Patients are not plane crash victims and hospitals are not plummeting jetliners

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

When advocates want to emphasize the human impact of a deadly disease or a treatment that might cure it, they sometimes draw an analogy to jumbo jets falling out of the sky. It happened recently in this NPR story about a potential new treatment for sepsis, which the story says kills about 300,000 people each year (estimates vary widely depending on the data sources used).

“So that’s the equivalent of three jumbo jets crashing every single day,” said Paul Marik, MD, an ICU physician quoted in the story.

I recall another time the analogy came up when I was reporting on public health recommendations related to salt intake. An American Medical Association official said that deaths due to excess sodium exceeded 150,000 per year.

“This is the equivalent of a jumbo jet with 400 people on it crashing every day,” said Dr. Steven Havas, then the AMA’s vice president of public health, in the Washington Post.

Hold on a minute: Are these deaths really comparable?

People dying of sepsis or heart disease don’t share a lot in common with plane crash victims. Hurtling to earth in a blaze of fire is an exceedingly rare and spectacular event – precisely the reason we can’t look away from such catastrophes.

Sepsis, by contrast, is common. And while it can occur suddenly in almost anyone who develops an infection, it most often occurs in the hospital among older patients or those with compromised immune systems and serious health problems. In many cases it’s the end of a long, slow descent rather a sudden plummet from the sky.

And salt? There is no conclusive evidence that lowering salt intake would prevent anyone from dying, and so those planeloads of victims simply can’t be said to exist. They are phantom blips on a radar screen tracking a flight that never left the ground.

“Conceptually useless and confusing at best and manipulative at worst”

Risk communication experts I spoke with panned the plane crash analogy and advised journalists to avoid it.

Steven Woloshin, MD

“I don’t think [planes crashing] is a particularly good analogy,“ says Steven Woloshin, MD, Professor of Medicine at the Dartmouth Institute for Health Policy and Clinical Practice and author of Know Your Chances, a book about understanding medical risks.

“Good risk messages communicate both magnitude and context,” he says. But the crash analogy delivers only on the magnitude side of the equation. “If you’re a journalist writing a story, instead of looking at plane crashes, you’d want to compare it with your chances of dying from something similar like cancer or heart disease.”

Mirjam Jenny, PhD, Head Research Scientist at the Harding Center for Risk Literacy, seconded Woloshin’s call for context. She said it was useful to compare risks for diseases with other causes of death that people are familiar with. However, she said the crash analogy didn’t serve this purpose because daily plane crashes simply don’t happen.

Mirjam Jenny, PhD

“In the ‘falling planes format,’ real risks (sepsis killing 300,000 people each year) are compared with completely made up risks (3 jumbo jets crashing every day),” she said. “Planes crash, of course, but at a very different rate, which makes the risk comparison conceptually useless and confusing at best and manipulative at worst. In my view, comparisons of different risks can provide the reader with useful context, but only if all risks in the comparison are real.”

She offered examples of real comparisons that could put large numbers — which are difficult for people to grasp — in context. For example: “To put this into perspective, roughly 11 out of 100 people who died in 2015 in the US, died of sepsis.” Or, “that is roughly 3 times the number of people that can be seated in the Texas Longhorns football stadium.”

Why should you care about this?

Comparing medical deaths to plane crashes is an attempt to overcome your cognitive defenses. It’s designed to shock you into paying attention to something you wouldn’t ordinarily care about. It’s a play to fear and emotion.

The intent may sound noble – “Do something about sepsis!” – but in practice I find that these appeals are a prelude to oversimplified solutions. Having accepted that the problem is like jets crashing all around us, we are primed to accept whatever answer is being proposed to stop the carnage. Our decision-making is impaired.

In the case of sepsis, the quoted researcher was publicizing preliminary evidence showing that vitamin C may be an effective treatment. But instead of carefully calibrating that message to fit the early stage of the research, he trumpeted the treatment as a potential “cure” in a news release that received worldwide media attention.

The benefits of sodium reduction initiatives are similarly exaggerated. While it’s certain that cutting sodium would reduce population-wide blood pressures, it’s not clear that desalinating the food supply would yield the promised lifesaving benefits. Gold-standard clinical trials haven’t shown  that cutting sodium prevents the outcomes that people care about like heart attacks and death (few such studies have been done). Some observational studies show that reducing sodium intake to levels recommended by public health groups is associated with adverse effects.

