I remember the first research project I was genuinely proud of. It was clumsy, small, and took far longer than it should have. But I learned something real: how to ask a question, wrestle with data, and accept what the evidence actually showed. That experience shaped how I practice medicine today.
I worry that this kind of formative struggle is disappearing. In its place, something uglier has taken root: a transactional model of research where the point is not discovery but decoration, where a bibliography line matters more than what it says, and where citations have replaced curiosity as the currency of academic worth.
The numbers are staggering. Medical students matching into neurosurgery in 2024 reported an average of 37 research products on their applications. Even those who failed to match averaged 31. Across 22 of 24 specialties, research output among first-year residents jumped nearly 20% in a single year following USMLE Step 1's shift to pass/fail scoring. This is not a renaissance of medical science. It is, as one medical educator put it bluntly, "just the latest battleground in the Residency Selection Arms Race."
The COVID-19 pandemic accelerated this trend. With clinical rotations canceled and travel restricted, students, particularly IMGs navigating an already complex and unpredictable pathway to the Match, turned to remote research as the one thing they could still do. Many had no choice. But the infrastructure around them, the mentorship, the methodological rigor, the institutional oversight, did not scale with the demand. The result was predictable: a flood of low-yield publications, surveys with dismal response rates, case series from single departments, p-hacked EMR data-mining exercises, and an ocean of "COVID-19" in the title.
About a quarter of all medical student-authored publications now receive zero citations. Not low citations. Zero. The science did not matter. The line on the CV did.
And then there is the darker side. On WhatsApp and Telegram, groups openly sell authorship positions on medical publications for a fee, with pricing that varies by author sequence. First authorship costs more. These are not shadowy operations; they recruit in broad daylight. On Reddit, medical students discuss the ethics of paid authorship with a disquieting casualness.
This is not an isolated phenomenon. A 2025 study identified over 32,000 articles likely originating from paper mills in a single journal. Researchers estimate that approximately 108,000 fake biomedical publications enter the literature annually, representing nearly 6% of all indexed biomedical papers, fueling an industry worth an estimated $1 billion per year. The number of fake papers doubled every 1.5 years between 2016–2020. Publisher Hindawi shut down entirely because paper mills had overrun its journals.
The International Committee of Medical Journal Editors criteria for authorship are clear, and payment is not among them. But when the system rewards counting and not reading, the market responds.
But this is more than an academic integrity problem. Those low-quality studies do not just sit harmlessly in PubMed. They get scooped into systematic reviews and meta-analyses, which sit at the top of the evidence hierarchy and shape clinical guidelines. John Ioannidis warned nearly a decade ago that the production of systematic reviews had reached "epidemic proportions," with the vast majority being "unnecessary, misleading, and/or conflicted." By 2014, more systematic reviews of trials were being published than new randomized controlled trials themselves.
Cochrane has since developed new policies for managing problematic studies, acknowledging that retracted articles are "only the tip of the iceberg." When the bricks are faulty, the house cannot stand, no matter how carefully you stack them. The downstream consequence is that clinicians making evidence-based decisions may be relying on evidence that was never meant to inform care; it was meant to fill a CV.
I am not writing this to cast blame on medical students, especially IMGs who face extraordinary structural barriers. The problem is not their ambition. It is a system that incentivizes volume over value and then acts surprised when it gets exactly what it asked for.
We need program directors to articulate what they truly value in research, substance over spreadsheets, and to align selection criteria with that message. We need ERAS to separate abstracts from manuscripts, categorize research by type, and make it harder to inflate numbers through double-counting. We need journals to invest in detecting paper mill submissions before they contaminate the literature. And we need to build genuine mentorship pathways for students from underresourced institutions, so that "doing research" means learning to think like a scientist rather than learning to game a system.
The goal of academic medicine was never to produce the longest bibliography. It was to produce physicians who could look at the evidence, know what it means, and care enough to do it right. If we lose that, no number of citations will bring it back.
Muhammad Sameed is a pulmonary and critical care physician at Jefferson Einstein Hospital Philadelphia (JEHP). He is a clinical assistant professor of medicine at Sidney Kimmel Medical College, Thomas Jefferson University and serves as the regional director for the ILD clinic at JEHP.
Image by AMR BO SHANAB/SCIENCE PHOTO LIBRARY / Getty Images
