Table 4.

Prognostic effect estimate for a 1‐unit change in microvessel density, for each method of measurement in each study, where the outcome is mortality.

	Chalkley method		All vessels method
Study ID	Log hazard ratio	SE of log hazard ratio	Log hazard ratio	SE of log hazard ratio	Within‐study covariance*	Patient‐level correlation
1	0.122	0.087
3	0.039	0.06
4			−0.02	0.09
5			0.058	0.063
6	0.104	0.065	0.02	0.091	0.0033	0.55
7	0.039	0.038
8			0.239	0.039
9			−0.211	0.221
10			0.03	0.061
11			−0.01	0.02
12			0.307	0.252
13	0.02	0.066	0	0.025	0.0012	0.74
14			−0.693	0.758
15			−0.174	0.142
16			−0.02	0.025
17	0.03	0.047	0.049	0.037	0.00048	0.27

^*Assuming within‐study correlation is equal to the patient‐level correlation.

NB data derived from the reported hazard ratios and CIs in Figures 1 and 2 of Trivella et al. 44 SE, standard error.