Probabilistic Distributions of Time to Relapse Since Quitting Cigarette Smoking
Li, L. (2011, July). Probabilistic Distributions of Time to Relapse Since Quitting Cigarette Smoking. Presented at JSM 2011, Miami Beach, FL.
To estimate relapse rates among cigarette smokers who have ever quitted smoking we analyzed data in the 1998-1999 Tobacco Use Supplement to the Current Population Survey on quitting smoking among smokers who reported having smoked at least 100 cigarettes in their lifetime and smoked daily. The mixture Weibull survival functions of the form S(t) = (1-p) x exp(-mtk) + p, where p is the assumed proportion of permanent quitters, were fitted to the data on time to relapse. The maximum likelihood estimates of the parameters m, k, and p were obtained via the EM algorithm. The estimated mixture Weibull survival functions were close to the nonparametric Kaplan-Meier estimates. The hazard rate of relapsing is clearly not a constant. Most of the relapsing events occurred within the first 100 days since quitting smoking and the relapse rate was quite small after the first 200 days.