A 70-year-old man is evaluated for a persistent leukocytosis. He was hospitalized 10 days ago for a severe exacerbation of chronic obstructive pulmonary disease. He was intubated for 3 days, was diagnosed with a left lower lobe pneumonia, and was treated with antibiotics. His white blood cell count on admission was 20,000 per mcL. It dropped as low as 15,000 on day 6 but is now 25,000, with 23,000 polymorphonuclear leukocytes (10% band forms). He is on oral prednisone 15 mg once daily. Chest x-ray shows no infiltrate. Urinalysis without WBCs.

What is the most likely cause of his leukocytosis?

A) Pulmonary embolus.

B) Lung abscess.

C) Perinephric abscess.

D) Prednisone.

E) Clostridium difficile infection.

CDC/Jennifer Hulsey

This isn’t an uncommon scenario, in which your hospitalized patient has a climbing WBC without a clear cause. Often, the patient may well be improving from the condition that they were originally hospitalized for, but the climbing WBC count is concerning and often delays discharge. What should we think of in the patient whose WBC climbs in the hospital, and the cause isn’t readily apparent?

The most likely diagnosis in otherwise unexplained leukocytosis in a hospitalized patient is C. difficile.

Anna Wanahita, MD, of the St. John Clinic in Tulsa, Okla., and her colleagues prospectively studied 60 patients admitted to a VA hospital who had unexplained leukocytosis.¹ All patients had stool specimens sent for C. difficile toxin; in addition, 26 hospitalized control patients without leukocytosis also had stool sent for C. difficile toxin. For study purposes, leukocytosis was defined as a WBC greater than 15,000 per mcL. Any patient for whom C. difficile toxin was sent because of clinical suspicion and who was positive was excluded from the study results.

Almost 60% of the patients with unexplained leukocytosis (35 of 60) had a positive C. difficile toxin, compared with 12% of the controls (P less than .001). More than half of the patients with a positive C. difficile test had the onset of leukocytosis prior to any symptoms of colitis. Leukocytosis responded to treatment with metronidazole in 83% of the patients with a positive C. difficile toxin, and 75% of the patients who had leukocytosis did not have a positive C. difficile toxin.

In another study, Mamatha Bulusu, and colleagues did a retrospective study of 70 hospitalized patients who had diarrhea and underwent testing for C. difficile.² They evaluated the pattern of white blood cell counts in patients who were positive and negative for C. difficile toxin. The mean WBC for C. difficile–positive patients was 15,800, compared with 7,700 for the patients who were C. difficile negative (P less than .01). They described three patterns: one in which leukocytosis occurred at the onset of diarrhea; a pattern in which unexplained leukocytosis occurred days prior to diarrhea; and a pattern in which patients treated for infection with leukocytosis had a worsening of their leukocytosis at the onset of diarrheal symptoms. Treatment with metronidazole led to a resolution of leukocytosis in all the C. difficile–positive patients.

Another possibility in this case was WBC elevation because of the patient’s prednisone. Prednisone can increase WBC as early as the first day of therapy.³ The elevation and rapidity of increase are dose related. The important pearl is that steroid-induced leukocytosis involves an increase of polymorphonuclear white blood cells with a rise in monocytes and a decrease in eosinophils and lymphocytes.

Importantly, increased band forms (greater than 6%) and toxic granulation rarely ever occur with steroid-induced leukocytosis, and the presence of these features should strongly suggest a different cause.⁴

Pearl: Think of underlying C. difficile infection in your hospitalized patient with unexplained leukocytosis.
 

Dr. Paauw is professor of medicine in the division of general internal medicine at the University of Washington, Seattle, and he serves as third-year medical student clerkship director at the University of Washington. Contact Dr. Paauw at [email protected].

References

1. Am J Med. 2003 Nov;115(7):543-6.

2. Am J Gastroenterol. 2000 Nov;95(11):3137-41.

3. J Clin Invest. 1975 Oct;56(4):808-13.

4. Am J Med. 1981 Nov;71(5):773-8.

 

A 70-year-old man is evaluated for a persistent leukocytosis. He was hospitalized 10 days ago for a severe exacerbation of chronic obstructive pulmonary disease. He was intubated for 3 days, was diagnosed with a left lower lobe pneumonia, and was treated with antibiotics. His white blood cell count on admission was 20,000 per mcL. It dropped as low as 15,000 on day 6 but is now 25,000, with 23,000 polymorphonuclear leukocytes (10% band forms). He is on oral prednisone 15 mg once daily. Chest x-ray shows no infiltrate. Urinalysis without WBCs.

What is the most likely cause of his leukocytosis?

A) Pulmonary embolus.

B) Lung abscess.

C) Perinephric abscess.

D) Prednisone.

E) Clostridium difficile infection.

CDC/Jennifer Hulsey

This isn’t an uncommon scenario, in which your hospitalized patient has a climbing WBC without a clear cause. Often, the patient may well be improving from the condition that they were originally hospitalized for, but the climbing WBC count is concerning and often delays discharge. What should we think of in the patient whose WBC climbs in the hospital, and the cause isn’t readily apparent?

The most likely diagnosis in otherwise unexplained leukocytosis in a hospitalized patient is C. difficile.

Anna Wanahita, MD, of the St. John Clinic in Tulsa, Okla., and her colleagues prospectively studied 60 patients admitted to a VA hospital who had unexplained leukocytosis.¹ All patients had stool specimens sent for C. difficile toxin; in addition, 26 hospitalized control patients without leukocytosis also had stool sent for C. difficile toxin. For study purposes, leukocytosis was defined as a WBC greater than 15,000 per mcL. Any patient for whom C. difficile toxin was sent because of clinical suspicion and who was positive was excluded from the study results.

Almost 60% of the patients with unexplained leukocytosis (35 of 60) had a positive C. difficile toxin, compared with 12% of the controls (P less than .001). More than half of the patients with a positive C. difficile test had the onset of leukocytosis prior to any symptoms of colitis. Leukocytosis responded to treatment with metronidazole in 83% of the patients with a positive C. difficile toxin, and 75% of the patients who had leukocytosis did not have a positive C. difficile toxin.

In another study, Mamatha Bulusu, and colleagues did a retrospective study of 70 hospitalized patients who had diarrhea and underwent testing for C. difficile.² They evaluated the pattern of white blood cell counts in patients who were positive and negative for C. difficile toxin. The mean WBC for C. difficile–positive patients was 15,800, compared with 7,700 for the patients who were C. difficile negative (P less than .01). They described three patterns: one in which leukocytosis occurred at the onset of diarrhea; a pattern in which unexplained leukocytosis occurred days prior to diarrhea; and a pattern in which patients treated for infection with leukocytosis had a worsening of their leukocytosis at the onset of diarrheal symptoms. Treatment with metronidazole led to a resolution of leukocytosis in all the C. difficile–positive patients.

Another possibility in this case was WBC elevation because of the patient’s prednisone. Prednisone can increase WBC as early as the first day of therapy.³ The elevation and rapidity of increase are dose related. The important pearl is that steroid-induced leukocytosis involves an increase of polymorphonuclear white blood cells with a rise in monocytes and a decrease in eosinophils and lymphocytes.

Importantly, increased band forms (greater than 6%) and toxic granulation rarely ever occur with steroid-induced leukocytosis, and the presence of these features should strongly suggest a different cause.⁴

Pearl: Think of underlying C. difficile infection in your hospitalized patient with unexplained leukocytosis.
 

Dr. Paauw is professor of medicine in the division of general internal medicine at the University of Washington, Seattle, and he serves as third-year medical student clerkship director at the University of Washington. Contact Dr. Paauw at [email protected].

References

1. Am J Med. 2003 Nov;115(7):543-6.

2. Am J Gastroenterol. 2000 Nov;95(11):3137-41.

3. J Clin Invest. 1975 Oct;56(4):808-13.

4. Am J Med. 1981 Nov;71(5):773-8.

 

A 70-year-old man is evaluated for a persistent leukocytosis. He was hospitalized 10 days ago for a severe exacerbation of chronic obstructive pulmonary disease. He was intubated for 3 days, was diagnosed with a left lower lobe pneumonia, and was treated with antibiotics. His white blood cell count on admission was 20,000 per mcL. It dropped as low as 15,000 on day 6 but is now 25,000, with 23,000 polymorphonuclear leukocytes (10% band forms). He is on oral prednisone 15 mg once daily. Chest x-ray shows no infiltrate. Urinalysis without WBCs.

What is the most likely cause of his leukocytosis?

A) Pulmonary embolus.

B) Lung abscess.

C) Perinephric abscess.

D) Prednisone.

E) Clostridium difficile infection.

CDC/Jennifer Hulsey

This isn’t an uncommon scenario, in which your hospitalized patient has a climbing WBC without a clear cause. Often, the patient may well be improving from the condition that they were originally hospitalized for, but the climbing WBC count is concerning and often delays discharge. What should we think of in the patient whose WBC climbs in the hospital, and the cause isn’t readily apparent?

The most likely diagnosis in otherwise unexplained leukocytosis in a hospitalized patient is C. difficile.

Anna Wanahita, MD, of the St. John Clinic in Tulsa, Okla., and her colleagues prospectively studied 60 patients admitted to a VA hospital who had unexplained leukocytosis.¹ All patients had stool specimens sent for C. difficile toxin; in addition, 26 hospitalized control patients without leukocytosis also had stool sent for C. difficile toxin. For study purposes, leukocytosis was defined as a WBC greater than 15,000 per mcL. Any patient for whom C. difficile toxin was sent because of clinical suspicion and who was positive was excluded from the study results.

Almost 60% of the patients with unexplained leukocytosis (35 of 60) had a positive C. difficile toxin, compared with 12% of the controls (P less than .001). More than half of the patients with a positive C. difficile test had the onset of leukocytosis prior to any symptoms of colitis. Leukocytosis responded to treatment with metronidazole in 83% of the patients with a positive C. difficile toxin, and 75% of the patients who had leukocytosis did not have a positive C. difficile toxin.

In another study, Mamatha Bulusu, and colleagues did a retrospective study of 70 hospitalized patients who had diarrhea and underwent testing for C. difficile.² They evaluated the pattern of white blood cell counts in patients who were positive and negative for C. difficile toxin. The mean WBC for C. difficile–positive patients was 15,800, compared with 7,700 for the patients who were C. difficile negative (P less than .01). They described three patterns: one in which leukocytosis occurred at the onset of diarrhea; a pattern in which unexplained leukocytosis occurred days prior to diarrhea; and a pattern in which patients treated for infection with leukocytosis had a worsening of their leukocytosis at the onset of diarrheal symptoms. Treatment with metronidazole led to a resolution of leukocytosis in all the C. difficile–positive patients.

Another possibility in this case was WBC elevation because of the patient’s prednisone. Prednisone can increase WBC as early as the first day of therapy.³ The elevation and rapidity of increase are dose related. The important pearl is that steroid-induced leukocytosis involves an increase of polymorphonuclear white blood cells with a rise in monocytes and a decrease in eosinophils and lymphocytes.

Importantly, increased band forms (greater than 6%) and toxic granulation rarely ever occur with steroid-induced leukocytosis, and the presence of these features should strongly suggest a different cause.⁴

Pearl: Think of underlying C. difficile infection in your hospitalized patient with unexplained leukocytosis.
 

Dr. Paauw is professor of medicine in the division of general internal medicine at the University of Washington, Seattle, and he serves as third-year medical student clerkship director at the University of Washington. Contact Dr. Paauw at [email protected].

References

1. Am J Med. 2003 Nov;115(7):543-6.

2. Am J Gastroenterol. 2000 Nov;95(11):3137-41.

3. J Clin Invest. 1975 Oct;56(4):808-13.

4. Am J Med. 1981 Nov;71(5):773-8.

 

WASHINGTON – Clinicians who treat patients with emotional and psychiatric problems must put risk management interventions in place for their safety, Jeffrey N. Younggren, PhD, said at the annual convention of the American Psychological Association.

“Many times, people lose sight of the nature of their therapeutic relationship,” said Dr. Younggren, professor of psychology at the University of Missouri in Columbia. To stay safe, clinicians must overreact, he said, just as they do with suicide risk assessments.

Dr. Younggren critiqued an American Psychological Association article on safety and offered his own recommendations. Among them:

• Think about evacuation strategies. “Don’t get between that individual and the door,” said Dr. Younggren, a clinical and forensic psychologist.
• Refuse to see patients who are inebriated or intoxicated. If such a patient shows up for an appointment in one of these conditions and refuses to leave, call the police.
• Remove yourself from physical danger. “I’m a very good ‘fall on the ground’ person,” said Dr. Younggren, who said he has been attacked by patients three times in his career. “That’s a risk management strategy.”
• Terminate patients appropriately in the absence of threats. However, “if someone threatens you, write them a letter, and you’re done,” he said.

Dr. Younggren suggested that other recommendations in the article were unrealistic, such as, don’t work alone at night, install security cameras, and learn self-defense techniques. “What does [learn self-defense techniques] mean,” he asked. “My best one is to fall down.”

Mismanagement of the therapeutic alliance can careen out of control, as it did in the case of Ensworth vs. Mullvain.

In that case, decided in 1990, Heather Ensworth, PhD, a psychologist who practiced in California, treated a patient named Cynthia Mullvain for just short of 2 years and then terminated the treatment. But Ms. Mullvain did not accept the termination and persuaded Dr. Ensworth to see her again “to resolve the termination issues to help [Mullvain] disengage from [Ensworth].”

After several harassing incidents, Dr. Ensworth terminated contact with Ms. Mullvain a second time. At this point, Dr. Ensworth sought and was granted a restraining order against the patient. Despite the restraining order, Ms. Mullvain’s harassing behavior continued. Among other things, she stalked Dr. Ensworth, sent her threatening letters, and started doing community service work at a library located about 150 feet away from Dr. Ensworth’s home, according to Dr. Ensworth’s petition seeking a second restraining order. Ultimately, the court ruled that Ms. Mullvain had “willfully engaged in a course of conduct that seriously alarmed, annoyed, or harassed Ensworth, and that Ensworth actually suffered substantial emotional distress.”

Ernest J. Bordini, PhD said that, beyond private offices, nurses and aides are at greatest risk when it comes to workplace violence. According to a report by the Occupational Health and Safety Administration (OSHA), in 2013, psychiatric aides had the highest rate among health care workers of violent injuries that led to days away from work: 590 per 10,000 full-time employees, compared with 55 such injuries per 10,000 for nursing assistants. The report said the highest risk areas were emergency departments, geriatrics, and behavioral health.

Psychiatric patients are more likely to be the victims of violence than perpetrators, but Dr. Bordini, a neuropsychologist with expertise in forensic assessment, said in an interview that he wanted to add a point.

“It is important to dismiss the notion that all psychiatric patients do not have elevated risks of assault,” he said. “Those who present with psychoses or bipolar disorder can have elevated risk, especially if they develop delusional thoughts or obsessions about the therapist or another individual. Paranoid individuals already feel threatened, and hence can strike out in anticipation.”

He said he and his colleagues are not advocating that all clinicians train in self defense or arm themselves. However, it is essential to be proactive. Falling down can work for some, Dr. Bordini said, but “experience teaches us that playing possum does not always cease an attack. I recommend de-escalation, escape, and/or self-defense plans that one has practiced, feels comfortable with, and feels confident that they can execute under stress.”

At the meeting, he said some patients are able to sense fear from the clinician. “If you’re skittish, [this will] put you at higher risk,” said Dr. Bordini, executive director of Clinical Psychology Associates of North Central Florida in Gainesville. “That sense of intuition is something you should tend to,” he said, citing The Gift of Fear (New York: Dell, 1999) by Gavin de Becker as an example of a book that explores recognizing and reacting to subtle signs of danger. “If you’re not comfortable seeing a patient, listen to that.”

Neither Dr. Younggren nor Dr. Bordini had financial disclosures.



To access OSHA’s guidelines for workplace violence in health care settings, visit https://www.osha.gov/Publications/OSHA3826.pdf. The American Medical Association’s latest policy on workplace violence can be found at https://www.ama-assn.org/ama-adopts-new-public-health-policies-improve-health-nation.

 

[email protected]

On Twitter @ginalhenderson

 

WASHINGTON – Clinicians who treat patients with emotional and psychiatric problems must put risk management interventions in place for their safety, Jeffrey N. Younggren, PhD, said at the annual convention of the American Psychological Association.

“Many times, people lose sight of the nature of their therapeutic relationship,” said Dr. Younggren, professor of psychology at the University of Missouri in Columbia. To stay safe, clinicians must overreact, he said, just as they do with suicide risk assessments.

Dr. Younggren critiqued an American Psychological Association article on safety and offered his own recommendations. Among them:

• Think about evacuation strategies. “Don’t get between that individual and the door,” said Dr. Younggren, a clinical and forensic psychologist.
• Refuse to see patients who are inebriated or intoxicated. If such a patient shows up for an appointment in one of these conditions and refuses to leave, call the police.
• Remove yourself from physical danger. “I’m a very good ‘fall on the ground’ person,” said Dr. Younggren, who said he has been attacked by patients three times in his career. “That’s a risk management strategy.”
• Terminate patients appropriately in the absence of threats. However, “if someone threatens you, write them a letter, and you’re done,” he said.

Dr. Younggren suggested that other recommendations in the article were unrealistic, such as, don’t work alone at night, install security cameras, and learn self-defense techniques. “What does [learn self-defense techniques] mean,” he asked. “My best one is to fall down.”

Mismanagement of the therapeutic alliance can careen out of control, as it did in the case of Ensworth vs. Mullvain.

In that case, decided in 1990, Heather Ensworth, PhD, a psychologist who practiced in California, treated a patient named Cynthia Mullvain for just short of 2 years and then terminated the treatment. But Ms. Mullvain did not accept the termination and persuaded Dr. Ensworth to see her again “to resolve the termination issues to help [Mullvain] disengage from [Ensworth].”

After several harassing incidents, Dr. Ensworth terminated contact with Ms. Mullvain a second time. At this point, Dr. Ensworth sought and was granted a restraining order against the patient. Despite the restraining order, Ms. Mullvain’s harassing behavior continued. Among other things, she stalked Dr. Ensworth, sent her threatening letters, and started doing community service work at a library located about 150 feet away from Dr. Ensworth’s home, according to Dr. Ensworth’s petition seeking a second restraining order. Ultimately, the court ruled that Ms. Mullvain had “willfully engaged in a course of conduct that seriously alarmed, annoyed, or harassed Ensworth, and that Ensworth actually suffered substantial emotional distress.”

Ernest J. Bordini, PhD said that, beyond private offices, nurses and aides are at greatest risk when it comes to workplace violence. According to a report by the Occupational Health and Safety Administration (OSHA), in 2013, psychiatric aides had the highest rate among health care workers of violent injuries that led to days away from work: 590 per 10,000 full-time employees, compared with 55 such injuries per 10,000 for nursing assistants. The report said the highest risk areas were emergency departments, geriatrics, and behavioral health.

Psychiatric patients are more likely to be the victims of violence than perpetrators, but Dr. Bordini, a neuropsychologist with expertise in forensic assessment, said in an interview that he wanted to add a point.

