Table 5.

Log-logistic inverse Gaussian shared frailty model analysis of obstetric fistula patients at Mekelle Hamlin Fistula Center, January 2015 to January 2020.

Variables	Estimate ($\widehat{\beta }$)	EXP ($\widehat{\beta }$)	95% CI for exp ($\widehat{\beta }$)	SE ($\widehat{\beta }$)	p value
Height
<150 cm	1	1	1	1
>150 cm	-0.3730	0.688	0.517, 0.860	0.0875	0.047∗
Age at delivery
<18	1	1	1	1
18-30	-0.35650	10.700	0.457, 0.943	0.1238	0.004∗
> 30	-0.65023	0.522	0.259, 0.784	0.1338	0.000∗
Residence
Urban	1	1	1	1
Rural	0.3057	1.357	1.236, 1.479	0.0621	0.000∗
Educational status
No education	1	1	1	1
Primary education	-0.08094	0.922	0.830, 1.014	0.0469	0.085
Secondary and above	-0.18593	0.830	0.693, 0.967	0.0699	0.008∗
Physiotherapy use
No	1	1	1	1
Yes	-0.22146	0.801	0.662, 0.940	0.0708	0.002∗
Place of delivery
Health institution	1	1	1	1
Home	0.15264	1.165	1.034, 1.296	0.0668	0.022∗
Fistula type
VVF	1	1	1	1
RVF	-0.51585	0.597	0.433, 0.761	0.0838	0.000∗
Fistula size
Small	1	1	1	1
Medium	-0.03473	0.966	0.836, 1.096	0.0664	0.601
Large	0.19990	1.221	1.110, 1.332	0.0567	0.000∗
Extensive	0.24832	1.282	1.175, 1.388	0.0544	0.000∗

θ = 0.0762, τ = 0.0367, λ = 0.29, ρ = 2.5, and AIC = 229.677. Likelihood ratio test of θ = 0, chi-square = 8.97, and p value = 0.001^∗. $\text{Exp}\left(\widehat{\beta }\right)$ indicates acceleration factor. ^∗Significant at 5% level. 95% CI for $\exp \left(\widehat{\beta }\right)$: confidence interval for $\exp \left(\widehat{\beta }\right)$; SE ($\widehat{\beta }$): standard error for $\widehat{\beta }$.