Effects of a performance and quality improvement intervention on the work environment in HIV-related care: a quasi-experimental evaluation in Zambia

Table 5 Percent of providers who rated quality as good in several HIV/AIDS service areas ^a

HIV/AIDS service area	Intervention group			Comparison group			Adjusted model – interaction P value^c
HIV/AIDS service area	Baseline, n = 27	Endline, n = 29	P value^b	Baseline, n = 16	Endline, n = 29	P value^b	Adjusted model – interaction P value^c
ART treatment readiness	80.0	100.0	0.01	60.0	91.3	0.02	–^d
ART initiation	72.0	92.6	0.05	61.5	96.2	0.004	0.463
ART follow-up	68.0	96.6	0.005	73.3	76.9	0.80	0.121
PMTCT	80.0	100.0	0.01	57.1	81.5	0.10	–^d
Laboratory	48.0	79.3	0.02	7.1	50.0	0.007	0.402
Infection prevention	92.0	89.7	0.77	73.3	82.1	0.50	0.169
Medical recording keeping	84.0	93.1	0.29	73.3	85.2	0.35	0.921

^aResponse scale was 1 to 5, with 5 meaning “strongly agree” and 1 meaning “strongly disagree”. This table reflects the percent of respondents who “agreed” and “strongly agreed”.
^b P value from t-test at baseline or at endline.
^cInteraction term P value from multivariate logistic regression models of each result on intervention status, time point, and interaction of these two variables, while controlling for provider cadre and ZDF branch and accounting for clustering of responses within each facility.
^dThe dash means that the Odds Ratio coefficient for the interaction term = 1, thus no P value is associated with it.
ART, Antiretroviral therapy; PMTCT, Prevention of mother-to-child transmission of HIV.

ISSN: 1478-4491