“It is important to dismiss the notion that all psychiatric patients do not have elevated risks of assault,” he said. “Those who present with psychoses or bipolar disorder can have elevated risk, especially if they develop delusional thoughts or obsessions about the therapist or another individual. Paranoid individuals already feel threatened, and hence can strike out in anticipation.”

He said he and his colleagues are not advocating that all clinicians train in self defense or arm themselves. However, it is essential to be proactive. Falling down can work for some, Dr. Bordini said, but “experience teaches us that playing possum does not always cease an attack. I recommend de-escalation, escape, and/or self-defense plans that one has practiced, feels comfortable with, and feels confident that they can execute under stress.”

At the meeting, he said some patients are able to sense fear from the clinician. “If you’re skittish, [this will] put you at higher risk,” said Dr. Bordini, executive director of Clinical Psychology Associates of North Central Florida in Gainesville. “That sense of intuition is something you should tend to,” he said, citing The Gift of Fear (New York: Dell, 1999) by Gavin de Becker as an example of a book that explores recognizing and reacting to subtle signs of danger. “If you’re not comfortable seeing a patient, listen to that.”

Neither Dr. Younggren nor Dr. Bordini had financial disclosures.



To access OSHA’s guidelines for workplace violence in health care settings, visit https://www.osha.gov/Publications/OSHA3826.pdf. The American Medical Association’s latest policy on workplace violence can be found at https://www.ama-assn.org/ama-adopts-new-public-health-policies-improve-health-nation.

 

[email protected]

On Twitter @ginalhenderson

 

WASHINGTON – Clinicians who treat patients with emotional and psychiatric problems must put risk management interventions in place for their safety, Jeffrey N. Younggren, PhD, said at the annual convention of the American Psychological Association.

“Many times, people lose sight of the nature of their therapeutic relationship,” said Dr. Younggren, professor of psychology at the University of Missouri in Columbia. To stay safe, clinicians must overreact, he said, just as they do with suicide risk assessments.

Dr. Younggren critiqued an American Psychological Association article on safety and offered his own recommendations. Among them:

• Think about evacuation strategies. “Don’t get between that individual and the door,” said Dr. Younggren, a clinical and forensic psychologist.
• Refuse to see patients who are inebriated or intoxicated. If such a patient shows up for an appointment in one of these conditions and refuses to leave, call the police.
• Remove yourself from physical danger. “I’m a very good ‘fall on the ground’ person,” said Dr. Younggren, who said he has been attacked by patients three times in his career. “That’s a risk management strategy.”
• Terminate patients appropriately in the absence of threats. However, “if someone threatens you, write them a letter, and you’re done,” he said.

Dr. Younggren suggested that other recommendations in the article were unrealistic, such as, don’t work alone at night, install security cameras, and learn self-defense techniques. “What does [learn self-defense techniques] mean,” he asked. “My best one is to fall down.”

Mismanagement of the therapeutic alliance can careen out of control, as it did in the case of Ensworth vs. Mullvain.

In that case, decided in 1990, Heather Ensworth, PhD, a psychologist who practiced in California, treated a patient named Cynthia Mullvain for just short of 2 years and then terminated the treatment. But Ms. Mullvain did not accept the termination and persuaded Dr. Ensworth to see her again “to resolve the termination issues to help [Mullvain] disengage from [Ensworth].”

After several harassing incidents, Dr. Ensworth terminated contact with Ms. Mullvain a second time. At this point, Dr. Ensworth sought and was granted a restraining order against the patient. Despite the restraining order, Ms. Mullvain’s harassing behavior continued. Among other things, she stalked Dr. Ensworth, sent her threatening letters, and started doing community service work at a library located about 150 feet away from Dr. Ensworth’s home, according to Dr. Ensworth’s petition seeking a second restraining order. Ultimately, the court ruled that Ms. Mullvain had “willfully engaged in a course of conduct that seriously alarmed, annoyed, or harassed Ensworth, and that Ensworth actually suffered substantial emotional distress.”

Ernest J. Bordini, PhD said that, beyond private offices, nurses and aides are at greatest risk when it comes to workplace violence. According to a report by the Occupational Health and Safety Administration (OSHA), in 2013, psychiatric aides had the highest rate among health care workers of violent injuries that led to days away from work: 590 per 10,000 full-time employees, compared with 55 such injuries per 10,000 for nursing assistants. The report said the highest risk areas were emergency departments, geriatrics, and behavioral health.

Psychiatric patients are more likely to be the victims of violence than perpetrators, but Dr. Bordini, a neuropsychologist with expertise in forensic assessment, said in an interview that he wanted to add a point.

“It is important to dismiss the notion that all psychiatric patients do not have elevated risks of assault,” he said. “Those who present with psychoses or bipolar disorder can have elevated risk, especially if they develop delusional thoughts or obsessions about the therapist or another individual. Paranoid individuals already feel threatened, and hence can strike out in anticipation.”

He said he and his colleagues are not advocating that all clinicians train in self defense or arm themselves. However, it is essential to be proactive. Falling down can work for some, Dr. Bordini said, but “experience teaches us that playing possum does not always cease an attack. I recommend de-escalation, escape, and/or self-defense plans that one has practiced, feels comfortable with, and feels confident that they can execute under stress.”

At the meeting, he said some patients are able to sense fear from the clinician. “If you’re skittish, [this will] put you at higher risk,” said Dr. Bordini, executive director of Clinical Psychology Associates of North Central Florida in Gainesville. “That sense of intuition is something you should tend to,” he said, citing The Gift of Fear (New York: Dell, 1999) by Gavin de Becker as an example of a book that explores recognizing and reacting to subtle signs of danger. “If you’re not comfortable seeing a patient, listen to that.”

Neither Dr. Younggren nor Dr. Bordini had financial disclosures.



To access OSHA’s guidelines for workplace violence in health care settings, visit https://www.osha.gov/Publications/OSHA3826.pdf. The American Medical Association’s latest policy on workplace violence can be found at https://www.ama-assn.org/ama-adopts-new-public-health-policies-improve-health-nation.

 

[email protected]

On Twitter @ginalhenderson

 

DENVER – Physicians are by and large doing “pretty poorly” at counseling women about the use of marijuana during pregnancy, Torri D. Metz, MD, observed at the annual meeting of the Teratology Society.

This is of particular concern because the increasing legalization of recreational marijuana across the United States means growing use, possibly including use by pregnant women. National surveys indicate a high percentage of pregnant women believe there is slight or no harm in using marijuana once or twice per week, said Dr. Metz, an ob.gyn. at the University of Colorado, Denver, who is researching the effects of marijuana in pregnancy.

Bruce Jancin/Frontline Medical News

As an ob.gyn. practicing in Colorado – one of the first states to legalize recreational marijuana and a state where medical marijuana is protected under the state constitution – Dr. Metz often fields questions from obstetric care providers frustrated by patients who say, “I’m going to use it. You can’t show me data that says it’s not safe.”

Here’s how she likes to handle that situation: She starts out by freely admitting that that’s true. The available evidence is limited, mixed, and often flawed.

“I say, ‘I can’t give you data that says absolutely it’s not safe, but I also absolutely cannot give you data saying it is safe.’ I would favor saying, ‘I can’t tell you it’s safe. And if there’s any possible risk, let’s talk about things we know are safe we can use as alternatives for whatever you’re using cannabis for,’ ” she explained.

A Colorado survey of more than 1,700 mothers in the WIC (Women, Infants, and Children) nutrition program shed light on the reasons women use marijuana while pregnant or breastfeeding. Sixty-three percent of current users cited as a perceived benefit that it helped with depression, anxiety, and/or stress. Sixty percent reported it helped with pain. Nearly half used marijuana for nausea and vomiting. Just 39% did so for recreation.

Dr. Metz’s anecdotal experience has been that many health care providers are flubbing the opportunity to counsel women about marijuana use in pregnancy. This impression was bolstered by a recent study by investigators at the University of Pittsburgh who audio-recorded 468 first prenatal visits.

In total, 19% of patients disclosed marijuana use to 47 health care providers. In nearly half of those encounters, the providers didn’t respond to the disclosure at all. And when they did respond, it typically wasn’t by providing thoughtful, informed counseling on the risks or outcomes of using marijuana in pregnancy. Instead, the response was most often punitive: for example, a warning that evidence of use at delivery would result in a call to child protective services (Obstet Gynecol. 2016 Apr;127[4]:681-7).

Because of Colorado’s lengthy experience with legalized marijuana, the state Department of Public Health and Environment has endeavored to create resources of value for health care providers and patients (www.colorado.gov/cdphe/marijuana-clinical-guidelines). The website contains a fact sheet for patients regarding marijuana in pregnancy and breastfeeding. For physicians, there is plain-language guidance on how to talk effectively about marijuana with patients, including suggested responses to selected commonly voiced misconceptions.

The website also includes the results of a 2014 marijuana-in-pregnancy literature review by a state advisory committee composed of Colorado specialists in pediatrics, ob.gyn., family medicine, public health, and addiction medicine.

The committee determined that there is moderate evidence that the use of marijuana in pregnancy is associated with increased risk of reduced fetal growth, lower IQ scores in young children, adverse effects on a child’s cognitive function and academic ability, and an increase in attention problems. There was deemed to be limited evidence of an association with stillbirth and isolated ventricular septal defects. There is also “mixed” evidence for associations with preterm delivery, reduced birth weight, and selected congenital anomalies.

Since that 2014 review, a new signal of potential harm stemming from maternal marijuana use in pregnancy has appeared: a possible increased risk of neonatal ICU admission. In one retrospective study including 361 marijuana users and 6,107 nonusers, the users had a 1.54-fold increased risk for neonatal ICU admission in an analysis adjusted for maternal demographics and tobacco use (J Perinatol. 2015 Dec;35[12]:991-5).

Moreover, investigators at the University of Arizona in Tucson performed a meta-analysis of 24 studies and concluded that infants exposed to cannabis in utero were at 2.02-fold increased likelihood of neonatal ICU admission, a 1.77-fold increased risk of low birth weight, and 1.36-fold increased odds of anemia (BMJ Open. 2016 Apr 5;6[4]:e0009986. doi: 10.1136/bmjopen-2015-009986).

“That obviously would have a big public health impact,” Dr. Metz said.

In marked contrast, however, just a few months later investigators at Washington University in St. Louis reported finding no significantly increased risk of neonatal ICU admission or any other adverse neonatal outcome after adjustment for tobacco use and other potential confounders in a meta-analysis of 31 studies (Obstet Gynecol. 2016 Oct;128[4]:713-23).

These contradictory meta-analyses underscore a key point about the existing literature on the safety of marijuana use in pregnancy: It provides few, if any, definitive answers. The studies conducted in the 1980s and 1990s are of limited generalizability because concentrations of tetrahydrocannabinol were so small, compared with today’s products. Ascertainment of exposure to marijuana in pregnancy is unreliable in the absence of confirmatory biologic sampling. Self-reported use is unreliable and is typically an underestimate. Adjustment for confounders associated with adverse neonatal outcomes is challenging.

“Biologic sampling is critical,” Dr. Metz said. “We actually don’t know who’s using, and we lack information on the timing and quantity of exposure.

“Part of the problem is the data are so mixed that you can really find whatever you want in the literature to support your bias,” she added.

Still, in light of the signals of possible harm, she urged her colleagues to advise patients not to use marijuana in pregnancy. Patients need to understand that there are no known benefits of marijuana use in pregnancy, there are possible risks, and there is no known safe amount of cannabis in pregnancy.

Dr. Metz reported having no financial conflicts related to her presentation.

 

[email protected]

 

DENVER – Physicians are by and large doing “pretty poorly” at counseling women about the use of marijuana during pregnancy, Torri D. Metz, MD, observed at the annual meeting of the Teratology Society.

This is of particular concern because the increasing legalization of recreational marijuana across the United States means growing use, possibly including use by pregnant women. National surveys indicate a high percentage of pregnant women believe there is slight or no harm in using marijuana once or twice per week, said Dr. Metz, an ob.gyn. at the University of Colorado, Denver, who is researching the effects of marijuana in pregnancy.

Bruce Jancin/Frontline Medical News

As an ob.gyn. practicing in Colorado – one of the first states to legalize recreational marijuana and a state where medical marijuana is protected under the state constitution – Dr. Metz often fields questions from obstetric care providers frustrated by patients who say, “I’m going to use it. You can’t show me data that says it’s not safe.”

Here’s how she likes to handle that situation: She starts out by freely admitting that that’s true. The available evidence is limited, mixed, and often flawed.

“I say, ‘I can’t give you data that says absolutely it’s not safe, but I also absolutely cannot give you data saying it is safe.’ I would favor saying, ‘I can’t tell you it’s safe. And if there’s any possible risk, let’s talk about things we know are safe we can use as alternatives for whatever you’re using cannabis for,’ ” she explained.

A Colorado survey of more than 1,700 mothers in the WIC (Women, Infants, and Children) nutrition program shed light on the reasons women use marijuana while pregnant or breastfeeding. Sixty-three percent of current users cited as a perceived benefit that it helped with depression, anxiety, and/or stress. Sixty percent reported it helped with pain. Nearly half used marijuana for nausea and vomiting. Just 39% did so for recreation.

Dr. Metz’s anecdotal experience has been that many health care providers are flubbing the opportunity to counsel women about marijuana use in pregnancy. This impression was bolstered by a recent study by investigators at the University of Pittsburgh who audio-recorded 468 first prenatal visits.

In total, 19% of patients disclosed marijuana use to 47 health care providers. In nearly half of those encounters, the providers didn’t respond to the disclosure at all. And when they did respond, it typically wasn’t by providing thoughtful, informed counseling on the risks or outcomes of using marijuana in pregnancy. Instead, the response was most often punitive: for example, a warning that evidence of use at delivery would result in a call to child protective services (Obstet Gynecol. 2016 Apr;127[4]:681-7).

Because of Colorado’s lengthy experience with legalized marijuana, the state Department of Public Health and Environment has endeavored to create resources of value for health care providers and patients (www.colorado.gov/cdphe/marijuana-clinical-guidelines). The website contains a fact sheet for patients regarding marijuana in pregnancy and breastfeeding. For physicians, there is plain-language guidance on how to talk effectively about marijuana with patients, including suggested responses to selected commonly voiced misconceptions.

The website also includes the results of a 2014 marijuana-in-pregnancy literature review by a state advisory committee composed of Colorado specialists in pediatrics, ob.gyn., family medicine, public health, and addiction medicine.

The committee determined that there is moderate evidence that the use of marijuana in pregnancy is associated with increased risk of reduced fetal growth, lower IQ scores in young children, adverse effects on a child’s cognitive function and academic ability, and an increase in attention problems. There was deemed to be limited evidence of an association with stillbirth and isolated ventricular septal defects. There is also “mixed” evidence for associations with preterm delivery, reduced birth weight, and selected congenital anomalies.

Since that 2014 review, a new signal of potential harm stemming from maternal marijuana use in pregnancy has appeared: a possible increased risk of neonatal ICU admission. In one retrospective study including 361 marijuana users and 6,107 nonusers, the users had a 1.54-fold increased risk for neonatal ICU admission in an analysis adjusted for maternal demographics and tobacco use (J Perinatol. 2015 Dec;35[12]:991-5).

Moreover, investigators at the University of Arizona in Tucson performed a meta-analysis of 24 studies and concluded that infants exposed to cannabis in utero were at 2.02-fold increased likelihood of neonatal ICU admission, a 1.77-fold increased risk of low birth weight, and 1.36-fold increased odds of anemia (BMJ Open. 2016 Apr 5;6[4]:e0009986. doi: 10.1136/bmjopen-2015-009986).

“That obviously would have a big public health impact,” Dr. Metz said.

In marked contrast, however, just a few months later investigators at Washington University in St. Louis reported finding no significantly increased risk of neonatal ICU admission or any other adverse neonatal outcome after adjustment for tobacco use and other potential confounders in a meta-analysis of 31 studies (Obstet Gynecol. 2016 Oct;128[4]:713-23).

These contradictory meta-analyses underscore a key point about the existing literature on the safety of marijuana use in pregnancy: It provides few, if any, definitive answers. The studies conducted in the 1980s and 1990s are of limited generalizability because concentrations of tetrahydrocannabinol were so small, compared with today’s products. Ascertainment of exposure to marijuana in pregnancy is unreliable in the absence of confirmatory biologic sampling. Self-reported use is unreliable and is typically an underestimate. Adjustment for confounders associated with adverse neonatal outcomes is challenging.

“Biologic sampling is critical,” Dr. Metz said. “We actually don’t know who’s using, and we lack information on the timing and quantity of exposure.

“Part of the problem is the data are so mixed that you can really find whatever you want in the literature to support your bias,” she added.

Still, in light of the signals of possible harm, she urged her colleagues to advise patients not to use marijuana in pregnancy. Patients need to understand that there are no known benefits of marijuana use in pregnancy, there are possible risks, and there is no known safe amount of cannabis in pregnancy.

Dr. Metz reported having no financial conflicts related to her presentation.

 

[email protected]

 

DENVER – Physicians are by and large doing “pretty poorly” at counseling women about the use of marijuana during pregnancy, Torri D. Metz, MD, observed at the annual meeting of the Teratology Society.

This is of particular concern because the increasing legalization of recreational marijuana across the United States means growing use, possibly including use by pregnant women. National surveys indicate a high percentage of pregnant women believe there is slight or no harm in using marijuana once or twice per week, said Dr. Metz, an ob.gyn. at the University of Colorado, Denver, who is researching the effects of marijuana in pregnancy.

Bruce Jancin/Frontline Medical News

As an ob.gyn. practicing in Colorado – one of the first states to legalize recreational marijuana and a state where medical marijuana is protected under the state constitution – Dr. Metz often fields questions from obstetric care providers frustrated by patients who say, “I’m going to use it. You can’t show me data that says it’s not safe.”

Here’s how she likes to handle that situation: She starts out by freely admitting that that’s true. The available evidence is limited, mixed, and often flawed.

“I say, ‘I can’t give you data that says absolutely it’s not safe, but I also absolutely cannot give you data saying it is safe.’ I would favor saying, ‘I can’t tell you it’s safe. And if there’s any possible risk, let’s talk about things we know are safe we can use as alternatives for whatever you’re using cannabis for,’ ” she explained.

A Colorado survey of more than 1,700 mothers in the WIC (Women, Infants, and Children) nutrition program shed light on the reasons women use marijuana while pregnant or breastfeeding. Sixty-three percent of current users cited as a perceived benefit that it helped with depression, anxiety, and/or stress. Sixty percent reported it helped with pain. Nearly half used marijuana for nausea and vomiting. Just 39% did so for recreation.

Dr. Metz’s anecdotal experience has been that many health care providers are flubbing the opportunity to counsel women about marijuana use in pregnancy. This impression was bolstered by a recent study by investigators at the University of Pittsburgh who audio-recorded 468 first prenatal visits.

