Table 6. Effect of the intervention on CMAM enrollment, treatment, and recovery outcomes for AM episodes assessed by longitudinal study.
| Comparison | Intervention | Δa | 95% CI | p-Value | |
|---|---|---|---|---|---|
| AM episodes | n = 595 | n = 452 | |||
| Enrolled in CMAM program | 265 (45%) | 154 (34%) | −10 | (−18, −1.9) | 0.016 |
| n = 265 | n = 154 | ||||
| Treatment coverage (primary outcome)b | 128 (48%) | 75 (49%) | 0.90 | (−9.2, 11) | 0.86* |
| Treatment initiatedc | 223 (84%) | 130 (85%) | 1.9 | (−7.0, 11) | 0.67 |
| Recovery within 3 months after enrollment | 219 (83%) | 130 (85%) | 2.2 | (−5.2, 9.6) | 0.56 |
| Length of enrolled episodes, dd | 74 ± 67 | 66 ± 50 | −7.6e | (−22, 6.4) | 0.29 |
| MAM episodes | n = 508 | n = 389 | |||
| Enrolled in CMAM program | 226 (45%) | 137 (35%) | −8.3 | (−16, −0.10) | 0.047 |
| n = 226 | n = 137 | ||||
| Treatment coverageb | 96 (42%) | 53 (39%) | −1.6 | (−11, 8.0) | 0.74 |
| Treatment initiatedc | 161 (71%) | 96 (70%) | 0.37 | (−11, 12) | 0.95 |
| Recovery within 3 months after enrollment | 199 (88%) | 121 (88%) | −0.76 | (−8.7, 7.1) | 0.85 |
| Length of enrolled episodes, dd | 60 ± 47 | 60 ± 41 | 0.83 e | (−10, 12) | 0.88 |
| SAM episodes | n = 120 | n = 78 | |||
| Enrolled in CMAM program | 58 (48%) | 25 (32%) | −15 | (−32, 1.2) | 0.07 |
| n = 58 | n = 25 | ||||
| Treatment coverageb | 5 (8.1%) | 6 (24%) | 15 | (−3.4, 34) | 0.11 |
| Treatment initiatedc | 23 (40%) | 12 (50%) | 11 | (−14, 36) | 0.39 |
| Recovery within 3 months after enrollment | 39 (67%) | 17 (67%) | 0.56 | (−19, 20) | 0.96 |
| Length of enrolled episodes, dd | 98 ± 90 | 82 ± 66 | −5.5e | (−48, 37) | 0.80 |
Data are n(%) or mean ± SD. Abbreviations: AM, acute malnutrition; CI, confidence interval; CMAM, community-based management of acute malnutrition; ICC, intracluster correlation coefficient; MAM, moderate acute malnutrition; pp, percentage point; SAM, severe acute malnutrition.
*Not statistically significant after correcting for multiple testing of primary outcomes, using a pcritical = 0.025 calculated using the Benjamini–Hochberg method. ICC for primary outcomes are presented in S2 Table.
aDifference between intervention and comparison arm expressed in pp (unless specified otherwise), analyzed using a mixed-effect linear probability regression model with health center and child as random effects and sampling strata, health district, month of inclusion, child sex, whether the child was the first liveborn, child age at the start of an episode, and intervention as fixed effects, unless specified otherwise.
bTreatment coverage defined as the proportion of AM, MAM, SAM children that received continuous treatment from CMAM enrollment onwards over the total number of enrolled AM, MAM, SAM children, respectively.
cTreatment initiated implies that AM children received either a MAM or SAM treatment, MAM children received MAM treatment, and SAM children received SAM treatment.
dEpisode length is measured from the onset of the AM, MAM, or SAM episode until the moment the child is free from AM for at least one monthly measurement.
eDifference in mean episode length (days) between intervention and comparison arm analyzed using a linear mixed-effects regression model with health center and child as random effects and sampling strata, health district, month of inclusion, child sex, whether the child was the first liveborn, child age at the start of an episode, and intervention as fixed effects.