For decades, The BMJ (formerly the British Medical Journal) was the kind of publication clinicians brought up when they wanted to impress people at conferences. It stood for careful methods, evidence-based medicine, and the occasional sharp piece of investigative journalism that actually made bad actors sweat.
That’s why many science-minded physicians did a double take in 2021–2023. Suddenly, The BMJ wasn’t just publishing clinical trials and thoughtful editorialsit was pushing “investigations” that looked tailor-made to fuel COVID-19 vaccine skepticism. Science-Based Medicine managing editor David Gorski summed up the collective head tilt in his 2023 article, essentially asking, “What the heck happened to The BMJ?”
In this deep dive, we’ll unpack what changed, why those changes matter for vaccine discourse, and how the broader ecosystem of misinformation, journalism, and evidence-based medicine turned one respected journal into a case study in how not to communicate about vaccines.
From fraud-busting hero to antivax talking-point generator
The BMJ’s glory days: exposing real fraud
To understand the disappointment, you have to remember where The BMJ started. More than a decade ago, the journal published Brian Deer’s landmark investigation into Andrew Wakefield’s infamous 1998 Lancet paper that linked the MMR vaccine to autism. That work helped expose Wakefield’s work as fraudulent and was a turning point in dismantling one of the most persistent myths in vaccine history.
This was the BMJ everyone admired: skeptical, methodical, and willing to call out bad science even when it was politically messy. It embodied the best of evidence-based medicineusing data, context, and transparent reasoning to correct the record.
Fast-forward to 2023: a very different vibe
By late 2023, Science-Based Medicine and other science communicators were noticing a different pattern. Instead of carefully demolishing bad claims, The BMJ was publishing pieces that:
- Overstated problems in vaccine safety systems like VAERS.
- Relied heavily on cherry-picked anecdotes with minimal context.
- Framed regulatory bodies like the CDC and FDA as evasive or untrustworthy, often without strong evidence.
The problem wasn’t that The BMJ dared to scrutinize vaccinesthat’s good science. The problem was that the journal’s 2021–2023 “investigations” on COVID-19 vaccines and safety systems frequently read like they were built to confirm a narrative rather than test a hypothesis.
The VAERS “broken system” story: how a real issue got weaponized
What VAERS is actually for
The Vaccine Adverse Event Reporting System (VAERS) is the U.S. “early warning” system for vaccine safety. Co-run by the CDC and FDA, VAERS is designed to accept reports of anything that happens after vaccinationwhether or not the vaccine caused it. It’s a signal-detection tool, not a verdict machine.
VAERS has two key traits that matter for this discussion:
- Passive reporting: anyone can file a report (patients, parents, clinicians, manufacturers).
- Built to be noisy: it intentionally collects rough, messy data so that statisticians can look for unusual patterns“safety signals”which then get validated using more robust systems like the Vaccine Safety Datalink (VSD) and other active surveillance tools.
What The BMJ’s VAERS “investigation” claimed
In November 2023, The BMJ ran a high-profile story by journalist Jennifer Block arguing that VAERS was “broken.” Her article described:
- Anecdotes of individuals struggling with a clunky reporting process.
- Delays in follow-up by CDC staff amid a flood of COVID-19 vaccine reports.
- Separation between public VAERS data and back-end updates, framed as a lack of transparency.
None of these concerns are completely fictional. The COVID-19 vaccination campaign massively increased VAERS volume, and staffing didn’t magically scale overnight. But the way the article was written made it easy for antivaccine activists to spin as “proof” that regulators were hiding harms and that VAERS data confirmed their worst fears.
What the evidence actually says about vaccine safety
While The BMJ was painting a picture of a system in chaos, multiple large datasets were telling a very different story. U.S. and international research has consistently found that COVID-19 vaccines:
- Substantially reduce severe disease, hospitalization, and death.
- Carry rare but real risks (such as myocarditis in certain young male populations), which were identified quickly through VAERS and active systems like VSD.
- Do not increase long-term all-cause mortality. A nationwide French study of 28 million adults, for example, found a 25% lower mortality rate overall in vaccinated people and no sign of increased four-year all-cause mortality.
In other words: yes, the safety systems were stressed and imperfect, but they did what they were supposed to dogenerate signals, confirm or refute them with better data, and adjust guidance accordingly.
How The BMJ became useful to antivax narratives
Cherry-picking and anecdote inflation
Science-Based Medicine and other critics pointed out a recurring pattern in The BMJ’s newer investigations:
- Start with a small number of emotionally charged anecdotes.