In total, 19% of patients disclosed marijuana use to 47 health care providers. In nearly half of those encounters, the providers didn’t respond to the disclosure at all. And when they did respond, it typically wasn’t by providing thoughtful, informed counseling on the risks or outcomes of using marijuana in pregnancy. Instead, the response was most often punitive: for example, a warning that evidence of use at delivery would result in a call to child protective services (Obstet Gynecol. 2016 Apr;127[4]:681-7).

Because of Colorado’s lengthy experience with legalized marijuana, the state Department of Public Health and Environment has endeavored to create resources of value for health care providers and patients (www.colorado.gov/cdphe/marijuana-clinical-guidelines). The website contains a fact sheet for patients regarding marijuana in pregnancy and breastfeeding. For physicians, there is plain-language guidance on how to talk effectively about marijuana with patients, including suggested responses to selected commonly voiced misconceptions.

The website also includes the results of a 2014 marijuana-in-pregnancy literature review by a state advisory committee composed of Colorado specialists in pediatrics, ob.gyn., family medicine, public health, and addiction medicine.

The committee determined that there is moderate evidence that the use of marijuana in pregnancy is associated with increased risk of reduced fetal growth, lower IQ scores in young children, adverse effects on a child’s cognitive function and academic ability, and an increase in attention problems. There was deemed to be limited evidence of an association with stillbirth and isolated ventricular septal defects. There is also “mixed” evidence for associations with preterm delivery, reduced birth weight, and selected congenital anomalies.

Since that 2014 review, a new signal of potential harm stemming from maternal marijuana use in pregnancy has appeared: a possible increased risk of neonatal ICU admission. In one retrospective study including 361 marijuana users and 6,107 nonusers, the users had a 1.54-fold increased risk for neonatal ICU admission in an analysis adjusted for maternal demographics and tobacco use (J Perinatol. 2015 Dec;35[12]:991-5).

Moreover, investigators at the University of Arizona in Tucson performed a meta-analysis of 24 studies and concluded that infants exposed to cannabis in utero were at 2.02-fold increased likelihood of neonatal ICU admission, a 1.77-fold increased risk of low birth weight, and 1.36-fold increased odds of anemia (BMJ Open. 2016 Apr 5;6[4]:e0009986. doi: 10.1136/bmjopen-2015-009986).

“That obviously would have a big public health impact,” Dr. Metz said.

In marked contrast, however, just a few months later investigators at Washington University in St. Louis reported finding no significantly increased risk of neonatal ICU admission or any other adverse neonatal outcome after adjustment for tobacco use and other potential confounders in a meta-analysis of 31 studies (Obstet Gynecol. 2016 Oct;128[4]:713-23).

These contradictory meta-analyses underscore a key point about the existing literature on the safety of marijuana use in pregnancy: It provides few, if any, definitive answers. The studies conducted in the 1980s and 1990s are of limited generalizability because concentrations of tetrahydrocannabinol were so small, compared with today’s products. Ascertainment of exposure to marijuana in pregnancy is unreliable in the absence of confirmatory biologic sampling. Self-reported use is unreliable and is typically an underestimate. Adjustment for confounders associated with adverse neonatal outcomes is challenging.

“Biologic sampling is critical,” Dr. Metz said. “We actually don’t know who’s using, and we lack information on the timing and quantity of exposure.

“Part of the problem is the data are so mixed that you can really find whatever you want in the literature to support your bias,” she added.

Still, in light of the signals of possible harm, she urged her colleagues to advise patients not to use marijuana in pregnancy. Patients need to understand that there are no known benefits of marijuana use in pregnancy, there are possible risks, and there is no known safe amount of cannabis in pregnancy.

Dr. Metz reported having no financial conflicts related to her presentation.

 

[email protected]

 

BOSTON – More than a third of 63 patients with unilateral papillary thyroid carcinoma and ipsilateral lymph node metastases also had occult, contralateral, lateral-neck lymph nodes with metastases, but the low reported rate of contralateral neck recurrence has raised the question of whether routine resection of these contralateral lymph nodes benefits patients.

“Prophylactic contralateral lateral-neck dissection may not be relevant for patients with unilateral N1b tumors,” Dana Hartl, MD, PhD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

Routine, prophylactic surgical resection of contralateral, lateral-neck lymph nodes in these patients “was our protocol for the past 20 years. I think we will change our practice” based on these findings, predicted Dr. Hartl, chief of thyroid surgery at Gustave Roussy in Paris.

She envisioned a more targeted approach to using surgical resection of contralateral, lateral-neck lymph nodes. “We would still do this surgery for patients with lots of big, bulky nodules; it just takes an additional 10-20 minutes. It’s not that long or morbid. For patients with lots of [affected] nodes, you’ve got to clear it out. But for a patient with a small, level III node on one side I will no longer do [prophylactic, contralateral, lateral-neck lymph node removal].”

Dr. Hartl and her associates reviewed records for 63 patients with unilateral, unifocal papillary thyroid carcinoma who underwent prophylactic, contralateral, lymph node removal during 1997-2016. They ranged from 11-84 years old, and 60% had extrathyroidal extension of their primary tumor. The patients averaged having four lymph nodes with metastatic cells in the ipsilateral lateral neck, an average of five affected lymph nodes in the ipsilateral central neck, and an average of two affected lymph nodes in the contralateral central neck.

Among the 63 patients, 23 (37%) had metastases-containing lymph nodes at levels III and IV in the contralateral lateral neck, with a range of 1-17 metastases per node. All nodes were less than 1 cm in diameter. Extracapsular spread had occurred in 11 of the 23 cases.

The relatively frequent presence of metastases in the contralateral, lateral-neck lymph nodes contrasted with the reported low 5%-15% rate of recurrence in these lymph nodes in patients with N1b disease, Dr. Hartl said.

The analysis identified two factors that significantly linked with having contralateral, lateral-neck metastases: having a bilateral tumor with contralateral microcancer, or having at least four lymph nodes positive for metastases in the ipsilateral central compartment, she reported. When patients had fewer than five positive lymph nodes in the ipsilateral central neck, their risk for occult metastases in contralateral lymph nodes was 26%.

“It’s not very strong data; we were a little disappointed” with the results, Dr. Hartl said.

[email protected]

On Twitter @mitchelzoler

 

BOSTON – More than a third of 63 patients with unilateral papillary thyroid carcinoma and ipsilateral lymph node metastases also had occult, contralateral, lateral-neck lymph nodes with metastases, but the low reported rate of contralateral neck recurrence has raised the question of whether routine resection of these contralateral lymph nodes benefits patients.

“Prophylactic contralateral lateral-neck dissection may not be relevant for patients with unilateral N1b tumors,” Dana Hartl, MD, PhD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

Routine, prophylactic surgical resection of contralateral, lateral-neck lymph nodes in these patients “was our protocol for the past 20 years. I think we will change our practice” based on these findings, predicted Dr. Hartl, chief of thyroid surgery at Gustave Roussy in Paris.

She envisioned a more targeted approach to using surgical resection of contralateral, lateral-neck lymph nodes. “We would still do this surgery for patients with lots of big, bulky nodules; it just takes an additional 10-20 minutes. It’s not that long or morbid. For patients with lots of [affected] nodes, you’ve got to clear it out. But for a patient with a small, level III node on one side I will no longer do [prophylactic, contralateral, lateral-neck lymph node removal].”

Dr. Hartl and her associates reviewed records for 63 patients with unilateral, unifocal papillary thyroid carcinoma who underwent prophylactic, contralateral, lymph node removal during 1997-2016. They ranged from 11-84 years old, and 60% had extrathyroidal extension of their primary tumor. The patients averaged having four lymph nodes with metastatic cells in the ipsilateral lateral neck, an average of five affected lymph nodes in the ipsilateral central neck, and an average of two affected lymph nodes in the contralateral central neck.

Among the 63 patients, 23 (37%) had metastases-containing lymph nodes at levels III and IV in the contralateral lateral neck, with a range of 1-17 metastases per node. All nodes were less than 1 cm in diameter. Extracapsular spread had occurred in 11 of the 23 cases.

The relatively frequent presence of metastases in the contralateral, lateral-neck lymph nodes contrasted with the reported low 5%-15% rate of recurrence in these lymph nodes in patients with N1b disease, Dr. Hartl said.

The analysis identified two factors that significantly linked with having contralateral, lateral-neck metastases: having a bilateral tumor with contralateral microcancer, or having at least four lymph nodes positive for metastases in the ipsilateral central compartment, she reported. When patients had fewer than five positive lymph nodes in the ipsilateral central neck, their risk for occult metastases in contralateral lymph nodes was 26%.

“It’s not very strong data; we were a little disappointed” with the results, Dr. Hartl said.

[email protected]

On Twitter @mitchelzoler

 

BOSTON – More than a third of 63 patients with unilateral papillary thyroid carcinoma and ipsilateral lymph node metastases also had occult, contralateral, lateral-neck lymph nodes with metastases, but the low reported rate of contralateral neck recurrence has raised the question of whether routine resection of these contralateral lymph nodes benefits patients.

“Prophylactic contralateral lateral-neck dissection may not be relevant for patients with unilateral N1b tumors,” Dana Hartl, MD, PhD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

Routine, prophylactic surgical resection of contralateral, lateral-neck lymph nodes in these patients “was our protocol for the past 20 years. I think we will change our practice” based on these findings, predicted Dr. Hartl, chief of thyroid surgery at Gustave Roussy in Paris.

She envisioned a more targeted approach to using surgical resection of contralateral, lateral-neck lymph nodes. “We would still do this surgery for patients with lots of big, bulky nodules; it just takes an additional 10-20 minutes. It’s not that long or morbid. For patients with lots of [affected] nodes, you’ve got to clear it out. But for a patient with a small, level III node on one side I will no longer do [prophylactic, contralateral, lateral-neck lymph node removal].”

Dr. Hartl and her associates reviewed records for 63 patients with unilateral, unifocal papillary thyroid carcinoma who underwent prophylactic, contralateral, lymph node removal during 1997-2016. They ranged from 11-84 years old, and 60% had extrathyroidal extension of their primary tumor. The patients averaged having four lymph nodes with metastatic cells in the ipsilateral lateral neck, an average of five affected lymph nodes in the ipsilateral central neck, and an average of two affected lymph nodes in the contralateral central neck.

Among the 63 patients, 23 (37%) had metastases-containing lymph nodes at levels III and IV in the contralateral lateral neck, with a range of 1-17 metastases per node. All nodes were less than 1 cm in diameter. Extracapsular spread had occurred in 11 of the 23 cases.

The relatively frequent presence of metastases in the contralateral, lateral-neck lymph nodes contrasted with the reported low 5%-15% rate of recurrence in these lymph nodes in patients with N1b disease, Dr. Hartl said.

The analysis identified two factors that significantly linked with having contralateral, lateral-neck metastases: having a bilateral tumor with contralateral microcancer, or having at least four lymph nodes positive for metastases in the ipsilateral central compartment, she reported. When patients had fewer than five positive lymph nodes in the ipsilateral central neck, their risk for occult metastases in contralateral lymph nodes was 26%.

“It’s not very strong data; we were a little disappointed” with the results, Dr. Hartl said.

[email protected]

On Twitter @mitchelzoler

Atypical vascular lesions (AVLs) of the breast are rare cutaneous vascular proliferations that present as erythematous, violaceous, or flesh-colored papules, patches, or plaques in women who have undergone radiation treatment for breast carcinoma.^1,2 These lesions most commonly develop in the irradiated area within 3 to 6 years following radiation treatment.³

Various terms have been used to describe AVLs in the literature, including atypical hemangiomas, benign lymphangiomatous papules, benign lymphangioendotheliomas, lymphangioma circumscriptum, and acquired progressive lymphangiomas, suggesting benign behavior.^4-10 However, their identity as benign lesions has been a source of controversy, with some investigators proposing that AVLs may be a precursor lesion to postirradiation angiosarcoma.² Research has addressed if there are markers that can predict AVL types that are more likely to develop into angiosarcomas.¹ Although most clinicians treat AVLs with complete excision, there currently are no specific guidelines to direct this practice.

We report the case of a patient with a history of 1 AVL that was excised who developed 3 additional AVLs in the same breast over the course of 15 months.

Case Report

A 55-year-old woman with a history of obesity, hypertension, and infiltrating ductal carcinoma in situ of the right breast (grade 2, estrogen receptor and progesterone receptor positive) underwent a right breast lumpectomy and sentinel lymph node dissection. Three months later, she underwent re-excision for positive margins and started adjuvant hormonal therapy with tamoxifen. One month later, she began external beam radiation therapy and received a total dose of 6040 cGy over the course of 9 weeks (34 total treatments).

The patient presented to an outside dermatology clinic 2 years after completing external beam radiation therapy for evaluation of a new pink nodule on the right mid breast. The nodule was biopsied and discovered to be an AVL. Pathology showed an anastomosing proliferation of thin-walled vascular channels mainly located in the superficial dermis with notable endothelial nuclear atypia and hyperchromasia. There were several tiny foci with the beginnings of multilayering with prominent endothelial atypia (Figure 1). She underwent complete excision for this AVL with negative margins.

Six months after the initial AVL diagnosis, she presented to our dermatology clinic with another asymptomatic red bump on the right breast. On physical examination, a 4-mm firm, erythematous, well-circumscribed papule was noted on the medial aspect of the right breast along with a similar-appearing 4-mm papule on the right lateral aspect of the right breast (Figure 2). The patient was unsure of the duration of the second lesion but felt that it had been present at least as long as the other lesion. Both lesions clinically resembled typical capillary hemangiomas. A 6-mm punch biopsy of the right medial breast was performed and revealed enlarged vessels and capillaries in the upper dermis lined by endothelial cells with focal prominent nuclei without necrosis, overt atypia, mitosis, or tufting (Figure 3). Immunostaining was positive for CD34, factor VIII antigen, podoplanin (D2-40), and CD31, and negative for cytokeratin 7 and pankeratin. This staining was compatible with a lymphatic-type AVL.¹ A diagnosis of AVL was made and complete excision with clear margins was performed. At the time of this excision, a biopsy of the right lateral breast was performed revealing thin-walled, dilated vascular channels in the superficial dermis with architecturally atypical angulated outlines, mild endothelial nuclear atypia, and hyperchromasia without endothelial multilayering. Clear margins were noted on the biopsy, but the patient subsequently declined re-excision of this third AVL.

During a subsequent follow-up visit 9 months later, the patient was noted to have a 2-mm red, vascular-appearing papule on the right upper medial breast (Figure 2). A 6-mm biopsy was performed and revealed thin-walled vascular channels in the superficial dermis with endothelial nuclear atypia consistent with an AVL.

 

Comment

Fineberg and Rosen⁸ were the first to describe AVLs in their 1994 study of 4 women with cutaneous vascular proliferations that developed after radiation and chemotherapy for breast cancer. They concluded that these AVLs were benign lesions distinct from angiosarcomas.⁸ However, further research has challenged the benign nature of AVLs. In 2005, Brenn and Fletcher² studied 42 women diagnosed with either angiosarcoma or atypical radiation-associated cutaneous vascular lesions. They suggested that AVLs resided on the same spectrum as angiosarcomas and that AVLs may be precursor lesions to angiosarcomas.² Furthermore, Hildebrandt et al¹¹ in 2001 and Di Tommaso and Fabbri¹² in 2003 published case reports of individual patients who developed an angiosarcoma from a preexisting AVL.

The controversy continued when Patton et al¹ published a study in 2008 in which 32 cases of AVLs were reviewed. In this study, 2 histologic types of AVLs were described: vascular type and lymphatic type. Vascular-type AVLs are characterized by irregularly dispersed, pericyte-invested, capillary-sized vessels within the papillary or reticular dermis that often are associated with extravasated erythrocytes or hemosiderin. On the other hand, lymphatic-type AVLs display thin-walled, variably anastomosing, lymphatic vessels lined by attenuated or slightly protuberant endothelial cells. These subtypes have been suggested based on the antigens known to be present in certain tissues, specifically vascular and lymphatic tissue. Despite these seemingly distinct histologies, 6 lesions classified as vascular type displayed some histologic overlap with the lymphatic-type AVLs. The authors concluded that the vascular type showed greater potential to develop into an angiosarcoma based on the degree of endothelial atypia.¹

In 2011, Santi et al¹³ found that both AVLs and angiosarcomas share inactivation mutations in the tumor suppressor gene TP53, providing further evidence to suggest that AVLs may be precursors to angiosarcomas.

Although the malignant potential of AVLs remains questionable, research has shown that they do have a propensity to recur.³ In 2007, Gengler et al³ determined that 20% of patients with AVLs experienced recurrence after a biopsy or excision with varying margins; however, the group stated that these new vascular lesions might not be recurrences but rather entirely new lesions in the same irradiated field (field-effect phenomenon). Several other studies demonstrated that more than 30% of patients with 1 AVL developed more lesions within the same irradiated area.^3,14-16 Despite the high rate of recurrence documented in the literature, only 5 of more than 100 diagnosed AVLs have progressed to angiosarcoma.^1,3

Many differences can be noted when comparing the histology of AVLs versus angiosarcomas, though some are subtle (Table). Angiosarcomas display poorly circumscribed vascular infiltration into the subcutaneous tissue, multilayering of endothelial cells, prominent nucleoli, hemorrhage, mitoses, and notable aytpia. Atypical vascular lesions lack these features and tend to be wedge shaped and display chronic inflammation.^8,15,17-19 Atypical vascular lesions show superficial localized growth without destruction of adjacent adnexa, display dilated vascular spaces, and exhibit large endothelial cells.^{5,6,8,14,15,19,20} However, there is overlap between AVLs and angiosarcomas that can make diagnosis difficult.^{2,14,16,17,19} Areas within or just outside of an angiosarcoma, especially in well-differentiated angiosarcomas, can appear histologically identical to AVLs, and multiple biopsies may be required for diagnosis.^17,19,21

Conclusion

More research is needed in the arenas of classification, diagnosis, treatment, and follow-up recommendations for AVLs. In particular, more specific histologic markers may be needed to identify those AVLs that may progress to angiosarcomas. Although most AVLs are treated with excision, a consensus needs to be reached on adequate surgical margins. Lastly, due to the tendency of AVLs to recur coupled with their unknown malignant potential, recommendations are needed for consistent follow-up examinations.

Atypical vascular lesions (AVLs) of the breast are rare cutaneous vascular proliferations that present as erythematous, violaceous, or flesh-colored papules, patches, or plaques in women who have undergone radiation treatment for breast carcinoma.^1,2 These lesions most commonly develop in the irradiated area within 3 to 6 years following radiation treatment.³

Various terms have been used to describe AVLs in the literature, including atypical hemangiomas, benign lymphangiomatous papules, benign lymphangioendotheliomas, lymphangioma circumscriptum, and acquired progressive lymphangiomas, suggesting benign behavior.^4-10 However, their identity as benign lesions has been a source of controversy, with some investigators proposing that AVLs may be a precursor lesion to postirradiation angiosarcoma.² Research has addressed if there are markers that can predict AVL types that are more likely to develop into angiosarcomas.¹ Although most clinicians treat AVLs with complete excision, there currently are no specific guidelines to direct this practice.

We report the case of a patient with a history of 1 AVL that was excised who developed 3 additional AVLs in the same breast over the course of 15 months.

