Negotiating the fog of health care fantasy
Before we can fully understand why medical dogma has gone so far astray, and why Medicare’s quality indicators push us to purse a costly and aggressive medical course that has not been shown to be beneficial to the elderly, we need to take a step back and talk about something called relative risk/benefit. So much of what we know about the efficacy of tests and treatments is built upon a faulty foundation that has led to erroneous and often dangerous conclusions. That is because virtually every outcome measure in studies and in the press is presented in the form of relative numbers. What is relative risk/benefit? Most simply, it is the percentage of change that a specified intervention will achieve. For instance, when we claim that using statin cholesterol medicines reduces the risk of heart attack by 30%, that is relative benefit of using a statin. But the absolute benefit—the real chance that any individual will benefit from using a statin—is often much less impressive. Take the example of a fabricated disease that causes horns to grow on your head. Let’s say that the horn disease gets a lot of press because a few famous people contract it. Then let’s say that a company develops a drug that cuts the risk of someone developing the horn disease by 50%. That sounds pretty impressive! But if the horn disease is very rare (only 2 out of a billion people get it), and if the drug cuts that risk down to 1 out of a billion, that is a 50% relative risk reduction in the chance someone taking the drug will get the disease, something the press may celebrate as a major breakthrough on the front page of its paper! The absolute risk reduction of disease, though, is only one out of a billion. In other words, out of a billion people taking the drug, only one will benefit. Who knows how many others will suffer side effects; that information may not be published! Much of what we know about health care rests upon a temple of relative risk numbers. How good are stents, statins, aggressive treatment of diabetes and hypertension, memory drugs, vitamins, annual exams, stress tests, hospital admissions; even a tiny improvement can be magnified into something that seems much more significant once we speak in the language of relative risk. In fact, very few studies have ever shown meaningful improvement in virtually any interventions in the elderly once absolute risk numbers are used, and many show very high risks in absolute risk terms with our vaulted interventions. But still, the press and the medical community chatter only in the language of relative risk, and so we are misled into thinking we should always be doing more. So as we continue talking about number fixing, remember: much of the fiction that drives us to want to fix numbers is based on relative benefits, and behind those faulty calculations are absolute risks of those very same interventions that the press and others have buried beneath a wishful script that has made us believe that with aggressive posturing we can cure aging itself.