User login
ESTES PARK, COLO. – These days there’s a smartphone app for everything. But not all apps for clinical decision rules are truly helpful in making better patient care decisions in frontline primary care medicine, Dr. Robert E. Burke cautioned at a conference on internal medicine sponsored by the University of Colorado.
Five clinical decision rules have proven merit, each accessible via app as well as by an online website, he said. All five address medical issues that are common, important, and complex. What these rules have in common is excellent – not merely good – predictive value, thorough validation in multiple populations, a high likelihood that their use will influence patient management in individual cases, and an enthusiastic thumbs-up vote by physician-users.
These top-tier clinical decision rules address the following situations: determining which patients with community-acquired pneumonia can safely be managed at home, assessing whether a patient with chronic liver disease is likely to have advanced fibrosis or cirrhosis without resorting to liver biopsy, identifying who can be sent home safely after an episode of syncope, determining whether to prescribe oral anticoagulation for atrial fibrillation, and stratifying risk after a transient ischemic attack (TIA) or stroke.
These clinical decision rules are not a replacement for clinical judgment, but they do make risk/benefit calculations a lot easier, said Dr. Burke, a hospitalist at the Denver VA Medical Center and the university.
Triage of pneumonia patients. Pneumonia is the third most common reason for hospital admission as well as early readmission in the United States. So lots of different pneumonia severity scores have been developed. But few of them are any good, according to Dr. Burke.
Many of these subpar scores measure the likelihood that a patient with pneumonia will be admitted to the intensive care unit. That’s not what primary care physicians really want to know.
"Unfortunately, ICU admission rates vary widely in the U.S. and have nothing to do with how sick patients are or their comorbid conditions. It’s really not clear why the rates vary so much, but the predictive value of those scores is pretty poor, since ICU admission is a subjective thing," he explained.
The score he recommends is the Pneumonia Severity Index. It has been validated as an admission rule in several randomized trials totaling more than 4,000 patients. The score places patients into one of five risk categories. In the randomized trials, patients in the three lowest-risk categories had no negative outcomes when treated as outpatients. Plus, the use of the Pneumonia Severity Index reduced hospital admissions by up to 20%.
The score is calculated using 20 factors readily obtainable from the patient history, physical exam, and lab work. It’s a laborious calculation that’s perfect for an app. And the app exists: It’s called Simpli PSI.
Assessing fibrosis/cirrhosis risk in patients with chronic liver disease. Liver biopsy is painful, costly, decidedly not risk free, and often misleading due to the small tissue sample obtained. Transient elastography, a noninvasive ultrasound-based alternative, is promising but not ready for prime time. Yet knowing whether a patient has advanced fibrosis or cirrhosis is important in terms of prognosis, treatment, and surveillance strategy.
This is where the FIB-4 Index, a serum score for fibrosis, is of value. The FIB-4 Index is easily calculated from basic laboratory data. Findings from multiple studies have shown excellent discrimination regardless of whether patients had hepatitis C, hepatitis B, or nonalcoholic fatty liver disease.
"It has a c-statistic [equivalent to the area under the receiver operating characteristic curve] of 0.84. That’s about as good as it gets," according to the hospitalist.
The app is called Liver Calc. Like all the other apps Dr. Burke highlighted, it is available to iPhone users at the App Store.
"Or if you were to type in ‘FIB-4 Index’ on Google, you’ll find a website that will actually calculate the score online," he noted.
Whom to hospitalize for syncope. "Syncope is one of the most common things I see as a physician of in-patients," Dr. Burke said. "It accounts for maybe 1%-3% of emergency department visits and 6% of hospitalizations in the United States."
The San Francisco Syncope Rule is of proven value in helping to identify patients at low risk of serious outcomes in the 30 days following a syncopal episode. These are the patients who don’t need to undergo emergency department evaluation. The rule has been prospectively validated in tens of thousands of patients. It had a negative predictive value of 97% in a systematic review. The rule works less well in patients over age 65, where the likelihood of arrhythmia as the cause of syncope increases sharply. And the rule works best when applied only after a thorough history and physical exam aimed at finding a cause for the patient’s syncope.