Case Report

A 55-year-old woman with a history of obesity, hypertension, and infiltrating ductal carcinoma in situ of the right breast (grade 2, estrogen receptor and progesterone receptor positive) underwent a right breast lumpectomy and sentinel lymph node dissection. Three months later, she underwent re-excision for positive margins and started adjuvant hormonal therapy with tamoxifen. One month later, she began external beam radiation therapy and received a total dose of 6040 cGy over the course of 9 weeks (34 total treatments).

The patient presented to an outside dermatology clinic 2 years after completing external beam radiation therapy for evaluation of a new pink nodule on the right mid breast. The nodule was biopsied and discovered to be an AVL. Pathology showed an anastomosing proliferation of thin-walled vascular channels mainly located in the superficial dermis with notable endothelial nuclear atypia and hyperchromasia. There were several tiny foci with the beginnings of multilayering with prominent endothelial atypia (Figure 1). She underwent complete excision for this AVL with negative margins.

Six months after the initial AVL diagnosis, she presented to our dermatology clinic with another asymptomatic red bump on the right breast. On physical examination, a 4-mm firm, erythematous, well-circumscribed papule was noted on the medial aspect of the right breast along with a similar-appearing 4-mm papule on the right lateral aspect of the right breast (Figure 2). The patient was unsure of the duration of the second lesion but felt that it had been present at least as long as the other lesion. Both lesions clinically resembled typical capillary hemangiomas. A 6-mm punch biopsy of the right medial breast was performed and revealed enlarged vessels and capillaries in the upper dermis lined by endothelial cells with focal prominent nuclei without necrosis, overt atypia, mitosis, or tufting (Figure 3). Immunostaining was positive for CD34, factor VIII antigen, podoplanin (D2-40), and CD31, and negative for cytokeratin 7 and pankeratin. This staining was compatible with a lymphatic-type AVL.¹ A diagnosis of AVL was made and complete excision with clear margins was performed. At the time of this excision, a biopsy of the right lateral breast was performed revealing thin-walled, dilated vascular channels in the superficial dermis with architecturally atypical angulated outlines, mild endothelial nuclear atypia, and hyperchromasia without endothelial multilayering. Clear margins were noted on the biopsy, but the patient subsequently declined re-excision of this third AVL.

During a subsequent follow-up visit 9 months later, the patient was noted to have a 2-mm red, vascular-appearing papule on the right upper medial breast (Figure 2). A 6-mm biopsy was performed and revealed thin-walled vascular channels in the superficial dermis with endothelial nuclear atypia consistent with an AVL.

 

Comment

Fineberg and Rosen⁸ were the first to describe AVLs in their 1994 study of 4 women with cutaneous vascular proliferations that developed after radiation and chemotherapy for breast cancer. They concluded that these AVLs were benign lesions distinct from angiosarcomas.⁸ However, further research has challenged the benign nature of AVLs. In 2005, Brenn and Fletcher² studied 42 women diagnosed with either angiosarcoma or atypical radiation-associated cutaneous vascular lesions. They suggested that AVLs resided on the same spectrum as angiosarcomas and that AVLs may be precursor lesions to angiosarcomas.² Furthermore, Hildebrandt et al¹¹ in 2001 and Di Tommaso and Fabbri¹² in 2003 published case reports of individual patients who developed an angiosarcoma from a preexisting AVL.

The controversy continued when Patton et al¹ published a study in 2008 in which 32 cases of AVLs were reviewed. In this study, 2 histologic types of AVLs were described: vascular type and lymphatic type. Vascular-type AVLs are characterized by irregularly dispersed, pericyte-invested, capillary-sized vessels within the papillary or reticular dermis that often are associated with extravasated erythrocytes or hemosiderin. On the other hand, lymphatic-type AVLs display thin-walled, variably anastomosing, lymphatic vessels lined by attenuated or slightly protuberant endothelial cells. These subtypes have been suggested based on the antigens known to be present in certain tissues, specifically vascular and lymphatic tissue. Despite these seemingly distinct histologies, 6 lesions classified as vascular type displayed some histologic overlap with the lymphatic-type AVLs. The authors concluded that the vascular type showed greater potential to develop into an angiosarcoma based on the degree of endothelial atypia.¹

In 2011, Santi et al¹³ found that both AVLs and angiosarcomas share inactivation mutations in the tumor suppressor gene TP53, providing further evidence to suggest that AVLs may be precursors to angiosarcomas.

Although the malignant potential of AVLs remains questionable, research has shown that they do have a propensity to recur.³ In 2007, Gengler et al³ determined that 20% of patients with AVLs experienced recurrence after a biopsy or excision with varying margins; however, the group stated that these new vascular lesions might not be recurrences but rather entirely new lesions in the same irradiated field (field-effect phenomenon). Several other studies demonstrated that more than 30% of patients with 1 AVL developed more lesions within the same irradiated area.^3,14-16 Despite the high rate of recurrence documented in the literature, only 5 of more than 100 diagnosed AVLs have progressed to angiosarcoma.^1,3

Many differences can be noted when comparing the histology of AVLs versus angiosarcomas, though some are subtle (Table). Angiosarcomas display poorly circumscribed vascular infiltration into the subcutaneous tissue, multilayering of endothelial cells, prominent nucleoli, hemorrhage, mitoses, and notable aytpia. Atypical vascular lesions lack these features and tend to be wedge shaped and display chronic inflammation.^8,15,17-19 Atypical vascular lesions show superficial localized growth without destruction of adjacent adnexa, display dilated vascular spaces, and exhibit large endothelial cells.^{5,6,8,14,15,19,20} However, there is overlap between AVLs and angiosarcomas that can make diagnosis difficult.^{2,14,16,17,19} Areas within or just outside of an angiosarcoma, especially in well-differentiated angiosarcomas, can appear histologically identical to AVLs, and multiple biopsies may be required for diagnosis.^17,19,21

Conclusion

More research is needed in the arenas of classification, diagnosis, treatment, and follow-up recommendations for AVLs. In particular, more specific histologic markers may be needed to identify those AVLs that may progress to angiosarcomas. Although most AVLs are treated with excision, a consensus needs to be reached on adequate surgical margins. Lastly, due to the tendency of AVLs to recur coupled with their unknown malignant potential, recommendations are needed for consistent follow-up examinations.

Atypical vascular lesions (AVLs) of the breast are rare cutaneous vascular proliferations that present as erythematous, violaceous, or flesh-colored papules, patches, or plaques in women who have undergone radiation treatment for breast carcinoma.^1,2 These lesions most commonly develop in the irradiated area within 3 to 6 years following radiation treatment.³

Various terms have been used to describe AVLs in the literature, including atypical hemangiomas, benign lymphangiomatous papules, benign lymphangioendotheliomas, lymphangioma circumscriptum, and acquired progressive lymphangiomas, suggesting benign behavior.^4-10 However, their identity as benign lesions has been a source of controversy, with some investigators proposing that AVLs may be a precursor lesion to postirradiation angiosarcoma.² Research has addressed if there are markers that can predict AVL types that are more likely to develop into angiosarcomas.¹ Although most clinicians treat AVLs with complete excision, there currently are no specific guidelines to direct this practice.

We report the case of a patient with a history of 1 AVL that was excised who developed 3 additional AVLs in the same breast over the course of 15 months.

Case Report

A 55-year-old woman with a history of obesity, hypertension, and infiltrating ductal carcinoma in situ of the right breast (grade 2, estrogen receptor and progesterone receptor positive) underwent a right breast lumpectomy and sentinel lymph node dissection. Three months later, she underwent re-excision for positive margins and started adjuvant hormonal therapy with tamoxifen. One month later, she began external beam radiation therapy and received a total dose of 6040 cGy over the course of 9 weeks (34 total treatments).

The patient presented to an outside dermatology clinic 2 years after completing external beam radiation therapy for evaluation of a new pink nodule on the right mid breast. The nodule was biopsied and discovered to be an AVL. Pathology showed an anastomosing proliferation of thin-walled vascular channels mainly located in the superficial dermis with notable endothelial nuclear atypia and hyperchromasia. There were several tiny foci with the beginnings of multilayering with prominent endothelial atypia (Figure 1). She underwent complete excision for this AVL with negative margins.

Six months after the initial AVL diagnosis, she presented to our dermatology clinic with another asymptomatic red bump on the right breast. On physical examination, a 4-mm firm, erythematous, well-circumscribed papule was noted on the medial aspect of the right breast along with a similar-appearing 4-mm papule on the right lateral aspect of the right breast (Figure 2). The patient was unsure of the duration of the second lesion but felt that it had been present at least as long as the other lesion. Both lesions clinically resembled typical capillary hemangiomas. A 6-mm punch biopsy of the right medial breast was performed and revealed enlarged vessels and capillaries in the upper dermis lined by endothelial cells with focal prominent nuclei without necrosis, overt atypia, mitosis, or tufting (Figure 3). Immunostaining was positive for CD34, factor VIII antigen, podoplanin (D2-40), and CD31, and negative for cytokeratin 7 and pankeratin. This staining was compatible with a lymphatic-type AVL.¹ A diagnosis of AVL was made and complete excision with clear margins was performed. At the time of this excision, a biopsy of the right lateral breast was performed revealing thin-walled, dilated vascular channels in the superficial dermis with architecturally atypical angulated outlines, mild endothelial nuclear atypia, and hyperchromasia without endothelial multilayering. Clear margins were noted on the biopsy, but the patient subsequently declined re-excision of this third AVL.

During a subsequent follow-up visit 9 months later, the patient was noted to have a 2-mm red, vascular-appearing papule on the right upper medial breast (Figure 2). A 6-mm biopsy was performed and revealed thin-walled vascular channels in the superficial dermis with endothelial nuclear atypia consistent with an AVL.

 

Comment

Fineberg and Rosen⁸ were the first to describe AVLs in their 1994 study of 4 women with cutaneous vascular proliferations that developed after radiation and chemotherapy for breast cancer. They concluded that these AVLs were benign lesions distinct from angiosarcomas.⁸ However, further research has challenged the benign nature of AVLs. In 2005, Brenn and Fletcher² studied 42 women diagnosed with either angiosarcoma or atypical radiation-associated cutaneous vascular lesions. They suggested that AVLs resided on the same spectrum as angiosarcomas and that AVLs may be precursor lesions to angiosarcomas.² Furthermore, Hildebrandt et al¹¹ in 2001 and Di Tommaso and Fabbri¹² in 2003 published case reports of individual patients who developed an angiosarcoma from a preexisting AVL.

The controversy continued when Patton et al¹ published a study in 2008 in which 32 cases of AVLs were reviewed. In this study, 2 histologic types of AVLs were described: vascular type and lymphatic type. Vascular-type AVLs are characterized by irregularly dispersed, pericyte-invested, capillary-sized vessels within the papillary or reticular dermis that often are associated with extravasated erythrocytes or hemosiderin. On the other hand, lymphatic-type AVLs display thin-walled, variably anastomosing, lymphatic vessels lined by attenuated or slightly protuberant endothelial cells. These subtypes have been suggested based on the antigens known to be present in certain tissues, specifically vascular and lymphatic tissue. Despite these seemingly distinct histologies, 6 lesions classified as vascular type displayed some histologic overlap with the lymphatic-type AVLs. The authors concluded that the vascular type showed greater potential to develop into an angiosarcoma based on the degree of endothelial atypia.¹

In 2011, Santi et al¹³ found that both AVLs and angiosarcomas share inactivation mutations in the tumor suppressor gene TP53, providing further evidence to suggest that AVLs may be precursors to angiosarcomas.

Although the malignant potential of AVLs remains questionable, research has shown that they do have a propensity to recur.³ In 2007, Gengler et al³ determined that 20% of patients with AVLs experienced recurrence after a biopsy or excision with varying margins; however, the group stated that these new vascular lesions might not be recurrences but rather entirely new lesions in the same irradiated field (field-effect phenomenon). Several other studies demonstrated that more than 30% of patients with 1 AVL developed more lesions within the same irradiated area.^3,14-16 Despite the high rate of recurrence documented in the literature, only 5 of more than 100 diagnosed AVLs have progressed to angiosarcoma.^1,3

Many differences can be noted when comparing the histology of AVLs versus angiosarcomas, though some are subtle (Table). Angiosarcomas display poorly circumscribed vascular infiltration into the subcutaneous tissue, multilayering of endothelial cells, prominent nucleoli, hemorrhage, mitoses, and notable aytpia. Atypical vascular lesions lack these features and tend to be wedge shaped and display chronic inflammation.^8,15,17-19 Atypical vascular lesions show superficial localized growth without destruction of adjacent adnexa, display dilated vascular spaces, and exhibit large endothelial cells.^{5,6,8,14,15,19,20} However, there is overlap between AVLs and angiosarcomas that can make diagnosis difficult.^{2,14,16,17,19} Areas within or just outside of an angiosarcoma, especially in well-differentiated angiosarcomas, can appear histologically identical to AVLs, and multiple biopsies may be required for diagnosis.^17,19,21

Conclusion

More research is needed in the arenas of classification, diagnosis, treatment, and follow-up recommendations for AVLs. In particular, more specific histologic markers may be needed to identify those AVLs that may progress to angiosarcomas. Although most AVLs are treated with excision, a consensus needs to be reached on adequate surgical margins. Lastly, due to the tendency of AVLs to recur coupled with their unknown malignant potential, recommendations are needed for consistent follow-up examinations.

 

Epilepsy estimates available for the first time for every state show that the disorder is widespread, with at least 3.4 million people affected, according to the Centers for Disease Control and Prevention.

The CDC data also show that the number of people with epilepsy is increasing, probably as a result of population growth. The number of affected adults went from 2.3 million in 2010 to 3 million in 2015, and the number of children with epilepsy rose from 450,000 in 2007 to 470,000 in 2015, CDC investigators reported (MMWR. 2017 Aug 11;66[31]:821-5).

The state estimates show that Mississippi has the highest epilepsy rate in the country at 1,194 cases per 100,000 population in 2015, followed by West Virginia (1,174) and Louisiana (1,173). Utah has the lowest rate at 960 cases per 100,000 population, with North Dakota next at 963 per 100,000 and Alaska third at 970. (The CDC report provided the number of cases per state, so rates given here are Frontline calculations using population estimates from the Census Bureau.)

“Millions of Americans are impacted by epilepsy, and unfortunately, this study shows cases are on the rise,” CDC Director Brenda Fitzgerald said in a separate statement. “Proper diagnosis is key to finding an effective treatment – and at CDC we are committed to researching, testing, and sharing strategies that will improve the lives of people with epilepsy.”

The CDC investigators based their estimates for children under age 18 years on data from the 2011-2012 National Survey of Children’s Health; estimates for those age 18 and over are based on data from the 2015 National Health Interview Survey.

[email protected]

 

Epilepsy estimates available for the first time for every state show that the disorder is widespread, with at least 3.4 million people affected, according to the Centers for Disease Control and Prevention.

The CDC data also show that the number of people with epilepsy is increasing, probably as a result of population growth. The number of affected adults went from 2.3 million in 2010 to 3 million in 2015, and the number of children with epilepsy rose from 450,000 in 2007 to 470,000 in 2015, CDC investigators reported (MMWR. 2017 Aug 11;66[31]:821-5).

The state estimates show that Mississippi has the highest epilepsy rate in the country at 1,194 cases per 100,000 population in 2015, followed by West Virginia (1,174) and Louisiana (1,173). Utah has the lowest rate at 960 cases per 100,000 population, with North Dakota next at 963 per 100,000 and Alaska third at 970. (The CDC report provided the number of cases per state, so rates given here are Frontline calculations using population estimates from the Census Bureau.)

“Millions of Americans are impacted by epilepsy, and unfortunately, this study shows cases are on the rise,” CDC Director Brenda Fitzgerald said in a separate statement. “Proper diagnosis is key to finding an effective treatment – and at CDC we are committed to researching, testing, and sharing strategies that will improve the lives of people with epilepsy.”

The CDC investigators based their estimates for children under age 18 years on data from the 2011-2012 National Survey of Children’s Health; estimates for those age 18 and over are based on data from the 2015 National Health Interview Survey.

[email protected]

 

Epilepsy estimates available for the first time for every state show that the disorder is widespread, with at least 3.4 million people affected, according to the Centers for Disease Control and Prevention.

The CDC data also show that the number of people with epilepsy is increasing, probably as a result of population growth. The number of affected adults went from 2.3 million in 2010 to 3 million in 2015, and the number of children with epilepsy rose from 450,000 in 2007 to 470,000 in 2015, CDC investigators reported (MMWR. 2017 Aug 11;66[31]:821-5).

The state estimates show that Mississippi has the highest epilepsy rate in the country at 1,194 cases per 100,000 population in 2015, followed by West Virginia (1,174) and Louisiana (1,173). Utah has the lowest rate at 960 cases per 100,000 population, with North Dakota next at 963 per 100,000 and Alaska third at 970. (The CDC report provided the number of cases per state, so rates given here are Frontline calculations using population estimates from the Census Bureau.)

“Millions of Americans are impacted by epilepsy, and unfortunately, this study shows cases are on the rise,” CDC Director Brenda Fitzgerald said in a separate statement. “Proper diagnosis is key to finding an effective treatment – and at CDC we are committed to researching, testing, and sharing strategies that will improve the lives of people with epilepsy.”

The CDC investigators based their estimates for children under age 18 years on data from the 2011-2012 National Survey of Children’s Health; estimates for those age 18 and over are based on data from the 2015 National Health Interview Survey.

[email protected]

 

BOSTON – Under the newly revised Bethesda System for Reporting Thyroid Cytology, slated for official release in October 2017, the six cytology-based diagnostic categories for thyroid lesions stay exactly the same as in the 10-year-old first edition, but some associated malignancy risks have changed.

Important changes include molecular testing to further assess malignancy risk in thyroid nodules and the introduction of lobectomy as a treatment option, “which really wasn’t an option 10 years ago,” in the first iteration of the Bethesda System (New York: Springer US, 2010), its coauthor Edmund S. Cibas, MD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

He singled out reframing the malignancy risks for some of the six cytology categories as a top message of the revision, and he attributed these changes to two main factors: routine molecular testing, and creation of a new diagnostic category, the “noninvasive follicular thyroid neoplasm with papillary-like nuclear features” (NIFTP).

An Endocrine Pathology Society working group created the NIFTP designation in 2016 to describe an encapsulated follicular variant of papillary thyroid carcinoma that is characterized by lack of invasion, a follicular growth pattern, and nuclear features of papillary thyroid carcinoma with a very low risk of an adverse outcome (JAMA Oncology. 2016 Aug;2[8]:1023-9) (Cancer Cytopathol. 2016 Sep;124[9]:616-20).

NIFTP is not an overt malignancy. The revised Bethesda System “limits malignancy to cases with features of classic malignant papillary thyroid carcinoma,” explained Dr. Cibas, professor of pathology at Harvard Medical School and director of cytopathology at Brigham and Women’s Hospital, both in Boston.

Because the Bethesda System categories link to specific management recommendations, the new edition orients patients toward more conservative management decisions, specifically lobectomies instead of total thyroidectomies, he said in an interview.

