From: Reporting methods in studies developing prognostic models in cancer: a review
 | % (n = 47) articles |
---|---|
Study design | Â |
   RCT* | 11 (5) |
   Prospective cohort study | 21 (10) |
   Database or other retrospective cohort | 68 (32) |
Reason for sample size | Â |
   No reason given | 77 (36) |
   Justified time interval (clinical or technology) | 4 (2) |
   Trial size (RCT or cohort) | 19 (9) |
Power calculation reported | 0 (0) |
 | Median (IQR) [range] |
Eligible for study: | Â |
   Number of patients (n = 35) | 403 (148 to 833) |
   Number of events (n = 20) | 112 (62 to 289) |
Included in analysis: | Â |
   Number of patients (n = 45) | 342 (130 to 684) |
   Number of events (n = 33) | 110 (61 to 230) |
Number of events per candidate variable (n = 28) | 10 (5 to 33) [2 to 95] |