Friday, 20 March 2015 20:39

Leveraging Evidence-Based Medicine in the Hospital Setting

Written by

r hirschWhen caring for a patient, doctors use many resources to determine what care to provide. They first perform a thorough history and physical examination to help them develop a differential diagnosis – the list of possible conditions that may be present. For example, if the patient has chest pain, the differential diagnosis could range from simple heartburn or a pulled muscle to a blood clot in the lungs or a life-threatening tear in the aorta. They may develop this differential from memory or make use of a textbook, a medical journal such as the New England Journal of Medicine, or an authoritative, evidence-based resource, the best known of which is UpToDate. The doctor then uses that differential diagnosis and the estimated probability of each diagnosis to determine what tests are appropriate and in what order to perform them. If a heart problem is high on the list, the doctor would not want to perform a test for gallbladder disease first since heart disease is not only more likely, it is more life-threatening.

But physicians cannot possibly know the exact likelihood of every diagnosis for each individual patient, so there are also tools that help determine the likelihood of a disease being present (and therefore, the need for testing to find out for sure). There are tools for many conditions, such as the Ottawa Rules, which can be used to determine when imaging of the ankle, knee, cervical spine, and head are indicated in the emergency department. Evidence-based tools such as this are also used by insurers to make coverage determinations. The American College of Radiology produces a wide range of guidelines for imaging studies called ACR Appropriateness Criteria©, which have been adopted by many commercial insurers for use in pre-certifying advanced imaging studies. For fee-for-service Medicare, the Centers for Medicare & Medicaid Services (CMS) and the Medicare Administrative Contractors (MACs) use these tools to develop national and local coverage determinations and guide their reviewers when reviewing claims.

 

One of the most common diagnoses seen in hospitals is, in fact, chest pain, and as noted, the causes can range from trivial to immediately life-threatening. Performing coronary angiography is considered the gold standard for determining if a patient has significant coronary artery disease, but that test requires the use of intravenous contrast and exposure to ionizing radiation, both of which can have short- and long-term side effects. It is also an invasive test with potentially serious complications, and it is costly. Because of these factors, it is often preferable to perform risk stratification and perform coronary angiography only in those at high risk for significant coronary artery disease.

To help guide physicians to find those at highest risk of coronary artery disease (and therefore most in need of angiography), several cardiology specialty societies have developed appropriate use criteria that use the patient's age, sex, and results of prior testing to classify patients as "appropriate," "uncertain," or "inappropriate" for coronary angiography. The number of specialty societies that provided input to these criteria and the volume of literature reviewed in their creation suggested that their accuracy would be all but assured.

So it was disheartening to read a recent publication in the Annals of Internal Medicine by a group of Canadian researchers who did a simple thing: they applied the 2012 appropriate use criteria retrospectively to every patient who had coronary angiography at 18 hospitals in Ontario between 2008 and 2011. In that time period, over 48,000 patients underwent angiography. Fifty-eight percent of angiographic studies were classified as appropriate, 10.8 percent were classified as inappropriate, and 31 percent were classified as uncertain. The results of the angiography in the 5,209 patients who, by criteria, should not have even had angiography demonstrated that 1,609 patients (31 percent) had obstructive CAD and 987 patients (19 percent) underwent revascularization, meaning angioplasty with or without stenting. But even more worrisome is that 372 patients actually had either left main coronary artery disease or triple vessel coronary artery disease, both of which are conditions that have a high likelihood of leading to heart attack or death within a short period of time.

These results must be interpreted with caution. There were no attempts to quantify other factors that the physician may have taken into consideration when deciding to perform angiography, such as trials of conservative care or other imaging such as coronary artery calcium scoring, so the population may be skewed. Performing coronary angiography is less common in Canada than in the United States, so these results are not generalizable to the U.S. Nonetheless, studies such as this remind us that even well-designed tools have limitations and should not be used by CMS, the MACs or commercial insurers to draw lines in the sand; there must be room left for rational clinical judgment and extenuating circumstances.

About the Author

Ronald Hirsch, M.D. serves as vice president of physician advisory services (AccretivePAS®) in the Regulation and Education Group ("the REG Specialists"). Prior to his employment at Accretive Health, Dr. Hirsch, a board-certified internist and HIV specialist, practiced and served as president of a multispecialty practice in Illinois and medical director of case management at Sherman Hospital in Elgin, Ill.

Contact the Author

This email address is being protected from spambots. You need JavaScript enabled to view it.

Last modified on Friday, 20 March 2015 20:49

Ronald Hirsch, M.D. serves as vice president of physician advisory services (AccretivePAS®) in the Regulation and Education Group ("the REG Specialists"). Prior to his employment at Accretive Health, Dr. Hirsch, a board-certified internist and HIV specialist, practiced and served as president of a multispecialty practice in Illinois and medical director of case management at Sherman Hospital in Elgin, Ill.