June 20, 2011
Avocado on the Salad Bar at Downstate
For my epidemiology class, one of our assignments this week was to discuss the Randomized Clinical Trial. Strengths, weaknesses. As usual, I took a somewhat lighthearted look at the issue. It's been just over a year since my magical day at SUNY Downstate with the *real scientists* and I was thinking about how much fun we had talking about low carb, physics and baseball (you can guess which one of those conversations I could actually contribute to, as opposed to watching in amazed bewilderment.) There were fresh avocado halves on the salad bar, where Dr. Feinman bought me lunch after promising to do so for nearly a year.
I love Randomized Clinical Trials. How can you not? The very idea that even a small portion of the world can be controlled, manipulated, observed and measured is so comforting in these times of profound uncertainty. It reminds me of a musing of a teenage heroine in my favorite teen romance novel, "Puppy Love" by Janice Harrell (written by my mother's former psychiatrist's wife some thirty years ago but never mind) when she (the heroine, not the psychiatrist or his wife) was looking at her brother's best friend (on whom she had a big crush) from across the dinner table: "You don't have to own Sequoia National Park to be glad it's there," I feel happy just knowing that out there Randomized Clinical Trials are being performed, even if I'm not the one conducting them.
The strength of an RCT is its ability to single out a specific factor or set of factors and study its effects. You want to know the effects of avocado consumption on medical school professors? Select a group of medical school professors and feed them avocados. Select a very similar group and don't feed them avocados. If you want to do a crossover trial, switch the groups. Does regular avocado consumption yield shinier coats in medical school professors? What better way to find out than to do an RCT? We don't have to sit around waiting to see if they eat avocados and develop shiny coats, nor do we have to just guess based on the numbers of avocado sales in the medical school cafeteria and the shininess of the hair of the medical school professors. We can proactively set up a situation to introduce avocado consumption and then just as we giveth the avocado, we take it away.
One of the weaknesses I think of the RCT is that it gives the researchers a false sense of security and therefore may encourage researchers to make conclusions regarding causality that are not merited. It's very hard to control for everything, or to anticipate what will cause what. What if the medical school professors really like avocados, and therefore they are happier while eating the avocados, which leads them to comb their hair more often (everyone knows that medical school professors groom more when they're happy and like cats, often cease to groom when depressed) which leads to greater distribution of the natural oils of the hair throughout their heads, leading the hair to appear shinier and to in fact be shinier. Researchers might have hypothesized that avocados, due to their concentration of fat, would make the hair shinier, but in fact this study did not prove for certain that the fat in the avocados caused the improved shininess of the hair. Going to press in the NEJM with the news that eating avocados makes your hair shiny might send medical school professors to the produce market in droves, but the ones who don't like avocados might quickly find that frequent avocado consumption isn't making their hair shinier at all.
RCTs are often impractical and often unethical. A problem with viewing an RCT as the gold standard is that it might discourage one from making recommendations based on other studies, even when those recommendations might be beneficial to the public's health. It's a dangerous game, to be sure, but the paper I am attaching about an RCT using the crossover method to determine the efficacy of parachute in preventing deaths among those who are thrown out of airplanes is a good example of where an RCT was truly not necessary to make a public health recommendation. Read it: it is the funniest thing EVER. I repeat: the funniest thing EVER!
Regarding equipoise, I can see that in the case of a horrible outbreak of a new virus, an RCT of a potential vaccine that also has pretty serious known side effects could come under fire, because the question of whether intervention or no intervention would benefit the patient is a tough one to answer. Ethically it seems to be always safest to do nothing, but in the case of a vaccine and a new and deadly or debilitating illness, it's harder to define the concept of "benefit the patient most."
However the clear examples from medical history already raised on the discussion board point to examples where the principle of equipoise would have ruled out research that was unethical.
I think that in the end it is safest to feed avocados to medical school professors. What's the worst thing that can happen? An allergic reaction here and there, a green stain on a white shirt... it could be worse. You could be thrown out of a plane without a parachute.
Posted by april at June 20, 2011 6:56 PM
Regarding your line about vaccines having serious side effects: you'd be amazed if you watched TV, to see the incredible number of ads that are on every day for a huge variety of meds. What is even more astounding is that the potential side effects of the meds are often much worse than the disease they are intended to treat! There is one asthma medication in particular whose side effects include (you guessed it) death from asthma! Creepy. I won't even mention the erectile disfunction meds.....
Posted by: Judith at June 21, 2011 6:50 PM
April, I think I'm going to refer people to this post in the future! :-D
Judith, the thing about side effects is that not everyone gets them or gets them severely. If you feed eggs to people, you'll get side effects from people who have egg allergies. Severe side effects. And even NOT taking a medication has a side effect... in the case of erectile dysfunction meds, a diminished sex life or even marital problems, and the psychological/physical consequences. If your side effect is a rash, then probably meds>no meds. If you are one of the, say, 20 who get heart palpitations, no meds>meds. Weigh your risks with either side and make a choice, I guess. :/
Inaction is also a choice and has consequences.
On the subject of silly studies: http://aem.asm.org/cgi/reprint/15/4/899.pdf
Posted by: Marie at June 23, 2011 6:24 AM
I've been following your blog a lot ever since I read it from almost beginning to end back in February 2010. I may have even left a comment back then, although I honestly can't remember and I might have used a different name, lol, since I think I tried to keep a CR blog for about 10 minutes before I shied away and deleted.
Anyway, I'm back to "moderate CR" I guess you would call it. I'm just dropping down to 1400-1500 (I'm a 5'2" female who exercises) and I do my best to avoid crappy food, but I still have some on occasion.
I just wanted to tell you something that happened today that just made me think of you. I was at a work conference and they served lunch. They had almost NOTHING I felt comfortable eating! There was lunch meats (I'm mostly vegetarian), lots of bread, mayonnaise-ridden potato salads, desserts, etc. So I was with another colleague who can't eat that kind of stuff either because of stomach problems. She and I looked at the garnishments for the sandwiches (some lettuce and tomato slices and pickles) and we descended on it like vultures!! I think everyone around us thought we were nuts, filling up our plate with that stuff! ;D
(I'm not perfect, though, I had chocolate candy during the meetings, lol)
Posted by: olehcat at June 27, 2011 5:15 PM