A graph showing the data and a forest plot for a meta-analysis of randomised trials comparing endarterectomy versus stenting for carotid artery stenosis to prevent death or stroke.
There are 10 trials. The data are the names of the trials, number in the endarterectomy group, number of events in that group, number in the artery stenting group, number of events in that group. For the 10 trials these are: Naylor et al 1998, 12, 0, 11, 5, Wallstent 2001, 112, 5, 107, 13, CAVATAS 2001, 253, 25, 251, 25, Brooks et al 2001, 51, 1, 53, 0, Brooks et al 2004, 42, 0, 43, SAPPHIRE 2004, 167, 8, 167, 7 EVA-3S 2006, 262, 10, 265, 25 SPACE 2006, 584, 38, 599, 46, BACASS 2007 10, 1, 10, 0, ICSS 2009, 857, 43, 853, 72.
The total numbers are given as 2350, 131, 2359, 193.
Odds ratios with 95% confidence intervals in parentheses are given for the ten trials as follows: 0.05 (0.00 to 0.99), 0.34 (10.12 to 0.98), 0.99 (0.55 to 1.78), 3.18 (01.3 to 78.83), blank row, 1.15 (0.41 to 3.25), 0.38 (0.18 to 0.81), 0.84 (0.54 to 1.31), 3.32 (0.12 to 91.60), 0.57 (0.39 to 0.85).
The forest plot has scale 0.2 to 5.0, so the first, second, fourth, seventh, and ninth studies have confidence intervals with arrows indicating that they go off the scale. There is no symbol or line for the fifth study, Brooks et al 2004.
Back to meta-analysis exercise.
This page maintained by Martin Bland.
Last updated: 18 February, 2009.