Novel microsimulation model of tobacco use behaviours and outcomes: calibration and validation in a US population

Abstract

Background and objectiveSimulation models can project effects of tobacco use and cessation and inform tobacco control policies. Most existing tobacco models do not explicitly include relapse, a key component of the natural history of tobacco use. Our objective was to develop, calibrate and validate a novel individual-level microsimulation model that would explicitly include smoking relapse and project cigarette smoking behaviours and associated mortality risks.MethodsWe developed the Simulation of Tobacco and Nicotine Outcomes and Policy (STOP) model, in which individuals transition monthly between tobacco use states (current/former/never) depending on rates of initiation, cessation and relapse. Simulated individuals face tobacco use-stratified mortality risks. For US women and men, we conducted cross-validation with a Cancer Intervention and Surveillance Modeling Network (CISNET) model. We then incorporated smoking relapse and calibrated cessation rates to reflect the difference between a transient quit attempt and sustained abstinence. We performed external validation with the National Health Interview Survey (NHIS) and the linked National Death Index. Comparisons were based on root-mean-square error (RMSE).ResultsIn cross-validation, STOP-generated projections of current/former/never smoking prevalence fit CISNET-projected data well (coefficient of variation (CV)-RMSE≤15%). After incorporating smoking relapse, multiplying the CISNET-reported cessation rates for women/men by 7.75/7.25, to reflect the ratio of quit attempts to sustained abstinence, resulted in the best approximation to CISNET-reported smoking prevalence (CV-RMSE 2%/3%). In external validation using these new multipliers, STOP-generated cumulative mortality curves for 20-year-old current smokers and never smokers each had CV-RMSE ≤1% compared with NHIS. In simulating those surveyed by NHIS in 1997, the STOP-projected prevalence of current/former/never smokers annually (1998–2009) was similar to that reported by NHIS (CV-RMSE 12%).ConclusionsThe STOP model, with relapse included, performed well when validated to US smoking prevalence and mortality. STOP provides a flexible framework for policy-relevant analysis of tobacco and nicotine product use.

Longitudinal Transitions in Initiation, Cessation, and Relapse of Smoking and E-Cigarette Use Among US Youth and Adults

“…In future work, simulation models can be populated with these estimates along with additional data on smoking and e-cig use-stratified mortality, noncommunicable disease incidence, quality of life, and health care costs to predict the outcomes and cost-effectiveness of proposed policy interventions targeting smoking and e-cig use. 23 Population Assessment of Tobacco and Health (PATH) Study Waves 1-4.5 respondents were classified into never, former, and current smoking/e-cig use. A single participant's classification could differ between the two products: e.g., former smoker/current e-cig user.…”

“…When future waves of PATH are released, there may be sufficient data to model time-variant smoking relapse using a semi-Markov multi-state model (which includes unobserved sub-states of former smokers representing high versus low relapse likelihood), rather than relying on external smoking relapse data. 22,23 Our estimates account for sex and age but not for race, socio-economic status, level of nicotine dependence, and exposure to product advertising, which may be associated with differences in cigarette smoking and e-cig use behaviors. Our inclusion of more states and transitions than previous models made the computational intensity of including such variables infeasible.…”

“…In future work, simulation models can be populated with these estimates along with additional data on smoking and e-cig use-stratified mortality, noncommunicable disease incidence, quality of life, and health care costs to predict the outcomes and cost-effectiveness of proposed policy interventions targeting smoking and e-cig use. 23…”

“…We utilized the Simulation of Tobacco and Nicotine Outcomes and Policy (STOP) microsimulation model, now updated to incorporate e-cig use, to assess how accurately a tobacco and nicotine policy model parametrized with our transition frequency estimates could predict prevalence of each of the nine smoking and e-cig use states over time (Supplemental Methods). 23 We populated STOP's initial age, sex, and cigarette smoking and e-cig use distributions with the corresponding estimates at PATH Wave 2 (PATH Wave 1 lacked the necessary questions to distinguish between established and experimental e-cig use among youth, Figure 1). We ran the STOP model for 24 months, extracted model-predicted smoking and e-cig use prevalence for the entire population at each month, and compared STOP projected prevalence at 12 months and at 24 months to empirical data from PATH Waves 3 and 4.…”

Development and validation of DeciBHAL-US: A novel microsimulation model of hearing loss across the lifespan in the United States

“…We used coefficient of variance root mean square error (CV-RMSE) to compare model projected outcomes with the published literature, and considered CV-RMSE≤15% to indicate adequate model fit. 52 , 53 We also compared model-projected results with published 95% confidence intervals whenever possible. For select model outcomes, we present low and high ranges based on input value computed 95% confidence intervals in Appendix 7.…”