Characteristics of self-identified case–control studies (CCs) in 6 surgical journals, 1995–2000, and factors associated with their being verified as such
| Variable | Self-ID’d CCs, no. (& %) | Verified as true epidemiologic CCs* | |||
|---|---|---|---|---|---|
| No. (& %) | OR | 95% CI | p value | ||
| Journal | |||||
| American Journal of Surgery | 4 (7) | 1 (25) | 0.3 | 0–3.8 | |
| Annals of Surgery | 4 (7) | 2 (50) | 0.8 | 0.1–8.8 | |
| Archives of Surgery | 30 (55) | 9 (30) | 0.3 | 0.1–1.7 | |
| British Journal of Surgery† | 7 (13) | 4 (57) | 1.0 | — | |
| Journal of the ACS | 6 (11) | 2 (33) | 0.4 | 0–3.6 | |
| Surgery | 4 (7) | 1 (25) | 0.3 | 0–3.8 | |
| Total case–control studies | 55 (100) | 19 (35) | |||
| Number of authors | 0.88 | ||||
| 2 or 3† | 13 (24) | 4 (31) | 1.0 | — | |
| 4 or 5 | 18 (33) | 7 (39) | 1.4 | 0.3–6.5 | |
| 6 or more | 24 (44) | 8 (33) | 1.1 | 0.3–4.8 | |
| Focus on a surgical procedure | 28 (51) | 5 (18) | 0.2 | 0.1–0.7 | 0.011 |
| Not on a surgical procedure† | 27 (49) | 14 (52) | 1.0 | — | |
| Logistic regression analysis used | 14 (25) | 8 (57) | 3.6 | 0.1–12.9 | 0.045 |
| Not used† | 41 (75) | 11 (27) | 1.0 | — | |
| Odds ratios reported | 11 (20) | 9 (82) | 15.3 | 2.8–82.6 | 0.002 |
| Not reported† | 44 (80) | 10 (23) | 1.0 | — | |
↵* Percentages are of self-identified case–control studies; the odds ratio and 95% confidence interval (CI) were estimated for each variable by logistic regression; p values are for the Wald test of the null hypothesis that none of the categories of the variable was associated with being a true case–control study.
↵† The referent category for the regression model.
ACS = American College of Surgeons; CCs = case–control studies; ID’d = identified; OR = odds ratio.
Percentages may not add to 100 because of rounding.