Scholarlink

Background <p>Child mortality remains a significant public health concern in Ethiopia, particularly among children under five. Despite recent improvements in healthcare and economic development, mortality rates remain high due to infectious diseases, malnutrition, and limited access to healthcare. This study aimed to identify key determinants of child mortality among ever-married women aged 15–49 in Ethiopia using advanced multi-level count regression models.</p> Methods <p>This study used data from the Ethiopia Demographic and Health Survey (EDHS), Data from 16,650 women were analyzed using single-level and multi-level count models, including Poisson, negative binomial, zero-inflated, and hurdle models. The multilevel hurdle negative binomial (HNB) model was selected for its ability to distinguish between factors influencing the occurrence and frequency of child deaths.</p> Result <p>The truncated negative binomial part of the selected multilevel HNB model demonstrated a significant reduction in the risk of child mortality for babies born to mothers with the following characteristics: older age at first birth (IRR = 0.9398; 95% CI: 0.9139–0.9655), primary education level (IRR = 0.9298; 95% CI: 0.6763–1.2783), secondary education or above (IRR = 0.5961; 95% CI: 0.4103–0.8658), breastfeeding (IRR = 0.7886; 95% CI: 0.6484–0.9592), and contraceptive use (IRR = 0.8038; 95% CI: 0.6593–0.9799). Additionally, the study revealed significant regional differences in child mortality (<i>p</i> = 0.0401 and 0.0418), indicating clustering effects. There were notable regional variations in how mother occupation, father education, family size, contraceptive use, and mother education level influenced child mortality in Ethiopia. The multilevel HNB model revealed strong regional variations, with statistically significant clustering effects.</p> Conclusion <p>The random coefficients model was the most accurate of the three multilevel HNB regression models for predicting newborn mortality per mother. Significant factors identified included family size, age at first birth, birth order, contraceptive use, maternal education level, maternal occupation, breastfeeding practices, residence, and multiple births. Regional disparities in child mortality rates highlight the need for localized interventions. Strategies should focus on improving maternal education, access to healthcare in rural and nomadic areas, promoting breastfeeding, and increasing contraceptive use. Policymakers should prioritize high-mortality regions like Afar and Gambela for targeted support.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Child mortality among ever married women aged 15 to 49 years in Ethiopia using multilevel count regression models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月19日

查看全文
  • Buzuneh Tasfa Marine,
  • Dagne Tesfaye Mengistie,
  • Mihiret Genene Zewde,
  • Kitesa Biresa Duftu,
  • Habitamu Wudu,
  • Maru Zewdu Kassie,
  • Abraham Lomboro Dimore,
  • Chekol Alemu,
  • Firew Tiruneh Tiyare

摘要

Background

Child mortality remains a significant public health concern in Ethiopia, particularly among children under five. Despite recent improvements in healthcare and economic development, mortality rates remain high due to infectious diseases, malnutrition, and limited access to healthcare. This study aimed to identify key determinants of child mortality among ever-married women aged 15–49 in Ethiopia using advanced multi-level count regression models.

Methods

This study used data from the Ethiopia Demographic and Health Survey (EDHS), Data from 16,650 women were analyzed using single-level and multi-level count models, including Poisson, negative binomial, zero-inflated, and hurdle models. The multilevel hurdle negative binomial (HNB) model was selected for its ability to distinguish between factors influencing the occurrence and frequency of child deaths.

Result

The truncated negative binomial part of the selected multilevel HNB model demonstrated a significant reduction in the risk of child mortality for babies born to mothers with the following characteristics: older age at first birth (IRR = 0.9398; 95% CI: 0.9139–0.9655), primary education level (IRR = 0.9298; 95% CI: 0.6763–1.2783), secondary education or above (IRR = 0.5961; 95% CI: 0.4103–0.8658), breastfeeding (IRR = 0.7886; 95% CI: 0.6484–0.9592), and contraceptive use (IRR = 0.8038; 95% CI: 0.6593–0.9799). Additionally, the study revealed significant regional differences in child mortality (p = 0.0401 and 0.0418), indicating clustering effects. There were notable regional variations in how mother occupation, father education, family size, contraceptive use, and mother education level influenced child mortality in Ethiopia. The multilevel HNB model revealed strong regional variations, with statistically significant clustering effects.

Conclusion

The random coefficients model was the most accurate of the three multilevel HNB regression models for predicting newborn mortality per mother. Significant factors identified included family size, age at first birth, birth order, contraceptive use, maternal education level, maternal occupation, breastfeeding practices, residence, and multiple births. Regional disparities in child mortality rates highlight the need for localized interventions. Strategies should focus on improving maternal education, access to healthcare in rural and nomadic areas, promoting breastfeeding, and increasing contraceptive use. Policymakers should prioritize high-mortality regions like Afar and Gambela for targeted support.