Skip to main content

Table 1 Summary of reported data scenarios from McCormick Matthews [24]

From: A non-linear optimisation method to extract summary statistics from Kaplan-Meier survival plots using the published P value

Scenario Reported Data Number Required Method
1 i) P value/chi-square statistic for ln (HR) or HR
ii) Total number of events
6 (17%) Equations 8 and 10, Parmar [8]
2 i) IPD 0 (0%) Available statistical methods
3 i) KM survival plot
ii) Number at risk included at regular intervals
4 (11%) Estimation from the KM plots, Guyot [16]
4a i) KM survival plot
ii) No total number of events
iii) No number at risk included
iv) P-value/chi-square statistic for ln (HR) or HR (exact)
19 (53%) Estimations from the KM plots, Parmar [8]
4b v) P-value/chi-square statistic for ln (HR) or HR (non-exact) 4 (11%)
5 i) KM survival plot
ii) No total number of events
iii) No number at risk included
iv) No p value
3 (8%) Estimations from the KM plots, Parmar [8]
  Total 36 (100%)  
  1. The frequency of reported data scenarios from a meta-analysis published by McCormick Matthews et al. [24] in which studies did not explicitly state both the ln (HR) and var. ln (HR). The required equations to extract ln (HR) and var. ln (HR) given that set of data is also given