Impact of strategic use of antiretroviral therapy intervention to the HIV continuum of care in 13 cities in Indonesia: an interrupted time series analysis

Table 3 Level change^a, pre and post SUFA rates, and the slope change^b for HIV cases at the point of SUFA implementation

		Model 1		Time interaction^d	Model 2		Time interaction^d
		IRR (95% CI)	p-value^c	Time interaction^d	IRR (95% CI)	p-value^c	Time interaction^d
Level change	Predicted cases (n)
Pre-SUFA	58.7 (Dec 2013)	1.00			1.00
Post-SUFA	60.5 (Jan 2014)	1.04 (0.95, 1.13)	0.445		0.77 (0.69, 0.86)	< 0.001
Slope	Predicted cases/month (n)
Pre-SUFA (per month)	47.1 (Jan 2011-Dec 2013)	1.01 (1.01, 1.02)	< 0.001		0.97 (0.97, 0.98)	< 0.001
Post-SUFA (per month)	56.3 (Jan 2014-Dec 2016)	1.00 (1.00, 1.00)	< 0.001		0.99 (0.98, 0.99)	< 0.001
Slope change
Time × SUFA		0.98 (0.98–0.99)		< 0.001	1.01 (1.01–1.02)		< 0.001

IRR Incidence rate ratio
^aLevel change assesses the relative change in tests per month immediately post SUFA intervention
. ^bSlope change tests the relative change in the monthly trend between pre and post SUFA. Model 1 includes fixed effects for SUFA, time and SUFA x time interaction and a random effect for site. Model 2 = Model 1 + additional adjustment for population at risk (the HIV tests performed for each district site)
^cEstimated using a mixed effects negative binomial regression model
. ^dP-value for the SUFA x Time interaction

ISSN: 1742-6405