Informing humanitarians worldwide 24/7 — a service provided by UN OCHA

Myanmar

The 2014 Myanmar Population and Housing Census - Thematic Report on Mortality - Census Report Volume 4-B

Attachments

Executive Summary

The main objectives of this thematic report are to estimate early-age and adult mortality using data from the 2014 Population and Housing Census, and to conduct analyses describing and explaining mortality levels and trends within the socioeconomic context of Myanmar, using appropriate concepts and taking relevant policy issues into consideration.
The vital registration system is not well developed in Myanmar. This situation is common in developing countries and a census is a major alternative source of mortality data in such contexts. Different census data are used to measure early-age and adult mortality.
In the 2014 Census, early-age mortality was measured from the responses to two simple retrospective questions on childbearing addressed to ever-married women aged 15 and over.
These questions referred to how many live children they had ever given birth to, and how many had died (or survived). Adult mortality was measured by using a question on the number of household members who had died during the 12 months preceding the Census.
According to the 2014 Census, infant and child mortality, which comprises under-five mortality, was high compared to other countries in the region. Previous estimates indicated a rapid decline during the 1960s and 1970s, with a substantial deceleration starting in the early 1980s. The decline has accelerated again during recent years.
An important issue revealed by the data is that substantial differences between sexes were observed in under-five mortality. The probability of dying among males is almost one third higher than that of females. Male infant mortality rates are universally higher than female rates. Biological factors are usually considered responsible for this differential. Nevertheless, child mortality sex differentials tend to disappear or even reverse in most countries. In Myanmar, sex differentials continue to be higher among males during child mortality. This is not easy to explain with census data alone, or even with household survey data. This topic can only be analysed by an in-depth qualitative study.
Adult mortality was found to be high; relatively much higher than under-five mortality. The main reason for this is the particularly high level of male adult mortality. It is probable that these sex differences could be caused by behavioural factors, in particular, the prevalence of unsafe and risky life styles among males. An important related finding is that, contrary to experiences in most countries, male mortality rates in urban areas were higher than in rural areas. However, female mortality rates were lower in urban than in rural areas. Studies that go beyond the interpretation of the 2014 Census data would be necessary to explain these patterns.
In order to better understand under-five mortality levels and patterns, a differential analysis was conducted. Several variables were considered as differentials of under-five mortality rates. All variables selected showed different effects on under-five mortality rates. The most important variable was women’s parity: the higher the number of children already born, the lower the probability of survival of the child. In spite of a low level of fertility in Myanmar, there is still scope to improve infant and child mortality through a decline in fertility. An important decline, especially in infant mortality, would be achieved if women gave birth to fewer children; studies have shown that the fertility rate is directly related to early-age mortality. The other differentials suggest that substantial reductions of under-five mortality could be achieved by improving the standard of living of the population.
A spatial analysis of early-age mortality was conducted. This analysis was conducted using Townships as the unit of analysis, and the proportion of children who had died among those ever born to women aged 20-34 as a mortality indicator. Although substantial variations were found, clusters of Townships with similar mortality rates were revealed. There are two clear clusters of Townships with medium-high and high early-age mortality. The first follows the Ayeyawady River, starting in the north-eastern part of the country and descending south to the delta. The second cluster is in the north-central part of the country and descends towards the first cluster. There are also smaller clusters of low early-age mortality in border areas. The Census does not, however, provide information to show the cause of this spatial distribution of early-age mortality, and an analysis of this type goes beyond the purpose of this thematic report. It would be important, however, to use this information to conduct a study that includes environmental and geopolitical characteristics of the States/Regions.
Twelve variables relating to the characteristics of the population in the Townships were identified. These variables were closely associated with early-age mortality. All the correlations were found to be statistically significant, although the magnitude of some was low.
The only two variables that were found to closely affect early-age mortality were the degree of development or under-development of Townships, and household composition. The other variable was an indicator of fertility. A more equal distribution of early-age mortality rates (and a subsequent overall decline) should be based not on a vertical expansion of health care but in improving the living conditions of the population, in making health care widely accessible, and in better understanding the family role in health care. This last indicator deserves more attention in future studies. In areas where large families prevail, children’s survival probabilities appear to be higher than in places where household extension is limited.
In the analysis of mortality differentials, indicators of household extension were also related to under-five mortality. It is important to get a better understanding of the mechanism through which this variable improves survival probabilities.
These results suggest the need for health policies based on the expansion of conventional health services and infrastructure that reach more marginalized populations, especially those living in hard-to reach areas. Policies directed to further reduce fertility may also have an important impact on under-five mortality. These results also call for unconventional propositions such as considering household composition in the formulation of health policies, as well as the importance of interventions that aim to change behaviours for the adoption of more healthy lifestyles, especially among males.