I once asked an executive in a pharmaceutical company who was intimately involved in touting the benefits of the drugs the company made how the numbers were derived.
“If a million people take a drug and instead of, say, 10,000 heart attacks that a typical million people would have, only 8,000 people have heart attacks, do they report the 2,000 fewer heart attacks as a 20 percent reduction in heart attacks?”
“Yes, that’s basically it.”
“Do they offset that benefit by noting that, say, 200,000 people experience side effects and maybe 20,000 are serious, including some that cause permanent harm?” (Numbers are for illustration only.)
“No, they just report a 20% reduction in heart attacks.”
I was reminded of that conversation when I read a news article recently about a new calculation the federal government is proposing to use in its analysis of the benefits of health improvements such as quitting smoking. The idea is that if one is going to say that not smoking saves $xx per year per person, through lower medical costs, one also should acknowledge that people get pleasure from smoking that they give up when they quit, and discount the savings by, the government is saying, 70 percent.
Economists are up in arms, saying that such a big discount will make smoking cessation programs look as if they simply lose money and provide no net benefit, and the nation’s health will be worse as a result.
Just to be clear about where I stand on smoking: my father died of lung cancer due to smoking when I was in my 20s; I have never smoked but have seen others battle this addiction nearly all their lives, and that is a life I would not wish on anyone.
That said, the government has a point in trying to bring both sides of the story into the discussion. The current debate is about factoring lost happiness into the equation. It’s also important to capture unintended but predictable harm. For example, as suggested above, if a drug reduces heart attacks, but side effects cause serious and permanent harm to a notable number of people, shouldn’t those facts be part of the conversation when you are trying to decide whether to take the drug for many years?
I haven’t yet seen anybody come up with a great way to be sure to capture and offset against each other all the significant benefits and risks of treatments and health management choices, including those that surface only over time. But it’s good to know that some effort is being made to bring such questions to light.
