Table 3.

Impact of the study and follow up on treated clinicians

	Dependent Variable: Percentage point increase in protocol adherence above the baseline
	Overall	primed items	non-primed items	Overall effect size
	(1)	(2)	(3)	(4)
Impact of initial interventions
Scrutiny	0.045 (0.020)**	0.056 (0.023)**	0.03 (0.014)**	0.045 (0.019)**
Post-Scrutiny	0.006 (0.022)	0.007 (0.027)	0.022 (0.016)	0.009 (0.022)
Impact of encouragement intervention
Short-term post enc.	0.051 (0.021)**	0.107 (0.025)***	0.055 (0.014)***	0.062 (0.020)***
Feedback received	0.002 (0.018)	0.013 (0.022)	0.001 (0.013)	0.011 (0.018)
Medium-term post enc.	0.114 (0.023)***	0.187 (0.027)***	0.084 (0.016)***	0.125 (0.023)***
Long-term post enc.	0.073 (0.023)***	0.067 (0.028)**	0.097 (0.016)***	0.066 (0.023)***
Impact of patient order (testing for detection of research team during data visit)
patient order by day	0.000 (0.002)	0.006 (0.003)**	0.003 (0.002)*	0.002 (0.002)
patient order after enc.	−0.005 (0.003)*	−0.008 (0.003)**	−0.005 (0.002)**	−0.005 (0.003)*
Clinician effects	clinician fixed effects included
Patient effects	patient and illness type characteristics included			patient chars included
N (patients)	2556	2555	2554	2557

Protocol adherence is the percentage of items required by protocol used, as reported by the patient. Each coefficient is the percentage point increase in adherence from a baseline average of 73%. Sample includes health workers seen in each of both rounds.

The model is a mixed effects regression with health worker fixed effects and facility random effects. Standard errors in parenthesis:

^***indicates

^***p<0.01,

^**p<0.05,

^*p<0.1