"For example, the rule assumes you’ve already checked for hypoglycemia, a common cause of syncope," the hospitalist explained.
The San Francisco Syncope Rule is available in app form as Doctor Tools of the Trade.
Atrial fibrillation: To anticoagulate or not? This is a decision that entails balancing the reduction in stroke risk obtained with anticoagulation against the potential harm in the form of bleeding. The best tool for assessing the potential stroke-risk benefit is the CHA2DS2-VASc score, which represents a significant advance over the CHADS2 score, in Dr. Burke’s view.
"What the CHA2DS2-VASc score is really useful for, I think, is in putting people into low- or high-risk buckets. It takes people out of that intermediate-risk category in CHADS2," he explained.
It turns out that when physicians rely upon clinical intuition to estimate the risk of clinically important bleeding in patients on oral anticoagulation, they tend to overestimate the true risk. This observation has led to a proliferation of scales aimed at predicting who is likely to bleed when placed on warfarin. The best performer among them, regardless of whether the endpoint is any clinically relevant bleeding, major bleeding, or all-cause mortality, is the HAS-BLED score.
"I would encourage you to use both the CHA2DS2-VASc and HAS-BLED, and to use them at the same time. In some cases you’ll find the numbers are really discordant; there may be much more benefit than you thought, or much more risk," Dr. Burke said.
The focus-AF calculator is the app that will do the work.
Risk stratification after TIA or stroke. The ABCD2 score provides an estimate of stroke risk within 7 or 90 days after a TIA. This information helps establish the urgency of patient evaluation and risk factor management. A patient with an ABCD2 score of 4 or more should go straight to the emergency department, while a score of 3 or less indicates outpatient evaluation is appropriate. The Neuro Toolkit app will run the numbers.
Dr. Burke is keeping a watchful eye on a number of other clinical decision rules that, while promising, aren’t quite ready for prime time in his view. These include the FRAX score, the TIMI score, and the Marburg Heart Score. The Marburg score, for example, has been extensively validated as a tool to help primary care physicians decide whether chest pain is cardiac or noncardiac. But in the clinical trials, the score wasn’t compared to clinical intuition. That’s a problem.
"I think our clinical intuition here is relatively good. So I’d like to see data showing the rule adds something to clinical intuition before I recommend it," he said.
In a head-to-head comparative trial, the Pittsburgh Knee Rule outperformed the older Ottawa Knee Rule as an aid in figuring out who needs imaging after a knee injury. Impressive, in Dr. Burke’s view, but he’d like to see the results confirmed in a second study.
He reported having no financial conflicts.
ESTES PARK, COLO. – These days there’s a smartphone app for everything. But not all apps for clinical decision rules are truly helpful in making better patient care decisions in frontline primary care medicine, Dr. Robert E. Burke cautioned at a conference on internal medicine sponsored by the University of Colorado.
Five clinical decision rules have proven merit, each accessible via app as well as by an online website, he said. All five address medical issues that are common, important, and complex. What these rules have in common is excellent – not merely good – predictive value, thorough validation in multiple populations, a high likelihood that their use will influence patient management in individual cases, and an enthusiastic thumbs-up vote by physician-users.
These top-tier clinical decision rules address the following situations: determining which patients with community-acquired pneumonia can safely be managed at home, assessing whether a patient with chronic liver disease is likely to have advanced fibrosis or cirrhosis without resorting to liver biopsy, identifying who can be sent home safely after an episode of syncope, determining whether to prescribe oral anticoagulation for atrial fibrillation, and stratifying risk after a transient ischemic attack (TIA) or stroke.
These clinical decision rules are not a replacement for clinical judgment, but they do make risk/benefit calculations a lot easier, said Dr. Burke, a hospitalist at the Denver VA Medical Center and the university.
Triage of pneumonia patients. Pneumonia is the third most common reason for hospital admission as well as early readmission in the United States. So lots of different pneumonia severity scores have been developed. But few of them are any good, according to Dr. Burke.
