Sunday, January 20, 2008

A note on studies

Anybody can type a few keywords into PubMed and have great fun worrying oneself and annoying one's doctor. On the positive side, even if you can't understand the bulk of what is written, knowing about a potential treatment's possible effects and side-effects can make it easier to make a difficult choice.

I am not an expert at evaluating studies, but here are a few tips:

1) Check the sample. More, obviously, is better, though most soft-tissue sarcoma studies involve small numbers of patients due to the rarity of the disease. Remember that in a small sample, even one additional response can skew the number in the absract dramatically. I tend also to be biased toward reports from major American and European cancer centers, as well as those published in big journals like Cancer.

2) Read -- OK, skim -- the whole thing. Most libraries, especially at universities, can access medical papers via their institutional subscriptions. (You can, of course, also find many on paper.) The biggest reason to do this is that soft-tissue sarcomas, which come in at least 66 very different flavors, are quite a diverse group, and what is applicable to one might not be applicable to yours. Usually there's a chart that breaks down the sample by disease and response where you can look for what happened to patients with epithelioid sarcoma, rhabdomyosarcoma, or whatever.

3) Check the date. Both when the study was published and when the patients were actually enrolled and treated. This is particularly crucial for so-called prognostic studies that tell you how other people with your disease did. To gather a statistically valid series of outcomes, the researchers may have to aggregate records going back decades. The progress of cancer treatments is overrated in many ways, but there's no point in worrying yourself sick about what happened to a small group of people 10 or 20 years ago in another country. Things have gotten better.

4) Bring your doctor in. Print at least the abstract, ideally the whole text, of anything that particularly worries or excites you. Then get a copy to your doc so she can help you evaluate it. It's safe to assume that your doctor is quite familiar with the clinical literature (if you don't feel this way, find another oncologist), so it's not like you're delivering news flashes, but sarcoma is rare enough that she might not have seen everything, especially if she spends most of her time treating other illnesses. I would use this tactic sparingly, out of respect for your doctor's time and expertise, but I think most physicians would rather spend a little time helping out occasionally rather than having you base decisions on information that they don't know you have, information that may not be fully relevant or up-to-date.

I'll add to this as I gather more thoughts that might seem helpful.

No comments: