Superficial coverage of medical errors could leave erroneous impression with readers

Kevin Lomangino is managing editor of He tweets as @KLomangino.

iStock_000023477801_SmallThe health news media were hopping yesterday with a big story about medical errors.

Seemingly all the major outlets carried the story, with headlines so alarming that they’d have any conscious hospital patient demanding an immediate discharge.

These headlines all bore a striking resemblance to that of a Johns Hopkins news release about the study: Study Suggests Medical Errors Now Third Leading Cause of Death in the U.S.

I call attention to the word “now” used in all of these headlines — which suggests to the casual reader that the rate of medical errors is the United States is on the rise.

But that’s not what the BMJ paper found. And it’s not how the paper described its conclusion in its own title:  Medical error—the third leading cause of death in the US.

Using data from several older studies, the BMJ paper used a mathematical model to extrapolate a new and different total for medical error related deaths. And that total — more than 250,000 — is higher than estimates from previous research.

But the paper did not attempt to assess trends in the rate of medical error-related deaths, nor did it collect any new data. So it’s dubious to suggest — without explanation — that errors are “now” coming into 3rd place.

Out of the many stories I read on this topic, only Bloomberg’s coverage explicitly addressed whether medical errors are actually increasing as implied in these headlines. Their story noted:

That doesn’t mean deaths from medical errors have increased since the 1990s. Because different methodologies were used to come up with the numbers, it’s hard to say what the trend looks like.

I’ll grant that this is a fine point and maybe even a little nitpicky. And if that were the only problem with the coverage, then I wouldn’t be writing this post.

But the problems run deeper than ambiguous headlines. As the blogger known as Skeptical Scalpel (a self-described former chairman of surgery) points out, the BMJ paper takes some pretty big leaps to come up with its new national total for error-related deaths. And the idea that all of these deaths are entirely preventable is also open to debate, he observes.

Makary’s review extrapolated that figure [~250,000 deaths] from three papers published before 2009 which had a combined 35 supposedly preventable deaths. That’s not a typo—35 deaths in all. One of the papers stated that all 9 deaths in three tertiary care hospitals were preventable. In his BMJ paper, Makary says, “some argue that all iatrogenic deaths are preventable.”

I disagree. I have analyzed other papers on this subject and pointed out that certain complications and deaths are not 100% preventable. For example, no study of deep venous thrombosis and pulmonary embolism shows total efficacy of any prevention strategy. And some patients will suffer myocardial infarctions and die even when they are properly treated.

Thirty five deaths extrapolated to 250,000?

I’d like to have seen more reporters taking a closer look at those calculations.

Vinay Prasad, MD, MPH, an Assistant Professor of Medicine at the Oregon Health and Sciences University, told me in a telephone interview that he was also underwhelmed by the news media coverage of the study. He said it’s “a very uncertain science” bordering on “rampant speculation” to ascertain whether a particular medical intervention or error would or would not have resulted in the death of a patient.

This is not like the Vietnam War where you know exactly who was a victim of the conflict. When the public hears a statistic like ‘251,454 deaths,’ they get the impression that these are people whose names you could put on a wall — but that’s not the case at all.

And he said that the framing of medical errors as the “third leading cause of death” is also likely to leave a misleading impression. Whereas any error is regrettable, he said, his suspicion is that some of these deaths are happening in very sick, elderly patients whose death was imminent regardless of what care they received.

When someone is 19 and they’re missing out on 70 years of lost life because of a medical error, that’s much different than when someone is 85 years old, already on a ventilator, and is likely to die any day now.

None of this is meant to excuse or diminish the problem of medical errors. Whether the total is 50,000 or 250,000, the number is too high and arguably under-appreciated.

But when news outlets join in lockstep to promote a sensational and possibly inflated figure without carefully vetting the underlying evidence or providing context, that’s also a significant problem whose ultimate consequences are unknown.

Prasad said that it’s possible that any attention to medical errors — even if the problem is overstated — would ultimately promote useful change in practices that would lead to better care for patients.

