In assessing clinical trials, it is more powerful to use prospective data than retrospective data. But just like our lives, there is more retrospective data with which to run analysis.  Retrospective analysis is the mere compilation of many trials using the same drugs and nearly the same parameters to determine if outcomes are reproducible and have “power”. Translation: Does the drug do what they claim it does? Occasionally, the researchers will widen the scope of their inquiry beyond those of the original, experiments and ask different questions. For example, for decades drug companies had developed and marketed drugs to lower cholesterol. The parameters are often very basic: does Drug X lower cholesterol? Period. It is useful to know if Drug X lowers cholesterol more than a sugar pill. And in comparison to the placebo, does it cause side effects? Interestingly, placebos can sometimes have more effect than the drug being developed and can cause side effects in equal measure. (This is an entirely different blog post) Sometimes the drug trial asks level 2 questions: Does Drug X lower cholesterol more than Drug Y. Does it have fewer side effects than Drug Y. Braver still and at the risk of getting  Drug X permanently booted from FDA consideration, they ask if Drug X lowers “outcomes”. This is a euphemism for morbid complications or death. When you get clean outcome data (limbs don’t fall off and people don’t die), it is a win-win-win. Win 1 means you get FDA approval and all the years and $$$ poured into the R&D can be recouped. Win 2 means the recouping of R&D costs makes shareholders happy. Win 3, Drug X may actually lower the risk of a heart attack or a stroke by lowering cholesterol 40%. Bingo! The real point technically is the long term benefit for the populace in terms of lower their chances of, well, basically dying. The commercials say, “Drug X lowers cholesterol and plaque in your arteries.” And as we sit the $50 cartoon of cigarettes on the check out conveyor with the bag of Doritos, the pack of bratwursts, the 2 dozen eggs with a pound of bacon and the 12-pack of beer, we think “If I take Drug X, I won’t have a heart attack and I won’t die.” Drug X is supposed to take care of the cholesterol for us. Ou own behavior or input is not seen as powerful or potent  enough to be considered a therapeutic method. If anything, telling a patient that they can lower their own cholesterol by 10% just by changing their diet….that feels punitive.

We all want something for nothing. We want someone else (or something else) to free us from personal responsibilities. I no longer have to be responsible for my own safety while driving: I have seat belts, anti-lock brakes, air bags and crumple zones. I do not have to be responsible for my health or longevity because I can take medications that clinically improves “outcomes”.

Then some wise guys goes back and does a reanalysis of the data. They widen the scope of the original analysis and just let the data speak for itself. Just like a patient: you usually only get the answer to the question you ask. Ask if Drug X lowers cholesterol, you get a binary answer. Ask an “open ended” question and you get a very different answer. A recent example of this happened with hormone replacement protocols for women. Estrogen was the panacea against aging. Take HRT and you preserve your skeleton. You lower your cholesterol. You won’t get old and look like a sagging hag except…..whoa….when the scope of analysis widened we learned HRT CAUSED heart attacks, strokes and breast cancer. [Soapbox warning: I think I’d rather get wrinkles and be really ornery through menopause and keep my boobs. No one medicates puberty yet we medicalize menopause. And it is all because we fear aging and dying.]

In medical training, when we teach medical students to take a patient’s medical history, we teach them to ask “close-ended” questions. Does it hurt here? Yes or no. You break ranks when you ask, “Where does it hurt?” Patients are stupefied when I ask them to “narrate” their symptoms. They think I am supposed to figure it out. I encourage them by saying, “You have all the answers. You know what is wrong with you, you just don’t know how to tell me.” It is not my job to find the cause of the problem. It is my job to ask the right questions and to ask them in an open ended way. It is also my job to widen the scope of analysis. Just because Mary comes to me complaining of her stomach hurting, I cannot focus only on her stomach and her digestion. There are other non-digestive organs in her abdominal cavity. And psychology also influences the gut through the autonomic nervous system. With whom does she eat her meals and where? What is she eating? Nothing can be assumed. Nothing can be edited, censored or excluded until it is all compiled.

I can look backwards on my life and interpret  it only by the parameters I set out to define. It can be treated as a linear data set. And by that protocol, I have poor outcomes [divorced]. But a curious scientist widens the scope and asks other questions and accidental discoveries happen in science even in a day of profit driven research. Sometimes, the benefit you are seeking from a clinical trial turns out to be a failure, but when re-analyzed a new, unexpected outcome has occurred. [Minoxidil is a cool example of this. Most people know minoxidil by its commercial name:Rogaine. Well, minoxidil is a blood pressure agent. It comes in pill form. It lowers blood pressure. In the ICU, it is used in a paste form to lower the blood pressure of people in hypertensive crisis. Except…..the paste made hair grow….anywhere. And a patent was born and some serious profits were made off the baby-boomers’ balding instead of their heart attacks.]

What other outcomes resulted from this life that were unexpected, unpredicted. Are there positive outcomes?  Are there morbidity risks? And did anyone ever ask the Level 2 questions? I like the open ended questions. I love the inquiry. I like the bigger pictures. There is a difference between rumination and analysis. Rumination is circular and non-productive. Analysis means formulation and learning. And that means growth and change.

Leave a Comment

Your email address will not be published. Required fields are marked *