Skip to main content

Table 3 Sensitivity analysis of handling missing data on the primary outcome

From: Update of statistical analysis plan for: Integration of smoking cessation into standard treatment for patients receiving opioid agonist therapy who are smoking tobacco: protocol for a randomised controlled trial (ATLAS4LAR)

Outcome

Handling of missing data at end of trial

Events n (%)

Absolute difference between arms (n, 95%CI)

Logistic regressiona odds ratio (95% CI)

Intervention

Control

No. of smokersb at 16 weeks, ITTc

Equal to baselined

    

Person excludede

    

No. of smokersb at 16 weeks, PPf

Equal to baselined

    

Person excludede

    

At least 50% reduction numberg of cigarettes at 16 weeks, ITT

Equal to baselined

    

Person excludede

    

At least 50% reduction numberg of cigarettes at 16 weeks, PP

Equal to baselined

    

Person excludede

    
  1. aUnadjusted analysis unless Table 1 indicates substantial differences between arms at baseline
  2. bA person smoking at least one cigarette per day or seven cigarettes per week
  3. cITT, intention to treat population: participants assessed according to randomisation regardless of adherence to trial
  4. dIf data on primary outcome is missing at 16 weeks, the results are set equal to baseline
  5. eIf data on primary outcome is missing at 16 weeks the person is excluded from the analysis (complete case)
  6. f PP, per protocol population: all participants who completed at least 50% of the trial visits
  7. gThe average daily number of cigarettes smoked, as reported by the participant