26-06-2025
Patient Survival Jumps If In-Hospital MD Aced Assessment Test: Study
Patients are significantly likelier to survive a hospital stay if their hospitalist scored in the ... More top quartile on a knowledge test linked to board certification.
The odds that a patient survives a hospital stay sharply increase if the physician overseeing in-hospital care aced a special test designed to assess doctor knowledge and judgment, according to a new study.
The study found that patients of hospitalists who placed in the top quartile of a professional exam from the American Board of Internal Medicine were nearly 8% less likely to die within a week than the patients of doctors with lower scores.
In an interview, Furman McDonald, a hospital medicine specialist who's president and chief executive officer of ABIM, emphasized that the actual number of patient lives affected by the relative difference among physicians was 'massively significant.'
'The absolute difference in mortality is about four people per thousand hospital admissions,' said McDonald. 'By way of comparison, the mortality from cardiovascular disease, the number one killer in this country, is about two patients per thousand. So this is a massively significant result.'
Assessing Knowledge
The professional exam, known as the Longitudinal Knowledge Assessment, is given quarterly and is designed to both assess physician clinical knowledge and encourage learning. Every five years, ABIM determines whether the physician maintains board certification – a prestigious designation that also has concrete economic value. The study examined the medical records of more than 260,000 Medicare patients and the test scores of more than 4,000 physicians specializing in hospital medicine.
If all hospitalized patients had fared as well as those cared for by physicians who achieved the top quartile in knowledge and judgment, 1,069 lives would have been saved yearly just considering the first week of care, the study's lead author, ABIM health economist Bradley Gray, said in an interview. Top scorers also had fewer readmissions, suggesting fewer complications.
'The patient has a better prognosis if the doctor knows more,' said McDonald. The expertise of the doctor really matters."
The predictive value of tests of doctor knowledge has known for decades and has been repeatedly reaffirmed in focused studies involving hundreds of thousands of patients and many different specialties, McDonald said. This study, published as a research letter in JAMA Internal Medicine, was designed to assess whether the results of this newer exam correlated with better patient outcomes.
The connection between ABIM board certification and patient outcomes, if substantiated for all the subspecialties the ABIM certifies, could easily affect many millions of Americans. The group oversees some 270,000 physicians practicing in 22 different subspecialties. A previous study sponsored by the ABIM found that the score on its exam for internal medicine trainees applying for certification for the first time was also associated with improved patient outcomes and reduced readmissions. McDonald says the group plans to continue its test assessment effort.
One notable feature of this study was a methodology that allowed researchers to compare different doctors as if they were working in the same hospital so as to minimize any influence the facility might have on outcomes. In addition, various adjustments to reduce the risk of misleading conclusions meant the researchers were 'underestimating the extent to which the higher scores are flags of quality,' said Gray. 'We really wanted to have a believable research design.'
Open Book Beats Closed Room
Before the Longitudinal Knowledge Assessment exam was instituted as an option in 2022, all physicians seeking to be recertified by the ABIM took a detailed, all-day test in a controlled environment every ten years. The LKA, in contrast, is a 30-question, open-book exam given every quarter, with a dashboard telling the doctor the correct answer to each question and, in detail, how they fared in comparison to the ABIM certification standard and their peers. That continual feedback is meant to prompt focused improvement before the recertification test, which is given every five years.
McDonald acknowledged he was 'stunned' when a clinical trial the group ran to evaluate alternative approaches to administering the new test showed that the open-book approach most produced results most likely to accurately discriminate among different clinicians. 'It was better able to tell the doctors who knew less from the doctors who knew more,' said McDonald. 'It was amazing.'
The approach worked, McDonald suggested, because of the way the questions are designed. ABIM spends a substantial amount of time and money developing and validating realistic clinical vignettes. The goal is for the answers to reflect reliable judgment, not rote knowledge.
'It turns out the doctors who know more are even able to know what to look up and how to look it up,' McDonald said. Thus far, use of artificial intelligence chatbots has not been a problem, but the group is closely monitoring the potential for abuse.
Physicians seeking board certification are highly motivated to learn how to be better doctors, Gray and McDonald both noted. Over time, looking at how doctors' scores change and evolve, 'we'll be able to see whether learning results in improvement in patient care,' Gray said.