Many of these subpar scores measure the likelihood that a patient with pneumonia will be admitted to the intensive care unit. That’s not what primary care physicians really want to know.
"Unfortunately, ICU admission rates vary widely in the U.S. and have nothing to do with how sick patients are or their comorbid conditions. It’s really not clear why the rates vary so much, but the predictive value of those scores is pretty poor, since ICU admission is a subjective thing," he explained.
The score he recommends is the Pneumonia Severity Index. It has been validated as an admission rule in several randomized trials totaling more than 4,000 patients. The score places patients into one of five risk categories. In the randomized trials, patients in the three lowest-risk categories had no negative outcomes when treated as outpatients. Plus, the use of the Pneumonia Severity Index reduced hospital admissions by up to 20%.
The score is calculated using 20 factors readily obtainable from the patient history, physical exam, and lab work. It’s a laborious calculation that’s perfect for an app. And the app exists: It’s called Simpli PSI.
Assessing fibrosis/cirrhosis risk in patients with chronic liver disease. Liver biopsy is painful, costly, decidedly not risk free, and often misleading due to the small tissue sample obtained. Transient elastography, a noninvasive ultrasound-based alternative, is promising but not ready for prime time. Yet knowing whether a patient has advanced fibrosis or cirrhosis is important in terms of prognosis, treatment, and surveillance strategy.
This is where the FIB-4 Index, a serum score for fibrosis, is of value. The FIB-4 Index is easily calculated from basic laboratory data. Findings from multiple studies have shown excellent discrimination regardless of whether patients had hepatitis C, hepatitis B, or nonalcoholic fatty liver disease.
"It has a c-statistic [equivalent to the area under the receiver operating characteristic curve] of 0.84. That’s about as good as it gets," according to the hospitalist.
The app is called Liver Calc. Like all the other apps Dr. Burke highlighted, it is available to iPhone users at the App Store.
"Or if you were to type in ‘FIB-4 Index’ on Google, you’ll find a website that will actually calculate the score online," he noted.
Whom to hospitalize for syncope. "Syncope is one of the most common things I see as a physician of in-patients," Dr. Burke said. "It accounts for maybe 1%-3% of emergency department visits and 6% of hospitalizations in the United States."
The San Francisco Syncope Rule is of proven value in helping to identify patients at low risk of serious outcomes in the 30 days following a syncopal episode. These are the patients who don’t need to undergo emergency department evaluation. The rule has been prospectively validated in tens of thousands of patients. It had a negative predictive value of 97% in a systematic review. The rule works less well in patients over age 65, where the likelihood of arrhythmia as the cause of syncope increases sharply. And the rule works best when applied only after a thorough history and physical exam aimed at finding a cause for the patient’s syncope.
"For example, the rule assumes you’ve already checked for hypoglycemia, a common cause of syncope," the hospitalist explained.
The San Francisco Syncope Rule is available in app form as Doctor Tools of the Trade.
Atrial fibrillation: To anticoagulate or not? This is a decision that entails balancing the reduction in stroke risk obtained with anticoagulation against the potential harm in the form of bleeding. The best tool for assessing the potential stroke-risk benefit is the CHA2DS2-VASc score, which represents a significant advance over the CHADS2 score, in Dr. Burke’s view.
"What the CHA2DS2-VASc score is really useful for, I think, is in putting people into low- or high-risk buckets. It takes people out of that intermediate-risk category in CHADS2," he explained.
It turns out that when physicians rely upon clinical intuition to estimate the risk of clinically important bleeding in patients on oral anticoagulation, they tend to overestimate the true risk. This observation has led to a proliferation of scales aimed at predicting who is likely to bleed when placed on warfarin. The best performer among them, regardless of whether the endpoint is any clinically relevant bleeding, major bleeding, or all-cause mortality, is the HAS-BLED score.
"I would encourage you to use both the CHA2DS2-VASc and HAS-BLED, and to use them at the same time. In some cases you’ll find the numbers are really discordant; there may be much more benefit than you thought, or much more risk," Dr. Burke said.
