After two back-to-back mass shootings this weekend, which left over 30 people dead in the United States, science communicator Neil deGrasse Tyson tweeted a series of statistics on nationwide fatality.
In the past 48hrs, the USA horrifically lost 34 people to mass shootings.— Neil deGrasse Tyson (@neiltyson) August 4, 2019
On average, across any 48hrs, we also lose…
500 to Medical errors
300 to the Flu
250 to Suicide
200 to Car Accidents
40 to Homicide via Handgun
Often our emotions respond more to spectacle than to data.
While many were quick to call the comparison insensitive and tone deaf, health professionals and science communicators had another bone to pick - this time with the legitimacy of the actual data.
The first statistic about medical errors, they say, is not only dangerously outdated, it's also misleading. In the words of Hank Green, the creator of YouTube's SciShow: "It's the kind of potentially damaging statistic a serious science communicator would never share without context."
You may have seen a tweet today saying that there are 500 deaths every two days in the us due to medical errors. If that sounds impossibly high to you...it is.— Hank Green (@hankgreen) August 4, 2019
It's the kind of potentially damaging statistic a serious science communicator would never share without context.
While the number wasn't plucked from thin air, it does have a long and controversial history that is easily searchable. It comes, originally, from a 1999 Institute of Medicine (IOM) report, which found up to 98,000 people were dying in US hospitals each year from preventable medical errors.
Within a year, other health professionals had formally disputed the idea, starting a two-decade long debate. In 2000, researchers at Indiana University pointed out that many of these 98,000 people were too sick to be compared to the general public. What's more, the IOM report did not account for the baseline rate of death without medical errors, which was found to be quite similar overall.
Even though patient safety reporting systems "proliferated in health care", the error rate didn't appear to decrease: in 2016, a review from Johns Hopkins hit on an unimaginably high number. Looking at studies published since the IOM report, many of which were based only on insurance claims, these researchers concluded that 251,454 hospitalised patients died from medical errors each year.
If that were true, it would mean that medical errors cause well over half of all hospital deaths in the US annually. Soon after it was published, the unfathomable statistic garnered widespread media attention, with sensational headlines claiming medical errors were the third most common cause of death in the United States.
In all likelihood, however, that's just not true. Physician and scientist David Gorski has been pushing back against this myth for years, which he says has been picked up and publicised by anti-vaxxers and those who sell alternative medicine, as an attack against conventional and evidence-based treatment.
As the managing editor of Science-Based Medicine, he has written several detailed rebuttals of these studies, and yet the myth continues to persist.
"Part of the reason, I think, is that these numbers are easy and tell a dire story," Gorski told ScienceAlert over email, "whereas explaining the context and why these numbers are almost certainly massively inflated is hard."
He and other researchers and physicians have long argued that the Hopkins review lacks any formal methodology and relies on wildly varying definitions of medical error that too often conflate these mistakes with adverse events in general.
"Adverse events happen even in the absence of medical errors," Gorski explained in an article for Science-Based Medicine.
"Many adverse events are not preventable and do not imply medical errors or substandard medical care. Moreover, determining whether a given medical error directly caused or contributed to a given death in the hospital is far from straightforward in most cases."
In other words, these estimates are too high, at least in part, because they pin far too many fatalities upon medical errors - a diagnosis that even experts find hard to agree on.
"Most of those errors are what is called 'failure to rescue' or not recognising a problem, rather than active error, it's delay in diagnosis, or delay in treatment," explains trauma surgeon Mark Hoofnagle on Twitter. "Decisions that, in retrospect delayed appropriate care - which isn't medicine killing, but failing to save."
Besides, looking outside the US, there's evidence to suggest the number is indeed much lower.
A 2013 report in the United Kingdom, for example, found that while five percent of deaths in hospitals were deemed to be more than 50 percent preventable, more than half of these occurred in older and sicker patients who weren't likely to live longer than a year.
I'll give him minor credit for using the IOM estimate of 90K deaths/yr from medical error instead of the ridiculously innumerate estimates of 250K to 400K (36-57% of ALL in-hospital deaths), but even that one is likely way too high. https://t.co/VBu4hC72em— David Gorski, MD, PhD (@gorskon) August 4, 2019
A recent and more rigorous study came up with a far more conservative number than either the IOM study or the Hopkins research.
Rather than simply looking at 'medical errors', authors of this study examined all adverse events and their link to patient mortality, whether a mistake or not.
Using data from the Global Burden of Diseases, Injuries, and Risk Factors (GBD) between 1990 and 2016, instead of simply using insurance claims, the new study settles on a number 50 to nearly 80-fold smaller than the Hopkins review.
Across the entire study period, the authors found 123,603 deaths in which adverse events were determined to be the underlying cause of death. And after controlling for population growth and ageing over those 26 years, they found that those rates had actually fallen by over 20 percent.
Of all the deaths related to adverse events, the study found 8.5 percent could be attributed to misadventure, or medical errors such as accidental laceration or incorrect dosage, and 14 percent could be attributed to adverse events associated with medical management.
This isn't to say that mistakes made by medical professionals are not a problem, or that they shouldn't be fixed. Merely that grossly exaggerating this number to the point of a crisis, where a huge number of people who are hospitalised could die from a medical mistake, is demonstrably false and dangerous.
"They see these numbers bandied about and, not having delved deeply into the issue, accept them," Gorski told ScienceAlert.
"To find more realistic numbers takes work and knowledge of the literature and methodology used in these studies, something possessed by a limited number of people."
Neil deGrasse Tyson has since apologised for what he called a "true but unhelpful" tweet.