Trends and determinants of mortality in women of reproductive age in rural Guinea-Bissau, West Africa - a cohort study

Background: There are few studies reporting mortality of women of reproductive age (WRA) in developing countries. The trend and patterns of their mortality may be important for documenting the health status of the population in general. Methods: We used a prospective open cohort of women aged 12 to 49 years living in the Bandim Health Project's rural Health and Demographic Surveillance System (HDSS) in 5 regions of Guinea-Bissau from 1996 to 2007. Information on in-and out-migration and deaths were collected through the HDSS routine procedures. We assessed the trends in mortality and the associated determinants using Cox regression models. Results: We followed 27,185 WRA for 141,693 person-years-at-risk (PYO) among whom 9,093 moved out and 1,006 died. Overall standardized mortality rate was 759 per 100,000 PYO. WRA mortality did not decline, but three periods could be distinguished: a stable mortality between 1996-2000 followed by 14% increase in mortality [Hazard rate ratio (HRR) = 1.14; 95% confidence interval (CI): 0.98-1.32; p = 0.08] between 2001-2003, and then in the last period from 2004-2007 a 25% decline (HRR = 0.75; 95% CI: 0.64-0.87; p < 0.001) in relation to the first period. Compared with the years 1990-1996 mortality increased in the first two periods until 2003; only in the last period did mortality reach the same level as in 1990-1996 (HRR = 0.96; 95% CI: 0.82-1.13; p = 0.62). The level of mortality differed between regions. In the adjusted analysis the eastern regions Bafata (HRR = 1.79; 95% CI: 1.38-2.32; p < 0.001) and Gabu (HRR = 1.70; 95% CI: 1.28-2.26; p < 0.001) had significantly higher mortality, but the hazard rate did not differ by ethnic group. As expected the rate increased with increasing age. Conclusions: Over the twelve-year period mortality of WRA did not decline. A stable mortality in the beginning was followed by an increase and then a return to the previous levels. Further monitoring of mortality is needed to identify the risk factors for the striking regional differences. Advantage should be taken of the HDSS to monitor progress towards the MDGs and beyond.