- Generalize them to the entire system (e.g., “VAERS is broken”).
- Understate or omit context that would blunt the dramalike the existence of multiple overlapping vaccine safety systems or the scale of the vaccination campaign.
This is classic narrative framing: the reader never quite gets overt falsehoods, but the omissions and emphasis are arranged so that the “obvious” conclusion is that regulators are incompetent or hiding something.
Asymmetrical skepticism: regulators vs. activists
Another red flag: The BMJ’s scrutiny was often harsh and forensic when directed at public health agencies, yet comparatively soft toward sources with long histories of vaccine skepticism or conspiracy-leaning activism.
While BMJ-affiliated content has also hosted work analyzing misinformation and its spread in English-language news media, those findings rarely seem to meaningfully inform the tone of some of its own high-profile investigations.
The result is a lopsided picture: regulators are portrayed as cagey and conflicted, while sources with strong ideological priors are framed as brave truth-tellers.
The social media megaphone effect
The problem is not just what The BMJ printsit’s how those articles are used. Once online, a piece with a dramatic headline about “broken” systems and “hidden” data is rapidly:
- Amplified by antivax influencers on X, Facebook, and Telegram.
- Stripped of nuance into memes and out-of-context screenshots.
- Cited as “from a top medical journal” to lend authority to claims that vaccines are killing thousands.
Even if the original article never explicitly says “COVID-19 vaccines are dangerous,” the combination of selective framing and downstream amplification ensures that this is precisely the takeaway large audiences receive.
Why this shift matters for evidence-based medicine
Trust is hard-earned and easily repurposed
The BMJ’s reputation was built over decades of publishing high-quality clinical trials, guidelines, and careful critiques. That trust is now being borrowedand sometimes squanderedevery time a borderline-responsible investigation gets turned into propaganda fodder.
Meanwhile, the actual infrastructure for vaccine safety monitoring is robust and multi-layered: VAERS, VSD, v-safe, CISA, the FDA’s BEST initiative, and independent academic groups all work together to detect and investigate potential risks.
When a major journal repeatedly implies that these systems are fundamentally untrustworthywithout providing strong evidence to back that claimit doesn’t just “ask hard questions.” It torpedoes confidence in the entire idea of post-marketing surveillance.
Misinformation vs. critical inquiry: there is a line
Not every controversial or critical article is misinformation. Questioning policy decisions, analyzing data limitations, and even exposing failures in safety systems are all essential. The difference is:
- Critical inquiry starts with evidence, acknowledges uncertainty, and presents context that might weaken the narrative.
- Misinformation-adjacent storytelling starts with a narrative, cherry-picks supporting evidence, and downplays context that complicates the story.
The concern raised by Science-Based Medicine is that some recent BMJ investigations sit uncomfortably close to the second category, even if they never cross the line into outright falsehood.
How journals can do better with high-stakes topics
1. Pair investigations with transparent methodological standards
Journals routinely demand strict methods for randomized trialssample size calculations, pre-specified outcomes, statistical plans. But investigative features often get a much lighter touch.
For high-stakes topics like vaccines and public health systems, journals could:
- Require clear criteria for source selection (who was interviewed, and why?).
- Include explicit discussion of limitations and potential biases in the reporting itself.
- Commission paired commentaries from independent experts in epidemiology and risk communication.
2. Treat epidemiologic context as non-optional
If you’re going to describe a “broken” safety system, readers deserve base rates and denominators. How many doses were given? How many serious adverse events would we expect in that population even without vaccination? How do VAERS signals compare with findings from active surveillance systems?
Without epidemiologic context, anecdotes will always feel like smoking guns, even when they’re just statistical noise.
3. Anticipate misuse in a polarized environment
We live in a world where anti-vaccine content is organized, motivated, and extremely good at clipping and reframing mainstream outputs. Studies of news coverage show how selective emphasis and repetition can magnify misinformation even when journalists don’t intend to spread it.
That means editors should ask, before publishing:
- “If this is screenshotted without the nuance, what impression will a casual reader walk away with?”
- “Are we providing enough caveats and data that bad-faith actors will have to work harder to weaponize this?”
Experiences from the trenches: how The BMJ’s shift feels on the ground
Beyond the high-level analysis, it’s worth looking at what this evolution in The BMJ’s coverage feels like to the people who live with the consequencesclinicians, researchers, and patients navigating a post-pandemic information mess.
