Skip to main content

Table 3 Reporting of follow-up time, plotting of survival curves and survival regression analyses

From: A systematic review of the quality of conduct and reporting of survival analyses of tuberculosis outcomes in Africa

  Number of articles n (%)
Sample size estimation reported
 Yes 9 (12)
 Not reported 67 (88)
Reporting of follow-up time
 Median time 28 (37)
 Person time 26 (34)
 Not reported 22 (29)
Survival curves
 Kaplan-Meier 51 (67)
 Nelson-Aalen 14 (18)
 Not reported 11 (15)
Comparison of survival curves
 Log-rank test 44 (58)
 Weighted log-rank test (Wilcoxon-Breslow-Gehan) 1 (1.3)
 Not reported 31 (41)
Survival Regression models (N = 74)
 Cox proportional hazard 67 (91)
 Competing risk analysis 3 (4.1)
 Parametric proportional hazard 2 (2.7)
 Parametric accelerated failure time 2 (2.7)
Reported regression models assumptions testeda
 Cox PH (N = 67)  
 Visual (graphical log-log plots) 11 (16)
 Schoenfeld residuals test 21 (31)
 Not reported 35 (52)
Competing risk analysis (N = 3)
 Schofield residuals test 1 (33)
 Not reported 2 (67)
 Parametric Methods (N = 4)  
 Information theory (LL, AIC, BIC) 4 (100)
  1. aThese are reported methods used to test the underlying assumptions in the statistical methods section and not the actual number of studies that reported the test results, PH-Proportional Hazard, LL-likelihood values, AIC-Akaike Information Criteria, BIC-Bayesian Information Criteria