Impact of the intervention package on behavioural outcomes at follow-up 1 (OLS)
| Mean at follow-up 1 | Partially adjusted | Adjusted | |||
| Beta (95% CI) | P value | Beta (95% CI) | P value | ||
| Overall compliance | |||||
| Percentage share of targeted behaviours performed | |||||
| QPM only | 41.3% | ||||
| QPM+targeting | 62.6% | 21.8 pp (18.2 pp to 25.4 pp) | <0.01 | 20.5 pp (16.7 pp to 24.4 pp) | <0.01 |
| Panel A: storing and selling QPM | |||||
| QPM grain unmixed during storage*† | |||||
| QPM only | 39.8% | ||||
| QPM+ targeting | 83.1% | 42.8 pp (35.3 pp to 50.2 pp) | <0.01 | 39.7 pp (32.0 pp to 47.5 pp) | <0.01 |
| QPM flour unmixed during storage† | |||||
| QPM only | 26.0% | ||||
| QPM+targeting | 74.1% | 48.0 pp (40.2 pp to 55.8 pp) | <0.01 | 46.5 pp (38.3 pp to 54.7 pp) | <0.01 |
| Not sold QPM since beginning of season† | |||||
| QPM only | 92.1% | ||||
| QPM+ targeting | 94.7% | 4.3 pp (0.5 pp to 8.1 pp) | 0.03 | 4.9 pp (0.7 pp to 9.0 pp) | 0.02 |
| Panel B: cooking | |||||
| Cooked QPM food specifically for young children† | |||||
| QPM only | 9.9% | ||||
| QPM+targeting | 24.8% | 15.0 pp (9.0 pp to 21.0 pp) | <0.01 | 14.4 pp (7.9 pp to 20.9 pp) | <0.01 |
| Days cooked QPM food specifically for young children | |||||
| QPM only | 0.2 | ||||
| QPM+targeting | 0.8 | 0.6 (0.4 to 0.8) | <0.01 | 0.6 (0.4 to 0.8) | <0.01 |
| Panel C: feeding and consumption | |||||
| Index child consumed any QPM in last 7 days† | |||||
| QPM only | 62.8% | ||||
| QPM+targeting | 81.6% | 19.3 pp (12.2 pp to 26.4 pp) | <0.01 | 17.3 pp (9.4 pp to 25.1 pp) | <0.01 |
| Days index child consumed QPM last week | |||||
| QPM only | 3.6 | ||||
| QPM+targeting | 4.5 | 1.0 (0.5 to 1.4) | <0.01 | 0.9 (0.4 to 1.4) | <0.01 |
| Index child consumed porridge with QPM last week† | |||||
| QPM only | 31.6% | ||||
| QPM+ targeting | 55.6% | 24.8 pp (17.1 pp to 32.5 pp) | <0.01 | 22.0 pp (13.6 pp to 30.4 pp) | <0.01 |
| Days index child consumed porridge with QPM last week | |||||
| QPM only | 0.6 | ||||
| QPM+targeting | 1.2 | 0.5 (0.4 to 0.7) | <0.01 | 0.4 (0.3 to 0.6) | <0.01 |
| Index child ate QPM for more days than household head† | |||||
| QPM only | 3.9% | ||||
| QPM+targeting | 17.3% | 13.0 pp (8.0 pp to 18.1 pp) | <0.01 | 12.2 pp (6.8 pp to 17.6 pp) | <0.01 |
| Difference in no of days QPM consumed between index child and household head | |||||
| QPM only | 0.0 | ||||
| QPM+targeting | 0.6 | 0.6 (0.4 to 0.8) | <0.01 | 0.5 (0.3 to 0.7) | <0.01 |
| Index child ate from own plate† | |||||
| QPM only | 64.5% | ||||
| QPM+targeting | 69.9% | 7.1 pp (−0.8 pp to 15.1 pp) | 0.08 | 7.4 pp (−1.1 pp to 15.9 pp) | 0.09 |
*Questions refer to how QPM was previously stored if household had already run out of QPM at the time of the first follow-up survey. Coefficients from ordinary least squares models are reported. Partially adjusted models only control for kebele to account for stratification and are clustered at the community health group level; adjusted models additionally control for household, caregiver and index child characteristics shown in table 1.
†Only items marked with '†' are in the overall compliance measure, which include (1) stored QPM grain separately, (2) stored QPM flour separately, (3) did not sell QPM, (4) cooked QPM specifically for young children, (5) index child ate QPM last week, (6) index child consumed QPM porridge last week, (7) index child ate QPM for more days than head of household, and (8) index child ate from own plate.
Beta, linear regression coefficient; PP, percentage points; QPM, Quality Protein Maize.