Table 2.
Mortality by time period and cohort strata
| No wasting (n=1120) | Moderate wasting (n=763) | Severe wasting or kwashiorkor (n=1218) | |
|---|---|---|---|
| All deaths | |||
| Number of deaths (%) | 39 (3·5%) | 62 (8·1%) | 249 (20·4%) |
| Risk ratio (95% CI)* | 1 (ref) | 2·54 (1·77–3·66) | 5·89 (3·68–9·43) |
| Deaths during index admission | |||
| Number of deaths (%) | 22 (2·0%) | 32 (4·2%) | 128 (10·5%) |
| Risk ratio (95% CI)* | 1 (ref) | 2·37 (1·64–3·41) | 5·43 (3·55–8·31) |
| 30-day mortality† | |||
| Number of deaths (%) | 26 (2·3%) | 40 (5·2%) | 168 (13·8%) |
| Mortality rate, deaths per 1000 child-months (95% CI) | 24·1 (16·5–36·4) | 55·0 (40·5–76·5) | 156 (134–183) |
| Rate ratio (95% CI)‡ | 1 (ref) | 2·37 (1·42–3·95) | 6·38 (4·16–9·78) |
| Post-discharge mortality§ | |||
| Number of deaths (%) | 17/1072 (1·6%) | 30/724 (4·1%) | 121/1078 (11·2%) |
| Mortality rate, deaths per 1000 child-months (95% CI) | 2·72 (1·72–4·57) | 7·32 (5·16–10·7) | 21·0 (17·6–25·3) |
| Rate ratio (95% CI)‡ | 1 (ref) | 2·73 (1·52–4·89) | 7·05 (4·32–11·5) |
*
Risk ratios from multi-level mixed effects log-binomial regression models adjusted for sex, age, and site and inverse probability weighted for selection and loss to follow-up.
†
Includes inpatient and post-discharge deaths within 30 days of hospital admission (fixed period).
‡
Rate ratios estimated using the Mantel-Cox method adjusted for sex, age, and site and inverse probability weighted for selection and loss to follow-up.
§
Includes deaths within 180 days of index hospital discharge (fixed period). The number of children discharged alive were 1072 children with no wasting, 724 with moderate wasting, and 1078 with severe wasting or kwashiorkor.