But he said it’s also possible that such attention could lead people to avoid the medical system entirely or seek out alternative forms of care that are not effective. And that would be a net loss. He added that he could also envision scenarios where, in reaction to the perceived massive problem of medical errors, hospitals “create some nightmarish bureaucratic solution” that ends up harming patients in unintended ways.

Prasad said it’s impossible to predict these eventual long-term repercussions, which is why journalists have to focus on more immediate objectives: accuracy and context.

“Communicating accurately is a virtue no matter what the downstream consequences are,” he said.

You might also like

Comments (23)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Antonio Piccolboni

May 5, 2016 at 6:11 pm

Thanks for covering this story. I agree the number reported is subject to large errors and that it does not imply any trend. But I don’t understand why you try to belittle it by criticising the “cause of death count” as a flawed statistics. It’s been used for at least a century and the CDC routinely reports it. It’s true that “years of potential life lost” may be a better statistics for many purposes, but that’s totally unrelated with the issue at hand. The idea that fatal medical errors affect prevalently the elderly and very sick is reasonable speculation and in any event the fact that a less sick patient can survive an error, at the cost of his or her quality of life, is little consolation. The idea that a statistics can be dismissed because it doesn’t provide names and last names for each unit is typical of doctors’ contempt for numbers. It’s like saying that the quarter million deaths estimated as a consequence of the Chernobyl accident or the half million, by some accounts, caused by the Iraq war never happened because we could only estimate them, not count, independent of the accuracy of the estimate.

    Kevin Lomangino

    May 6, 2016 at 8:26 am


    Thanks for reading and commenting. I think your analogy about Chernobyl and the Iraq War is off base. No one is saying that medical errors don’t happen or that they aren’t serious even when they happen to elderly very sick people. But I do think we as news consumers deserve some context when a study serves up a statistic that many experts would disagree with and which seems to be based on questionable assumptions. That’s what I was asking for, and that’s what I think news outlets had an obligation to provide with their coverage. But I did not see much of it.

    Kevin Lomangino
    Managing Editor

    Jack Reacher

    May 8, 2016 at 4:01 am

    @ Antonio
    An elderly is much more fragile than younger patients because of all the ageing process: more susceptible to infections, late immune response, late erythroipoiesis, less tolerance to stress, multi organ failure, etc.
    The fact that there are more deaths in the elderly doesn’t mean doctors make more medical errors at that age.
    In geriatrics, doctors are more careful to avoid fatal side effects treatment. However some deaths can’t be prevented even to the best of care as any treatment can have a serious adverse side effects in the elderly. To treat or not to treat is a common question geriatric doctors ask themselves, weighing all the risk benefit. However some study would find it a medical error, and preventable.
    So yes, the cause of death count is a flawed statistics.

Lisa LaMagna

May 6, 2016 at 5:41 pm

If the criticism is that the report needs to extrapolate data, then perhaps it is upon providers to document medical errors, report them to families and publicly, and expand coding on death certificates as the researchers recommend. Additionally, hospitals and care centers should provide open reporting systems and emotional and professional support for medical professionals who are part of this tragedy. However your criticism that this is subject to errors is seeking to discredit the media reporting, shows a basic misunderstanding of how media covers news stories. Even the Bloomberg story ran this headline: “Medical Errors Are Leading Killer After Heart Disease and Cancer, Study Finds.”

    Kevin Lomangino

    May 8, 2016 at 8:25 am


    Your suggestions regarding better documentation of medical errors seem reasonable. My post was about the lack of analysis and context in the media reports covering the BMJ study.

    Kevin Lomangino

Emily Paterson

May 6, 2016 at 9:58 pm

Not all groups covered this “big story.” Ironically, I didn’t see anything about this on KevinMD. In my state, I have been told that when the state went tort reform, attorneys left the state, switched to personal injury or only take specific malpractice cases, representing less than 1% of harm. The problems definitely seem out-of-control in some states.

Mel Caleb

May 7, 2016 at 4:24 am

My family deaths this year: two cases of kidney failure due to overuse of contrast dyes. One hospital acquired MRSA infection. One fatal medication error. All in one year, just my immediate family. Representative sample? Crap for luck?

David Mangen

May 7, 2016 at 8:01 am

I have not read the original article. Did the author provide standard errors for his estimate? This would provide some much needed context.

