Task shifting in maternal and newborn care: a non-inferiority study examining delegation of antenatal counseling to lay nurse aides supported by job aids in Benin

Table 4 Differences in maternal knowledge by topic and provider type (per protocol)

Percentage (%) of women with correct responses	Nurse-midwives	Lay Nurse Aides	Difference (β) 95% CI	Inference^a
No. pregnant women (N = 409)	206	203
Adjusted Scores^b
≥3 messages in prenatal care	56.0	79.8	23.8 (15.7, 32.0)*	S
≥3 messages in birth preparedness	39.3	52.0	12.7 (5.2, 20.1)*	S
≥3 danger signs during pregnancy	76.9	85.5	8.6 (3.3, 13.9)*	S
= 2 messages in clean delivery	54.7	52.6	-2.1 (-14.1, 9.9)	U
≥3 messages in newborn care^c	63.1	73.0	9.9 (-0.3, 20.1)	NI
Mean # correct responses	11.4	12.6	1.2 (0.4, 2.0)*	-

[a] Non-inferiority margin (∆) = -10 where inference drawn is designated by: S = superior; NI = non-inferior; U = insufficient evidence. [b] Scores adjusted for correlation of observations; site- and provider-level characteristics (random effects); counseling mode; and patient age, education, first prenatal visit, and total number of prenatal visits (fixed effects). [c] Includes only women at six to nine months of pregnancy. *Significant at p < 0.05. Note: Upper limits of the confidence interval are not interpreted for non-inferiority analyses.

ISSN: 1748-5908