Nicotine replacement therapy: evidence from observational studies versus clinical trials

To the Editor: George,1 who has received a research grant from Pfizer on smoking cessation, makes the manifestly incorrect statement that “there is no evidence for the effectiveness of cold turkey cessation, especially in moderate to heavy smokers”. In fact, it has always been the case that the most common method used by ex-smokers in their final successful attempt to quit is cold turkey.2

Those promoting pharmacological approaches to cessation typically base their recommendations on results from clinical trials. But clinical trial results do not reflect “real-world” results; people using pharmaceutical aids tend to have lower smoking cessation rates than people who quit unaided, because of indication bias (in which smokers with a high nicotine dependency and the worst cessation prognosis self-select to use pharmaceutical aids).3

A meta-analysis of nicotine replacement therapy (NRT) trials concluded that “... the true overall impact of NRT ... is similarly modest and represents success for only about 7% of all those treated in these trials”.4

Smoking cessation trials exclude large numbers of smokers, have participant “blindness” integrity problems (those on placebo often accurately guess that they have not been allocated to the treatment arm), and those in the treatment arm, by virtue of knowing that they are being studied, have higher rates of treatment completion than smokers in the community because of the Hawthorne effect.

Before the introduction of NRT over 20 years ago, the American Cancer Society estimated that “over 90% of the estimated 37 million people who have stopped smoking in this country since the Surgeon General’s first report linking smoking to cancer have done so unaided”.5 Recent reviews have concluded that the advent of smoking cessation pharmaco-therapies has not translated into increased rates of cessation in the population.6