The mortality and health report from the national 2012 census was released in July 2015.
For the first time in the national census history, the report includes information on maternal mortality and with data disaggregated to the regional level.
Maternal mortality ratio (MMR) for Tanzania is 432 maternal deaths per 100,000 live births. This translates to about 4 women deaths from pregnancy-related causes for every 1,000 live births, indicating that maternal mortality ratio has not declined fast enough to meet the MDG5 target of reducing maternal mortality by three quarters from the 1990 levels by 2015.
Maternal mortality ratio is higher in Tanzania Mainland than in Zanzibar (435 vs 307 deaths per 100,000 live births) and in urban than rural areas: 443 vs 336 deaths per 100,000 live births.
The finding may be surprising as previous analyses, including the Sharpened One Plan, suggested that availability and accessibility of most maternal health services were relatively better in urban than the rural settings.
Although the difference may be from differential in over-reporting or under-reporting of deaths, age misreporting and wrong dating of events with more urban deaths likely to be identified and reported correctly than those occurring in rural areas, sub-optimal quality of care cannot be ruled out.
Women may still die in facilities due to lack of timely appropriate quality care.
Overall, understanding how the various adjustments were made before maternal mortality ratios for various categories were calculated will be worthwhile to understand the results, especially in view of the importance of this milestone in understanding maternal mortality in the country.
This was the only census ever in our country history to include data on maternal mortality and previous data were largely from surveys and modelled estimates.
Teenager births accounted for almost 10% of all births in the country, a level lower than 20% in the 2010 DHS.
The data further highlight the high level of maternal mortality ratio (MMR) among teenagers compared to women aged 20-24 and 25-29 (341 vs 226 and 257 per 100,000 live births) but less than the levels among women aged 30 and over.
A trend similar to the overall national MMR is also observed for the urban-rural levels among teenagers with 401 and 252 deaths per 100,000 respectively. The evidence reaffirms a higher risk among teenagers and women aged 30 and over.
Teenage maternal mortality ratio was lower in Tanzania Mainland 339 death per 100,000 than in Zanzibar 424 death per 100,000. Likewise the proportion of maternal deaths to all deaths among teenagers was approximately 21% in Tanzania Mainland compared to almost 27% in Zanzibar. Of all pregnancy-related deaths in the country, teenager deaths account for about 8%.
Also, of note is the fact that among women aged 20-24 and 25-29 years, the proportional of maternal deaths to all deaths is almost 40% and 32% respectively, indicating a disproportionately higher deaths. Two things could explain this finding.
These are the age groups with the highest number of births in the country, thus the likelihood of more deaths occurring from the groups. Nonetheless, given equal care, these are among the age groups expected to be relatively safe in terms of obstetric risks yet the ones with the highest pregnancy-related deaths. But as long as many women continue delivering outside the health facilities without skilled delivery assistance and the quality of maternity care remains sub-optimal, those perceived to be of less risk will not be spared.
Marked differences are seen in maternal mortality ratios across regions ranging from 860 deaths per 100,000 in Rukwa to 187 per 100,000 in Simiyu.
Thirteen out of the total 30 regions in the country had MMR above the national average. Regions with MMR above 500 per 100,000 were (Rukwa (860), Njombe (788), Mbeya (776), Pwani (687), Katavi (670), Tanga (593), Arusha (585), Mtwara (579) and Dodoma (512).
None of these regions was among the eight priority regions identified during the Sharpened One Plan analysis as worst performing in most maternal health coverage indicators such as family planning, delivering in health facilities and skilled delivery care. In fact, three (Pwani from the Eastern Zone and Tanga and Arusha from the Northern Zone) are from the best performing zones in the coverage of most maternal health indicators.
It is difficult to understand and reliably explain the possible reasons for the regional differences based on the census data alone. Multiple and different explanations may exist to explain the observed pattern. One possible explanation for the finding could be poor quality of services such that women experience suboptimal survival even when they deliver in health facilities. Nevertheless, it is unlikely that such care is only more pronounced in these regions than those identifies as poor performing during the Sharpened one Plan assessment.
Contextual factors in health care seeking, especially in situations of pregnancy, labour or delivery-related complications may be another explanation.
The method used to estimate maternal mortality involved reporting information on the demise of family members, which also brings a possibility of differential reporting due to recall of causes of deaths and socio-cultural and normative differences across regions on death reporting. But no robust evidence exists in the Tanzanian contexts to support this assertion, especially at the level of census data.
Underlying disease conditions such as AIDS are known to contribute significantly to the high level of maternal mortality in sub-Saharan countries but for Tanzania, except for Njombe and Mbeya regions, Arusha, Dodoma and Tanga regions have one of the lowest levels of HIV infection so AIDS may not be a plausible explanation.
To conclude, the high level of maternal mortality remains a huge challenge in Tanzania requiring expedited progress in improving maternal survival and wellbeing.
Furthermore, relying on coverage indicators alone may not be enough to understand progress in survival and hence the need for periodic maternal mortality estimates to provide robust evidence to inform planning and implementation of key interventions.
Concurrently, improving vital registration should be the interim and long term goal.