Clinicians facing confused patients
Imagine a primary care physician who has spent the last few years counseling patients about COVID-19 vaccines. Early in the rollout, they leaned heavily on data from regulatory agencies, large observational studies, and yes, respected journals like The BMJ. Their go-to message: “The evidence shows these vaccines dramatically reduce your risk of severe COVID-19, and we’re watching safety very closely.”
Now picture a patient walking in holding a printout or screenshot of a BMJ investigation suggesting that VAERS is overwhelmed, opaque, and missing serious safety signals. The headline and pull quotes are alarming. The nuanced bits about limitations and context mostly disappeared once the story hit social media.
The physician has to do the slow, painstaking work of repair: explaining how VAERS actually functions, what “safety signals” mean, how active surveillance systems like VSD and CISA backstop passive reporting, and how large-scale cohort studies show no increase in long-term mortality among vaccinated people.
That’s not just a time sinkit’s a trust sink. Patients are left wondering why a prestigious medical journal and their own doctor seem to be telling two very different stories about the same system.
Researchers watching their work get overshadowed
On the research side, epidemiologists and statisticians who design vaccine safety studies are often frustrated to see careful, multi-year analyses boiled down to one line“vaccines are safe”while emotionally gripping anecdotes from an investigative feature dominate the conversation.
Many of these scientists are publishing work that directly addresses the fears hinted at in The BMJ’s investigations: large datasets showing stable or reduced all-cause mortality after vaccination, analyses of how safety signals are detected and managed, and reviews of misinformation patterns in news coverage.
Yet when a journal with a big megaphone chooses to highlight dramatic narratives about overwhelmed systems without balancing them with the weight of that evidence, it sends a quiet but powerful signal: the drama is more newsworthy than the data.
Public health communicators juggling mixed messages
Public health communicatorspeople working for health departments, hospital systems, or nonprofitsoften report that their work has become a kind of “translation triage.” They’re not just countering fringe blogs or anonymous social media accounts anymore. They’re also constantly contextualizing stories from mainstream outlets and journals whose headlines overshoot their own evidence.
When The BMJ publishes a piece that suggests VAERS is “not meeting its own standards,” communicators have to decide how to respond. Ignore it, and the story spreads unchecked. Engage with it, and they risk amplifying it further. Either way, they have to explain (again) how safety monitoring actually works, why multiple systems exist, and how academic and regulatory checks interact.
Patients stuck in the middle
For patients and families, the end result is a kind of epistemic whiplash. They see long-term safety studies, reassuring summary pieces from sources like JAMA or KFF, and statements from professional societies emphasizing that vaccines remain one of the safest and most effective interventions in modern medicine.
Then they stumble across a BMJ investigation framed around systemic failure and hidden information. It doesn’t say, “Vaccines are killing people,” but it strongly suggests that the systems we rely on are unreliable, underpowered, or opaque. For someone already anxious or distrustful, that can be the nudge that pushes them away from vaccinationdespite millions of doses and mountains of reassuring data.
Why the BMJ story matters beyond The BMJ
The story of “what the heck happened to The BMJ” isn’t just about one journal losing its way. It’s about how even reputable institutions can become unwitting amplifiers of distorted narratives when they lean too hard into drama, underplay context, or forget how their work will be used in a hyper-polarized environment.
Evidence-based medicine depends not only on good data, but also on responsible storytelling. When the storytelling slipseven slightlyin a prestigious outlet, the downstream effects ripple through clinics, public health campaigns, and dinner-table arguments for years. The challenge now isn’t just to call out missteps, but to rebuild norms that align high-impact journalism with the same rigor we demand from clinical research.
Conclusion: can The BMJ course-correct?
So, what the heck happened to The BMJ? In short: a respected journal with a proud history of exposing real medical fraud drifted into publishing investigations that, intentionally or not, lend credibility to antivaccine narratives by overplaying system flaws, under-explaining context, and underestimating how their stories will be weaponized.
None of this means The BMJ is “an antivax journal”it still publishes solid research, editorials, and reviews. But when its investigative arm treats complex safety systems and pandemic-era policy decisions as fodder for dramatic, context-light narratives, the journal stops being part of the solution and becomes part of the confusion.
The fix isn’t mysterious. It looks a lot like evidence-based medicine itself: transparent standards, balanced skepticism, careful use of data, and a deep awareness of how information can be misused. If The BMJ leans back into those principles, it can rebuild its position as a trustworthy voice in vaccine debates. If not, it risks being remembered less for debunking Wakefieldand more for accidentally giving his modern successors fresh material.
