Why We Don’t Take Our Meds: The Conclusion

“If you torture data long enough, it will confess anything.” –Nobel Laureate Ronald Coase

“[Medical] science is really hard to turn around. People get emotionally invested, financially invested, professionally invested.” –Dr. Jonathan Moreno, Professor of Bioethics, University of Pennsylvania

Last week I pointed out that a ridiculously expensive drug ($15,000/year) that had to be taken for life, that produced trivial improvements in heart troubles, and that didn’t prevent a single death was actually going to be paid for by the health insurance industry. In a world desperate to reduce health care costs, how can this happen?

Let’s go back to the “headline” number trumpeted by the drug industry—that the P-inhibitors reduced heart troubles “by 15%.” As I pointed out, that was the relative improvement, not the absolute improvement—the latter being what patients care about.

To understand why that’s the case, let’s take an extreme example of a relative improvement. Suppose that in the placebo group heart troubles occurred in this percentage of patients: 0.000005. If the miracle drug reduced heart troubles by 15%, then in the treated group this percentage of patients would experience heart problems: 0.00000425. Big deal. You would have to treat thousands and thousands of patients to prevent even one minor heart “episode.”

Of course, in the heart drug business it’s not quite that bad. Earlier in this series of posts I pointed out that even among the small minority of patient categories who benefit from statins (people who’ve already had heart attacks or strokes), only one patient in every 200 benefitted and those patients lived a mere three months longer than placebo patients. As Clint Eastwood might ask, “Do I feel lucky?”

Suppose, though, that we didn’t really care about benefits to individual patients. Suppose that all we really cared about was keeping health care costs down. That’s when we start caring about the relative reductions in bad outcomes. With P-inhibitors, individual patients reduce their LDL levels to almost nothing, but they experience trivial improvements in heart outcomes and don’t live even one minute longer. (In fact, they die slightly sooner.)

But that 15% reduction still matters. I don’t know what the annual cost of heart disease is in the US—the American Heart Association, which isn’t exactly objective, puts the cost at north of $315 billion/year. But, whatever, it’s a big number. If a drug can claim to reduce that number by 15%, health insurers are going to have to be interested, even if individual patients receive virtually no benefit and even if the drug costs $15,000/year for life.

Which raises a paradox. Virtually every drug paid for by the health insurance industry over the past few decades has promised to reduce health care costs – that’s why the insurers agreed to pay for them. Yet health care costs continue to soar. How can that be?

There are a few possible answers. One is that despite all the hoopla over drug costs, those costs actually represent less than 20% of overall health care spending. So if they promise a lot, what the hell, let’s try them.

But the main answer to the question is this: almost every new drug over-promises and under-delivers. A drug that reduces disease “by 15%” in the rigorously controlled environment of a clinical trial isn’t going to have anything like that impact in the real world. Here’s why:

• People selected to participate in clinical drug trials aren’t remotely representative of the actual patient population. Most people have conditions that eliminate them – heart troubles, diabetes, cancer, smokers, over-65, etc. But these people have to be treated in real life, and the outcomes for them will be far worse. If the improvement in the clinical trial was a measly 1%, we can be sure that effect will disappear entirely in the broader population.
• Side effects of the drugs are minimized in the clinical trials in all sorts of ways, but those side effects crop up far more often when real people start taking the drugs.
• People in the trials are monitored to make sure they take the drugs regularly and as directed. In the real world people take the drugs improperly or stop taking them altogether. (Half of people who’ve been prescribed statins stop taking them.)
• There is incredible financial and reputational pressure to produce good results during clinical trials. Am I suggesting that drug companies and medical scientists fudge the numbers? You bet. More than half of all studies published in reputable scientific journals can’t be replicated. And the most promising heart treatment in the past thirty years—stem cells—turns out to have been a fraud.

The overall result is that, far too often, and especially in the field of heart disease, we pay too much for drugs that offer too little to individual patients and that, in the real world, don’t even save money.

Earlier in this series of posts I noted that an American male in his sixties who is diagnosed with congestive heart failure can expect to live, on average, four and a half years—a number that has barely budged in half a century. For me that anniversary is coming up in a few weeks—bizarrely, on Christmas Eve.

Last month I visited the one of my docs who knows I haven’t been taking my meds. He certainly would have fired me as a patient except that he’s retiring at the end of the year anyway, so what the hell.

At the appointment we reminisced about our long professional relationship, stretching back more than two decades. We talked about his plans for retirement. Eventually, of course, the conversation turned to my unfortunate anniversary, coming up over the holidays.

My doc seized the occasion to plead with me to go back on my meds—all fourteen of them. “Maybe it’s not too late!” he said. He is utterly convinced that, one day soon I will be calling him from my hospice bed to tell him how right he was.

Maybe so. But as for me, I have other plans. First, to continue to feel as good as I can for as long as I can. Second, to bump my heart’s ejection fraction back up as close to 65% as I can get it. Third, to outlive my docs.