Study Author Year | Country | Newcastle-Ottawa Scale items addressing different methodological components of the study and potential sources of bias | |||||||
Representativeness of the exposed cohort | Selection of the non-exposed cohort | Ascertainment of exposure | Outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design/analysis (confounding) | Assessment of outcome blind/record linkage | Follow-up long enough for outcome to occur | Adequacy of follow-up of cohorts (attrition) | ||
Pruchno et al 199050 | USA | ✓ | ✓ | ✓ | ✓ | ✗* | ✗ | ✓ | ✗ |
Vitaliano et al 199344 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Wright 199442 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | ✓ | ✓ |
Markiewicz et al 199749 | Canada | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | ✓ | ✓ |
Bannister et al 199845 | UK | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Wells and Over 199847 | Australia | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✗ |
Wright et al 199843 | USA | ✗ | ✗ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Caron et al 199936 | USA | ✗ | ✗ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Fisher and Lieberman 199939 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Torossian and Ruffins 199948 | USA | ✗ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✗ |
Spruytte et al 200138 | Belgium | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✗ | ✓ |
Burgener and Twigg 200220 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
de Vugt et al 200446 | Netherlands | ✓ | ✓ | ✓ | ✗ | ✗ | ✗ | ✓ | ✓ |
McClendon et al 200451 | USA | ✗ | ✓ | ✓ | ✓ | ✗ | ✓ | ✓ | ✗ |
Stevens et al 200440 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Perren et al 200741 | Switzerland | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✓ |
Kunik et al 2010 (four papers)32–35 | USA | ✓ | ✓ | ✓ | ✓ | ✗* | ✗ | ✓ | ✓ |
Clare et al 201437 | UK | ✗ | ✓ | ✓ | ✓ | ✗ | ✗ | ✓ | ✗ |
Shroff 201552 | USA | ✓ | ✓ | ✓ | ✓ | ✗ | ✗ | ✗ | ✗ |
Snyder 201653 | USA | ✓ | ✓ | ✓ | ✓ | ✗* | ✓ | ✓ | ✗ |
✓=study dealt with this adequately. ✗=study was at risk of bias in this area, or provided no information to demonstrate otherwise.
*These studies adjusted for three out of the four key confounders.