Eshetu E Chaka1,2, Mulugeta Mekurie2, Ahmed Abdulahi Abdurahman3, Mahboubeh Parsaeian4, Reza Majdzadeh4,5. 1. Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences International Campus (TUMS-IC), Tehran, Iran. 2. Department of Public Health, College of Medicine and Health Sciences, Ambo University, Ambo, Ethiopia. 3. Department of Community Nutrition, School of Nutritional Sciences and Dietetics, Tehran University of Medical Sciences, Tehran, Iran. 4. Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran. 5. Department of Epidemiology and Biostatistics, School of Public Health, Knowledge Utilization Research Centre, Tehran University of Medical Sciences, Tehran, Iran.
Abstract
BACKGROUND: Neonatal mortality is a major health problem mainly in the developing countries and its reduction was remain stagnant during the era of Millennium Development Goal. Current global health policies emphasize institutional deliveries as a pathway to achieving reductions in neonatal mortality in developing countries. There are inconsistent conclusions of evidence about this fact, mainly in developing countries. Therefore, this study was conducted to assess the association between health facility delivery and neonatal mortality. METHODS: We systematically searched EMBASE, PubMed, ISI Web of Science and Scopus through 18 March 2018 and then updated on 14 February 2019. I2 test statistic was used to assess heterogeneity. Publication bias was checked using a funnel plot and meta-bias test. Random-effects model was used to determine the pooled effect size. RESULTS: Nineteen articles were included in the meta-analysis. The pooled odds ratios (ORs) indicated that health facility delivery was significantly associated with the odds of neonatal mortality (OR = 0.48; 95% CI: 0.38, 0.58). This significant inverse association was consistently found regardless of study design, geographical region and quality of the study. There is evidence of publication bias with high heterogeneity between studies (I2 = 84.5%). CONCLUSION: The odds of neonatal mortality were much likely lower among those delivered at a health facility than those delivery at home. Therefore, encouraging delivery in a health facility and minimizing any barrier to health facility are important. Further longitudinal studies based on larger, more representative samples are therefore needed to further assess the underlying relationships.
BACKGROUND: Neonatal mortality is a major health problem mainly in the developing countries and its reduction was remain stagnant during the era of Millennium Development Goal. Current global health policies emphasize institutional deliveries as a pathway to achieving reductions in neonatal mortality in developing countries. There are inconsistent conclusions of evidence about this fact, mainly in developing countries. Therefore, this study was conducted to assess the association between health facility delivery and neonatal mortality. METHODS: We systematically searched EMBASE, PubMed, ISI Web of Science and Scopus through 18 March 2018 and then updated on 14 February 2019. I2 test statistic was used to assess heterogeneity. Publication bias was checked using a funnel plot and meta-bias test. Random-effects model was used to determine the pooled effect size. RESULTS: Nineteen articles were included in the meta-analysis. The pooled odds ratios (ORs) indicated that health facility delivery was significantly associated with the odds of neonatal mortality (OR = 0.48; 95% CI: 0.38, 0.58). This significant inverse association was consistently found regardless of study design, geographical region and quality of the study. There is evidence of publication bias with high heterogeneity between studies (I2 = 84.5%). CONCLUSION: The odds of neonatal mortality were much likely lower among those delivered at a health facility than those delivery at home. Therefore, encouraging delivery in a health facility and minimizing any barrier to health facility are important. Further longitudinal studies based on larger, more representative samples are therefore needed to further assess the underlying relationships.