Retreatment with nicotine replacement therapy: Predictors of re-enrolment in a cohort of people who smoke seeking treatment in ambulatory settings.

Abstract

People who re-enrol in smoking cessation treatment repeatedly may benefit from more intensive or alternative interventions. We aimed to measure differences across participants in propensity to re-enrol in a large treatment program, to describe re-enrolment patterns, and to examine specific participant characteristics that predict re-enrolment. We analyzed 172,050 enrolments into the Ontario-wide Smoking Treatment for Ontario Patients (STOP) program between 2014 and 2023. STOP provides up to 26 weeks of nicotine replacement therapy (NRT) and allows re-enrolment after one year. We used probabilistic deduplication to identify 128,481 unique individuals. To analyze re-enrolment, we fit a mixed-effects Cox proportional hazards survival model for recurrent events, with demographic, clinical, and contextual variables included as fixed effects. Of all enrolments, 25.3 % (43,569) were repeat entries. Participant-level variance was very high (σ2 =1.66, median hazard ratio = 3.4), indicating large residual individual-level differences in propensity to re-enrol. The strongest predictors of reenrolment were time to first cigarette after waking (p < 0.001), lifetime quit attempts (p < 0.001), type of NRT provided (p < 0.001) and total visits in previous enrolment (HR 1.10, 95 % CI 1.10-1.11), clinic type (p < 0.001), schizophrenia (HR 1.51, 95 % CI 1.41-1.60), and enrolment during the COVID-19 era (HR 1.31, 95 % CI 1.20-1.44). Propensity to re-enrol varied markedly across individuals, implying that there is a subgroup of treatment-seeking people who are less likely to achieve longterm abstinence. Alternative interventions or long-term nicotine replacement may be justified for these individuals.