵ It may also decrease the incidence of pulmonary hypertension secondary to chronic thromboembolic disease.²⁶ It should be considered in patients with massive pulmonary embolism.²³

Whether fibrinolysis is appropriate for all patients with submassive pulmonary embolism remains controversial. Currently, it is not recommended for minor right ventricular dysfunction or myocardial necrosis if the patient has no signs of overt clinical decline.²³ The Pulmonary Embolism Thrombolysis trial²⁷ is an ongoing prospective randomized comparison of tenecteplase in a single bolus plus heparin vs heparin alone in normotensive patients with submassive pulmonary embolism, such as our patient. This trial may elucidate the benefit of fibrinolytic therapy in patients with submassive pulmonary embolism.

Patients at low risk are generally treated with heparin and warfarin anticoagulation alone. Fibrinolysis is not recommended in these patients, as the risk of bleeding outweighs the potential benefits.²³

Metoprolol may not be advisable for our patient, as her tachycardia is likely compensatory, and beta-blocker therapy could blunt this compensatory response, leading to inadequate systemic perfusion.

Dabigatran is an oral direct thrombin inhibitor that does not require laboratory monitoring. It is currently approved for the prevention of stroke in patients with atrial fibrillation. It has been shown to be as effective as warfarin in the treatment of acute venous thromboembolism²⁸ and may be a viable option in the future, but as of this writing it has not yet been approved in the United States for this indication. Furthermore, dabigatran inhibits thrombin immediately, so continued heparin bridging would not be necessary.

An inferior vena cava filter may prevent recurrent pulmonary embolism for patients who have absolute contraindications to anticoagulation, most significantly in the short term, ie, in the first few weeks after placement. However, these devices have not yet been shown to improve long-term mortality rates.

Embolectomy, percutaneous or surgical, is also an option. For patients in whom thrombolytic therapy is not effective, “rescue” surgical embolectomy has been associated with better outcomes compared with secondary thrombolysis and so should be considered.²⁹

Back to our patient

An intravenous infusion of alteplase is started, and the patient’s tachycardia improves. Her oxygen requirements normalize, and she is transferred to the general medical floor the next day. She receives subcutaneous dalteparin as a bridge therapy, and warfarin is titrated to a goal INR of 2.0 to 3.0. Because of the acute deep vein thrombosis in her right lower leg, she is instructed to wear knee-high fitted compression hose for primary prevention of postphlebitic syndrome.

HOW LONG TO TREAT? IS GENETIC TESTING INDICATED?

Patients with a first episode of unprovoked venous thromboembolism should receive oral anticoagulants for 6 months, while those with recurrent unprovoked venous thromboembolism require lifelong oral anticoagulation.²³

Whether to test for inherited thrombophilia after a first episode of venous thromboembolism to guide the duration of anticoagulation is controversial.³⁰ Indiscriminate testing has not been recommended in these patients,³¹ but the American College of Medical Genetics recommends genetic screening for factor V Leiden in patients who have an unprovoked incident of venous thromboembolism before age 50.³²

No randomized controlled trial has assessed whether thrombophilia testing decreases the recurrence rate of venous thromboembolism.³³ One uncontrolled study suggested that testing for inherited thrombophilias in patients with a first episode does not affect the risk of recurrence.³⁴ Testing is costly and may cause psychological distress for patients and family members.

Our patient is discharged home on warfarin for 6 months with subsequent follow-up evaluation in the thrombophilia clinic.

WHEN SHOULD WARFARIN BE RESTARTED?

4. If our patient were to discontinue oral anticoagulation in 6 months, which of the following, if present 1 month afterwards, would be a reason to restart oral anticoagulation?

• Elevated serum cotinine
• Positive pregnancy test
• Elevated follicle-stimulating hormone and luteinizing hormone and low estradiol levels
• Elevated D-dimer

Cotinine is a nicotine metabolite, and serum levels are elevated in smokers. Smoking and pregnancy both increase the risk of venous thromboembolism. However, smoking or pregnancy alone would not be a reason to increase the duration of anticoagulation.

Warfarin is contraindicated in pregnancy because of its teratogenic effects, particularly in the first trimester. Follicle-stimulating hormone and luteinizing hormone levels increase in response to decreased estradiol at menopause. Postmenopausal women are not at increased risk of venous thromboembolism unless they are taking oral estrogen hormone replacement therapy.

An elevated D-dimer level 1 month after stopping of oral anticoagulation has been associated with a higher rate of recurrence of venous thromboembolism, which is reduced by a resumption of anticoagulation.³⁵ Therefore, consideration should be given to resuming anticoagulation in patients who have an elevated D-dimer level.

TAKE-HOME POINTS

It is important to distinguish between massive, submassive, and low-risk pulmonary embolism, since each type has a different treatment and prognosis.

Fibrinolytic therapy is indicated in patients with massive pulmonary embolism when no contraindication is present, whereas it is not indicated in those with low-risk pulmonary embolism.

Management of submassive pulmonary embolism continues to be an area of considerable debate. Current American Heart Association guidelines recommend consideration of fibrinolysis initially in patients with submassive acute pulmonary embolism if there is concurrent clinical evidence of adverse prognosis—eg, new hemodynamic instability, worsening respiratory insufficiency, severe right ventricular dysfunction, or major myocardial necrosis.²³ On the other hand, the American College of Chest Physicians recommends against initial systemic fibrinolysis in submassive acute pulmonary embolism based on biomarkers or findings on ECG, transthoracic echocardiography, or CT, recommending it only in therapeutically anticoagulated patients deemed to be at high risk of hypotension.³⁶

Since the optimal treatment of submassive pulmonary embolism is still not known, it is important that clinicians remain up to date on the evidence and guidelines.