Medical errors: calculating deaths and assigning blame

The plane crash analogy also has been extensively used to call attention to deaths caused by medical errors. One recent study found that medical errors contribute to the deaths of 250,000 people each year, making it the third leading cause of death in the United States.

Vox recently described the tally from a different study as “the equivalent of nearly 10 jumbo jets crashing every week — or the entire population of Birmingham, Alabama dying every year.”

Once again, readers should tread carefully and question whether the numbers here really add up.

The eye-popping 250,000 figure is extrapolated from data that includes just 35 total deaths in all – hardly a representative sample to draw conclusions about the entire U.S. population.

Moreover, assigning blame for causes of death is a notoriously tricky business. Medical errors occur more frequently in older, sicker individuals who are receiving care that’s more complicated and more prone to mistakes. When an error occurs in someone who’s already on death’s door, is it fair to assign 100% of the blame for that death on a medical mistake? How much of the death is due to the error and how much to the underlying condition?

As a starting point, stories addressing this issue could make it clearer to whom the increased risk applies rather than simply passing along a massive estimate of deaths (the true size of which is vigorously disputed).

Unintended consequences of fear-inducing messages

This doesn’t excuse medical errors or suggest that errors shouldn’t be fully investigated and addressed — no matter how old or sick the patient is or how close to death he or she might be.

Nor is it a knock on patient safety groups and watchdog journalists who are pressing for systemic reforms in this area to reduce medical mistakes. Their work is vitally important.

It’s a call for everyone to exercise care when communicating such big numbers to the public — and in choosing the imagery that accompanies the statistics.

There is potential for unintended consequences when we equate hospitals with crashing jetliners.

Might some people frightened by such messages decide to avoid medical care altogether and die unnecessarily from treatable conditions?

The answer is, we don’t know. And that brings up another concern that Woloshin raised with respect to the plane crash message: He doubts it’s ever been tested to see how people interpret it.

“Maybe it works and they hear it as meaning a big number – but maybe it introduces cognitive dissonance, as in plane crashes are really rare, so maybe this isn’t worth paying attention to?”

Most reporters, of course, are never going to test the messages they use in their stories. But they can follow best practices for messaging that are based on risk communication research.

Here’s a crash course.

You might also like

Comments (10)

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

Skeptical Scalpel

May 1, 2017 at 3:06 pm

Where does the figure of 300,000 people dying of sepsis every year come from? I can’t find it. Please explain. Thanks.

    Kevin Lomangino

    May 1, 2017 at 4:09 pm

    The estimate of sepsis deaths comes from the NPR story. I checked it against this CDC report that suggested the number was a reasonable guesstimate:

    I thought the number was needed to set up the math in the subsequent “crashing planes” equation, but there’s certainly some uncertainty around the figure and I will update the post to reflect that.

    Thanks for commenting.

    Kevin Lomangino

      Skeptical Scalpel

      May 2, 2017 at 9:40 am

      Thanks for the link. I’m sorry to beat a dead horse here but 2,470,666 deaths divided by 16 years = 154,417. These are also listed as “sepsis-related deaths; for 22% of these decedents, sepsis was listed as the underlying cause of death.” 154,417 X 22% = 33, 983 deaths from sepsis per year. Many people die with sepsis but not necessarily because of sepsis. This is more in line with the National Vital Statistics Reports, Vol. 65, No. 5, June 30, 2016 which does not list septicemia as one of the top 10 causes of death in the US. At about 33,000 deaths per year, it would be 12th.

      I forgot to mention in my comment yesterday that I completely agree with you about the 747s crashing analogy. It is used to inflame rather than enlighten.

Karl Davis

May 1, 2017 at 3:15 pm

I also disliked the plane crash analogy. It didn’t add anything, and was distracting from the actual protocol comparison.
On the other hand, Dr. Marik’s data is compelling. He used hydrocortisone, which is approved for sepsis, vitamin C, which is being researched for sepsis, and vitamin B, which is known to be deficient in the cells of sepsis patients, and used them together. He saw dramatic declines in all-cause mortality, sepsis mortality, hospital re-admissions, and the time patients spend in the ICU. Did he make it up? I doubt it, because the hospital administrators confirmed that they are billing far less to insurance companies for sepsis patients now. It is one thing to wonder about fraud, but it is mostly used to get more money, not less. Other doctors who have tried this have also liked the results. I urge other hospitals to give it a shot. I would also like to see a multi-leg study comparing variations on this. We don’t know what dosages are optimal. Are the doses the same for men/women/races/weight of patients? Would it be better to have a high initial dose then back it down?
The plane crashes, however, are just hype.

