Table 7.

Logistic regression results showing impact of free drugs on treatment outcomes using matched dataset; N = 23,242

	Model A	Model B	Model C	Model D	Model E	Model F
Free drugs (Odds ratio)	1.482***	1.4945***	1.5028***	1.5364***	1.4381***	1.4519***
95% C.I.	(1.326, 1.657)	(1.336 1.672)	(1.342, 1.682)	(1.370, 1.722)	(1.278, 1.618)	(1.288, 1.637)
All Covariates		Yes	Yes	Yes		Yes
Diagnosing Quarter FE			Yes	Yes	Yes	Yes
District FE				Yes
Treatment Coordinator FE					Yes	Yes
Observations	23,242	23,242	23,242	23,242	23,242	23,242
Log Likelihood	-5,152.60	-4,978.19	-4,976.30	-4,855.65	-4,861.64	-4,690.78
Akaike Inf. Crit.	10,309.20	9,972.39	9,976.60	9,747.30	10,021.30	9,691.57

Note: (a) 95% C.I. based on robust standard errors; (b) All models were fit on the matched dataset; b) *p < 0.1; **p < 0.05; ***p < 0.01