Sample characteristics
A total of 180 pastoralist women participated in the baseline survey. Their average age was 27.44 ± 5.13 years, and the majority were in the age group of 20–29, comprising 108 people (60.0%). Of all the participants, 169 (93.9%) were married, 33.3% practiced polygamous and 166 (92.2%) were illiterate. Only 26 (14.4%) had a mobile phone, and most of them, 51 (28.3%), came from Yaballo Godha village. The median distance traveled to reach the health facility was 15.00 km with Inter Quartile Ranges (IQR) of (10.00 and 74.00). Most of the participants, 60 (33.3%), visit the Mansile dispensary. Regarding the use of contraceptives, 146 (81.1%) reported that they had never used contraceptives as a form of family planning, and 127 (70.6%) had never received advice on their use or did not know where to obtain them. On assisted birth, most of them 106 (58.9%) were assisted by Traditional Birth Attendants (TBA), 45 (25%) by Health Care Workers (HCW), and 29 (16.1%) by both TBA and HCW. Approximately 75 individuals (41.7%) have visited ANC 4+; 60 (33.3%) had HFD, and only 77 (42.8%) received PNC (Table 1).
Results of bivariate analysis of maternal health care service
According to the bivariate analysis (Table 2), some of the independent variables show a significant association with the dependent ANC 4+ visit, health facility delivery, and postnatal care.
Women in monogamous marriages were significantly more likely to have ANC 4+ visits p =
0.054, (n = 56; 46.7%), and postnatal care, p = 0.033, (n = 58; 48.3%), compared to women in a polygamous marriage, however, no statistically significant difference was noted in the likelihood of health facility delivery.
Having a mobile phone for communication was significantly predictive of ANC 4+ visits, HFD, and receiving PNC. Women with mobile phones are likely to have ANC 4+ visits p = 0.001, (n = 23, 88.5%), HFD p = 0.020, (n = 14, 53.8%), and PNC p = 0.001, (n = 24, 92%).
The distance between the health facility and the village of residence was significantly associated with ANC 4 + visits, HFD, and PNC, with women living less than 15 km from the health facility being more likely to receive ANC 4 + visits p = 0.001, (n = 60, 55.0%), HFD p = 0.001, (n = 47, 43.1%), and PNC p = 0.001), (n = 60, 55%) than women who leave more than 16 km away from health facilities.
Attending ANC 4 + visits were significantly associated with HFD and receiving PNC, with women who attended ANC 4+ more likely to deliver at a health facility, p = 0.001, (n = 45, 60%) and receive PNC, p = 0.001, (n = 73, 97.3%). Variables such as maternal age, marital status, and literacy level did not indicate any association with ANC 4+ visits, HFD, and PNC.
Multivariable analysis of maternal health service utilization
The multivariate analysis results are presented in (Table 3). Many variables that demonstrated significant differences in the bivariate analyses shown also significance in the multivariable analysis. Notably, factors such as possession of a mobile phone, monogamous family, and distance to the health facility retained their importance with ANC 4+ visits, HFD, and PNC. After accounting for confounding factors, other variables did not exhibit a statistically significant association.
Women residing close to a health facility displayed a threefold higher likelihood of attending up to ANC 4+ visits (OR 3.10, 95% CI 1.47–6.53), a 2.8-fold higher likelihood of delivering to health facilities (OR 2.80, 95% CI 1.34–5.84), and a 2.5-fold higher likelihood of having PNC done (OR 2.49, 95% CI 1.19–5.22), compared to those women living more than 15 km away from a health facility.
Additionally, women with a mobile phone displayed a 30-fold higher likelihood of attending up to ANC 4+ visits (OR 29.88, 95% CI 6.68–133.62), threefold higher likelihood of HFD (OR 2.56, 95% CI 0.99–6.63), and a 60-fold higher likelihood of having PNC (OR 60.45, 95% CI 10.43–350.55), compared to those women who do not possess mobile phone.
Finally, women in monogamous marriage displayed a fivefold higher likelihood of attending up to ANC 4+ visits (OR 5.17, 95% CI 1.88–14.23), likelihood of HFD (OR 1.67, 95% CI 0.77–3.62), and a sevenfold higher likelihood of having PNC (OR 7.05, 95% CI 2.35–21.19), compared to those women in polygamous marriage.
Three goodness-of-fit assessment methods, the Hosmer Lemeshow test, Classification table, and Area under the Receiver Operating Characteristic (ROC) curve were applied for all three dependent variables.
The Omnibus tests of model coefficients showed p < 0.001 for ANC 4+, p = 0.001 for HFD, and p < 0.001 for PNC; these results showed all p-values are less than 0.05, which indicates a very good model fit.
Nagelkerke R Square indicated (R2N = 0.356) for ANC 4+ visits, (R2N = 0.124) for HFD, and (R2N = 0.382) for PNC, a moderate influence between the dependent and independent variables. Hosmer Lemeshow goodness-of-fit results show that the p-values for ANC 4+, HFD, and PNC were 0.790, 0.441, and 0.937, respectively. P-values greater than 0.05 indicated a good-fitting model.
ROC results for ANC 4+ visits, HFD, and PNC were 0.782 (0.715–0.849), 0.680 (0.602–0.765), and, 0.787(0.721–0.853) respectively. This means the area under the ROC curve was more than 70.0% correctly classified by the model, which showed more than 70% accuracy for ANC 4+ and PNC but slightly low for HFD. All three methods of model fit assessment indicated good model fit.
Add Comment