Skip to main content

Table 2 Analysis configurations arising from three decision categories

From: Changes to aspects of ongoing randomised controlled trials with fixed designs

Type of decisionDecision options
Choice of endpoint1. Coronary heart disease (CHD) deatha
2. Revascularisation
3. All-cause mortality
Analysis set composition1. No exclusions (i.e. all randomised patients)a
2. Exclude patients with eGFR < 45 (Stage 3a kidney disease or worse) (excludes 4%)
3. Exclude patients with qualifying events within 9 months of randomisation (excludes 29%)
Covariate specification1. Nonea
2. Stroke (2 levels: yes, no)
3. Smoking status (2 levels: never smoked, current/ex-smoker)
4. Index ACS (3 levels: unstable angina pectoris (UAP), single myocardial infarction (MI), multiple MI)
5. Stroke and smoke
6. Stroke and index ACS
7. Smoke and index ACS
8. Stroke, smoke and index ACS
  1. ACS acute coronary syndrome, eGFR estimated glomerular filtration rate
  2. aBase-case analysis: CHD death, no exclusions, no baseline covariate adjustment