Peoples Pharmacy Podcast

This weekend, I had the pleasure of being a guest on Peoples Pharmacy, a great public radio show that hails out of Durham NC.  We had a far ranging, rapidly shifting conversation about lots of things.  Check it out if you are interested.  The show even gave me a chance to learn a new skill.  I have a nasty little virus, and by the end of the show I had mastered what radio people call the cough button!

Dying for Facts: Conclusion

In the last few posts, I’ve told the story of a couple heated debates. One still ongoing: “Does capital punishment deter criminals?” Another ended: “Do antiarrythmia drugs save lives?” The latter debate is over because people on both sides of the debate got together to find out the answer!
If we really want to improve political discourse in this country, we need to do what the cardiologists did in the 1980’s-convince opponents to collaborate with each other to find the facts. The CAST study, after all, brought together people who had been aggressive adversaries and turned them into constructive collaborators.
Take school vouchers: Imagine if a group of anti-voucher Democrats got together with pro-voucher Republicans and funded a research program comparing their favorite approaches. Both groups want what is best for their children and grandchildren, after all, so they ought to be sufficiently motivated to spend money in ways that will determine what’s best for the kids.
Now let’s take their partisan partnership a step further, and imagine that these adversaries are not only teaming up to fund the research, but are even joining forces to design the research. Not that I expect the legislators, themselves, to design such studies. But I can envision their respective scientifically-trained staffs playing a role in the design of the research trial. Indeed, by involving opponents in the design of a research study, we’ll end up with a better study. For if the study was put together exclusively by opponents of vouchers, rather than a bipartisan group, the study would hardly be likely to end up as a fair test of their efficacy.
Suppose that a bipartisan research committee-one that also includes social scientists knowledgeable about education-decides to conduct an experiment: They pick, say, 80 school districts from around the country, and randomly decide that 40 will have a voucher program for the next three years.
After intense debate the committee comes to consensus about what outcome measures they will use to evaluate the voucher program-some combination of, say, standardized tests scores and graduations rates. They even agree on the statistical methods they will use to assess these outcomes. What, then, is the likely outcome of their experiment?
Did you think I’d have an opinion on whether vouchers would benefit students? Well I hate to disappoint, but that is not at all what I meant by “the outcome of the experiment.” What I meant was: What will happen once the experiment is complete? And that, perhaps, can best be answered by looking at what will not happen…
Let’s imagine for a moment that a pro-voucher interest group funded a study which demonstrated that vouchers improve students’ test scores. It is pretty easy to envision the anti-voucher crowd criticizing the study-complaining about the outcome measures, for example, or about the use of an inappropriate control group. If the anti-voucher crowd, by contrast, helped design the study, it would be much more difficult for them to oppose vouchers on the grounds that they harm educational outcomes. Moreover, important legislators would be on board with the facts, having played a role in obtaining the data. As I’ll demonstrate in later chapters, the progress of science depends as much on psychology as it does on the nuts and bolts of any given research study. Partisan partnerships can help overcome some of those psychological barriers that so persistently prevent us from basing our policies on the best information science has to offer us.
I know this must sound woefully idealistic. But middle-age hasn’t caused me to abandon my ideals. In fact, to the contrary-my career in science and medicine has taught me that social progress depends on people striving for ideals, and it has shown me, too, that the scientific method is one of the very best tools humankind has created to help pursue those ideals.
I recognize that science is rarely as clean as it may appeared to have been in the anti-arrhythmia trial. Often it is quite messy, in fact-plagued by some of the same problems that afflict our political system: Strong opinions and even stronger emotions; name-calling and lies; even the frequent inability of scientists to admit when they are wrong. Indeed, “scientific truth” is often a matter of opinion, frequently a function of political consensus rather than indisputable evidence. For years, after all, Newton’s laws were the “truth,” until later physicists, most notably Einstein, revealed the laws to have been only better approximations of the truth than those of Newton’s predecessors.
Nevertheless, even if science is often more ‘truthy’ than truth-determining, the scientific method remains the best method humans have created by which to inch themselves, bit-by-bit, toward the truth about how the world works and how it might be improved. After all, people used to think that epileptic seizures were evidence that a person was possessed by the devil; whereas now scientists can pinpoint the area of the brain responsible for seizures, reducing future seizures with medications or surgery. Similarly, people used to think that continents were immovable, and now scientists have discovered that the continents used to be joined as one super-continent and, indeed, are continuing to drift across the globe.
Science is awfully good at revealing certain kinds of truth about how the world works. I would like to find ways to bring more of these truths to bear upon our social policies.
What do you think?
View original post and comments at Scientocracy

Dying for Facts Part 4: Getting Evidence!

