Table 1

Checklist for eligible study designs (based on Reeves et al 22)

FeatureEligible for inclusion if…
1. Was the intervention/comparator:
 Allocated to (provided for/administered to/chosen by) individuals?Yes
 Allocated to (provided for/administered to/chosen by) clusters of individuals?Yes
 Clustered in the way it was provided (by practitioner or organisational unit)?Yes
2. Were outcome data available:
 After intervention/comparator only (same individuals)?No
 After intervention/comparator only (not all same individuals)?No
 Before (once) AND after intervention/comparator (same individuals)?Yes
 Before (once) AND after intervention/comparator (not all same individuals)?Yes
 Multiple times before AND multiple times after intervention/comparator (same individuals)?Yes
 Multiple times before AND multiple times after intervention/comparator (not all same individuals)?Yes
3. Was the intervention effect estimated by:
 Change over time (same individuals at different time points)?Yes
 Change over time (not all same individuals at different time points)?Yes
 Difference between groups (of individuals or clusters receiving either intervention or comparator)?Yes
4. Did the researchers aim to control for confounding (design or analysis)
 Using methods that control in principle for any confounding?Yes
 Using methods that control in principle for time-invariant unobserved confounding?Yes
 Using methods that control only for confounding by observed covariates?No
5. Were groups of individuals or clusters formed by
 Randomisation?Yes
 Quasi-randomisation?Yes
 Explicit rule for allocation based on a threshold for a variable measured on a continuous or ordinal scale or boundary (in conjunction with identifying the variable dimension, below)?Yes
 Some other action of researchers?Yes
 Time differences?Yes
 Location differences?Yes
 Healthcare decision makers/practitioners?Yes
 Participants' preferences?Yes
 Policymaker?Yes
 On the basis of outcome?No
 Some other process? (specify)-
6. Were the following features of the study carried out after the study was designed
 Characterisation of individuals/clusters before intervention?Yes
 Actions/choices leading to an individual/cluster becoming a member of a group?Yes
 Assessment of outcomes?Yes
7. Were the following variables measured before intervention:
 Potential confounders?Yes
 Outcome variable(s)?Yes