David Antoon

May 7, 2016 at 1:43 pm

As one of thousands of patients permanently harmed patients, I welcome Makary’s estimate as well the more accurate peer reviewed article by John James. My injuries and these estimates are NOT “superficial.” Unreported medical error caused by concurrent, not overlapping, surgeries performed by unsupervised doctors unknown by the patients and in violation of patient consent is a crime. Hospitals are maximizing volume and profit without concern of adverse outcomes. Salaried staff physicians are incentivized with bonuses for volume output. Until positive patient outcomes are incentivized, patients will continue to be needlessly harmed; until adverse negative events causing patient harm and death are accurately reported, estimates based upon the best information available must be taken seriously. Makary’s study and John James’ earlier study are reporting the “tip of the iceberg” experienced by hundreds of thousands of harmed patients.

    Kevin Lomangino

    May 8, 2016 at 8:33 am


    Thanks for reading and commenting. Please re-read my post and the headline. I called the news coverage of the BMJ study “superficial.” I did not and would not call anyone’s injuries superficial. I believe that the accuracy of the BMJ data should be scrutinized — and indeed it has been questioned by a number of experts as discussed in my post. But that context did not make it into the news coverage of the study. I think that’s a problem worth calling attention to.

    Kevin Lomangino
    Managing Editor

Beth Waldron

May 9, 2016 at 5:20 am

Indeed there is an issue with what is considered a ’cause of death’ and it goes beyond medical errors. Even if one does agree with the 250,000 error figure, that still does not place it as the true ‘3rd leading cause of death’. Pulmonary embolism (VTE blood clot) killls approx 300,000 Americans annually (per CDC). Yet VTE is also excluded as a defined ’cause of death’ in CDC Vital Statistic reporting (which are limited to 113 ‘select’ causes of death), so you will not see it in any HHS ‘top 10 causes of death’ list. The implication is huge: It is difficult to advocate for what isn’t reported. As a result, there is $0 in the federal budget to target VTE, even though it is the #1 cause of preventable hospital death (per AHRQ). VTE kills more Americans annually than AIDS, breast cancer and car crashes combined, yet receives so little attention. The selective presentation of statistics truly prevents us from making better health policy decisions targeting those conditions which genuinely cause the greatest death and disability. More

Barbara Duffy

May 9, 2016 at 6:52 am

OK, I do’t get it — on two counts. #1 This is not new news (see below). #2 And more importantly – where is the outrage? Why aren’t we demanding better as professionals and as patients?
Safety, issues, reform, quality, 440,000 deaths
Issues, safety, quality How many die of medical mistakes?
Safety, issues, quality, nursing homes One third of nursing home patients harmed
Hospital Safety Score – patient deaths – hospital errors are third leading cause of death – safety, patient safety!
New study focuses on medical errors in outpatient settings
Deaths by medical mistakes hit record, July 2014, error, harm, EMR
Medication Errors Seen in Half of All Operations in Study, 80 percent of the mistakes were preventable, October 25, 2015, medication, safety, surgery

John Galbraith Simmons

May 9, 2016 at 8:31 am

For several years I’ve worked as a medical writer in the area of patient safety and must commend Kevin Lomangino for this critique of coverage. Note that Bloomberg and Washington Post stories rightly cite the 1999 Institute of Medicine study, which was indeed a key turning point with respect to awareness of the extent and magnitude of errors in medicine. But the WP report uses problematic phrasing, calling errors “incredibly common.” Nothing incredible about it. From my point of view, John Raphael’s windup paragraph was the best and most accurate of all coverage cited. He noted: “Medical errors don’t necessarily mean that doctors or nurses are inadequate. Most of the errors represent systemic problems, including poorly coordinated care, fragmented insurance networks, the absence or underuse of safety nets, and other protocols.” Exactly.

william reichert

May 9, 2016 at 9:48 am

Physician errors are real and do cause harm. But I am puzzled by the fact that according to the CDC, smoking “causes 480,000 deaths in the USA each year” and yet, this number, twice the physician caused number, is not listed as it should be as the #1 Cause of death in the USA each year. Why is that? If I set my house on fire and the fire department fails to put it out ,as they are expected to, because they arrived 20 minutes late, did the fire department cause my house to burn down?