    Kevin Lomangino

    May 1, 2017 at 4:11 pm


    I don’t think he made up the data. But claims of “cure” require remarkable evidence to back them up. And this small before and after study doesn’t provide that evidence.

    Kevin Lomangino
    Managing Editor

      Karl Davis

      May 2, 2017 at 9:55 am

      I agree with the saying that “extraordinary claims require extraordinary evidence”, as the saying goes, and that Dr. Marik has not proven that his treatment works. I see two levels of evidence here.
      One is proving that it works, which implies that a hospital that does not use the method is liable for malpractice damages. That definitely has not been proven.
      A lesser amount of evidence is needed to decide that a treatment is reasonably safe to try when a patient faces a high risk of death or permanent organ damage. I think this protocol is safe enough to try, because (1) all the medicines used are approved, (2) no adverse reactions to combining the treatments have been reported, (3) Dr. Marik’s hospital has dramatically lowered mortality and kept it down from Jan 2016 to the present, and (4) other “early adopters” have tried this and anecdotally reported favorable results.
      By now, I assume places all over the world have tried it. Dr. Marik claims to have received feedback from over 100 doctors who tried it, although there is no way to verify this. I think the key thing at this point would be to create a central team and data storage point, perhaps at the NIH, to gather and analyse this feedback.

Christian Lillis

May 2, 2017 at 7:58 am

Perhaps the plane crash analogy could be better or more nuanced. We lose tens of thousands of Americans to just preventable healthcare associated infections every year. The controversy over the figures put forth by CDC, journalists and others is largely due to the refusal by healthcare facilities to accurately track and report medical harm. As someone who lost a parent to sepsis brought on by a C. diff infection, I find the blithe way in which the author says these deaths are common to be utterly offensive. While my mother didn’t plunge out of the sky, she went from being a healthy 56-year-old kindergarten teacher to dead in a few days. In that time, she violently ill, had her system shocked with drugs to try and reverse the sepsis, had a total colectomy and, ultimately, succumbed to DIC; all while her family watched. The healthcare community should spend its time focusing on minimizing preventable medical harm than hang wringing over analogies that advocates use.

    Kevin Lomangino

    May 2, 2017 at 9:12 am


    I am sorry that my description caused offense, which was certainly not intentional. By pointing out the flaws in how we communicate about death and disease, I do not wish to minimize the suffering of any individual or their loved ones.

    I disagree that minimizing preventable medical harm is the only thing that should be focused on. As I point out in the post, journalists have a responsibility to report accurately and with context, or they risk causing harm themselves. We have no idea to what extent hyperbolic descriptions may influence people not to seek needed medical care or cause other harm, and I think that’s an important issue to be aware of.

    Kevin Lomangino
    Managing Editor

Kathy Day RN, Patient Safety Advocate

May 2, 2017 at 9:17 am

The comparison between the number of patients who die, every year, because or PREVENTABLE healthcare harm, and jumbo jets crashing is important. Think of the attention that a crash, killing several hundred passengers gets. Think about the investigation, the accountability of the airlines, the questions and answer. None of that happens with the hundreds of thousands of families affected by preventable healthcare harm every year. There is a total lack of attention for the up to 440,000 lives lost gets when each of those unfortunate patients dies, quietly in several hundred hospitals all over the country. Until these preventable deaths of patients of all ages gets the attention they should, this comparison is important. It is graphic. My own father died because of a hospital acquired infection, that he caught in a facility that had an outbreak, that was kept secret, and swept under the rug. He was the third of three patients who contracted MRSA in a very short time there, and all 3 died. This happened in a rural Maine hospital, and nobody would have ever known about it, except that I broadcast it all over and made a State and Federal issue of it. This would have been one more quiet, unmarked preventable death out of 440,000 every year. Blaming patients and their human condition (age, sex, current health, etc) is unacceptable. Healthcare industry accountability is the way to go, and prevention must be the priority. If it takes a comparison to multiple airline crashes, then so be it.

Peter Lipowicz

May 13, 2017 at 6:04 am

There is a real difference between salt guidelines and sepsis treatment. There will probably never be a “right answer” for salt as the question has been around for decades. But we will know soon whether the Marik protocol for sepsis works or not.