The focus-AF calculator is the app that will do the work.
Risk stratification after TIA or stroke. The ABCD2 score provides an estimate of stroke risk within 7 or 90 days after a TIA. This information helps establish the urgency of patient evaluation and risk factor management. A patient with an ABCD2 score of 4 or more should go straight to the emergency department, while a score of 3 or less indicates outpatient evaluation is appropriate. The Neuro Toolkit app will run the numbers.
Dr. Burke is keeping a watchful eye on a number of other clinical decision rules that, while promising, aren’t quite ready for prime time in his view. These include the FRAX score, the TIMI score, and the Marburg Heart Score. The Marburg score, for example, has been extensively validated as a tool to help primary care physicians decide whether chest pain is cardiac or noncardiac. But in the clinical trials, the score wasn’t compared to clinical intuition. That’s a problem.
"I think our clinical intuition here is relatively good. So I’d like to see data showing the rule adds something to clinical intuition before I recommend it," he said.
In a head-to-head comparative trial, the Pittsburgh Knee Rule outperformed the older Ottawa Knee Rule as an aid in figuring out who needs imaging after a knee injury. Impressive, in Dr. Burke’s view, but he’d like to see the results confirmed in a second study.
He reported having no financial conflicts.
ESTES PARK, COLO. – These days there’s a smartphone app for everything. But not all apps for clinical decision rules are truly helpful in making better patient care decisions in frontline primary care medicine, Dr. Robert E. Burke cautioned at a conference on internal medicine sponsored by the University of Colorado.
Five clinical decision rules have proven merit, each accessible via app as well as by an online website, he said. All five address medical issues that are common, important, and complex. What these rules have in common is excellent – not merely good – predictive value, thorough validation in multiple populations, a high likelihood that their use will influence patient management in individual cases, and an enthusiastic thumbs-up vote by physician-users.
These top-tier clinical decision rules address the following situations: determining which patients with community-acquired pneumonia can safely be managed at home, assessing whether a patient with chronic liver disease is likely to have advanced fibrosis or cirrhosis without resorting to liver biopsy, identifying who can be sent home safely after an episode of syncope, determining whether to prescribe oral anticoagulation for atrial fibrillation, and stratifying risk after a transient ischemic attack (TIA) or stroke.
These clinical decision rules are not a replacement for clinical judgment, but they do make risk/benefit calculations a lot easier, said Dr. Burke, a hospitalist at the Denver VA Medical Center and the university.
Triage of pneumonia patients. Pneumonia is the third most common reason for hospital admission as well as early readmission in the United States. So lots of different pneumonia severity scores have been developed. But few of them are any good, according to Dr. Burke.
Many of these subpar scores measure the likelihood that a patient with pneumonia will be admitted to the intensive care unit. That’s not what primary care physicians really want to know.
"Unfortunately, ICU admission rates vary widely in the U.S. and have nothing to do with how sick patients are or their comorbid conditions. It’s really not clear why the rates vary so much, but the predictive value of those scores is pretty poor, since ICU admission is a subjective thing," he explained.
The score he recommends is the Pneumonia Severity Index. It has been validated as an admission rule in several randomized trials totaling more than 4,000 patients. The score places patients into one of five risk categories. In the randomized trials, patients in the three lowest-risk categories had no negative outcomes when treated as outpatients. Plus, the use of the Pneumonia Severity Index reduced hospital admissions by up to 20%.
The score is calculated using 20 factors readily obtainable from the patient history, physical exam, and lab work. It’s a laborious calculation that’s perfect for an app. And the app exists: It’s called Simpli PSI.
Assessing fibrosis/cirrhosis risk in patients with chronic liver disease. Liver biopsy is painful, costly, decidedly not risk free, and often misleading due to the small tissue sample obtained. Transient elastography, a noninvasive ultrasound-based alternative, is promising but not ready for prime time. Yet knowing whether a patient has advanced fibrosis or cirrhosis is important in terms of prognosis, treatment, and surveillance strategy.
