INFANT AND CHILD MORTALITY 8 his chapter presents levels, trends, and differentials in perinatal, neonatal, postneonatal, infant, child, and under-5 mortality in Uganda. The information enhances understanding of population dynamics and will assist in the planning and evaluation of health policies and programmes. Estimates of infant and child mortality rates can be used to develop population projections. Information on childhood mortality also serves the need of the health sector to identify population groups that are at high risk. One of the targets of the Millennium Development Goals (MDGs) is to reduce the under-5 mortality rate by two-thirds between 1990 and 2015. Results from the 2011 UDHS can be used to monitor the impact of major interventions, strategies, and policies at the national level. Policies that affect the under 5 mortality rate are the National Health Policy (NHP II 2010/19) and the Health Sector Strategic and Investment Plan (HSSIP 2010/11-2014/15). The data used to estimate mortality were collected in the birth history section of the Woman’s Questionnaire. The birth history section begins with questions about the respondent’s experience with childbearing (i.e., the number of sons and daughters who live with the mother, the number who live elsewhere, and the number who have died). These questions are followed by a retrospective birth history, in which each respondent is asked to list each of her births, starting with the first birth. For each birth, data are obtained on sex, month and year of birth, survivorship status, and current age or, if the child is dead, age at death. This information is used to directly estimate mortality rates. In this report age-specific mortality rates are categorised and defined as follows:
• Neonatal mortality (NN): the probability of dying within the first month of life • Postneonatal mortality (PNN): the arithmetic difference between neonatal and infant mortality • Infant mortality (1q0): the probability of dying before the first birthday • Child mortality (4q1): the probability of dying between the first and the fifth birthdays •
Under-5 mortality (5q0): the probability of dying between birth and the fifth birthday

8.1 DATA QUALITY Estimates of infant and child mortality that are based on retrospective birth histories are subject to possible reporting errors that may adversely affect the quality of the data. The estimates may be affected by the completeness with which births and deaths are reported and recorded as well as the accuracy of information on current age and age at death for children who died. A lack of accurate information on the age at death may distort the age pattern of mortality. If age at death is misreported and the net effect of this age misreporting results in transference from one age bracket to another, it will bias the estimates. For example, a net transfer of deaths from an age of less than 1 month to a higher age will affect the estimates of neonatal and postneonatal mortality. To minimise errors in reporting age at death, interviewers were instructed to record age at death in days if the death took place in the month following the birth, in months if the child died before age 2, and in years if the child died at age 2 or older. Interviewers were also asked to probe for deaths reported at age 1 year to determine a more precise age at death in terms of months. Despite the emphasis during interviewer training and fieldwork monitoring on probing for accurate age at death, Appendix Table C.6 shows that, for the five years preceding the survey, there is considerable heaping of deaths at age 12 months, which is likely to lead to some underestimation of infant mortality. Another potential data quality problem is the selective omission from the birth histories of births that did not survive, which can lead to underestimation of mortality rates. When selective omission of childhood deaths occurs, it is usually most severe for deaths occurring early in infancy. One way that such omissions can be detected is by examining the proportion of infant deaths that are neonatal deaths. Generally, if there is substantial underreporting of deaths, the result is an abnormally low ratio of neonatal deaths to infant deaths. In the 2011 UDHS, the proportion of infant deaths occurring in the first month of life is 53 percent for the period zero to four years preceding the survey (Appendix Table C.6), which is within the normal range. Appendix Table C.5 shows death heaping at 7 and 14 days, which indicates rounding of age at death to one and two weeks, respectively. The age heaping at seven days leads to lower estimates of early neonatal mortality and perinatal mortality. However, it appears that early neonatal deaths among births that occurred in the first month of life have not been seriously underreported, since 76 percent of neonatal deaths were early neonatal deaths for the period zero to four years before the survey. Displacement of birth dates may distort mortality trends. This can occur if an interviewer knowingly records a death as occurring in an earlier year, which could happen if an interviewer were trying to cut down on the overall workload, because a lengthy set of additional questions must be asked about live births occurring during the five years preceding the interview. Appendix Table C.4 shows considerable year-of-birth transference for deceased children from 2006 to 2005, but relatively little transference for living children. This suggests that under-5 mortality is likely to be underestimated to some extent for the five-year period before the survey.

