Yesterday, I saw this graphic linked by Kevin Drum, purporting to show that the US hospitals are killing their patients in horrifying numbers:


The Hazards of Hospitals

The thing set off huge alarm bells.  Not because I think that it's impossible that US hospitals are dropping the ball on hospital acquired infections--indeed, my column next month, which is already at the printers, will be on antibiotic resistance.  And hospital procedures like hand-washing play a huge role in the spread of resistant infection.


But these numbers were wildly out-of-sync with the ones I'd seen.  While some European nations have made really admirable progress on hospital acquired infections (the Netherlands, for example, seems to have virtually eliminated MRSA through strict isolation procedures), others are as bad or worse.  What makes a country's health care system vulnerable to hospital acquired infections or other iatrogenic deaths is not how their health care is financed; it's a large number of idiosyncratic differences in hospital procedure, treatment guidelines, and drug distribution.

So I started looking at the links, which (at least the ones I looked at) didn't say what they were purported to say; for example, the graphic says that the US ranks last out of 19 countries in preventable deaths at hospitals, but the source seems to be a Commonwealth Fund report that ranks the US last in "amenable death", which basically includes any death that could have been prevented by more "timely and effective healthcare".

I was going to write all this up . . . but Kevin beat me to it:

I guess it was too good to check, so I didn't check it. But a reader emailed this morning to suggest that this was preposterous, and he seems to be right. I checked the references at the bottom of the MBCC chart, and none of them seemed to back up their numbers. What's more, a few years ago the CDC estimated 99,000 deaths per year out of 1.7 million HAIs, a mortality rate of 5.8%. For the EU, the European Centre for Disease Prevention and Control estimates 146,000 deaths per year out of 4.5 million HAIs (see p. 27), a mortality rate of 3.3%.

That's a modest difference, and it gets even more modest when you read more about these estimates, which are very, very rough and depend strongly on exactly how you count infections and how you attribute deaths. You can read much more about it in this WHO report if you're interested. The chart below, from the WHO report (with U.S. figures added from here), shows HAI prevalence rates in various high-income countries, and on this score the U.S. does pretty well. Most likely, the U.S. is about average both in prevalence of HAI and in mortality rates from HAI. Apologies for the error.

The problem of hospital acquired infection is quite bad enough without inflating it to ludicrous proportions. I have no idea why medicalbillingandcodingcertification.net wants to turn a collection of deliberate half-truths into an infographic, but they ought to be ashamed of themselves.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.