When I last posted to this site, I was recounting my experience as a medical resident in the late 80s, presumably saving my patients’ lives with wonder drugs.
A few weeks into my cardiology rotation, the senior cardiologist who had been mentoring me took leave of the hospital and was replaced by another experienced cardiologist. On his first day, this new cardiologist listened as I told him about one of my patients, who had benefited from these amazing new drugs. He wasn’t impressed: “We are treating patients,” he sniffed, “not ECGs. What evidence do you have that the drug is improving his life?”
I was stumped. Many of my arrhythmia patients did not even notice when they were experiencing these premature beats. “Well he may not be feeling any better because of the meds,” I recall saying “but at least he should live longer.” The cardiologist peered at me skeptically: “Are you sure?” he said. “What study has proven that these drugs save lives?”
I had no idea. Humbled, I trudged over to the library later that day to find the answer, all of that World Book training finally coming in handy. A glance at the medical literature quickly showed me that there were plenty of research studies showing that these drugs reduced arrhythmias, but, as my mentor anticipated, none showing that they saved lives.
No proof? Then why were these drugs on the market, I wondered? As it turns out, the FDA found itself under great pressure in the 1980’s to speed up the pace of drug approvals, particularly for patients with dangerous or incurable illnesses. In the face of such pressure, the FDA decided to approve the arrhythmia drugs, Encainide and Flecainide, based simply on their ability to reduce how often patients experience premature beats. The FDA, then, was settling for evidence that the drugs influenced an intermediate outcome-premature beats-without demanding proof that the drugs influenced the outcome that really mattered-whether patients live or die.
The FDA’s reliance on this intermediate outcome was not particularly controversial at the time. Patients die from arrhythmias, after all, and these drugs reduced the frequency of such arrhythmias; so it was logical to conclude that these drugs would reduce deaths from arrhythmia.
But my Mayo training had now shown me that not all doctors agreed. Indeed, cardiology meetings in those days were often the scene of heated debates about the merits of these drugs, sometimes devolving into shouting matches. Had the leaders of the American College of Cardiology been the Supreme Court of medical care, they might have decided 6 to 3 in favor of the drugs. But the minority would have no doubt written passionate dissenting opinions bemoaning the lack of evidence supporting these drugs.
Once again-people divided not so much by irresolvable moral conflicts as by a simple matter of fact: The drugs either saved lives or they didn’t, but experts could not decide what was true.
Is there any hope?
OK-In these last few posts, I’ve introduced you to two disputes, one from the 70’s over whether the death penalty saves lives by deterring future murders, and another from the 80’s over whether new heart drugs save lives by preventing fatal arrhythmias. Both disputes were characterized by passionate opinions. Supporters of the death penalty, and of the arrhythmia drugs, were convinced they were saving lives. Opponents, on the other hand, believed that the benefits in either case were unproven and, therefore, that it was wrong to put these interventions into practice.
Let me put it plainly: We do not need to let decades pass by without trying to figure out what is true; indeed, we cannot afford to do so. The Supreme Court remains divided over deterrence forty-plus years after the Gregg case. But such has not been the case for the second dispute I’ve introduced you to. Because cardiologists-so vehemently divided in the 1980’s-finally joined together to embark upon a fact-finding mission: Proponents of the arrhythmia drugs, even the companies who made the drugs, got together with leading critics of these same drugs, and conducted an experiment. They recruited patients with arrhythmias and decided, randomly, who would get the drugs and who would not, and then observed the patients closely over the following months to assess the results.
What they discovered shocked almost everyone involved in the study. Flecainide and Encainide not only failed to prevent people from dying of arrhythmias, they actually killed patients: Out of 743 people receiving placebo, 26 died during the first 10 months of the research trial, the majority of them dying of arrhythmias just as predicted. But out of 755 people receiving the arrhythmia drugs, 63 died, a 120% increase over the placebo group. A person with an arrhythmia who wanted to stay alive, then, was better off taking a sugar pill than one of these supposed miracle drugs!
If a few dozen deaths seems small to you in a trial involving almost 1500 patients, consider the huge number of people who were taking these drugs at that time. These were blockbuster drugs, prescribed for literally hundreds of thousands of patients. By one estimate, some 42,000 patients in the U.S. died from using Encainide and Flecainide. Just to put that in perspective, that’s about the same number of U.S. soldiers killed in the entire course of the Vietnam war.6
To wit, then: A controversy: do these drugs benefit patients?
A resolution: a study showing that they do not.
And a change in decision making: doctors stopped prescribing these drugs to these patients.
Cardiologists in the 1980’s didn’t simply settle into opinion camps and joust with each other over the merits of anti-arrhythmia drugs. They disagreed about a question of fact, and joined together to determine the truth. Why, then, can’t we join together in a similar manner to solve some of the socio-political disputes that have held us grid-locked for two or more generations?
Gun control advocates believe that stricter gun control laws will save lives, by reducing crimes of passion and accidental deaths, and by keeping guns out of the hands of criminals. Gun rights advocates, on the other hand, believe that stricter gun controls will increase death from guns, because criminals will be worried that potential victims will be armed. Can’t we figure this out?
Many pro-life advocates claim that abortions are bad for the mental and physical health of the women who have the abortions, causing them to not only experience depression, but also to suffer an increased risk of breast cancer. Many pro-choice advocates dispute these facts. Can’t science give us an answer?
Free market enthusiasts oppose minimum wage laws as being obtrusive policies that harm more people than they help, hurting employers and ultimately reducing employment. More intervention-minded politicians, on the other hand, claim that raising the minimum wage will put cash into people’s hands without bringing on such ill-effects.
Why argue, when we can run an experiment?
Stay tuned for Part 5, when I will throw out some ideas about what medical experiments can teach us about politics.
View original post and comments at Scientocracy

PeterUbel