in the nicotine replacement therapy literature: as- itself but rather of their belief that the 2. Eisenberg MJ, Filion KB, Yavin D, et al. Pharma-
sessment of the double-blind in clinical trials. Ad- cotherapies for smoking cessation: a meta-analysis
dict Behav 2004;29:673-84. placebo is not efficacious. of randomized controlled trials. CMAJ 2008;179:
3. Helliker K. Nicotine fix. Behind antismoking pol- Blinding plays an important role in 135-44.
icy, influence of drug company. Government guide- 3. Hróbjartsson A, Forfang E, Haahr MT, et al.
lines don’t push cold turkey; advisers’ company ensuring the validity of clinical trial re- Blinded trials taken to the test: an analysis of ran-
ties. Wall Street Journal 2007 Feb 8.Sect. A:1. sults. However, improved methods are domized clinical trials that report tests for the suc-
4. Dar R, Stronguin F, Etter JF. Assigned versus per- cess of blinding. Int J Epidemiol 2007;36:654-63.
ceived placebo effects in nicotine replacement needed to assess the success of blind- 4. Fergusson D, Glass KC, Waring D, et al. Turning a
therapy for smoking reduction in Swiss smokers. J ing. Until such methods are developed, blind eye: the success of blinding reported in a ran-
Consult Clin Psychol 2005;73:350-3. dom sample of randomised, placebo controlled tri-
5. US Centers for Disease Control and Prevention.
incomplete blinding is a potential limi- als. BMJ 2004;328:432.
Tobacco use among adults: United States, 2005. tation to most clinical trials. 5. Haynes RB, Sackett DL, Guyatt GH, et al. Clinical
MMWR Morb Mortal Wkly Rep 2006;55:1145-8. epidemiology: how to do clinical practice
research. 3rd ed. Philadelphia (PA): Lippincott
DOI:10.1503/cmaj.1080096 Kristian B. Filion MSc Williams and Wilkins; 2005.
Department of Epidemiology, Biostatistics 6. Sackett DL. Measuring the success of blinding in
and Occupational Health, McGill RCTs: Don’t, must, can’t or needn’t? Int J Epi-
[Six of the authors respond:] University demiol 2007;36:664-5.
Patrick Bélisle MSc DOI:10.1503/cmaj.1080105
We thank John Polito for his discus- Division of Clinical Epidemiology, McGill
sion of the importance of blinding in University Health Centre
clinical trials. Blinding is undoubtedly Lawrence Joseph PhD Correction
a key component of the validity of in- Gilles Paradis MD MSc
Department of Epidemiology, Biostatistics
ferences drawn from randomized con- and Occupational Health, McGill
In a recent research article on stable
trolled trials. For this reason, it is in- University, Montréal, Que. angina,1 a sentence in the “symptoms
cluded in tools to assess the quality of Stéphane Rinfret MD MSc and prognosis” section of Results that
clinical trials, such as the Jadad scale.1 Division of Cardiology, Laval Hospital, read “Among South Asian people with
In our meta-analysis of randomized Québec City, Que.