Are morbidity and mortality estimates from randomized controlled trials externally valid? A comparison of outcomes among infants enrolled into an RCT or a cohort study in Botswana

Table 4 Sensitivity analysis of treatment effects excluding infants experiencing outcomes within 30 days of enrollment

	Coefficient	P >\|z\|	95% CI
Risk of Mortality in the cohort study^a	0.02	< 0.001	(0.01, 0.02)
Difference in Mortality caused by RCT^b	0.00	0.947	(-0.01, 0.01)
Risk of Morbidity in the cohort study^a	0.12	< 0.001	(0.10, 0.13)
Difference in Morbidity caused by RCT^b	-0.05	< 0.001	(-0.07, -0.02)

^aIndicates potential-outcome mean (POM) had all children been enrolled into the cohort study
^bIndicates average treatment effect (ATE) attributable to enrollment into the RCT

ISSN: 1471-2288