The International Cytology Congress held a symposium during its meeting in Yokohama, Japan, in 2016, which resulted in the second edition of the Bethesda System (ACTA Cytol. 2016 Sep-Oct; 60[5]:399-405).

The changes in risk of malignancy occurred primarily in two categories, either “atypia of undetermined significance” (AUS) or “follicular lesions of undetermined significance” (FLUS). The risk of malignancy jumped from 5%-15% in the Bethesda System first edition up to 10%-30% in the revision. A smaller bump-up hit the category of “follicular neoplasm” or “suspicious for follicular neoplasm,” in which the risk of malignancy increased from 20%-30% in the first edition to 25%-40% in the revision. And, in the suspicion of malignancy category, the risk of malignancy actually lowered modestly, easing from 60%-75% in the first edition to 50%-75% in the revision.

Dr. Cibas highlighted the AUS/FLUS category with further notable features. The limit on laboratories reporting this category increased to 10% of total reports, up from 7% in the first edition. Management changed from the single options of a repeat fine-needle aspiration specimen to either that or molecular testing. Also, “the first edition was not clear that AUS and FLUS are synonyms. That will be a lot clearer” in the second edition, Dr. Cibas promised. The revision “will encourage labs that currently use [the terms] AUS and FLUS to mean two different things to make a choice between them.”

Another quirk of the AUS and FLUS category is that the risk of malignancy estimates are based on what Dr. Cibas called “flawed” data from only the selected subset of AUS or FLUS patients who have their nodule resected. “The reality is that most of the nodules are not resected” from patients with AUS or FLUS, so conclusions about the risk of malignancy come from a subset with considerable selection bias.

The definition of “follicular neoplasm” or “suspicious for follicular neoplasm” category also added “mild nuclear changes,” which can include increased nuclear size, contour irregularity, or chromatin clearing. The “suspicious for malignancy” category made a modest tweak to the risk of malignancy. Plus, “some of these patients will now undergo lobectomy rather than total thyroidectomy, which has been usual management.

The “suspicious for malignant” and “malignant” categories had little change aside from wider use of lobectomy, now feasible for any patient except those with metastatic disease, Dr. Cibas said.

[email protected]

On Twitter @mitchelzoler

 

BOSTON – Under the newly revised Bethesda System for Reporting Thyroid Cytology, slated for official release in October 2017, the six cytology-based diagnostic categories for thyroid lesions stay exactly the same as in the 10-year-old first edition, but some associated malignancy risks have changed.

Important changes include molecular testing to further assess malignancy risk in thyroid nodules and the introduction of lobectomy as a treatment option, “which really wasn’t an option 10 years ago,” in the first iteration of the Bethesda System (New York: Springer US, 2010), its coauthor Edmund S. Cibas, MD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

He singled out reframing the malignancy risks for some of the six cytology categories as a top message of the revision, and he attributed these changes to two main factors: routine molecular testing, and creation of a new diagnostic category, the “noninvasive follicular thyroid neoplasm with papillary-like nuclear features” (NIFTP).

An Endocrine Pathology Society working group created the NIFTP designation in 2016 to describe an encapsulated follicular variant of papillary thyroid carcinoma that is characterized by lack of invasion, a follicular growth pattern, and nuclear features of papillary thyroid carcinoma with a very low risk of an adverse outcome (JAMA Oncology. 2016 Aug;2[8]:1023-9) (Cancer Cytopathol. 2016 Sep;124[9]:616-20).

NIFTP is not an overt malignancy. The revised Bethesda System “limits malignancy to cases with features of classic malignant papillary thyroid carcinoma,” explained Dr. Cibas, professor of pathology at Harvard Medical School and director of cytopathology at Brigham and Women’s Hospital, both in Boston.

Because the Bethesda System categories link to specific management recommendations, the new edition orients patients toward more conservative management decisions, specifically lobectomies instead of total thyroidectomies, he said in an interview.

The International Cytology Congress held a symposium during its meeting in Yokohama, Japan, in 2016, which resulted in the second edition of the Bethesda System (ACTA Cytol. 2016 Sep-Oct; 60[5]:399-405).

The changes in risk of malignancy occurred primarily in two categories, either “atypia of undetermined significance” (AUS) or “follicular lesions of undetermined significance” (FLUS). The risk of malignancy jumped from 5%-15% in the Bethesda System first edition up to 10%-30% in the revision. A smaller bump-up hit the category of “follicular neoplasm” or “suspicious for follicular neoplasm,” in which the risk of malignancy increased from 20%-30% in the first edition to 25%-40% in the revision. And, in the suspicion of malignancy category, the risk of malignancy actually lowered modestly, easing from 60%-75% in the first edition to 50%-75% in the revision.

Dr. Cibas highlighted the AUS/FLUS category with further notable features. The limit on laboratories reporting this category increased to 10% of total reports, up from 7% in the first edition. Management changed from the single options of a repeat fine-needle aspiration specimen to either that or molecular testing. Also, “the first edition was not clear that AUS and FLUS are synonyms. That will be a lot clearer” in the second edition, Dr. Cibas promised. The revision “will encourage labs that currently use [the terms] AUS and FLUS to mean two different things to make a choice between them.”

Another quirk of the AUS and FLUS category is that the risk of malignancy estimates are based on what Dr. Cibas called “flawed” data from only the selected subset of AUS or FLUS patients who have their nodule resected. “The reality is that most of the nodules are not resected” from patients with AUS or FLUS, so conclusions about the risk of malignancy come from a subset with considerable selection bias.

The definition of “follicular neoplasm” or “suspicious for follicular neoplasm” category also added “mild nuclear changes,” which can include increased nuclear size, contour irregularity, or chromatin clearing. The “suspicious for malignancy” category made a modest tweak to the risk of malignancy. Plus, “some of these patients will now undergo lobectomy rather than total thyroidectomy, which has been usual management.

The “suspicious for malignant” and “malignant” categories had little change aside from wider use of lobectomy, now feasible for any patient except those with metastatic disease, Dr. Cibas said.

[email protected]

On Twitter @mitchelzoler

 

BOSTON – Under the newly revised Bethesda System for Reporting Thyroid Cytology, slated for official release in October 2017, the six cytology-based diagnostic categories for thyroid lesions stay exactly the same as in the 10-year-old first edition, but some associated malignancy risks have changed.

Important changes include molecular testing to further assess malignancy risk in thyroid nodules and the introduction of lobectomy as a treatment option, “which really wasn’t an option 10 years ago,” in the first iteration of the Bethesda System (New York: Springer US, 2010), its coauthor Edmund S. Cibas, MD, said at the World Congress on Thyroid Cancer.

Mitchel L. Zoler/Frontline Medical News

He singled out reframing the malignancy risks for some of the six cytology categories as a top message of the revision, and he attributed these changes to two main factors: routine molecular testing, and creation of a new diagnostic category, the “noninvasive follicular thyroid neoplasm with papillary-like nuclear features” (NIFTP).

An Endocrine Pathology Society working group created the NIFTP designation in 2016 to describe an encapsulated follicular variant of papillary thyroid carcinoma that is characterized by lack of invasion, a follicular growth pattern, and nuclear features of papillary thyroid carcinoma with a very low risk of an adverse outcome (JAMA Oncology. 2016 Aug;2[8]:1023-9) (Cancer Cytopathol. 2016 Sep;124[9]:616-20).

NIFTP is not an overt malignancy. The revised Bethesda System “limits malignancy to cases with features of classic malignant papillary thyroid carcinoma,” explained Dr. Cibas, professor of pathology at Harvard Medical School and director of cytopathology at Brigham and Women’s Hospital, both in Boston.

Because the Bethesda System categories link to specific management recommendations, the new edition orients patients toward more conservative management decisions, specifically lobectomies instead of total thyroidectomies, he said in an interview.

The International Cytology Congress held a symposium during its meeting in Yokohama, Japan, in 2016, which resulted in the second edition of the Bethesda System (ACTA Cytol. 2016 Sep-Oct; 60[5]:399-405).

The changes in risk of malignancy occurred primarily in two categories, either “atypia of undetermined significance” (AUS) or “follicular lesions of undetermined significance” (FLUS). The risk of malignancy jumped from 5%-15% in the Bethesda System first edition up to 10%-30% in the revision. A smaller bump-up hit the category of “follicular neoplasm” or “suspicious for follicular neoplasm,” in which the risk of malignancy increased from 20%-30% in the first edition to 25%-40% in the revision. And, in the suspicion of malignancy category, the risk of malignancy actually lowered modestly, easing from 60%-75% in the first edition to 50%-75% in the revision.

Dr. Cibas highlighted the AUS/FLUS category with further notable features. The limit on laboratories reporting this category increased to 10% of total reports, up from 7% in the first edition. Management changed from the single options of a repeat fine-needle aspiration specimen to either that or molecular testing. Also, “the first edition was not clear that AUS and FLUS are synonyms. That will be a lot clearer” in the second edition, Dr. Cibas promised. The revision “will encourage labs that currently use [the terms] AUS and FLUS to mean two different things to make a choice between them.”

Another quirk of the AUS and FLUS category is that the risk of malignancy estimates are based on what Dr. Cibas called “flawed” data from only the selected subset of AUS or FLUS patients who have their nodule resected. “The reality is that most of the nodules are not resected” from patients with AUS or FLUS, so conclusions about the risk of malignancy come from a subset with considerable selection bias.

The definition of “follicular neoplasm” or “suspicious for follicular neoplasm” category also added “mild nuclear changes,” which can include increased nuclear size, contour irregularity, or chromatin clearing. The “suspicious for malignancy” category made a modest tweak to the risk of malignancy. Plus, “some of these patients will now undergo lobectomy rather than total thyroidectomy, which has been usual management.

The “suspicious for malignant” and “malignant” categories had little change aside from wider use of lobectomy, now feasible for any patient except those with metastatic disease, Dr. Cibas said.

[email protected]

On Twitter @mitchelzoler

 

ESTES PARK, COLO. – Everyone diagnosed with osteoporosis deserves a laboratory assessment to rule out unsuspected secondary causes, according to Sterling West, MD. And he’s got a doozy of a workup he recommends to primary care physicians as “incredibly cost effective.”

“With this workup you’ll identify 98% of abnormalities at a mean cost of $366 per diagnosis. That’s incredibly cost effective. You’re going to get a lot of information with actually not very much outlay at all,” he said at a conference on internal medicine sponsored by the University of Colorado.

Bruce Jancin/Frontline Medical News

Applying this laboratory screening regimen to all patients diagnosed with osteoporosis is warranted because unsuspected secondary causes of the skeletal disease are so common. In various studies, laboratory screening has revealed a secondary cause in up to one third of postmenopausal women with osteoporosis, in up to half of men, and in 50%-80% of premenopausal osteoporosis patients, noted Dr. West, professor of medicine at the university.

The tests he advocates that primary care physicians order in all their patients with osteoporosis include a complete blood count, a complete metabolic panel, a 24-hour urine calcium/sodium/creatinine, a serum 25-hydroxyvitamin D level, and a serum phosphorus. In addition, men with osteoporosis should have their serum testosterone measured. A thyroid-stimulating hormone level should be obtained in patients who are taking thyroxine or if they look clinically hyperthyroid.

A measurement of parathyroid hormone is warranted as part of the screen in patients with an abnormal serum calcium. If the parathyroid hormone is normal, hyperparathyroidism can be ruled out.

Ordering a serum protein electrophoresis to check for multiple myeloma is appropriate in osteoporotic patients over age 50 years with an abnormal complete blood count.

This basic laboratory workup will identify patients with the relatively common secondary causes of low bone mineral density which account for 98% of all cases. These causes include vitamin D deficiency, malabsorption, hypogonadism, hypercalciuria, and myeloma.

“Leave the other 2% to me,” the rheumatologist suggested.

Special laboratory tests Dr. Sterling recommended that are best left to bone disease specialists include bone turnover markers, a serum tryptase/urine N-methylhistamine to screen for systemic mastocytosis, antitransglutaminase antibodies for celiac disease, a 24-hour urinary free cortisol and/or overnight dexamethasone suppression test to identify patients with Cushing syndrome, and bone biopsy.

Who should be referred to a bone specialist for a more extensive workup?

“If somebody is losing bone or fracturing and they’re on appropriate therapy and you believe they’re taking their medication, that’s for sure somebody that we should see. Also, a premenopausal woman with a high Z score who has had a fracture that’s atypical. And patients with stage 4 or 5 chronic kidney disease; those are some of the toughest cases and are best referred to a bone expert,” Dr. Sterling said.

On the other hand, if an osteoporotic patient simply can’t tolerate guideline-recommended initial therapy with an oral bisphosphonate such as alendronate (Fosamax) or risedronate (Actonel), there’s no need to bring in a specialist. Simply switch the patient to denosumab (Prolia), a monoclonal antibody against receptor activator of nuclear factor kappa-B ligand, administered by subcutaneous injection once every 6 months. The cost is about $2,200 per year, but the drug is covered by Medicare Part B. Clinical trials have demonstrated that denosumab boosts bone mineral density by 6%-9%, with an absolute 5% reduction in fractures and a 40%-68% relative risk reduction, he noted.

Dr. West reported having no financial conflicts of interest regarding his presentation.

[email protected]
 

 

ESTES PARK, COLO. – Everyone diagnosed with osteoporosis deserves a laboratory assessment to rule out unsuspected secondary causes, according to Sterling West, MD. And he’s got a doozy of a workup he recommends to primary care physicians as “incredibly cost effective.”

“With this workup you’ll identify 98% of abnormalities at a mean cost of $366 per diagnosis. That’s incredibly cost effective. You’re going to get a lot of information with actually not very much outlay at all,” he said at a conference on internal medicine sponsored by the University of Colorado.

Bruce Jancin/Frontline Medical News

Applying this laboratory screening regimen to all patients diagnosed with osteoporosis is warranted because unsuspected secondary causes of the skeletal disease are so common. In various studies, laboratory screening has revealed a secondary cause in up to one third of postmenopausal women with osteoporosis, in up to half of men, and in 50%-80% of premenopausal osteoporosis patients, noted Dr. West, professor of medicine at the university.

The tests he advocates that primary care physicians order in all their patients with osteoporosis include a complete blood count, a complete metabolic panel, a 24-hour urine calcium/sodium/creatinine, a serum 25-hydroxyvitamin D level, and a serum phosphorus. In addition, men with osteoporosis should have their serum testosterone measured. A thyroid-stimulating hormone level should be obtained in patients who are taking thyroxine or if they look clinically hyperthyroid.

A measurement of parathyroid hormone is warranted as part of the screen in patients with an abnormal serum calcium. If the parathyroid hormone is normal, hyperparathyroidism can be ruled out.

Ordering a serum protein electrophoresis to check for multiple myeloma is appropriate in osteoporotic patients over age 50 years with an abnormal complete blood count.

This basic laboratory workup will identify patients with the relatively common secondary causes of low bone mineral density which account for 98% of all cases. These causes include vitamin D deficiency, malabsorption, hypogonadism, hypercalciuria, and myeloma.

“Leave the other 2% to me,” the rheumatologist suggested.

Special laboratory tests Dr. Sterling recommended that are best left to bone disease specialists include bone turnover markers, a serum tryptase/urine N-methylhistamine to screen for systemic mastocytosis, antitransglutaminase antibodies for celiac disease, a 24-hour urinary free cortisol and/or overnight dexamethasone suppression test to identify patients with Cushing syndrome, and bone biopsy.

Who should be referred to a bone specialist for a more extensive workup?

“If somebody is losing bone or fracturing and they’re on appropriate therapy and you believe they’re taking their medication, that’s for sure somebody that we should see. Also, a premenopausal woman with a high Z score who has had a fracture that’s atypical. And patients with stage 4 or 5 chronic kidney disease; those are some of the toughest cases and are best referred to a bone expert,” Dr. Sterling said.

On the other hand, if an osteoporotic patient simply can’t tolerate guideline-recommended initial therapy with an oral bisphosphonate such as alendronate (Fosamax) or risedronate (Actonel), there’s no need to bring in a specialist. Simply switch the patient to denosumab (Prolia), a monoclonal antibody against receptor activator of nuclear factor kappa-B ligand, administered by subcutaneous injection once every 6 months. The cost is about $2,200 per year, but the drug is covered by Medicare Part B. Clinical trials have demonstrated that denosumab boosts bone mineral density by 6%-9%, with an absolute 5% reduction in fractures and a 40%-68% relative risk reduction, he noted.

Dr. West reported having no financial conflicts of interest regarding his presentation.

[email protected]
 

 

ESTES PARK, COLO. – Everyone diagnosed with osteoporosis deserves a laboratory assessment to rule out unsuspected secondary causes, according to Sterling West, MD. And he’s got a doozy of a workup he recommends to primary care physicians as “incredibly cost effective.”

“With this workup you’ll identify 98% of abnormalities at a mean cost of $366 per diagnosis. That’s incredibly cost effective. You’re going to get a lot of information with actually not very much outlay at all,” he said at a conference on internal medicine sponsored by the University of Colorado.

Bruce Jancin/Frontline Medical News

Applying this laboratory screening regimen to all patients diagnosed with osteoporosis is warranted because unsuspected secondary causes of the skeletal disease are so common. In various studies, laboratory screening has revealed a secondary cause in up to one third of postmenopausal women with osteoporosis, in up to half of men, and in 50%-80% of premenopausal osteoporosis patients, noted Dr. West, professor of medicine at the university.

The tests he advocates that primary care physicians order in all their patients with osteoporosis include a complete blood count, a complete metabolic panel, a 24-hour urine calcium/sodium/creatinine, a serum 25-hydroxyvitamin D level, and a serum phosphorus. In addition, men with osteoporosis should have their serum testosterone measured. A thyroid-stimulating hormone level should be obtained in patients who are taking thyroxine or if they look clinically hyperthyroid.

A measurement of parathyroid hormone is warranted as part of the screen in patients with an abnormal serum calcium. If the parathyroid hormone is normal, hyperparathyroidism can be ruled out.

Ordering a serum protein electrophoresis to check for multiple myeloma is appropriate in osteoporotic patients over age 50 years with an abnormal complete blood count.

This basic laboratory workup will identify patients with the relatively common secondary causes of low bone mineral density which account for 98% of all cases. These causes include vitamin D deficiency, malabsorption, hypogonadism, hypercalciuria, and myeloma.

“Leave the other 2% to me,” the rheumatologist suggested.

Special laboratory tests Dr. Sterling recommended that are best left to bone disease specialists include bone turnover markers, a serum tryptase/urine N-methylhistamine to screen for systemic mastocytosis, antitransglutaminase antibodies for celiac disease, a 24-hour urinary free cortisol and/or overnight dexamethasone suppression test to identify patients with Cushing syndrome, and bone biopsy.

Who should be referred to a bone specialist for a more extensive workup?

“If somebody is losing bone or fracturing and they’re on appropriate therapy and you believe they’re taking their medication, that’s for sure somebody that we should see. Also, a premenopausal woman with a high Z score who has had a fracture that’s atypical. And patients with stage 4 or 5 chronic kidney disease; those are some of the toughest cases and are best referred to a bone expert,” Dr. Sterling said.

