Table 2.
Simple linear | Linear interaction | Quadratic interaction | |
---|---|---|---|
Utilisation (LPM) | |||
ITT effect (Age ≥ 5 years) | −0.0310 | −0.0197 | 0.0885 |
(0.060) | (0.061) | (0.098) | |
N | 1633 | 1633 | 1633 |
AIC | 2235.2 | 2236.5 | 2234.3a |
Probability of reporting zero consultation and prescription expenditure (LPM) | |||
ITT effect (Age ≥ 5 years) | −0.134* | −0.142** | −0.0938 |
(0.071) | (0.071) | (0.116) | |
N | 915 | 915 | 915 |
AIC | 1180.8a | 1182.3 | 1186.0 |
Probability of reporting zero consultation expenditure (LPM) | |||
ITT effect (Age ≥ 5 years) | −0.121* | −0.126* | −0.0359 |
(0.073) | (0.072) | (0.113) | |
N | 915 | 915 | 915 |
AIC | 1136.5a | 1138.3 | 1141.2 |
Probability of reporting zero prescription expenditure (LPM) | |||
ITT effect (Age ≥ 5 years) | −0.0410 | −0.0628 | −0.0600 |
(0.058) | (0.059) | (0.095) | |
N | 915 | 915 | 915 |
AIC | 1005.8 | 1003.9a | 1007.8 |
Total expenditure (two-part model showing marginal effects) | |||
ITT effect (Age ≥ 5 years) | −0.144 | −0.0827 | −0.623 |
(0.462) | (0.436) | (1.753) | |
N | 915 | 915 | 915 |
AIC | 686.0 | 685.94 | 685.93a |
Control variables includes region dummies, head of households’ age and education
Standard errors in parentheses
LPM linear probability model
* p < 0.1, ** p < 0.05
aOptimal model specification