Stata statistical software release v 17

Participant characteristics will be summarised using mean and SD or median and IQR for continuous variables, and number and percentages for the categorical variables. The proportion of participants recruited and lost to follow-up at the end point will be estimated with 95% CIs. As part of the secondary analyses, logistic regression models will be used to estimate the intervention effects with 95% CIs for the prespecified outcomes, tobacco abstinence and self-reported motivation and intention to quit after adjusting for the stratification factor (UPHC) and baseline values of the outcome where it is available. These analyses will be done on an intention to treat (participants as randomised with available outcome data) basis, and multiple imputation will be used to handle missing outcome values if considered appropriate. The extent of missing data for each variable and the percentage of participants adhering to the intervention will be reported. Attrition levels by randomised group and the characteristics of participants who are lost to follow-up will also be reported. All analyses will be done using STATA software V.17 (StataCorp. 2019. Stata Statistical Software: Release V.17. College Station, Texas: StataCorp LLC) or updated versions. A detailed statistical analysis plan will be drawn up nearer to the analysis stage, prior to the database lock.

For each disease or group of diseases (ARD-T), we calculated yearly (2000–2017) and overall crude and age-standardized rates (ASRs, per million person-years) for the 29 municipalities (H-ASB) and the reference group. The standard population was the Brazilian 2010 population ≥ 30 years of age in 10-year age categories.
Standardized rate ratios (SRR) and 95% confidence intervals (95% CI) of H-ASB against the rest of the country as well as of H-ASB municipalities that started consumption before 1970 against those that started from 1970 onwards were calculated. Statistical analysis was performed with Stata Statistical Software: Release v17 (StataCorp. LLC, College Station, TX, USA, 2021) using the command “distrate” [26 (link)].