Map of life expectancy at birth from Global Education Project.

Thursday, April 27, 2006

The New JAMA: The stuff the corporate media didn't cover

One reason I started this blog is because I was continually frustrated by the lack of correlation between the stuff I read in the medical literature that I thought it was important for people to know about, and the stuff that got reported in the general media. After a year and a half of writing the blog, I'm still not entirely sure what it is that makes reporters think something is worthy of your attention.

Anyhow, until I figure it out, here's the latest. The two items in the new JAMA that appear to have gotten some attention are a study about implantable defibrillators, and a study by people at Public Citizen about conflicts of interest on FDA advisory panels. Both of these got spun like tops in the mass media, and therein lies a hint: there are coporate interests (medical device manufacturers and drug companies) with a direct interest in both of these stories. They probably called up the reporters to spin the stories, and then of course the reporters got on them and talked to the authors and/or various potential counterspinners, and they had the "he said/she said" format they needed. Journalists don't tell you what actually happened, that's not their job. They have to be fair and balanced, which means they have to transcribe what other people say, especially people who have a personal interest in the matter and who can be presumed to be trying to manipulate you.

So here is a story that I think is important. Eric Peterson and sufficient collaborators to sing the Coronation Mass looked at 350 hospitals around the country that provided care for people who had suffered heart attacks. As you probably know, there are process of care guidelines for these and other common conditions that are treated in hospitals. For example, people with heart attacks (I won't go into the technicalities of the different kinds of events that are called heart attacks, but we're talking about the common, basic kind here) should be given aspirin and other anti-clotting agents including a relatively new one called a glycoprotein IIb/IIIA inhibitor, and beta blockers. They should also get some of these drugs prescribed at discharge, and get dietary and smoking cessation counseling if indicated. Seems simple enough but on average, only 74% of indicated treatments were provided. Unfortunately, this is typical: for whatever reason, such standards of care are not consistently followed.

The question has always been, however, whether these process indicators are really valid, in the sense that they predict patient outcomes. Maybe doctors are using appropriate judgment in not always following them. This study finds otherwise -- in the 25% of hospitals with the lowest adherence to the standards, patients with a documented myocardial infarct (death of heart muscle tissue) were nearly twice as likely to die in the hospital as in the 25% of hospitals with the best adherence. Some of the specific predictors had to do with discharge practices, which obviously could not have caused in-hospital mortality, but are probably indicative of an overall culture of diligence and quality.

This study is observational, not experimental, and hence cannot absolutely prove that better adherence to the standards will improve patient outcomes. There could be other explanations, e.g. the less adherent hospitals are seeing sicker and weaker patients. But the authors beat up the numbers every way they could to try to rule this out and the conclusions stand up to the pounding.

One important observation: the hospitals with inferior quality indicators had a higher percentage of non-white patients. This suggests, as have some other studies, that disparities in health care quality according to race and ethnicity may be at least partly explained by where people get their care, rather than differential treatment within a given institution. That requires more study.

This study gives considerable support for the use of these process of care standards as indicators of quality. That has been controversial, and it still will be to some extent, but I think we'll be seeing more and more attention paid to these indicators, and to figuring out why they aren't followed. (How hard is it to give somebody an aspirin, for crying out loud?) We will see more and more pressure to rate hospitals publicly on these indicators. I have been a skeptic, but maybe this is a good idea.

No comments: