The effect of participant nonresponse on HIV prevalence estimates in a population-based survey in two informal settlements in Nairobi city

Table 5 Observed, imputed and overall adjusted prevalence of HIV

		Females		Males		Total
Interview and testing status	N	Prevalence [95% CI]	N	Prevalence [95% CI]	Total	Prevalence [95% CI]
Interviewed & Tested-Observed prevalence (a)	1713	12.43 [10.87-14.00]	1008	09.82 [07.98-11.66]	2721	11.47 [10.27-12.67]
Interviewed only-imputed prevalence (b)	363	11.93 [10.89-13.03]	210	8.95 [7.77-10.26]	573	10.84 [10.04-11.67]
Refused both-imputed prevalence (c)	142	10.07 [8.55-11.75]	61	9.34 [7.15-11.94]	203	9.85 [8.59-11.23]
No contact made-imputed prevalence (d)	713	11.78 [11.04-12.55]	557	10.43 [9.64-11.26]	1270	11.19 [10.65-11.75]
Imputed prevalence all non-response -imputed prevalence-b,c & d (e)	1218	11.63 [9.86-13.46]	828	9.97 [7.97-12.07]	2046	10.96 [9.6-12.30]
Overall, corrected prevalence-adjusted (f)	2931	12.10 [10.93-13.29]	1836	9.89 [8.54-11.28]	4767	11.25 [10.34-12.14]
Ratio (e/a)		0.94		1.02		0.96
Ratio (f/a)		0.97		1.01		0.98

ISSN: 1478-7954