7. Primary outcome 2.1: Proportion of women with at least 4 ANC visits.
| Study | Measure of effect | Statistical approach1 | Intervention | Control |
| Bhutta 2011 | Adjusted | 302/2339 | 191/2135 | |
| Laken 1995 | Not cluster trial – two intervention arms added together. calculated ln (OR) and se (Ln OR) | 104/194 | 101/101 | |
| Le Roux 2013 | 1.00 (0.74, 1.34) OR Random effects logistic regression, adjusted for neighbourhood clustering. Models for outcomes among HIV‐positive mothers control for baseline employment. | Straight into RevMan | 474/608 | 439/549 |
| Morris 2004 both packages | Calculated from change scores, Table 2 Program effects p. 2034 of main report. Intervention 38.1% + 18.4% = 56.5% of the final sample size in this group which is 110; 48.9% ‐ 0.7% = 48.2 % in the control group (n = 313). Adjusted as below | 62/110 | 151/313 | |
| Manthip 2015 | Adjusted | 62/127 | 53/190 | |
| Wu 2011 | Calculated adjusted SE from P value 0.246 | 376/673 | 253/591 |
1Primary analysis using ICC of 0.02.