The number 210. It’s meaningless to most individuals. Is it an area code for a phone number (for San Antonio, to be exact)? A number on the side of a house? But if you are a medical student or physician, even decades out from your medical training, you know exactly what that number means. It is a score on the United States Medical Licensing Exam (USMLE) Step 1 examination. A score that is below average. A score that destines young doctors-to-be to a life of ennui in a job they do not enjoy, because they could not match into the competitive specialty of their dreams. If they are able to match into post-graduate medical training at all.
Two hundred and ten. I am now on medical faculty at a highly regarded academic medical center; double board certified in internal medicine and geriatric medicine. I’m in a job I love, that challenges me and excites me—and 210 was my score on the USMLE Step 1. I felt emboldened to “out” my less-than-stellar performance, on a daylong examination I took nearly 10 years ago, with the recent announcement by the National Board of Medical Examiners (NBME) that the USMLE Step 1 will only be reported as pass/fail, with no numerical score, starting as early as 2022.
This is a seismic shift in the medical education community, and one that has numerous downstream effects in training doctors. It was a move that was sudden and unexpected, though there had been rumblings of discontent amongst much of the medical community about the USMLE Step 1 examination for years. The two most recent presidents of the NBME, Donald Melnick and Peter Katsufrakis, are white men in their 60s—a demographic not particularly associated with radical change. The decision, announced on February 12, 2020, has engendered much praise—and a significant amount of criticism—in the medical community on social media.
Why is this transition so earth-shattering? The USMLE Step 1 is a one-day test, taken at the end of the second year of medical school, that caps the “preclinical” medical school experience—the culmination of all the knowledge doctors-in-training should know before they can safely set foot in a hospital, learning to care for real live patients. Step 1 is designed as a criterion-referenced test—that is, one that measures performance against predetermined learning standards. However, it has morphed and been universally misused as a norm-referenced test, which compares test-takers to one another, even though the test was never designed this way and not built on the standard “bell curve.”
The USMLE Step 1 long been used as a screening criterion for graduate medical education (residency) training programs. Highly competitive medical specialties, such as dermatology and various surgical subspecialties (neurosurgery, orthopedic surgery, otolaryngology) have required very high scores. Even highly regarded programs in “less competitive” specialties, like internal medicine or pediatrics, have required students to answer many questions correctly on a multiple-choice test to be granted the privilege of walking through the hallowed wings of their hospitals. In the nonmedical lay community, it is assumed this examination determines how “smart” or qualified a doctor is to be practicing medicine. It is assumed a poor performance on this test indicates the doctor is incompetent.
That would be well and good—if it were true. In reality, like many other multiple-choice examinations physicians take over the years, this test has next to nothing to do with practicing medicine. The discontent over the content and misuse of USMLE Step 1 was initially brought to the fore by J Bryan Carmody, a pediatric nephrologist at Eastern Virginia Medical School in Norfolk, Va. Carmody, preparing lectures for preclinical medical students, perused retired Step 1 questions to make his lectures relevant. He soon found that some Step 1 questions were testing analysis of Southern blot, a molecular biology testing mechanism, or the biochemical pathways that influence the position of a cell within an organ.
I am an academic physician, but my research interests lay in how to implement a universal health care system akin to other high-income countries, in an era of unprecedented congressional gridlock—not in pipetting. Is knowing about somatic hypermutation in a Southern blot really making me a better physician and health policy researcher and advocate? Carmody did not think so either. Therefore, he and his colleagues initiated a yearslong effort to investigate the adverse effects of Step 1, culminating in a scathing commentary recently accepted to the premier medical journal Academic Medicine.
There has been evidence that “Step 1 mania,” as Carmody calls it, has implications far beyond tests and residency selection. The first is psychological. The high stakes placed on a one-day exam, with little room for life events, can create inordinate stress and anxiety. In a 2016 paper in the Journal of the American Medical Association, researchers found that 27 percent of medical students have depression symptoms, and 11 percent have suicidal ideation, versus 9.6 percent and 3.7 percent in similar age groups in the general population.
A further consideration is financial. The cost of medical school has skyrocketed in recent years. The annual median tuition at a public in-state medical school has increased from $26,700 in 2009, the year I entered medical school, to $39,000 in 2019. It’s even worse for public out-of-state and private medical schools, with median tuition increasing from about $46,000 in 2009 (already an exorbitant amount) to about $63,000/year in 2019.
This is pricing out many students from nonprivileged backgrounds. The median parental income of the matriculating U.S. medical student is $130,000 per year. Only 20 percent of medical students come from families in the bottom 60 percent of parental income (currently less than $75,000/year), and a mere 5 percent of students come from families in the bottom 20 percent of income (less than $25,000/year).
Where is this tuition money going? A recent paper from the University of North Carolina at Chapel Hill showed that $495,000 was spent aligning the preclinical curriculum to test preparation, for example through subscriptions to popular study resources. Furthermore, separate studies have shown that the USMLE Step 1 demonstrates biases in favor of men and traditional age students versus women and non-traditional (older) students and against African American medical students, who were rejected from an internal medicine program at higher rates as a result of Step 1 scores.
What about a physician’s future career? There was no relationship between USMLE Step 1 scores and the odds of receiving disciplinary action in clinical practice [though interestingly, there is a correlation with the USMLE Step 2 CK, an exam taken in the fourth year of medical school that has significantly more clinical relevance]. Finally, the board certification examinations taken at the end of residency/fellowship training, the criteria often used by patients and medical boards to assess physician knowledge within their specialty, have always been criterion-referenced and pass/fail.
The demise of the scored USMLE Step 1 will not in itself solve many of the problems outlined with medical education; however, it could provide the impetus for innovation. Some ideas include significantly reducing the “preclinical” curriculum, now that professors do not have to “teach to the test.” This will allow doctors-in-training to get closer to the patients they will spend their entire career with earlier in their training.
Another idea, which is starting to gain traction, is shortening medical school to three years for physicians pursuing primary care specialties. This will significantly reduce financial costs, making a physician career achievable for those from lower-income backgrounds who want to return to serve their communities. As many Americans demand a more equitable health care system through calls for a universal health care system, physicians should demand the same of medical education.