Table 2.
Sensitivity analysis of the association between number of community health-care worker POM-guided CLIP visits and the primary outcome among 30 721 pregnancies in CLIP*
|
0 visits |
1–3 visits |
4–7 visits |
≥8 visits |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Event rate (n=10 754) | Adjusted odds ratio† | Event rate (n=4309) | Adjusted odds ratio† | p value | Event rate (n=8974) | Adjusted odds ratio† | p value | Event rate (n=6684) | Adjusted odds ratio† | p value | ||
| Primary outcome‡ | 2928 (27·2%) | 1 (ref) | 1251 (29·0%) | 1·07 (0·98–1·18) | 0·16 | 2340 (26·1%) | 0·95 (0·88–1·03) | 0·21 | 1242 (18·6%) | 0·88 (0·80–0·97) | 0·0079 | |
| Maternal outcome | 1246 (11·6%) | 1 (ref) | 539 (12·5%) | 0·96 (0.84–1·08) | 0·49 | 1040 (11·6%) | 0·92 (0·82–1·02) | 0·11 | 436 (6·5%) | 0·86 (0·74–0·99) | 0·032 | |
| Maternal mortality | 24 (0·2%) | 1 (ref) | 11 (0·3%) | 0·96 (0·42–2·18) | 0·92 | 14 (0·2%) | 0·57 (0·28–1·19) | 0·14 | 5 (0·1%) | 0·32 (0·11–0·92) | 0·035 | |
| Maternal morbidity | 1229 (11·5%) | 1 (ref) | 535 (12·4%) | 0·97 (0·86–1·10) | 0·63 | 1032 (11·5%) | 0·93 (0·83–1·03) | 0·16 | 435 (6·5%) | 0·87 (0·76–1·01) | 0·060 | |
| Fetal or neonatal adverse outcome | 2132 (19·9%) | 1 (ref) | 922 (21·5%) | 1·16 (1·05–1·28) | 0·0047 | 1634 (18·2%) | 0·96 (0·88–1·05) | 0·37 | 928 (13·9%) | 0·86 (0·77–0·95) | 0·0037 | |
| Stillbirth | 505 (4·7%) | 1 (ref) | 279 (6·5%) | 1·54 (1·30–1·81) | <0·0001 | 396 (4·4%) | 0·88 (0·76–1.02) | 0·098 | 142 (2·1%) | 0·41 (0·33–0·51) | <0·0001 | |
| Neonatal mortality | 545 (5·3%) | 1 (ref) | 260 (6·5%) | 1·34 (1·14–1·58) | 0·0005 | 414 (4·8%) | 0·93 (0·81–1·08) | 0·35 | 188 (2·9%) | 0·66 (0·54–0·81) | <0·0001 | |
| Neonatal morbidity | 1266 (11·8%) | 1 (ref) | 485 (11·3%) | 0·87 (0·76–0·99) | 0·034 | 986 (11·0%) | 0·99 (0·89–1·10) | 0·79 | 725 (10·9%) | 1·24 (1·09–1·41) | 0·0016 | |
Data are number of events and proportion of total events by visit number category, and adjusted odds ratio with 95% CI in parentheses, unless otherwise stated. CLIP=Community-Level Interventions for Pre-eclampsia. POM=Pre-eclampsia integrated estimate of risk-On-the-Move.
These analyses included the 30 721 women in intervention clusters who were followed-up, excluding the 1341 women who were followed-up (ie, excluding 1341 women from the total population who were recruited and had a miscarriage before 20 weeks).
Adjusted for maternal characteristics (as in the primary analysis: maternal age, parity, basic education, and gestational age at enrolment; enrolment timing in the trial; and distance from the household to facility.
Defined as one or more of maternal morbidity or mortality, stillbirth, neonatal mortality, or neonatal morbidity; the primary outcome in the CLIP Trials.