Thursday, April 24, 2003
Lies, Damned Lies, and Statistics
Yesterday I read this abstract from the New England Journal of Medicine about adverse drug events in outpatients. I got it through the ACP Online email service. Although I don't have the link to the entire article (subscription required), the ACP condensation and the abstract raised some questions in my mind, so I'd like to talk about it anyway. The data for the study was collected from chart reviews and by having patients fill out surveys asking them if they had had any symptoms related to medications (I'm assuming the patients just filled out the surveys without asking health care providers whether the symptoms were likely to be due to the medications or not).
Here is what the ACP had to say about it, and I emphasize this because any reports in the lay press about the study are likely to be similar:
A recent study found that one-quarter of primary care outpatients suffered an adverse drug event. That figure is four times the number of events reported in hospital settings.
A patient survey of four primary care practices in Boston found that the drugs most commonly cited in adverse events were selective serotonin-reuptake inhibitors, antihypertensives and nonsteroidal anti-inflammatory drugs. While none of the events were life threatening or fatal, 13% were serious and 11% were preventable.
Authors of the study, which was published in the April 17 New England Journal of Medicine, said that physicians could have avoided many of the preventable events by using computerized prescribing tools. Researchers also said that physicians could have prevented or ameliorated many errors through better communication strategies, such as using e-mail and translators.
The authors also recommended better monitoring of side effects by physicians, nurses or pharmacists.
To restate: this study says that 25 percent of all patients in outpatient primary care practices are having side effects or "adverse drug events." This means, most likely, that anything - diarrhea following antibiotic use, headache, mildly upset stomach, anything - was viewed as a side effect.
Now, I don't want to give patients medications that make them worse off than they already are. And it's true that some adverse drug effects can be serious. But one of the facts of life is that medications have side effects, pure and simple. There is no such thing as a "smart drug" that only affects one symptom or organ system. Nonsteroidals can cause stomach pain or ulcers; serotonin reuptake inhibitors can interfere with sleep patterns or make people jittery. The question to me is, don't these medications have more benefits than side effects? The other question is, was any attempt made to separate really significant side effects or avoidable problems from run-of-the-mill complaints? It seems to me that the only thing this study is likely to do is to make people more reluctant to take medication that they actually need.
Before I close this sounding like a hard-nosed "take your medicine and shut up" doc, I should say that I always tell my patients to call me if they're having problems with their medications. I also try to tell patients what the most likely side effects are with their meds and what they can do to ameliorate them (if that's possible). Communication is important and it can be lifesaving. When I see statistics like this, though, my first reaction is to doubt the study, not to believe it.