Effectiveness of the Gratuité user fee exemption policy on utilization and outcomes of maternal, newborn and child health services in conflict-affected districts of Burkina Faso from 2013 to 2018: a pre-post analysis

Table 2 Modelled parallel trends between the intervention and control areas

Outcome	Coefficient*	95% CI	p value	Parallel trends test^$
ANC 1 rate	0.029	− 0.012, 0.071	0.141	0.185
ANC 5+ rate	− 0.016	− 0.101, 0.068	0.659	0.497
Rates of women seen in the first trimester of pregnancy	0.047	− 0.037, 0.132	0.227	0.162
Delivery rate	0.015	− 0.025, 0.054	0.408	0.146
Caesarean section rate (excluding the Dori–Djibo pair)	− 0.0004	− 0.012, 0.011	0.936	n/a
Rates of postnatal visits in the 6th hour	0.015	− 0.122, 0.151	0.809	0.960
Rates of postnatal visits on the 6th day	0.151	0.010, 0.291	0.039	0.029
Rates of postnatal visits in the 6th week	0.089	− 0.026, 0.205	0.111	0.119
Rates of malaria cases in pregnancy	0.001	− 0.004, 0.006	0.586	0.262
Facility maternal mortality ratio (per 100,000 live births)	− 46.164	− 136.724, 44.396	0.267	n/a
Facility early neonatal mortality rate (per 1000 live births)	− 0.290	− 1.670, 1.091	0.635	n/a
Rates of new consultations in children under 1 year	1.799	1.128, 2.469	< 0.001	0.153
Rates of new consultations in children under 1–4 years	0.813	0.497, 1.129	0.001	0.231
Rates of uncomplicated malaria cases in children under 5 years	0.588	0.443, 0.734	< 0.001	0.002
Rates of severe malaria cases in children under 5 years	− 0.002	− 0.016, 0.013	0.788	0.166

*Coefficient for Average Treatment Effect on the Treated (ATET), which is adjusted for group effects and time effects. The coefficient value signifies how much the mean of the dependent variable changes given a one-unit shift in the independent variable while holding other variables in the model constant. A positive coefficient indicates that as the value of the independent variable increases, the mean of the dependent variable also tends to increase. A negative coefficient suggests that the dependent variable tends to decrease as the independent variable increases. The confidence interval shows the range of possible values that explain the relationship between rates comparing intervention and control areas
^$Parallel-trends test (pre-treatment period). H₀: Linear trends are parallel
n/a: Could not be estimated due to insufficient data

ISSN: 1752-1505