Pyara Chauhan

May 9, 2016 at 9:55 am

This article renders good service to the medical fraternity and to the community at large by putting a reasonable perspective on medical errors and the resulting number of deaths. The total of 39 actual deaths might sound as ridiculous as the unimaginable number of a quarter of a million seemingly implied by the article in BMJ. I practiced in this country for over thirty years and held a number of significant positions on the various Hospital Medical Staffs including Morbidity and Mortality Committees. That certainly does not make me an expert, but if 250,000 deaths or even half due to medical errors is true, surely I should have noticed. My extrapolation is from the incidence of prostate cancer of about 200,000 to about 220.000. I saw my share of prostate cancer yearly in my not-so-busy but reasonable urology practice. I should certainly have come across an equal number of deaths from medical errors considering our stringent review criteria. Something to think about before we throw numbers around.

Doug Woolley

May 9, 2016 at 10:35 am

I believe the intent of this article was to point out that this information was old and may be not accurate as the analysis was performed from a perspective of 20,000 feet. As a retired MD. who used revel in the opportunity to define the cause of death in ER patients I found that even with immediate access to all members of the health care team and the medical record and all tests, that the cause of death was not always clear and almost always was multifactorial as very few patients got the ultimate test… an autopsy. These long distance/high elevation reports are interesting and can suggest trends, but are hardly accurate and definitive.

Lynn Colwell

May 9, 2016 at 10:49 am

So glad to see this. When I saw the headline, I immediately thought of this site and wondered if you’d be commenting. Thanks for the clarification.

Bill Conder

May 9, 2016 at 12:22 pm

I expect media coverage to be sensationalist. But Makary and Daniel make the important point that the CDC report is based on death certificates from states using ICD codes as indication of cause of death. However, there’s no code for medical system error in these ICD codes (“ICD-10 coding system has limited ability to capture most types of medical error.”). There’s an issue of culpability here as well as improving patient care. And if the problem is as big as their extrapolations suggest their conclusion points to medicine’s collateral damage, which could/should shake up the whole system. Then media sensationalism becomes more accurate.

Bonnie Friedman

May 9, 2016 at 1:59 pm

Medical errors are a serious problem regardless of the mathematical analysis under or over estimating the true numbers. Whatever the numbers, families can play a significant role in protecting their loved ones by staying alert, tracking developments, speaking up when necessary and working closely with the medical team. Here is my Letter to the Editor of the Washington Post in response to last week’s story:

Yanling Yu

May 9, 2016 at 2:17 pm

My Dad was an 81-year old elderly when he died due to a negligent use of a medication that was contraindicated for his medical conditions. He wanted no advanced interventions to save his life but safe and evidence-based treatments to help him remain comfortable. Yes, he might not have had many years left in comparison to a 19-year old. But, to him and to my family, everyday of his remaining life was precious. To say an elderly is going to die anyway because of their illness is insensitive to our elderly patients to say the least. It is not when one dies. It is the cause of the death and how one dies that matters.

    Kevin Lomangino

    May 9, 2016 at 2:50 pm


    I am sorry about your dad, and certainly did not mean to diminish or be insensitive to anyone’s suffering. Part of the context for my comments is that the care of elderly people who are nearing death becomes much more complicated because of multiple medications and other interventions that are happening to keep them alive. This increases the likelihood that a medical error will occur. So, if young people hear the numbers being reported in this study and conclude that they face a similar risk, they are misreading the results and being misinformed.

    Kevin Lomangino
    Managing Editor

george lundberg

May 10, 2016 at 11:19 pm

With the very low autopsy rates in most America hospitals, it is difficult to conclude anything meaningful about errors and deaths.

Edward Thompson

May 26, 2016 at 12:00 pm

Medical errors have always happened and will continue to do so. Actually finding out if a particular instance is avoidable is a question for the legal system, should that route be pursued. My advice? If you lose a loved one through possible medical negligence, you need expert help. Oh, and a good litigator, of course. They may be harder to find!