Thursday, June 6, 2013

Medicines for Smoking Cessation: Failure and Cover-Up

The 20-year-old Cochrane Collaboration is widely recognized for compiling and analyzing thousands of clinical trials, providing health professionals with useful information, in the form of authoritative analyses and reports, for clinical practice.  Regrettably, the Cochrane’s reputation was tarnished last week when it issued a report on smoking cessation medications (here). 

The report conveys the inaccurate impression that FDA-approved medicines are “effective” in helping smokers quit.  Uncritical media like Time magazine (here) quickly parroted the message:

“Based on the findings, all the drugs were successful at improving the odds for smokers who wanted to quit.  Participants were 80% more likely to quit when using a single NRT [nicotine replacement therapy] or taking bupropion compared to those using a placebo.”

While technically accurate, that statement entirely obscures the medicines’ actual impact and implies a ringing endorsement when none is merited.   

While Cochrane boasts that its research “covered 267 studies, involving 101,804 participants,” there is no information on how many participants quit by using the placebo.  This is a serious omission, well below the standards of the institution’s past reports.

In a summary written for laymen, the Cochrane notes: “NRT and bupropion helped about 80% more people to quit than placebo; this means that for every 10 people who quit with placebo about 18 could be expected to quit with NRT or with bupropion.  Varenicline more than doubled the chances of quitting compared with placebo, so that for every 10 who quit with placebo about 28 could be expected to quit with varenicline.”

These numbers are not meaningful, because they are abstract and without context.  Mark Eisenberg and colleagues from McGill University in 2008 published a meta-analysis of smoking medicines in the Canadian Medical Association Journal (abstract here) that provides the necessary context, by reporting numbers of successes and failures for placebo and for medicines. 

In the Eisenberg analysis the quit rate for all 11,000 participants given a placebo was just over 10%.  This means that out of a hundred smokers using placebo, about 10 successfully quit.  Quit rates were 17% using nicotine gum, 19% for the patch, 21% for bupropion and 26% for varenicline. 

Here’s the flip side:  90 out of 100 smokers failed to quit with placebo, 83% failed with nicotine gum, 81% with the patch, 79% with bupropion, and 74% failed with varenicline. 

And that’s not the end of the story.  Placebo and medicine quit rates in clinical trials are typically grossly elevated compared to the real world.  The reasons: (1) clinical trials carefully screen participants, enrolling only highly motivated smokers who are likely to complete the trial, and succeed; (2) trials provide intensive motivation and follow-up, increasing success rates; (3) there is well documented evidence that pharmaceutical company funding inflates quit rates for medicines (abstract here). 

In the real world, quit rates are half of clinical trials, around 5%, according to the National Institutes of Health (here), meaning that 95% of smokers fail.  It’s likely that real-world failure rates are 91% for nicotine gum or patch, 90% for bupropion and 87% for varenicline. 

Smoking cessation medicines are effective for only for a tiny minority of smokers.  Suppressing this fact deters smokers from learning about and adopting safer tobacco alternatives.  By further obscuring the truth, the Cochrane collaborates, intentionally or not, with tobacco prohibitionists who are undermining public health and condemning countless smokers to untimely and entirely avoidable death.  That is a real-world disaster.   


Luc Dussart said...

A major flaw of all these controlled studies is that blinding the presence of the active substance is not possible.
Smokers recognize quicly if it is present or not. So the results are usually meaningless.

Dale said...

It makes you wonder why they are trying to pass off these false reports as truth.

Vincent Jones said...

I think Luc makes a valid point, that the results are usually meaningless. Also I think that these people are obviously trying to hide something, but why?