Medical Education Reform Part 2

Opinions from the Trenches

Today we are happy to post a follow - up to our September 20th post on reforming medical education. These two articles are from trainees still making their way through the years of undergraduate and graduate medical education in the US and UK. (Coincidentally, they are all also podcasters!).

As always please share your thoughts. We hope this will be an ongoing “substack brainstorming” series.

First up, an essay by Daniel and Mitch Belkin. Daniel Belkin is a transitional year intern at Memorial Sloan Kettering Cancer Center. He will complete his diagnostic radiology residency at Johns Hopkins. Mitch Belkin is a transitional year intern at Maryland General Hospital. He will complete his diagnostic radiology residency at University of Maryland Medical Center. Together, Daniel and Mitch co-host the External Medicine Podcast (@exmedpod).

Signaling in Med Ed

The goal of medical school is to develop caring physicians who have the skills to diagnose and treat patients. Medical education reformers have criticized specific teaching methodologies and the specific content of knowledge tested. For instance, shouldn’t the MCAT/USMLE Step 1 reflect the actual substance of being a doctor? Rather than, say, memorizing the right-hand rule of magnetism or enzymes in the Krebs cycle? Shouldn’t we teach future doctors to be better Bayesians or to better parse biomedical literature? While these criticisms are reasonable, and more reform is necessary along these lines, to truly understand medical training one needs to pause and look at competing economic theories of education.

The two major theories of education are the Human Capital Theory (HCT) and the Signaling Theory (ST). HCT argues the value of education is in skill development. Employers hire workers because they develop practical knowledge during their training. By contrast, ST contends that education is more about demonstrating valuable qualities which students already possess. So which theory better explains med ed?

Consider the Sheepskin effect. A student who completes only 95% of medical school or residency has 95% of the knowledge but receives little of the financial benefit. If educational value were skills-based, someone who completed most of these programs ought to get most of the benefits, right? But no, this isn’t true.

Of course, students do learn skills during medical school such as how to do a physical exam or an H&P. However, the selection pressures of the application cycle, the curricula of medical schools, and the evaluation methodology all suggest signaling is an essential part of med ed. 

So what qualities do students signal they possess during medical training?

1. Diligence. Standardized exams are crucial to medical school admissions (MCAT) and success in residency selection (USMLE Step 1-2). Testing in medical school is frequent and brutal. Exam results correlate poorly with most measured outcomes except other standardized exams. However, by going through the gauntlet of multiple-choice hell, students signal their commitment to medicine. The fact that the signal is expensive (high opportunity cost) makes it harder to fake.

2. Conformity. Medical training emphasizes obedience to formalized leadership structures. There is a clear totem pole from department chair to attending physician down to the 3rd year medical student. What better way to signal conformity than to jump through AAMC and LCME hoops, take orders from senior residents and attendings, and patiently talk to standardized patients who think they’re Al Pacino. The beating in of conformity might be a way to create doctors who obediently follow guidelines without doing the research themselves, and who won't question their slow corporatization and loss of autonomy.

3. Willingness to work for free. Much attention has been paid (pun intended) to how poorly resident doctors are remunerated relative to their value. However, a willingness to provide free labor is what medical schools select for when they expect hundreds of hours of volunteering during the admission process. This along with mounting expectations of research likely add further barriers to those from lower socioeconomic backgrounds.

4. Brown-nosing. Because subjective evaluations during hospital clerkships matter so much in residency selection, medical students obsess over their social interactions during 3rd and 4th year. This likely feeds into the culture of slurping in academic medicine, and selects for people who enjoy the process and against those who are neurodivergent.

Because medical training is in large part signaling, superficial curricular changes will not address the underlying traits we’re selecting for. While diligence is reasonable, if we want to improve physician quality, we need to select for more creativity, independent-mindedness, socioeconomic diversity, and neurodivergence.

---

Next up an article by George Milner. George is a final year medical student at the University of Cambridge. Interested in medical education, evidence-based medicine. He is co-host of the Pager Podcast (@PagerPodcast).

Bringing high-quality medical education to a wide audience

In the current system, a combination of competition and lack of organisation between universities and programs leads to much of the best medical education content staying in-house. The guarding of medical education resources and excessive replication of non-interactive content by different providers to their respective audiences serves the interests of universities (and some medical educators) rather than students and patients.

Adam Cifu’s proposal of a national, free online pre-clinical course can be expanded to include clinical education. Nationally funding a set of free shared resources selected from a range of institutions and integrating these resources into clinical courses could:

1. give local educators more time to run interactive teaching;

2. save students from suffering through poorly delivered and poorly prepared content;

3. allow more time and money to be used to generate resources and keep them up-to-date;

4. increase equality between different courses;

5. increase incentives for educators to deliver high quality content. (any teaching delivered should be competitive with the best alternative use of a student’s time).

There is a particular mandate to share resources in systems in which medical education receives large public subsidies.

One criticism of an increasing reliance on centralised resources is that this would stifle variation. A generation of doctors who all approach problems uniformly is not necessarily advantageous. Some counterpoints to this are:

1.      such a system would never provide all teaching – there is still huge space for local variation;

2.      variation can also be introduced in central resources – two lectures teaching different approaches to AKI is not impossible;

3.      students can be better exposed to practice outside their home institutions through resource sharing;

4.      not all variation is good variation!

For profit or non-profit organisations already provide widely accessible resources and, as Adam Cifu acknowledges, they are vastly popular amongst students (and doctors). They do, however, suffer from limitations. These include limited funding; excessive focus on passing exams versus excelling in clinical practice; prohibitive pricing structures and lack of appropriate detail for different stages of training.

The extent to which “mainstream” medical education organisations should interact with private ones is also an interesting question. However they view their appropriateness, medical educators should have an innate interest in the resources that their students are using.

Consider the following thought experiment faced by a professor of paediatrics: which of the following is likely to have the greatest positive effect on national cohort of medical students?

a)      Recording a lecture on childhood rashes and uploading it to the medical school’s online portal or,

b)      Reviewing and editing the questions on childhood rashes in a national question bank (none of which I have any affiliation to beyond a subscription)

I think most would pick the latter, despite an intrinsic aversion to it. It is likely that a free, public system integrating resources from different institutions would have more buy-in from the best medical educators. Resources could also be better integrated into local teaching programs. One compelling reason why students persist with low quality teaching is the ever-present worry that its content will appear in local exams. More frequently referring to high quality content centrally instead of generating low quality content locally could avoid this.

Overall, a vast amount of high-quality medical education does not reach as wide an audience as it should do. At the same time, most educators would be delighted to see students benefit from their work. Whilst some universities may feel that greater sharing would diminish their competitive advantage, such advantages are largely self-serving. Greater collaboration is in the interests of patients, students and all those interested in improving care.