Impact of maternal death reviews at a rural hospital in Zambia: a mixed methods study.

BACKGROUND

Maternal mortality in sub-Saharan Africa remains high despite programmatic efforts to improve maternal health. In 2007, the Zambian Ministry of Health mandated facility-based maternal death review (MDR) programs in line with World Health Organization recommendations. We assessed the impact of an [MDR program] at a district-level hospital in rural Zambia.

METHODS

We conducted a mixed methods convergent study using hospital data on maternal mortality and audit reports of 106 maternal deaths from 2007 to 2011. To evaluate the overall impact of MDR on maternal mortality, we compared baseline (2007) to late (2010-11) post-intervention inpatient maternal mortality indicators. MDR committee reports were coded and dominant themes were extracted in a qualitative analysis. We assessed potential risk factors for maternal mortality in a before-and-after design comparing the periods 2008-09 and 2010-11.

RESULTS

In-hospital maternal mortality declined from 23 per thousand live births in 2007 to 8 per thousand in 2010-11 (P < 0.01). Maternal case fatality for puerperal sepsis and uterine rupture decreased significantly from 63 and 32% in 2007 to 10 and 9% in 2010-11 (P < 0.01). No significant reduction was seen in case fatality due to postpartum hemorrhage. Qualitative analysis of risk factors for maternal mortality revealed four core themes: standards of practice, health systems, accessibility, and patient factors. Specific risk factors included delayed referral, missed diagnoses, intra-hospital delays in care, low medication inventory, and medical error. We found no statistically significant differences in the prevalence of risk factors between the before-and-after periods.

CONCLUSIONS

Implementation of MDR was accompanied by a significant decrease in maternal mortality with reductions in maternal death from puerperal sepsis and uterine rupture, but not postpartum hemorrhage. Qualitative analysis of audit reports identified several modifiable risk factors within four core areas. Comparisons of potential explanatory factors did not show any differences over time. These results imply that MDR offers a means for hospitals to curtail maternal deaths, except deaths due to postpartum hemorrhage, suggesting additional interventions are needed. Documentation of MDR meetings provides an instrument to guide further quality improvements.