From: Applying the ROBINS-I tool to natural experiments: an example from public health
Type of bias assessed | Cochrane risk of bias (RoB) domains | EPHPP tool domains | ROBINS-I domains | Comment |
---|---|---|---|---|
Selection | Sequence generation (Cochrane RoB) | Not applicable to NRS | ||
Allocation concealment (Cochrane RoB) | Study design (EPHPP) | Domain 1: confounding and Domain 2: selection | ||
Confounding | Baseline outcome characteristics similar (EPOC) | Control for confounding through analysis or design (EPHPP) | Domain 1: confounding | |
Baseline characteristics similar (EPOC) | ||||
Baseline response | Baseline response (EPHPP) | Selection (EPHPP) | ||
Attrition | Incomplete outcome data (Cochrane RoB) | Withdrawals at follow-up (EPHPP) | Domain 5: missing data | |
Contamination* | Contamination (EPOC) | |||
Reporting* | Selective reporting (Cochrane RoB) | Domain 7: selection of reported result | ||
Performance * | Blinding—participants (Cochrane RoB) | Blinding—participants and assessors (combined) (EPHPP) | Domain 6: measurement of outcome | Rarely applicable to housing improvement studies—no studies blinded participants |
Detection* | Blinding—assessors (Cochrane RoB) | |||
Blinding—analysts (EPHPP) | ||||
Performance* | Intervention implementation: within study variation of exposure to intervention (review authors) | Heterogeneity of exposure to intervention and potential to benefit from intervention (Review authors) | Domain 3: measurement of intervention | This measure used in the Cochrane RoB and EHPP was developed by the authors |
Domain 4: deviation from intended intervention | ||||
Outcome measure* | Data collection (EPHPP) | Designed to indicate appropriate data collection tools and outcomes |