On the other hand, if an osteoporotic patient simply can’t tolerate guideline-recommended initial therapy with an oral bisphosphonate such as alendronate (Fosamax) or risedronate (Actonel), there’s no need to bring in a specialist. Simply switch the patient to denosumab (Prolia), a monoclonal antibody against receptor activator of nuclear factor kappa-B ligand, administered by subcutaneous injection once every 6 months. The cost is about $2,200 per year, but the drug is covered by Medicare Part B. Clinical trials have demonstrated that denosumab boosts bone mineral density by 6%-9%, with an absolute 5% reduction in fractures and a 40%-68% relative risk reduction, he noted.

Dr. West reported having no financial conflicts of interest regarding his presentation.

[email protected]
 

 

PARIS – Five-year hemodynamic results of the first randomized trial of transcatheter versus surgical aortic valve replacement in low-surgical-risk patients with severe aortic stenosis showed continued superior valve performance in the TAVR group, Lars Sondergaard, MD, reported at the annual congress of the European Association of Percutaneous Cardiovascular Interventions.

“The durability results are very encouraging. We can’t see that the TAVR patients are doing worse. So I think this is setting the scene to try to move forward in patients at low risk and also in younger patients,” declared Dr. Sondergaard, professor of cardiology at the University of Copenhagen.

He presented an update from the Nordic Aortic Valve Intervention (NOTION) trial, a prospective, multicenter, randomized, all-comers clinical trial in which 280 patients with symptomatic severe aortic stenosis at low surgical risk were assigned to surgical aortic valve replacement (SAVR) or to TAVR with the self-expanding CoreValve. Their mean age was 79 years, with an average Society of Thoracic Surgeons projected risk of mortality score of 3%. Eighty-two percent of participants had an STS score below 4%. Roughly 40% of TAVR patients got the first-generation CoreValve in the 26-mm size, 40% received the 29-mm version, and the rest got the 31-mm CoreValve.

Bruce Jancin/Frontline Medical News

With 94% compliance with follow-up through 4 years post procedure, the primary clinical endpoint – a composite of all-cause mortality, MI, and stroke – had occurred in 29.1% of the TAVR group and was similar at 30.2% in the SAVR group. The all-cause mortality rate was 20% in the TAVR group, compared with 23% in the SAVR cohort, a nonsignificant difference.

Among patients in the lowest-surgical-risk and youngest subgroup – those aged 70-75 with a Society of Thoracic Surgeons risk score below 4% – the composite primary endpoint rate at 4 years was 15.6% with TAVR compared with 27.2% with SAVR. However, only 62 NOTION participants fell into this category, so the between-group difference, while sizable, didn’t achieve statistical significance, according to Dr. Sondergaard.

There was a trade-off between the two valve replacement strategies with regard to procedural complications. The rate of new-onset atrial fibrillation was far higher in the SAVR group: 59.4% at 1 year and 60.2% at 4 years of follow-up, compared with 21.2% and 24.5% at 1 and 4 years, respectively, in the TAVR group.

On the other hand, 38% of the TAVR patients got a new pacemaker within the first year of follow-up, compared with only 2.4% in the SAVR group. At 4 years, 43.7% of the TAVR group had a pacemaker, versus 9% of the SAVR group.

Turning to the hemodynamic data, the cardiologist noted that the effective orifice area in the TAVR group went from 0.71 cm² at baseline to 1.66 at 1 year and remained steady thereafter at 1.67 cm² through 5 years. The TAVR group’s mean gradient improved from 45.4 mm Hg at baseline to 8.6 mm Hg at 1 year and 7.9 mm Hg at 5 years. These outcomes were significantly better than in the SAVR group, where the effective orifice area went from 0.74 cm² at baseline to 1.32 at 1 year and 1.24 cm² at 5 years, while the mean gradient fell from 44.9 mm Hg to 12.5 at 1 year and 13.6 mm Hg at 5 years.

Moderate hemodynamic structural valve deterioration was significantly more common in the SAVR group: 20.7% at 5 years, compared with 2.9% in the TAVR patients. The opposite was true with regard to moderate paravalvular leak, which occurred in 20.9% of the TAVI group but only 1.5% of SAVR patients.

Late complications were rare following either procedure. There were no cases of valve thrombosis through 5 years. The incidence of endocarditis at 5 years was 4.3% in the TAVR patients and similar at 5.9% in the SAVR group.

Discussant Samer Mansour, MD, of the University of Montreal, remarked that the rate of new pacemaker implantation following TAVR seemed extraordinarily high.

“This was early days,” Dr. Sondergaard explained. “We had a lower threshold for putting in a pacemaker and we put the valves in a little deeper.”

About half of new pacemaker recipients didn’t use the device after the first year, he added. Also, neither getting a new pacemaker nor moderate paravalvular leak was associated with increased mortality in the TAVR group.

Dr. Mansour observed that subtle but real differences in mortality probably wouldn’t show up in a 280-patient trial. Dr. Sondergaard concurred.

“We designed the NOTION trial in 2008-2009. Knowing what we know now, we should have had a larger study, but at that time TAVR volume wasn’t that big and it wasn’t realistic as a Nordic trial to include 1,000 patients. This was the best we could do,” he said.

Follow-up in the NOTION study will continue out to 10 years.

The study is funded by Medtronic. Dr. Sondergaard reported serving as a consultant to and receiving research grant support from the company.

 

[email protected]

 

PARIS – Five-year hemodynamic results of the first randomized trial of transcatheter versus surgical aortic valve replacement in low-surgical-risk patients with severe aortic stenosis showed continued superior valve performance in the TAVR group, Lars Sondergaard, MD, reported at the annual congress of the European Association of Percutaneous Cardiovascular Interventions.

“The durability results are very encouraging. We can’t see that the TAVR patients are doing worse. So I think this is setting the scene to try to move forward in patients at low risk and also in younger patients,” declared Dr. Sondergaard, professor of cardiology at the University of Copenhagen.

He presented an update from the Nordic Aortic Valve Intervention (NOTION) trial, a prospective, multicenter, randomized, all-comers clinical trial in which 280 patients with symptomatic severe aortic stenosis at low surgical risk were assigned to surgical aortic valve replacement (SAVR) or to TAVR with the self-expanding CoreValve. Their mean age was 79 years, with an average Society of Thoracic Surgeons projected risk of mortality score of 3%. Eighty-two percent of participants had an STS score below 4%. Roughly 40% of TAVR patients got the first-generation CoreValve in the 26-mm size, 40% received the 29-mm version, and the rest got the 31-mm CoreValve.

Bruce Jancin/Frontline Medical News

With 94% compliance with follow-up through 4 years post procedure, the primary clinical endpoint – a composite of all-cause mortality, MI, and stroke – had occurred in 29.1% of the TAVR group and was similar at 30.2% in the SAVR group. The all-cause mortality rate was 20% in the TAVR group, compared with 23% in the SAVR cohort, a nonsignificant difference.

Among patients in the lowest-surgical-risk and youngest subgroup – those aged 70-75 with a Society of Thoracic Surgeons risk score below 4% – the composite primary endpoint rate at 4 years was 15.6% with TAVR compared with 27.2% with SAVR. However, only 62 NOTION participants fell into this category, so the between-group difference, while sizable, didn’t achieve statistical significance, according to Dr. Sondergaard.

There was a trade-off between the two valve replacement strategies with regard to procedural complications. The rate of new-onset atrial fibrillation was far higher in the SAVR group: 59.4% at 1 year and 60.2% at 4 years of follow-up, compared with 21.2% and 24.5% at 1 and 4 years, respectively, in the TAVR group.

On the other hand, 38% of the TAVR patients got a new pacemaker within the first year of follow-up, compared with only 2.4% in the SAVR group. At 4 years, 43.7% of the TAVR group had a pacemaker, versus 9% of the SAVR group.

Turning to the hemodynamic data, the cardiologist noted that the effective orifice area in the TAVR group went from 0.71 cm² at baseline to 1.66 at 1 year and remained steady thereafter at 1.67 cm² through 5 years. The TAVR group’s mean gradient improved from 45.4 mm Hg at baseline to 8.6 mm Hg at 1 year and 7.9 mm Hg at 5 years. These outcomes were significantly better than in the SAVR group, where the effective orifice area went from 0.74 cm² at baseline to 1.32 at 1 year and 1.24 cm² at 5 years, while the mean gradient fell from 44.9 mm Hg to 12.5 at 1 year and 13.6 mm Hg at 5 years.

Moderate hemodynamic structural valve deterioration was significantly more common in the SAVR group: 20.7% at 5 years, compared with 2.9% in the TAVR patients. The opposite was true with regard to moderate paravalvular leak, which occurred in 20.9% of the TAVI group but only 1.5% of SAVR patients.

Late complications were rare following either procedure. There were no cases of valve thrombosis through 5 years. The incidence of endocarditis at 5 years was 4.3% in the TAVR patients and similar at 5.9% in the SAVR group.

Discussant Samer Mansour, MD, of the University of Montreal, remarked that the rate of new pacemaker implantation following TAVR seemed extraordinarily high.

“This was early days,” Dr. Sondergaard explained. “We had a lower threshold for putting in a pacemaker and we put the valves in a little deeper.”

About half of new pacemaker recipients didn’t use the device after the first year, he added. Also, neither getting a new pacemaker nor moderate paravalvular leak was associated with increased mortality in the TAVR group.

Dr. Mansour observed that subtle but real differences in mortality probably wouldn’t show up in a 280-patient trial. Dr. Sondergaard concurred.

“We designed the NOTION trial in 2008-2009. Knowing what we know now, we should have had a larger study, but at that time TAVR volume wasn’t that big and it wasn’t realistic as a Nordic trial to include 1,000 patients. This was the best we could do,” he said.

Follow-up in the NOTION study will continue out to 10 years.

The study is funded by Medtronic. Dr. Sondergaard reported serving as a consultant to and receiving research grant support from the company.

 

[email protected]

 

PARIS – Five-year hemodynamic results of the first randomized trial of transcatheter versus surgical aortic valve replacement in low-surgical-risk patients with severe aortic stenosis showed continued superior valve performance in the TAVR group, Lars Sondergaard, MD, reported at the annual congress of the European Association of Percutaneous Cardiovascular Interventions.

“The durability results are very encouraging. We can’t see that the TAVR patients are doing worse. So I think this is setting the scene to try to move forward in patients at low risk and also in younger patients,” declared Dr. Sondergaard, professor of cardiology at the University of Copenhagen.

He presented an update from the Nordic Aortic Valve Intervention (NOTION) trial, a prospective, multicenter, randomized, all-comers clinical trial in which 280 patients with symptomatic severe aortic stenosis at low surgical risk were assigned to surgical aortic valve replacement (SAVR) or to TAVR with the self-expanding CoreValve. Their mean age was 79 years, with an average Society of Thoracic Surgeons projected risk of mortality score of 3%. Eighty-two percent of participants had an STS score below 4%. Roughly 40% of TAVR patients got the first-generation CoreValve in the 26-mm size, 40% received the 29-mm version, and the rest got the 31-mm CoreValve.

Bruce Jancin/Frontline Medical News

With 94% compliance with follow-up through 4 years post procedure, the primary clinical endpoint – a composite of all-cause mortality, MI, and stroke – had occurred in 29.1% of the TAVR group and was similar at 30.2% in the SAVR group. The all-cause mortality rate was 20% in the TAVR group, compared with 23% in the SAVR cohort, a nonsignificant difference.

Among patients in the lowest-surgical-risk and youngest subgroup – those aged 70-75 with a Society of Thoracic Surgeons risk score below 4% – the composite primary endpoint rate at 4 years was 15.6% with TAVR compared with 27.2% with SAVR. However, only 62 NOTION participants fell into this category, so the between-group difference, while sizable, didn’t achieve statistical significance, according to Dr. Sondergaard.

There was a trade-off between the two valve replacement strategies with regard to procedural complications. The rate of new-onset atrial fibrillation was far higher in the SAVR group: 59.4% at 1 year and 60.2% at 4 years of follow-up, compared with 21.2% and 24.5% at 1 and 4 years, respectively, in the TAVR group.

On the other hand, 38% of the TAVR patients got a new pacemaker within the first year of follow-up, compared with only 2.4% in the SAVR group. At 4 years, 43.7% of the TAVR group had a pacemaker, versus 9% of the SAVR group.

Turning to the hemodynamic data, the cardiologist noted that the effective orifice area in the TAVR group went from 0.71 cm² at baseline to 1.66 at 1 year and remained steady thereafter at 1.67 cm² through 5 years. The TAVR group’s mean gradient improved from 45.4 mm Hg at baseline to 8.6 mm Hg at 1 year and 7.9 mm Hg at 5 years. These outcomes were significantly better than in the SAVR group, where the effective orifice area went from 0.74 cm² at baseline to 1.32 at 1 year and 1.24 cm² at 5 years, while the mean gradient fell from 44.9 mm Hg to 12.5 at 1 year and 13.6 mm Hg at 5 years.

Moderate hemodynamic structural valve deterioration was significantly more common in the SAVR group: 20.7% at 5 years, compared with 2.9% in the TAVR patients. The opposite was true with regard to moderate paravalvular leak, which occurred in 20.9% of the TAVI group but only 1.5% of SAVR patients.

Late complications were rare following either procedure. There were no cases of valve thrombosis through 5 years. The incidence of endocarditis at 5 years was 4.3% in the TAVR patients and similar at 5.9% in the SAVR group.

Discussant Samer Mansour, MD, of the University of Montreal, remarked that the rate of new pacemaker implantation following TAVR seemed extraordinarily high.

“This was early days,” Dr. Sondergaard explained. “We had a lower threshold for putting in a pacemaker and we put the valves in a little deeper.”

About half of new pacemaker recipients didn’t use the device after the first year, he added. Also, neither getting a new pacemaker nor moderate paravalvular leak was associated with increased mortality in the TAVR group.

Dr. Mansour observed that subtle but real differences in mortality probably wouldn’t show up in a 280-patient trial. Dr. Sondergaard concurred.

“We designed the NOTION trial in 2008-2009. Knowing what we know now, we should have had a larger study, but at that time TAVR volume wasn’t that big and it wasn’t realistic as a Nordic trial to include 1,000 patients. This was the best we could do,” he said.

Follow-up in the NOTION study will continue out to 10 years.

The study is funded by Medtronic. Dr. Sondergaard reported serving as a consultant to and receiving research grant support from the company.

 

[email protected]

Hospitals typically allocate beds based on historical patient volumes. If funding decreases, hospitals will usually try to maximize resource utilization by allocating beds to attain occupancies close to 100% for significant periods of time. This will invariably cause days in which hospital occupancy exceeds capacity, at which time critical entry points (such as the emergency department and operating room) will become blocked. This creates significant concerns over the patient quality of care.

Hospital administrators have very few options when hospital occupancy exceeds 100%. They could postpone admissions for “planned” cases, bring in additional staff to increase capacity, or instigate additional methods to increase hospital discharges such as expanding care resources in the community. All options are costly, bothersome, or cannot be actioned immediately. The need for these options could be minimized by enabling hospital administrators to make more informed decisions regarding hospital bed management by knowing the likely number of discharges in the next 24 hours.

Predicting the number of people who will be discharged in the next day can be approached in several ways. One approach would be to calculate each patient’s expected length of stay and then use the variation around that estimate to calculate each day’s discharge probability. Several studies have attempted to model hospital length of stay using a broad assortment of methodologies, but a mechanism to accurately predict this outcome has been elusive^1,2 (with Verburg et al.³ concluding in their study’s abstract that “…it is difficult to predict length of stay…”). A second approach would be to use survival analysis methods to generate each patient’s hazard of discharge over time, which could be directly converted to an expected daily risk of discharge. However, this approach is complicated by the concurrent need to include time-dependent covariates and consider the competing risk of death in hospital, which can complicate survival modeling.^4,5 A third approach would be the implementation of a longitudinal analysis using marginal models to predict the daily probability of discharge,⁶ but this method quickly overwhelms computer resources when large datasets are present.

In this study, we decided to use nonparametric models to predict the daily number of hospital discharges. We first identified patient groups with distinct discharge patterns. We then calculated the conditional daily discharge probability of patients in each of these groups. Finally, these conditional daily discharge probabilities were then summed for each hospital day to generate the expected number of discharges in the next 24 hours. This paper details the methods we used to create our model and the accuracy of its predictions.

 

METHODS

Study Setting and Databases Used for Analysis

The study took place at The Ottawa Hospital, a 1000-bed teaching hospital with 3 campuses that is the primary referral center in our region. The study was approved by our local research ethics board.

The Patient Registry Database records the date and time of admission for each patient (defined as the moment that a patient’s admission request is registered in the patient registration) and discharge (defined as the time when the patient’s discharge from hospital was entered into the patient registration) for hospital encounters. Emergency department encounters were also identified in the Patient Registry Database along with admission service, patient age and sex, and patient location throughout the admission. The Laboratory Database records all laboratory studies and results on all patients at the hospital.

Study Cohort

We used the Patient Registry Database to identify all people aged 1 year or more who were admitted to the hospital between January 1, 2013, and December 31, 2015. This time frame was selected to (i) ensure that data were complete; and (ii) complete calendar years of data were available for both derivation (patient-days in 2013-2014) and validation (2015) cohorts. Patients who were observed in the emergency room without admission to hospital were not included.

Study Outcome

The study outcome was the number of patients discharged from the hospital each day. For the analysis, the reference point for each day was 1 second past midnight; therefore, values for time-dependent covariates up to and including midnight were used to predict the number of discharges in the next 24 hours.

Study Covariates

Baseline (ie, time-independent) covariates included patient age and sex, admission service, hospital campus, whether or not the patient was admitted from the emergency department (all determined from the Patient Registry Database), and the Laboratory-based Acute Physiological Score (LAPS). The latter, which was calculated with the Laboratory Database using results for 14 tests (arterial pH, PaCO2, PaO2, anion gap, hematocrit, total white blood cell count, serum albumin, total bilirubin, creatinine, urea nitrogen, glucose, sodium, bicarbonate, and troponin I) measured in the 24-hour time frame preceding hospitalization, was derived by Escobar and colleagues⁷ to measure severity of illness and was subsequently validated in our hospital.⁸ The independent association of each laboratory perturbation with risk of death in hospital is reflected by the number of points assigned to each lab value with the total LAPS being the sum of these values. Time-dependent covariates included weekday in hospital and whether or not patients were in the intensive care unit.

Analysis

We used 3 stages to create a model to predict the daily expected number of discharges: we identified discharge risk strata containing patients having similar discharge patterns using data from patients in the derivation cohort (first stage); then, we generated the preliminary probability of discharge by determining the daily discharge probability in each discharge risk strata (second stage); finally, we modified the probability from the second stage based on the weekday and admission service and summed these probabilities to create the expected number of discharges on a particular date (third stage).

