Table 3.
Effect of Development Media International's radio campaign on all-cause post-neonatal under-5 child mortality (intention-to-treat analysis)
| Mortality rate per 1000 livebirths (95% CI) |
Risk ratio (95% CI); p value |
|||
|---|---|---|---|---|
| Control group | Intervention group | Cluster-level analysis, adjusted for pre-intervention level* | Cluster-level analysis, adjusted for pre-intervention level and confounder score† | |
| March, 2010, to February, 2012 | 93·3 (72·5–114·2) | 125·1 (104·8–145·5) | .. | .. |
| March, 2012, to December, 2012 | 88·0 (66·3–109·8) | 114·5 (95·5–133·5) | 0·99 (0·83–1·18); 0·870 | 1·00 (0·81–1·24); 0·919 |
| January, 2013, to October, 2013 | 71·4 (52·4–90·4) | 105·2 (81·9–128·4) | 1·00 (0·82–1·22); 0·991 | 0·95 (0·76–1·18); 0·640 |
| November, 2013, to October, 2014 | 58·5 (44·8–72·3) | 85·1 (67·7–102·6) | 1·06 (0·84–1·32); 0·591 | 1·04 (0·80–1·36); 0·729 |
| March, 2012, to October, 2014 | 71·8 (53·5–90·2) | 100·5 (82·0–119·0) | 1·01 (0·86–1·20); 0·842 | 1·00 (0·82–1·22); 0·999 |
*
p value for effect modification by time=0·353.
†
p value for effect modification by time=0·353.