This is where the FIB-4 Index, a serum score for fibrosis, is of value. The FIB-4 Index is easily calculated from basic laboratory data. Findings from multiple studies have shown excellent discrimination regardless of whether patients had hepatitis C, hepatitis B, or nonalcoholic fatty liver disease.
"It has a c-statistic [equivalent to the area under the receiver operating characteristic curve] of 0.84. That’s about as good as it gets," according to the hospitalist.
The app is called Liver Calc. Like all the other apps Dr. Burke highlighted, it is available to iPhone users at the App Store.
"Or if you were to type in ‘FIB-4 Index’ on Google, you’ll find a website that will actually calculate the score online," he noted.
Whom to hospitalize for syncope. "Syncope is one of the most common things I see as a physician of in-patients," Dr. Burke said. "It accounts for maybe 1%-3% of emergency department visits and 6% of hospitalizations in the United States."
The San Francisco Syncope Rule is of proven value in helping to identify patients at low risk of serious outcomes in the 30 days following a syncopal episode. These are the patients who don’t need to undergo emergency department evaluation. The rule has been prospectively validated in tens of thousands of patients. It had a negative predictive value of 97% in a systematic review. The rule works less well in patients over age 65, where the likelihood of arrhythmia as the cause of syncope increases sharply. And the rule works best when applied only after a thorough history and physical exam aimed at finding a cause for the patient’s syncope.
"For example, the rule assumes you’ve already checked for hypoglycemia, a common cause of syncope," the hospitalist explained.
The San Francisco Syncope Rule is available in app form as Doctor Tools of the Trade.
Atrial fibrillation: To anticoagulate or not? This is a decision that entails balancing the reduction in stroke risk obtained with anticoagulation against the potential harm in the form of bleeding. The best tool for assessing the potential stroke-risk benefit is the CHA2DS2-VASc score, which represents a significant advance over the CHADS2 score, in Dr. Burke’s view.
"What the CHA2DS2-VASc score is really useful for, I think, is in putting people into low- or high-risk buckets. It takes people out of that intermediate-risk category in CHADS2," he explained.
It turns out that when physicians rely upon clinical intuition to estimate the risk of clinically important bleeding in patients on oral anticoagulation, they tend to overestimate the true risk. This observation has led to a proliferation of scales aimed at predicting who is likely to bleed when placed on warfarin. The best performer among them, regardless of whether the endpoint is any clinically relevant bleeding, major bleeding, or all-cause mortality, is the HAS-BLED score.
"I would encourage you to use both the CHA2DS2-VASc and HAS-BLED, and to use them at the same time. In some cases you’ll find the numbers are really discordant; there may be much more benefit than you thought, or much more risk," Dr. Burke said.
The focus-AF calculator is the app that will do the work.
Risk stratification after TIA or stroke. The ABCD2 score provides an estimate of stroke risk within 7 or 90 days after a TIA. This information helps establish the urgency of patient evaluation and risk factor management. A patient with an ABCD2 score of 4 or more should go straight to the emergency department, while a score of 3 or less indicates outpatient evaluation is appropriate. The Neuro Toolkit app will run the numbers.
Dr. Burke is keeping a watchful eye on a number of other clinical decision rules that, while promising, aren’t quite ready for prime time in his view. These include the FRAX score, the TIMI score, and the Marburg Heart Score. The Marburg score, for example, has been extensively validated as a tool to help primary care physicians decide whether chest pain is cardiac or noncardiac. But in the clinical trials, the score wasn’t compared to clinical intuition. That’s a problem.
"I think our clinical intuition here is relatively good. So I’d like to see data showing the rule adds something to clinical intuition before I recommend it," he said.
In a head-to-head comparative trial, the Pittsburgh Knee Rule outperformed the older Ottawa Knee Rule as an aid in figuring out who needs imaging after a knee injury. Impressive, in Dr. Burke’s view, but he’d like to see the results confirmed in a second study.
He reported having no financial conflicts.
EXPERT ANALYSIS FROM THE ANNUAL INTERNAL MEDICINE PROGRAM
