How soon should alternative smoking cessation treatments be offered after incomplete response to nicotine replacement therapy?

Abstract

Several effective interventions exist for tobacco dependence. These operate via different mechanisms, making treatment sequencing an important clinical option. As total time in treatment is often short, determining the length of each trial is crucial. In this analysis, we examine how cessation probability varies with time in treatment in a large program providing nicotine replacement therapy (NRT). We used data on 87,952 people enrolled in the Smoking Treatment for Ontario Patients (STOP) program between 2014 and 2024. We fit a log-logistic interval-censored survival model to examine cessation probability as a function of time in treatment. We adjusted for sociodemographic and clinical variables and included willingness to set a quit date as an ancillary parameter. We used multiple imputation to include cases with missing baseline variables. The predicted probability of cessation was 15.5 % for the first week, 9.4 % for the second, and decreased to 3 % per week by 12 weeks. People who had set a quit date were markedly more likely to quit soon after initiation, but the general pattern of decreasing probability of full response was present for all participants. People receiving NRT are most likely to quit very early in their course of treatment, and become unlikely to do so after several weeks without full response. Given the limited length of most courses of care, this implies that an offer of alternative interventions, if available, might best be made after a relatively short trial of NRT.