TABLE 3.

Results of weighted logistic regression with odds ratios of end-line relative to baseline and intervention relative to control group for selected items.

Item	Response†	End-line relative to baseline			Intervention group relative to control group
		Group	OR	p‡	Period	OR	p‡
A. Did a member of staff attend to you within 15 min of arriving at the unit?	Yes	Control	6.098	0.0007***	Baseline	1.204	0.2775
		Intervention	8.418	0.0001***	End-line	1.662	0.1434
B. Did the sister ask if it was okay to examine you?	Yes	Control	0.987	0.9599	Baseline	0.462	< 0.0001***
		Intervention	2.381	0.0018***	End-line	1.114	0.6748
C. Did any staff member say anything that upset you?	No	Control	0.904	0.9725	Baseline	0.838	0.7136
		Intervention	1.215	0.8760	End-line	1.127	0.8493
D. How did the staff speak to you during labour?	All staff spoke nicely	Control	1.801	0.1530	Baseline	0.568	< 0.0001***
		Intervention	3.172	0.0009***	End-line	1.000	0.9990
E. How respectfully do you think the sisters treated you during your stay in the labour ward?	A lot of respect	Control	1.690	0.1079	Baseline	0.443	< 0.0001***
		Intervention	4.334	< 0.0001***	End-line	1.137	0.3602
F. How satisfied were you with the treatment you received in the labour ward?	Completely satisfied	Control	1.763	0.0736	Baseline	0.525	< 0.0001***
		Intervention	4.044	< 0.0001***	End-line	1.204	0.1868

^***, Significant at 1% level.

^†, Modelling of responses: Item A: Yes versus No; Item B: Yes versus No; Item C: No versus Yes; Item D: All staff spoke nicely versus Not all staff spoke nicely; Item E: A lot of respect versus Not a lot of respect; Item F: Completely satisfied versus Not completely satisfied.

^‡, Adjusted p value = Exceedance probability for accepting H₀: OR = 1 of ad hoc tests (end-line relative to baseline) and (intervention relative to control).