. Author manuscript; available in PMC: 2009 Jan 23.

Published in final edited form as: Ann Appl Stat. 2008;2(4):1409–1431. doi: 10.1214/08-AOAS193

Table 4.

VAX003: Summary of the posterior distributions of the transmission probability and the vaccine efficacy per infectious needle-sharing act for the overall study population and by baseline risk level and HIV subtype, compared to the standard analysis

				p		VE (Bayesian)		VE (Cox^a)

Risk Level	Subtype	Total^b	Infected^c	Median	95% C.S.	Median	95% C.S.	Estimate	95% C.I.
Overall		13797	206	0.026	0.021, 0.031	-0.08	-0.43, 0.20	0.001	-0.31, 0.24
	E		160	0.028	0.022, 0.034	-0.12	-0.52, 0.17	-0.014	-0.38, 0.25
	B		32	0.019	0.012, 0.029	0.18	-0.57, 0.60
	E/B			1.45	0.91, 2.39
Low		6622	80	0.033	0.024, 0.045	0.06	-0.49, 0.41
	E		55	0.034	0.022, 0.048	0.04	-0.66, 0.42
	B		16	0.032	0.015, 0.058	0.18	-1.33, 0.67
	E/B			1.06	0.51, 2.54
High		7175	126	0.023	0.017, 0.029	-0.10	-0.60, 0.23
	E		105	0.025	0.019, 0.032	-0.21	-0.77, 0.19
	B		16	0.015	0.008, 0.026	0.34	-0.63, 0.77
	E/B			1.68	0.92, 3.31

Results based on Cox proportional hazards model in Pitisuttithum et al. (2006).

Total number of six-month intervals.

Intervals for 5 subjects infected by visit 0 (E:4, B:1) are excluded. The 14 untypeable infections are not shown.