The first stage identified discharge risk strata based on the covariates listed above. This was determined by using a survival tree approach⁹ with proportional hazard regression models to generate the “splits.” These models were offered all covariates listed in the Study Covariates section. Admission service was clustered within 4 departments (obstetrics/gynecology, psychiatry, surgery, and medicine) and day of week was “binarized” into weekday/weekend-holiday (because the use of categorical variables with large numbers of groups can “stunt” regression trees due to small numbers of patients—and, therefore, statistical power—in each subgroup). The proportional hazards model identified the covariate having the strongest association with time to discharge (based on the Wald X² value divided by the degrees of freedom). This variable was then used to split the cohort into subgroups (with continuous covariates being categorized into quartiles). The proportional hazards model was then repeated in each subgroup (with the previous splitting variable[s] excluded from the model). This process continued until no variable was associated with time to discharge with a P value less than .0001. This survival-tree was then used to cluster all patients into distinct discharge risk strata.

In the second stage, we generated the preliminary probability of discharge for a specific date. This was calculated by assigning all patients in hospital to their discharge risk strata (Appendix). We then measured the probability of discharge on each hospitalization day in all discharge risk strata using data from the previous 180 days (we only used the prior 180 days of data to account for temporal changes in hospital discharge patterns). For example, consider a 75-year-old patient on her third hospital day under obstetrics/gynecology on December 19, 2015 (a Saturday). This patient would be assigned to risk stratum #133 (Appendix A). We then measured the probability of discharge of all patients in this discharge risk stratum hospitalized in the previous 6 months (ie, between June 22, 2015, and December 18, 2015) on each hospital day. For risk stratum #133, the probability of discharge on hospital day 3 was 0.1111; therefore, our sample patient’s preliminary expected discharge probability was 0.1111.

To attain stable daily discharge probability estimates, a minimum of 50 patients per discharge risk stratum-hospitalization day combination was required. If there were less than 50 patients for a particular hospitalization day in a particular discharge risk stratum, we grouped hospitalization days in that risk stratum together until the minimum of 50 patients was collected.

The third (and final) stage accounted for the lack of granularity when we created the discharge risk strata in the first stage. As we mentioned above, admission service was clustered into 4 departments and the day of week was clustered into weekend/weekday. However, important variations in discharge probabilities could still exist within departments and between particular days of the week.¹⁰ Therefore, we created a correction factor to adjust the preliminary expected number of discharges based on the admission division and day of week. This correction factor used data from the 180 days prior to the analysis date within which the expected daily number of discharges was calculated (using the methods above). The correction factor was the relative difference between the observed and expected number of discharges within each division-day of week grouping.

For example, to calculate the correction factor for our sample patient presented above (75-year-old patient on hospital day 3 under gynecology on Saturday, December 19, 2015), we measured the observed number of discharges from gynecology on Saturdays between June 22, 2015, and December 18, 2015, (n = 206) and the expected number of discharges (n = 195.255) resulting in a correction factor of (observed-expected)/expected = (195.255-206)/195.206 = 0.05503. Therefore, the final expected discharge probability for our sample patient was 0.1111+0.1111*0.05503=0.1172. The expected number of discharges on a particular date was the preliminary expected number of discharges on that date (generated in the second stage) multiplied by the correction factor for the corresponding division-day or week group.

 

RESULTS

There were 192,859 admissions involving patients more than 1 year of age that spent at least part of their hospitalization between January 1, 2013, and December 31, 2015 (Table). Patients were middle-aged and slightly female predominant, with about half being admitted from the emergency department. Approximately 80% of admissions were to surgical or medical services. More than 95% of admissions ended with a discharge from the hospital with the remainder ending in a death. Almost 30% of hospitalization days occurred on weekends or holidays. Hospitalizations in the derivation (2013-2014) and validation (2015) group were essentially the same, except there was a slight drop in hospital length of stay (from a median of 4 days to 3 days) between the 2 periods.

Patient and hospital covariates importantly influenced the daily conditional probability of discharge (Figure 1). Patients admitted to the obstetrics/gynecology department were notably more likely to be discharged from hospital with no influence from the day of week. In contrast, the probability of discharge decreased notably on the weekends in the other departments. Patients on the ward were much more likely to be discharged than those in the intensive care unit, with increasing age associated with a decreased discharge likelihood in the former but not the latter patients. Finally, discharge probabilities varied only slightly between campuses at our hospital with discharge risk decreasing as severity of illness (as measured by LAPS) increased.

The TEND model contained 142 discharge risk strata (Appendix A). Weekend-holiday status had the strongest association with discharge probability (ie, it was the first splitting variable). The most complex discharge risk strata contained 6 covariates. The daily conditional probability of discharge during the first 2 weeks of hospitalization varied extensively between discharge risk strata (Figure 2). Overall, the conditional discharge probability increased from the first to the second day, remained relatively stable for several days, and then slowly decreased over time. However, this pattern and day-to-day variability differed extensively between risk strata.

The observed daily number of discharges in the validation cohort varied extensively (median 139; interquartile range [IQR] 95-160; range 39-214). The TEND model accurately predicted the daily number of discharges with the expected daily number being strongly associated with the observed number (adjusted R² = 89.2%; P < 0.0001; Figure 3). Calibration decreased but remained significant when we limited the analyses by hospital campus (General: R² = 46.3%; P < 0.0001; Civic: R² = 47.9%; P < 0.0001; Heart Institute: R² = 18.1%; P < 0.0001). The expected number of daily discharges was an unbiased estimator of the observed number of discharges (its parameter estimate in a linear regression model with the observed number of discharges as the outcome variable was 1.0005; 95% confidence interval, 0.9647-1.0363). The absolute difference in the observed and expected daily number of discharges was small (median 1.6; IQR −6.8 to 9.4; range −37 to 63.4) as was the relative difference (median 1.4%; IQR −5.5% to 7.1%; range −40.9% to 43.4%). The expected number of discharges was within 20% of the observed number of discharges in 95.1% of days in 2015.

DISCUSSION

Knowing how many patients will soon be discharged from the hospital should greatly facilitate hospital planning. This study showed that the TEND model used simple patient and hospitalization covariates to accurately predict the number of patients who will be discharged from hospital in the next day.

We believe that this study has several notable findings. First, we think that using a nonparametric approach to predicting the daily number of discharges importantly increased accuracy. This approach allowed us to generate expected likelihoods based on actual discharge probabilities at our hospital in the most recent 6 months of hospitalization-days within patients having discharge patterns that were very similar to the patient in question (ie, discharge risk strata, Appendix A). This ensured that trends in hospitalization habits were accounted for without the need of a period variable in our model. In addition, the lack of parameters in the model will make it easier to transplant it to other hospitals. Second, we think that the accuracy of the predictions were remarkable given the relative “crudeness” of our predictors. By using relatively simple factors, the TEND model was able to output accurate predictions for the number of daily discharges (Figure 3).

This study joins several others that have attempted to accomplish the difficult task of predicting the number of hospital discharges by using digitized data. Barnes et al.¹¹ created a model using regression random forest methods in a single medical service within a hospital to predict the daily number of discharges with impressive accuracy (mean daily number of discharges observed 8.29, expected 8.51). Interestingly, the model in this study was more accurate at predicting discharge likelihood than physicians. Levin et al.¹² derived a model using discrete time logistic regression to predict the likelihood of discharge from a pediatric intensive care unit, finding that physician orders (captured via electronic order entry) could be categorized and used to significantly increase the accuracy of discharge likelihood. This study demonstrates the potential opportunities within health-related data from hospital data warehouses to improve prediction. We believe that continued work in this field will result in the increased use of digital data to help hospital administrators manage patient beds more efficiently and effectively than currently used resource intensive manual methods.^13,14

Several issues should be kept in mind when interpreting our findings. First, our analysis is limited to a single institution in Canada. It will be important to determine if the TEND model methodology generalizes to other hospitals in different jurisdictions. Such an external validation, especially in multiple hospitals, will be important to show that the TEND model methodology works in other facilities. Hospitals could implement the TEND model if they are able to record daily values for each of the variables required to assign patients to a discharge risk stratum (Appendix A) and calculate within each the daily probability of discharge. Hospitals could derive their own discharge risk strata to account for covariates, which we did not include in our study but could be influential, such as insurance status. These discharge risk estimates could also be incorporated into the electronic medical record or hospital dashboards (as long as the data required to generate the estimates are available). These interventions would permit the expected number of hospital discharges (and even the patient-level probability of discharge) to be calculated on a daily basis. Second, 2 potential biases could have influenced the identification of our discharge risk strata (Appendix A). In this process, we used survival tree methods to separate patient-days into clusters having progressively more homogenous discharge patterns. Each split was determined by using a proportional hazards model that ignored the competing risks of death in hospital. In addition, the model expressed age and LAPS as continuous variables, whereas these covariates had to be categorized to create our risk strata groupings. The strength of a covariate’s association with an outcome will decrease when a continuous variable is categorized.¹⁵ Both of these issues might have biased our final risk strata categorization (Appendix A). Third, we limited our model to include simple covariates whose values could be determined relatively easily within most hospital administrative data systems. While this increases the generalizability to other hospital information systems, we believe that the introduction of other covariates to the model—such as daily vital signs, laboratory results, medications, or time from operations—could increase prediction accuracy. Finally, it is uncertain whether or not knowing the predicted number of discharges will improve the efficiency of bed management within the hospital. It seems logical that an accurate prediction of the number of beds that will be made available in the next day should improve decisions regarding the number of patients who could be admitted electively to the hospital. It remains to be seen, however, whether this truly happens.

In summary, we found that the TEND model used a handful of patient and hospitalization factors to accurately predict the expected number of discharges from hospital in the next day. Further work is required to implement this model into our institution’s data warehouse and then determine whether this prediction will improve the efficiency of bed management at our hospital.

 

Disclosure: CvW is supported by a University of Ottawa Department of Medicine Clinician Scientist Chair. The authors have no conflicts of interest

 

Hospitals typically allocate beds based on historical patient volumes. If funding decreases, hospitals will usually try to maximize resource utilization by allocating beds to attain occupancies close to 100% for significant periods of time. This will invariably cause days in which hospital occupancy exceeds capacity, at which time critical entry points (such as the emergency department and operating room) will become blocked. This creates significant concerns over the patient quality of care.

Hospital administrators have very few options when hospital occupancy exceeds 100%. They could postpone admissions for “planned” cases, bring in additional staff to increase capacity, or instigate additional methods to increase hospital discharges such as expanding care resources in the community. All options are costly, bothersome, or cannot be actioned immediately. The need for these options could be minimized by enabling hospital administrators to make more informed decisions regarding hospital bed management by knowing the likely number of discharges in the next 24 hours.

Predicting the number of people who will be discharged in the next day can be approached in several ways. One approach would be to calculate each patient’s expected length of stay and then use the variation around that estimate to calculate each day’s discharge probability. Several studies have attempted to model hospital length of stay using a broad assortment of methodologies, but a mechanism to accurately predict this outcome has been elusive^1,2 (with Verburg et al.³ concluding in their study’s abstract that “…it is difficult to predict length of stay…”). A second approach would be to use survival analysis methods to generate each patient’s hazard of discharge over time, which could be directly converted to an expected daily risk of discharge. However, this approach is complicated by the concurrent need to include time-dependent covariates and consider the competing risk of death in hospital, which can complicate survival modeling.^4,5 A third approach would be the implementation of a longitudinal analysis using marginal models to predict the daily probability of discharge,⁶ but this method quickly overwhelms computer resources when large datasets are present.

In this study, we decided to use nonparametric models to predict the daily number of hospital discharges. We first identified patient groups with distinct discharge patterns. We then calculated the conditional daily discharge probability of patients in each of these groups. Finally, these conditional daily discharge probabilities were then summed for each hospital day to generate the expected number of discharges in the next 24 hours. This paper details the methods we used to create our model and the accuracy of its predictions.

 

METHODS

Study Setting and Databases Used for Analysis

The study took place at The Ottawa Hospital, a 1000-bed teaching hospital with 3 campuses that is the primary referral center in our region. The study was approved by our local research ethics board.

The Patient Registry Database records the date and time of admission for each patient (defined as the moment that a patient’s admission request is registered in the patient registration) and discharge (defined as the time when the patient’s discharge from hospital was entered into the patient registration) for hospital encounters. Emergency department encounters were also identified in the Patient Registry Database along with admission service, patient age and sex, and patient location throughout the admission. The Laboratory Database records all laboratory studies and results on all patients at the hospital.

Study Cohort

We used the Patient Registry Database to identify all people aged 1 year or more who were admitted to the hospital between January 1, 2013, and December 31, 2015. This time frame was selected to (i) ensure that data were complete; and (ii) complete calendar years of data were available for both derivation (patient-days in 2013-2014) and validation (2015) cohorts. Patients who were observed in the emergency room without admission to hospital were not included.

Study Outcome

The study outcome was the number of patients discharged from the hospital each day. For the analysis, the reference point for each day was 1 second past midnight; therefore, values for time-dependent covariates up to and including midnight were used to predict the number of discharges in the next 24 hours.

Study Covariates

Baseline (ie, time-independent) covariates included patient age and sex, admission service, hospital campus, whether or not the patient was admitted from the emergency department (all determined from the Patient Registry Database), and the Laboratory-based Acute Physiological Score (LAPS). The latter, which was calculated with the Laboratory Database using results for 14 tests (arterial pH, PaCO2, PaO2, anion gap, hematocrit, total white blood cell count, serum albumin, total bilirubin, creatinine, urea nitrogen, glucose, sodium, bicarbonate, and troponin I) measured in the 24-hour time frame preceding hospitalization, was derived by Escobar and colleagues⁷ to measure severity of illness and was subsequently validated in our hospital.⁸ The independent association of each laboratory perturbation with risk of death in hospital is reflected by the number of points assigned to each lab value with the total LAPS being the sum of these values. Time-dependent covariates included weekday in hospital and whether or not patients were in the intensive care unit.

Analysis

We used 3 stages to create a model to predict the daily expected number of discharges: we identified discharge risk strata containing patients having similar discharge patterns using data from patients in the derivation cohort (first stage); then, we generated the preliminary probability of discharge by determining the daily discharge probability in each discharge risk strata (second stage); finally, we modified the probability from the second stage based on the weekday and admission service and summed these probabilities to create the expected number of discharges on a particular date (third stage).

The first stage identified discharge risk strata based on the covariates listed above. This was determined by using a survival tree approach⁹ with proportional hazard regression models to generate the “splits.” These models were offered all covariates listed in the Study Covariates section. Admission service was clustered within 4 departments (obstetrics/gynecology, psychiatry, surgery, and medicine) and day of week was “binarized” into weekday/weekend-holiday (because the use of categorical variables with large numbers of groups can “stunt” regression trees due to small numbers of patients—and, therefore, statistical power—in each subgroup). The proportional hazards model identified the covariate having the strongest association with time to discharge (based on the Wald X² value divided by the degrees of freedom). This variable was then used to split the cohort into subgroups (with continuous covariates being categorized into quartiles). The proportional hazards model was then repeated in each subgroup (with the previous splitting variable[s] excluded from the model). This process continued until no variable was associated with time to discharge with a P value less than .0001. This survival-tree was then used to cluster all patients into distinct discharge risk strata.

In the second stage, we generated the preliminary probability of discharge for a specific date. This was calculated by assigning all patients in hospital to their discharge risk strata (Appendix). We then measured the probability of discharge on each hospitalization day in all discharge risk strata using data from the previous 180 days (we only used the prior 180 days of data to account for temporal changes in hospital discharge patterns). For example, consider a 75-year-old patient on her third hospital day under obstetrics/gynecology on December 19, 2015 (a Saturday). This patient would be assigned to risk stratum #133 (Appendix A). We then measured the probability of discharge of all patients in this discharge risk stratum hospitalized in the previous 6 months (ie, between June 22, 2015, and December 18, 2015) on each hospital day. For risk stratum #133, the probability of discharge on hospital day 3 was 0.1111; therefore, our sample patient’s preliminary expected discharge probability was 0.1111.

To attain stable daily discharge probability estimates, a minimum of 50 patients per discharge risk stratum-hospitalization day combination was required. If there were less than 50 patients for a particular hospitalization day in a particular discharge risk stratum, we grouped hospitalization days in that risk stratum together until the minimum of 50 patients was collected.

The third (and final) stage accounted for the lack of granularity when we created the discharge risk strata in the first stage. As we mentioned above, admission service was clustered into 4 departments and the day of week was clustered into weekend/weekday. However, important variations in discharge probabilities could still exist within departments and between particular days of the week.¹⁰ Therefore, we created a correction factor to adjust the preliminary expected number of discharges based on the admission division and day of week. This correction factor used data from the 180 days prior to the analysis date within which the expected daily number of discharges was calculated (using the methods above). The correction factor was the relative difference between the observed and expected number of discharges within each division-day of week grouping.

For example, to calculate the correction factor for our sample patient presented above (75-year-old patient on hospital day 3 under gynecology on Saturday, December 19, 2015), we measured the observed number of discharges from gynecology on Saturdays between June 22, 2015, and December 18, 2015, (n = 206) and the expected number of discharges (n = 195.255) resulting in a correction factor of (observed-expected)/expected = (195.255-206)/195.206 = 0.05503. Therefore, the final expected discharge probability for our sample patient was 0.1111+0.1111*0.05503=0.1172. The expected number of discharges on a particular date was the preliminary expected number of discharges on that date (generated in the second stage) multiplied by the correction factor for the corresponding division-day or week group.

 

RESULTS

There were 192,859 admissions involving patients more than 1 year of age that spent at least part of their hospitalization between January 1, 2013, and December 31, 2015 (Table). Patients were middle-aged and slightly female predominant, with about half being admitted from the emergency department. Approximately 80% of admissions were to surgical or medical services. More than 95% of admissions ended with a discharge from the hospital with the remainder ending in a death. Almost 30% of hospitalization days occurred on weekends or holidays. Hospitalizations in the derivation (2013-2014) and validation (2015) group were essentially the same, except there was a slight drop in hospital length of stay (from a median of 4 days to 3 days) between the 2 periods.

Patient and hospital covariates importantly influenced the daily conditional probability of discharge (Figure 1). Patients admitted to the obstetrics/gynecology department were notably more likely to be discharged from hospital with no influence from the day of week. In contrast, the probability of discharge decreased notably on the weekends in the other departments. Patients on the ward were much more likely to be discharged than those in the intensive care unit, with increasing age associated with a decreased discharge likelihood in the former but not the latter patients. Finally, discharge probabilities varied only slightly between campuses at our hospital with discharge risk decreasing as severity of illness (as measured by LAPS) increased.

The TEND model contained 142 discharge risk strata (Appendix A). Weekend-holiday status had the strongest association with discharge probability (ie, it was the first splitting variable). The most complex discharge risk strata contained 6 covariates. The daily conditional probability of discharge during the first 2 weeks of hospitalization varied extensively between discharge risk strata (Figure 2). Overall, the conditional discharge probability increased from the first to the second day, remained relatively stable for several days, and then slowly decreased over time. However, this pattern and day-to-day variability differed extensively between risk strata.

