Wednesday, March 23, 2016

Clinical Trials and Errors Blur Consumer Tobacco Preferences

Doctors, myself included (here, here and here), value clinical trials, for good reason: They are powerful tools using sophisticated scientific methods to precisely determine the effectiveness of treatments for specific diseases.  But clinical trials are ill-suited to the measurement of consumers' preferences for products in the marketplace.  The appropriate tool for that is post-market surveillance.

The latter mechanism can be used, for example, to analyze whether coffee drinkers can be induced to switch to Coca Cola.  Looking at the American South, for example, observational research will show that some consumers prefer the cold caffeinated, carbonated beverage over coffee, giving rise to the jest, “Coke, it’s not just for breakfast anymore.”  In this case, one survey is more informative and accurate than a dozen clinical trials.

In reality, and contrary to the notions of many clinicians, there are distinct parallels in the consumption of smoke-free tobacco and coffee.  Both products are addictive but largely harmless. Most importantly, use of either product is not itself a disease.  The vast majority of smokers aren’t “sick” and don’t want to be “treated,” yet most smoking cessation efforts are designed around the clinical trial model where smoking is the “illness,” pharmaceutical nicotine, e-cigarettes or snus is the “treatment,” and nicotine/tobacco abstinence is the “cure.”  Given this, it is not surprising that most smokers don’t want to enroll in quit-smoking trials – leading researchers to select only smokers who are highly motivated to quit, then paid and pampered to remain in the trial.  Even so, most trials end with abysmally low compliance, completion and success rates.

The futility of the clinical trial model is illustrated by a new UK study appearing in Annals of Internal Medicine (abstract here).  Twenty-three nurses enrolled nearly 700 highly motivated smokers; the treatment was quitting abruptly or gradually two weeks later, accompanied by intense behavioral support and medicinal nicotine; the desired outcome (i.e., the “cure”) was abstinence.

In public comments, the lead author touted abrupt quitting as the best strategy, effectively celebrating a 73% failure rate at six months compared to an 82% failure rate among gradual quitters.  One-year rates, which are the gold standard for cessation trials, were not reported; they were likely higher.

Why does the medical research establishment demand clinical trial proof for quit-smoking aids, including e-cigarettes and smokeless tobacco products?  Perhaps because the clinical trial model enables the tobacco prohibitionist establishment to maintain the “illness-treatment-cure” perception of smoking and heathcare.  In fact, as I note in my book, For Smokers Only, tobacco harm reduction puts smokers in control of their health choices and consequences.

Clinical trials are the wrong way to evaluate cigarette substitutes as smoking cessation aids, and they should not be the standard by which these products are judged.  Instead, the FDA, along with other health authorities, should endorse e-cigarettes and smokeless tobacco products as safer cigarette substitutes.  The success or failure of smoke-free products in deterring smoking should be analyzed through post-market surveillance – that is, simply observing consumer activity.


Carl V Phillips said...

I totally agree with your thesis, that clinical trials are a misguided way to assess consumer preferences or what works in the free-living world. As you know, I have written extensively on that very point. But I am not sure this one is really the worst of them.

If interpreted properly modestly (which, of course, public health authors and the press never do -- but we can for them), it compares two options for cessation that are genuine rivals and for which people's knowledge of their own tastes might not be the best guide. This makes it far better than the utter junk clinical trials of, say, substitution of a low-risk product versus some standard abstinence method. Of course, the present case also has the problems of heterogeneity and the artificiality of the situation that such trials do. But...

The heterogeneity might not be so obvious to the individual. That is, someone who might like a particular substitute probably knows it or figures it out pretty fast. Someone who is just ready to quit to abstinence knows it. (We have jointly written about that, of course.) But people may or may not really know whether they are one of those who are better off with cold turkey versus tapering. That is not common intuition. So, yes, some people who failed with cold turkey might have succeeded with tapering even though cold turkey did better on average, and they would be getting bad advice from this. But if they are not identifiable, then playing the odds is not a terrible idea.

As for the artificiality, it is obvious that being handed an alternative product and ordered/encouraged to use it as a substitute for a month is radically different from real-world situations and thus might not have much bearing on them. But even though the specific numbers do not translate, it is not absurd to suggest that which of these rival tactics works better in the trial setting also works better in the real world.

So, yes, this study was ridiculously over-interpreted. But it is actually more valid than most clinical trials of behavioral choices.

Joel Nitzkin said...

It is important to consider that FDA, CDC and others involved in the "official" tobacco control movement have blinders on. Most (but not all) are acting in good faith, totally unaware that their framing of smoking as a disease (as opposed to a behavior) blinds them to the possibility that any non-pharmaceutical product might offer personal or public health benefits.

In the disease frame, the only scientific studies that are meaningful are randomized controlled trials. Survey and other data mean nothing to them. Since the behavioral dynamics of e-cig use for smoking cessation, involving free choice, possibly frequent changes in device, nicotine strength and flavor and use of a product visible to all -- these dynamics cannot be blinded or randomized. Therefore, in terms of development of clinical guidelines -- these data mean nothing.

Randomized clinical trials, by their very nature, tend to be short term and tend to ignore behavioral considerations -- attributing all results to the pharmacology of the product. Thus, randomized trials comparing a given e-cig product to a placebo or nicotine replacement therapy (NRT) product tend to show about the same efficacy of the NRT product. These studies totally ignore the extremely high relapse rates in the months and years following the initial study. Here is where e-cigs really stand out -- since the ex-smoker, if feeling the need to relapse, is far more likely to relapse to e-cig, not cigarette use.

It is also important to note that, at least for American physicians, the organizational culture within which they practice their medicine makes it far easier to recommend "evidence-based" pharmaceuticals than e-cigs. Their judgment is hemmed in by clinical guidelines, insurance company oversight, and liability concerns.

As I see it, the only way in which e-cigs and related vapor products can get their proper place in the sun for smoking cessation would be for federal authorities to openly acknowledge smoking as a behavior, and for FDA to authorize some of their research funds for exploration of potential benefits of adding a tobacco harm reduction component to current tobacco control programming, utilizing low-risk non-pharmaceutical nicotine delivery products.

Even though they do not use the term "tobacco harm reduction," they practice it when they recommend long-term use of an NRT product in a harm reduction mode. As they see it, this is not tobacco harm reduction because once a product has been licensed as a smoking cessation pharmaceutical, it is no longer a tobacco-related product.

As I recently stated in a letter to the editor of NTR , science has nothing to do with current American tobacco control policy as it relates to e-cigs and tobacco harm reduction. It continues to be based on traditional thinking within the tobacco control community.

Joel L. Nitzkin, MD