| |
Antenatal care attendance
|
Boiling the blade
|
Appropriate dressing of cord
|
Not discarding colostrum
|
---|
|
Intervention
|
n = 2535
|
n = 2454
|
n = 2454
|
n = 2409
|
---|
|
Control
|
n = 2838
|
n = 2762
|
n = 2762
|
n = 2711
|
---|
% BETTER
|
Intervention
|
19.1
|
21.8
|
17.9
|
29.7
|
|
Control
|
16.6
|
12.1
|
16.3
|
23.2
|
% GOOD
|
Intervention
|
36.0
|
32.5
|
63.2
|
41.7
|
|
Control
|
12.5
|
12.8
|
56.7
|
34.7
|
% BAD
|
Intervention
|
36.8
|
39.2
|
4.2
|
17.5
|
|
Control
|
65.6
|
68.1
|
9.3
|
26.5
|
% WORSE
|
Intervention
|
8.2
|
6.4
|
14.7
|
11.2
|
|
Control
|
5.2
|
7.0
|
17.8
|
15.5
|
TOTAL (%)
|
Intervention
|
100
|
100
|
100
|
100
|
|
Control
|
100
|
100
|
100
|
100
|
Intervention/control comparisons : Odds ratios (95% confidence interval)
| | | | | |
%BETTER/%BAD ratio *
| |
2.04 (1.82, 2.27)
|
3.13 (2.78, 3.45)
|
2.44 (1.92, 3.13)
|
1.92 (1.69, 2.22)
|
%BETTER/%WORSE ratio*
| |
0.73 (0.59, 0.91)
|
1.96 (1.59, 2.44)
|
1.33 (1.15, 1.54)
|
1.79 (1.52, 2.08)
|
%BETTER/%GOOD ratio*
| |
0.40 (0.35, 0.46)
|
0.71 (0.61, 0.81)
|
0.99 (0.88, 1.10)
|
1.06 (0.95, 1.19)
|
- *Results from multilevel multinomial models. The estimates and intervals are adjusted to take account of the correlations between pregnancies within the same women, women from the same household, households from the same VDC and VDCs within the same matched pair. All odds ratios are significantly different to 1. Coefficients greater than 1 indicate that the women in the intervention clusters had a more favourable distribution, those less than 1 are associated with a less favourable response for the women in the intervention compared with control clusters.