Predictors of maternal mortality in Ghana: Evidence from the 2017 GMHS verbal autopsy data

Joshua Sumankuuro, Joseph K. Wulifan, William Angko, Judith Crockett, Emmanuel K. Derbile, John K. Ganle

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)


Maternal mortality remains a significant public health challenge in many low and middle-income countries, including Ghana. From Ghana’s 2017 Maternal Health Survey verbal autopsy data, we examined the predictors of maternal mortality in Ghana.
Methods: A total of 1240 deaths of women aged 15-49 were involved in the survey across all regions in Ghana. Binary and multivariate logistic regression analyses were employed; confidence level was set at 95%.
Results: The results show that the prevalence of maternal death was 13.2% (164/1240). After adjusting for potential covariates, women aged 20-29 years (aOR = 4.270, 95%CI= 1.864 – 9.781, p=0.001), bled during labour/delivery (aOR= 0.241, 95%CI = 0.059 – 0.992, p=0.049), and those who used traditional/herbal medicines during pregnancy were more likely to die compared to non-users (aOR= 3.461, 95%CI = 1.651 – 7.258, p=0.001).
Conclusion: Our findings highlight the need to intensify maternal education regarding the value to be gained by increasing skilled healthcare during complications in pregnancy to allow effective management of complications during labour/delivery. Also, education for pregnant women and their families on possible adverse effects of using unapproved traditional/herbal medicines during pregnancy as well as a need to seek timely care before the onset of labour to allow healthcare providers ample opportunity to address labour and birth complications, is urgently required.
Original languageEnglish
Pages (from-to)1512-1531
Number of pages20
JournalInternational Journal of Health Planning and Management
Issue number6
Early online date09 Sept 2020
Publication statusPublished - 23 Nov 2020


Dive into the research topics of 'Predictors of maternal mortality in Ghana: Evidence from the 2017 GMHS verbal autopsy data'. Together they form a unique fingerprint.

Cite this