Inter-rater reliability for Jadad Scale and Newcastle–Ottawa Scale (NOS): by question
| Question—Jadad Scale | κ (95% CI) | Question—NOS cohort | κ (95% CI) | Question—NOS case–control | κ (95% CI) |
| Randomisation | 0.50 (−1.00 to 1.00) | Representativeness of exposed cohort | −0.13 (−0.36 to 0.11) | Case definition adequate | 1.00 (1.00 to 1.00) |
| Appropriate randomisation | 0.56 (0.29 to 0.83) | Selection of non-exposed cohort | −0.14 (−0.28 to 0.00) | Cases representative | −0.20 (−0.49 to 0.09) |
| Double blind | 0.41 (0.16 to 0.66) | Exposure ascertainment | 0.00 (0.00 to 0.00) | Control selection | 0.25 (−0.19 to 0.69) |
| Appropriate double blind | 0.17 (−0.07 to 0.41) | Outcome not present at baseline | 0.20 (−0.33 to 0.73) | Control definition | 0.14 (−0.54 to 0.82) |
| Description of withdrawals | 0.21 (−0.02 to 0.45) | Comparability of cohorts | 0.12 (−0.23 to 0.47) | Case and control comparability | 0.00 (0.00 to 0.00) |
| Description of inclusion/exclusion criteria | 0.27 (−0.03 to 0.57) | Outcome assessment | 0.31 (−0.08 to 0.69) | Exposure ascertainment | −0.11 (−0.68 to 0.46) |
| Description of adverse effects | 0.13 (−0.11 to 0.37) | Follow-up long enough | −0.09 (−0.22 to 0.04) | Same ascertainment method for cases and controls | 0.60 (−0.07 to 1.00) |
| Description of statistical analysis | 0.49 (0.21 to 0.77) | Follow-up adequate | 0.39 (−0.02 to 0.81) | Non-response rate | −0.11 (−0.65 to 0.43) |
κ, Kappa.