The observed daily number of discharges in the validation cohort varied extensively (median 139; interquartile range [IQR] 95-160; range 39-214). The TEND model accurately predicted the daily number of discharges with the expected daily number being strongly associated with the observed number (adjusted R² = 89.2%; P < 0.0001; Figure 3). Calibration decreased but remained significant when we limited the analyses by hospital campus (General: R² = 46.3%; P < 0.0001; Civic: R² = 47.9%; P < 0.0001; Heart Institute: R² = 18.1%; P < 0.0001). The expected number of daily discharges was an unbiased estimator of the observed number of discharges (its parameter estimate in a linear regression model with the observed number of discharges as the outcome variable was 1.0005; 95% confidence interval, 0.9647-1.0363). The absolute difference in the observed and expected daily number of discharges was small (median 1.6; IQR −6.8 to 9.4; range −37 to 63.4) as was the relative difference (median 1.4%; IQR −5.5% to 7.1%; range −40.9% to 43.4%). The expected number of discharges was within 20% of the observed number of discharges in 95.1% of days in 2015.

DISCUSSION

Knowing how many patients will soon be discharged from the hospital should greatly facilitate hospital planning. This study showed that the TEND model used simple patient and hospitalization covariates to accurately predict the number of patients who will be discharged from hospital in the next day.

We believe that this study has several notable findings. First, we think that using a nonparametric approach to predicting the daily number of discharges importantly increased accuracy. This approach allowed us to generate expected likelihoods based on actual discharge probabilities at our hospital in the most recent 6 months of hospitalization-days within patients having discharge patterns that were very similar to the patient in question (ie, discharge risk strata, Appendix A). This ensured that trends in hospitalization habits were accounted for without the need of a period variable in our model. In addition, the lack of parameters in the model will make it easier to transplant it to other hospitals. Second, we think that the accuracy of the predictions were remarkable given the relative “crudeness” of our predictors. By using relatively simple factors, the TEND model was able to output accurate predictions for the number of daily discharges (Figure 3).

This study joins several others that have attempted to accomplish the difficult task of predicting the number of hospital discharges by using digitized data. Barnes et al.¹¹ created a model using regression random forest methods in a single medical service within a hospital to predict the daily number of discharges with impressive accuracy (mean daily number of discharges observed 8.29, expected 8.51). Interestingly, the model in this study was more accurate at predicting discharge likelihood than physicians. Levin et al.¹² derived a model using discrete time logistic regression to predict the likelihood of discharge from a pediatric intensive care unit, finding that physician orders (captured via electronic order entry) could be categorized and used to significantly increase the accuracy of discharge likelihood. This study demonstrates the potential opportunities within health-related data from hospital data warehouses to improve prediction. We believe that continued work in this field will result in the increased use of digital data to help hospital administrators manage patient beds more efficiently and effectively than currently used resource intensive manual methods.^13,14

Several issues should be kept in mind when interpreting our findings. First, our analysis is limited to a single institution in Canada. It will be important to determine if the TEND model methodology generalizes to other hospitals in different jurisdictions. Such an external validation, especially in multiple hospitals, will be important to show that the TEND model methodology works in other facilities. Hospitals could implement the TEND model if they are able to record daily values for each of the variables required to assign patients to a discharge risk stratum (Appendix A) and calculate within each the daily probability of discharge. Hospitals could derive their own discharge risk strata to account for covariates, which we did not include in our study but could be influential, such as insurance status. These discharge risk estimates could also be incorporated into the electronic medical record or hospital dashboards (as long as the data required to generate the estimates are available). These interventions would permit the expected number of hospital discharges (and even the patient-level probability of discharge) to be calculated on a daily basis. Second, 2 potential biases could have influenced the identification of our discharge risk strata (Appendix A). In this process, we used survival tree methods to separate patient-days into clusters having progressively more homogenous discharge patterns. Each split was determined by using a proportional hazards model that ignored the competing risks of death in hospital. In addition, the model expressed age and LAPS as continuous variables, whereas these covariates had to be categorized to create our risk strata groupings. The strength of a covariate’s association with an outcome will decrease when a continuous variable is categorized.¹⁵ Both of these issues might have biased our final risk strata categorization (Appendix A). Third, we limited our model to include simple covariates whose values could be determined relatively easily within most hospital administrative data systems. While this increases the generalizability to other hospital information systems, we believe that the introduction of other covariates to the model—such as daily vital signs, laboratory results, medications, or time from operations—could increase prediction accuracy. Finally, it is uncertain whether or not knowing the predicted number of discharges will improve the efficiency of bed management within the hospital. It seems logical that an accurate prediction of the number of beds that will be made available in the next day should improve decisions regarding the number of patients who could be admitted electively to the hospital. It remains to be seen, however, whether this truly happens.

In summary, we found that the TEND model used a handful of patient and hospitalization factors to accurately predict the expected number of discharges from hospital in the next day. Further work is required to implement this model into our institution’s data warehouse and then determine whether this prediction will improve the efficiency of bed management at our hospital.

 

Disclosure: CvW is supported by a University of Ottawa Department of Medicine Clinician Scientist Chair. The authors have no conflicts of interest

 

Hospitals typically allocate beds based on historical patient volumes. If funding decreases, hospitals will usually try to maximize resource utilization by allocating beds to attain occupancies close to 100% for significant periods of time. This will invariably cause days in which hospital occupancy exceeds capacity, at which time critical entry points (such as the emergency department and operating room) will become blocked. This creates significant concerns over the patient quality of care.

Hospital administrators have very few options when hospital occupancy exceeds 100%. They could postpone admissions for “planned” cases, bring in additional staff to increase capacity, or instigate additional methods to increase hospital discharges such as expanding care resources in the community. All options are costly, bothersome, or cannot be actioned immediately. The need for these options could be minimized by enabling hospital administrators to make more informed decisions regarding hospital bed management by knowing the likely number of discharges in the next 24 hours.

Predicting the number of people who will be discharged in the next day can be approached in several ways. One approach would be to calculate each patient’s expected length of stay and then use the variation around that estimate to calculate each day’s discharge probability. Several studies have attempted to model hospital length of stay using a broad assortment of methodologies, but a mechanism to accurately predict this outcome has been elusive^1,2 (with Verburg et al.³ concluding in their study’s abstract that “…it is difficult to predict length of stay…”). A second approach would be to use survival analysis methods to generate each patient’s hazard of discharge over time, which could be directly converted to an expected daily risk of discharge. However, this approach is complicated by the concurrent need to include time-dependent covariates and consider the competing risk of death in hospital, which can complicate survival modeling.^4,5 A third approach would be the implementation of a longitudinal analysis using marginal models to predict the daily probability of discharge,⁶ but this method quickly overwhelms computer resources when large datasets are present.

In this study, we decided to use nonparametric models to predict the daily number of hospital discharges. We first identified patient groups with distinct discharge patterns. We then calculated the conditional daily discharge probability of patients in each of these groups. Finally, these conditional daily discharge probabilities were then summed for each hospital day to generate the expected number of discharges in the next 24 hours. This paper details the methods we used to create our model and the accuracy of its predictions.

 

METHODS

Study Setting and Databases Used for Analysis

The study took place at The Ottawa Hospital, a 1000-bed teaching hospital with 3 campuses that is the primary referral center in our region. The study was approved by our local research ethics board.

The Patient Registry Database records the date and time of admission for each patient (defined as the moment that a patient’s admission request is registered in the patient registration) and discharge (defined as the time when the patient’s discharge from hospital was entered into the patient registration) for hospital encounters. Emergency department encounters were also identified in the Patient Registry Database along with admission service, patient age and sex, and patient location throughout the admission. The Laboratory Database records all laboratory studies and results on all patients at the hospital.

Study Cohort

We used the Patient Registry Database to identify all people aged 1 year or more who were admitted to the hospital between January 1, 2013, and December 31, 2015. This time frame was selected to (i) ensure that data were complete; and (ii) complete calendar years of data were available for both derivation (patient-days in 2013-2014) and validation (2015) cohorts. Patients who were observed in the emergency room without admission to hospital were not included.

Study Outcome

The study outcome was the number of patients discharged from the hospital each day. For the analysis, the reference point for each day was 1 second past midnight; therefore, values for time-dependent covariates up to and including midnight were used to predict the number of discharges in the next 24 hours.

Study Covariates

Baseline (ie, time-independent) covariates included patient age and sex, admission service, hospital campus, whether or not the patient was admitted from the emergency department (all determined from the Patient Registry Database), and the Laboratory-based Acute Physiological Score (LAPS). The latter, which was calculated with the Laboratory Database using results for 14 tests (arterial pH, PaCO2, PaO2, anion gap, hematocrit, total white blood cell count, serum albumin, total bilirubin, creatinine, urea nitrogen, glucose, sodium, bicarbonate, and troponin I) measured in the 24-hour time frame preceding hospitalization, was derived by Escobar and colleagues⁷ to measure severity of illness and was subsequently validated in our hospital.⁸ The independent association of each laboratory perturbation with risk of death in hospital is reflected by the number of points assigned to each lab value with the total LAPS being the sum of these values. Time-dependent covariates included weekday in hospital and whether or not patients were in the intensive care unit.

Analysis

We used 3 stages to create a model to predict the daily expected number of discharges: we identified discharge risk strata containing patients having similar discharge patterns using data from patients in the derivation cohort (first stage); then, we generated the preliminary probability of discharge by determining the daily discharge probability in each discharge risk strata (second stage); finally, we modified the probability from the second stage based on the weekday and admission service and summed these probabilities to create the expected number of discharges on a particular date (third stage).

The first stage identified discharge risk strata based on the covariates listed above. This was determined by using a survival tree approach⁹ with proportional hazard regression models to generate the “splits.” These models were offered all covariates listed in the Study Covariates section. Admission service was clustered within 4 departments (obstetrics/gynecology, psychiatry, surgery, and medicine) and day of week was “binarized” into weekday/weekend-holiday (because the use of categorical variables with large numbers of groups can “stunt” regression trees due to small numbers of patients—and, therefore, statistical power—in each subgroup). The proportional hazards model identified the covariate having the strongest association with time to discharge (based on the Wald X² value divided by the degrees of freedom). This variable was then used to split the cohort into subgroups (with continuous covariates being categorized into quartiles). The proportional hazards model was then repeated in each subgroup (with the previous splitting variable[s] excluded from the model). This process continued until no variable was associated with time to discharge with a P value less than .0001. This survival-tree was then used to cluster all patients into distinct discharge risk strata.

In the second stage, we generated the preliminary probability of discharge for a specific date. This was calculated by assigning all patients in hospital to their discharge risk strata (Appendix). We then measured the probability of discharge on each hospitalization day in all discharge risk strata using data from the previous 180 days (we only used the prior 180 days of data to account for temporal changes in hospital discharge patterns). For example, consider a 75-year-old patient on her third hospital day under obstetrics/gynecology on December 19, 2015 (a Saturday). This patient would be assigned to risk stratum #133 (Appendix A). We then measured the probability of discharge of all patients in this discharge risk stratum hospitalized in the previous 6 months (ie, between June 22, 2015, and December 18, 2015) on each hospital day. For risk stratum #133, the probability of discharge on hospital day 3 was 0.1111; therefore, our sample patient’s preliminary expected discharge probability was 0.1111.

To attain stable daily discharge probability estimates, a minimum of 50 patients per discharge risk stratum-hospitalization day combination was required. If there were less than 50 patients for a particular hospitalization day in a particular discharge risk stratum, we grouped hospitalization days in that risk stratum together until the minimum of 50 patients was collected.

The third (and final) stage accounted for the lack of granularity when we created the discharge risk strata in the first stage. As we mentioned above, admission service was clustered into 4 departments and the day of week was clustered into weekend/weekday. However, important variations in discharge probabilities could still exist within departments and between particular days of the week.¹⁰ Therefore, we created a correction factor to adjust the preliminary expected number of discharges based on the admission division and day of week. This correction factor used data from the 180 days prior to the analysis date within which the expected daily number of discharges was calculated (using the methods above). The correction factor was the relative difference between the observed and expected number of discharges within each division-day of week grouping.

For example, to calculate the correction factor for our sample patient presented above (75-year-old patient on hospital day 3 under gynecology on Saturday, December 19, 2015), we measured the observed number of discharges from gynecology on Saturdays between June 22, 2015, and December 18, 2015, (n = 206) and the expected number of discharges (n = 195.255) resulting in a correction factor of (observed-expected)/expected = (195.255-206)/195.206 = 0.05503. Therefore, the final expected discharge probability for our sample patient was 0.1111+0.1111*0.05503=0.1172. The expected number of discharges on a particular date was the preliminary expected number of discharges on that date (generated in the second stage) multiplied by the correction factor for the corresponding division-day or week group.

 

RESULTS

There were 192,859 admissions involving patients more than 1 year of age that spent at least part of their hospitalization between January 1, 2013, and December 31, 2015 (Table). Patients were middle-aged and slightly female predominant, with about half being admitted from the emergency department. Approximately 80% of admissions were to surgical or medical services. More than 95% of admissions ended with a discharge from the hospital with the remainder ending in a death. Almost 30% of hospitalization days occurred on weekends or holidays. Hospitalizations in the derivation (2013-2014) and validation (2015) group were essentially the same, except there was a slight drop in hospital length of stay (from a median of 4 days to 3 days) between the 2 periods.

Patient and hospital covariates importantly influenced the daily conditional probability of discharge (Figure 1). Patients admitted to the obstetrics/gynecology department were notably more likely to be discharged from hospital with no influence from the day of week. In contrast, the probability of discharge decreased notably on the weekends in the other departments. Patients on the ward were much more likely to be discharged than those in the intensive care unit, with increasing age associated with a decreased discharge likelihood in the former but not the latter patients. Finally, discharge probabilities varied only slightly between campuses at our hospital with discharge risk decreasing as severity of illness (as measured by LAPS) increased.

The TEND model contained 142 discharge risk strata (Appendix A). Weekend-holiday status had the strongest association with discharge probability (ie, it was the first splitting variable). The most complex discharge risk strata contained 6 covariates. The daily conditional probability of discharge during the first 2 weeks of hospitalization varied extensively between discharge risk strata (Figure 2). Overall, the conditional discharge probability increased from the first to the second day, remained relatively stable for several days, and then slowly decreased over time. However, this pattern and day-to-day variability differed extensively between risk strata.

The observed daily number of discharges in the validation cohort varied extensively (median 139; interquartile range [IQR] 95-160; range 39-214). The TEND model accurately predicted the daily number of discharges with the expected daily number being strongly associated with the observed number (adjusted R² = 89.2%; P < 0.0001; Figure 3). Calibration decreased but remained significant when we limited the analyses by hospital campus (General: R² = 46.3%; P < 0.0001; Civic: R² = 47.9%; P < 0.0001; Heart Institute: R² = 18.1%; P < 0.0001). The expected number of daily discharges was an unbiased estimator of the observed number of discharges (its parameter estimate in a linear regression model with the observed number of discharges as the outcome variable was 1.0005; 95% confidence interval, 0.9647-1.0363). The absolute difference in the observed and expected daily number of discharges was small (median 1.6; IQR −6.8 to 9.4; range −37 to 63.4) as was the relative difference (median 1.4%; IQR −5.5% to 7.1%; range −40.9% to 43.4%). The expected number of discharges was within 20% of the observed number of discharges in 95.1% of days in 2015.

DISCUSSION

Knowing how many patients will soon be discharged from the hospital should greatly facilitate hospital planning. This study showed that the TEND model used simple patient and hospitalization covariates to accurately predict the number of patients who will be discharged from hospital in the next day.

We believe that this study has several notable findings. First, we think that using a nonparametric approach to predicting the daily number of discharges importantly increased accuracy. This approach allowed us to generate expected likelihoods based on actual discharge probabilities at our hospital in the most recent 6 months of hospitalization-days within patients having discharge patterns that were very similar to the patient in question (ie, discharge risk strata, Appendix A). This ensured that trends in hospitalization habits were accounted for without the need of a period variable in our model. In addition, the lack of parameters in the model will make it easier to transplant it to other hospitals. Second, we think that the accuracy of the predictions were remarkable given the relative “crudeness” of our predictors. By using relatively simple factors, the TEND model was able to output accurate predictions for the number of daily discharges (Figure 3).

This study joins several others that have attempted to accomplish the difficult task of predicting the number of hospital discharges by using digitized data. Barnes et al.¹¹ created a model using regression random forest methods in a single medical service within a hospital to predict the daily number of discharges with impressive accuracy (mean daily number of discharges observed 8.29, expected 8.51). Interestingly, the model in this study was more accurate at predicting discharge likelihood than physicians. Levin et al.¹² derived a model using discrete time logistic regression to predict the likelihood of discharge from a pediatric intensive care unit, finding that physician orders (captured via electronic order entry) could be categorized and used to significantly increase the accuracy of discharge likelihood. This study demonstrates the potential opportunities within health-related data from hospital data warehouses to improve prediction. We believe that continued work in this field will result in the increased use of digital data to help hospital administrators manage patient beds more efficiently and effectively than currently used resource intensive manual methods.^13,14

Several issues should be kept in mind when interpreting our findings. First, our analysis is limited to a single institution in Canada. It will be important to determine if the TEND model methodology generalizes to other hospitals in different jurisdictions. Such an external validation, especially in multiple hospitals, will be important to show that the TEND model methodology works in other facilities. Hospitals could implement the TEND model if they are able to record daily values for each of the variables required to assign patients to a discharge risk stratum (Appendix A) and calculate within each the daily probability of discharge. Hospitals could derive their own discharge risk strata to account for covariates, which we did not include in our study but could be influential, such as insurance status. These discharge risk estimates could also be incorporated into the electronic medical record or hospital dashboards (as long as the data required to generate the estimates are available). These interventions would permit the expected number of hospital discharges (and even the patient-level probability of discharge) to be calculated on a daily basis. Second, 2 potential biases could have influenced the identification of our discharge risk strata (Appendix A). In this process, we used survival tree methods to separate patient-days into clusters having progressively more homogenous discharge patterns. Each split was determined by using a proportional hazards model that ignored the competing risks of death in hospital. In addition, the model expressed age and LAPS as continuous variables, whereas these covariates had to be categorized to create our risk strata groupings. The strength of a covariate’s association with an outcome will decrease when a continuous variable is categorized.¹⁵ Both of these issues might have biased our final risk strata categorization (Appendix A). Third, we limited our model to include simple covariates whose values could be determined relatively easily within most hospital administrative data systems. While this increases the generalizability to other hospital information systems, we believe that the introduction of other covariates to the model—such as daily vital signs, laboratory results, medications, or time from operations—could increase prediction accuracy. Finally, it is uncertain whether or not knowing the predicted number of discharges will improve the efficiency of bed management within the hospital. It seems logical that an accurate prediction of the number of beds that will be made available in the next day should improve decisions regarding the number of patients who could be admitted electively to the hospital. It remains to be seen, however, whether this truly happens.

In summary, we found that the TEND model used a handful of patient and hospitalization factors to accurately predict the expected number of discharges from hospital in the next day. Further work is required to implement this model into our institution’s data warehouse and then determine whether this prediction will improve the efficiency of bed management at our hospital.

 

Disclosure: CvW is supported by a University of Ottawa Department of Medicine Clinician Scientist Chair. The authors have no conflicts of interest