EARLY CHILDHOOD MORTALITY RATES: LEVELS AND TRENDS
Table 8.1 shows neonatal, postneonatal, infant, child, and under-five mortality rates for successive five-year periods before the survey. For the five years preceding the survey, the infant mortality rate was 54 per 1,000 live births. This implies that one in every 19 babies born in Uganda does not live to the first birthday. Those who survive to the first birthday, 38 out of 1,000 would die before reaching their fifth birthday. This shows that one in 11 children dies before their fifth birthday. The under-five mortality rate was 90 per 1,000 live births. The first month of life is associated with the highest risk to survival. As childhood mortality declines, postneonatal mortality usually declines faster than the neonatal mortality Infant and Child Mortality • 99 because neonatal mortality is frequently caused by biological factors that are not easily addressed by primary health care interventions. The neonatal and postneonatal mortality rates were 27 deaths per 1,000 live births, each. Results from the 2011 UDHS data show a remarkable decline in all levels of childhood mortality over the 15-year period preceding the survey. Infant mortality declined by 39 percent, from 89 deaths per 1,000 live births to 54 deaths per 1,000 live births. Furthermore, under-5 mortality declined by 37 percent over the same period, from 143 deaths per 1,000 live births to 90 deaths per 1,000 live births. As childhood mortality declines, postneonatal mortality usually declines faster than neonatal mortality because neonatal mortality is frequently caused by biological factors that are not easily addressed by primary care interventions. This is corroborated in the data: the neonatal and postneonatal mortality declined over the 15-year period preceding the survey by 21 percent and 50 percent, respectively. Mortality trends can also be examined by comparing data from UDHS surveys conducted in 2000-01, 2006, and 2011. Figure 8.1 shows improvement in all components of early childhood mortality rates. Under-5 mortality declined from 152 deaths per 1,000 live births in the 2000-01 UDHS to 90 in the 2011 UDHS, infant mortality declined from 88 deaths to 54 deaths per 1,000 live births, and postneonatal mortality declined from 55 deaths to 27 deaths per 1,000 live births during the same period. The change in neonatal mortality rate is not as pronounced; it declined from 33 deaths per 1,000 live births in 2000-01 to 29 deaths per 1,000 live births in 2006, and it declined only slightly to 27 deaths per 1,000 deaths in 2011. Figure 8.1 Trends in childhood mortality 33 55 88 69 152 29 46 76 67 137 27 27 54 38 90 Neonatal Postneonatal Infant Child Under-five 0 20 40 60 80 100 120 140 160 180 Deaths per 1,000 2000-01 2006 2011 Note: In the 2000-2001 UDHS, areas making up the districts of Amuru, Nwoya, Bundibugyo, Ntoroko, Gulu, Kasese, Kitgum, Lamwo, Agago, and Pader were excluded from the sample. These areas contained about 5 percent of the national population of Uganda. Thus, the trends need to be viewed in that light. Data refer to the 5 y

8.3 EARLY CHILDHOOD MORTALITY RATES BY SOCIOECONOMIC CHARACTERISTICS Table 8.2 shows differentials in childhood mortality by socioeconomic characteristics of the mother for the 10-year period preceding the survey. All childhood mortality rates, except neonatal mortality, are lower in urban than in rural areas. For example, the infant and under-5 mortality rates in rural areas are 66 and 111 deaths per 1,000 live births compared with 54 and 77 deaths per 1,000 live births, respectively, in urban areas. There are substantial regional variations in early childhood mortality rates. With the exception of neonatal mortality, Kampala, an entirely urban region with a higher socioeconomic status than the other regions, has the lowest childhood mortality rates when compared with other regions. The infant mortality rate ranges from a low of 47 deaths per 1,000 live births in Kampala to 87 and 88 deaths per 1,000 live births in Karamoja and West Nile, respectively. Similarly, the under-5 mortality is lowest in Kampala (65 deaths per 1,000 live births) and highest in Karamoja (153 deaths per 1,000 live births). As expected, the mother’s level of education is associated with the child’s probability for survival. Generally, children born to mothers with secondary or higher education have much lower childhood mortality rates when compared with children of uneducated mothers. For example, child mortality among children born to mothers with no education (59 deaths per 1,000 live births) is more than double that of children born to mothers with secondary or higher education (23 deaths per 1,000 live births). Similarly, the under-5 mortality among children born to uneducated mothers is 133 deaths per 1,000 live births compared with 79 deaths per 1,000 live births among children born to mothers with secondary or higher education. The only exception is neonatal mortality, where there is no clear pattern by mother’s education. With the exception of neonatal mortality, all other childhood mortality rates are highest among children in the lowest or second lowest wealth quintile and lowest among those in the wealthiest quintile. For example, under-5 mortality ranges from 72 deaths per 1,000 live births among the richest children to 125 deaths per 1,000 live births among children in the second lowest quintile. Table 8.2 Early childhood mortality rates by socioeconomic characteristics Neonatal, postneonatal, infant, child, and under-5 mortality rates for the 10-year period preceding the survey, by background characteristics, Uganda 2011 Background characteristic Neonatal mortality (NN) Postneonatal mortality (PNN)1 Infant mortality (1q0) Child mortality (4q1) Under-5 mortality (5q0) Residence Urban 31 23 54 25 77 Rural 30 36 66 47 111 Region Kampala 27 20 47 19 65 Central 1 44 31 75 37 109 Central 2 31 23 54 35 87 East Central 23 38 61 48 106 Eastern 24 23 47 41 87 Karamoja 29 59 87 72 153 North 31 35 66 42 105 West Nile 38 50 88 41 125 Western 30 38 68 52 116 Southwest 33 42 76 57 128 Mother's education No education 32 46 78 59 133 Primary 29 34 63 45 105 Secondary+ 33 24 57 23 79 Wealth quintile Lowest 26 50 76 52 123 Second 31 38 69 60 125 Middle 30 34 64 38 100 Fourth 33 30 63 44 104 Highest 34 14 48 25 72 1 Computed as the difference between the infant and neonatal mortality rates

8.4 EARLY CHILDHOOD MORTALITY BY DEMOGRAPHIC CHARACTERISTICS The demographic characteristics of both mothers and children play an important role in the survival probability of children. Table 8.3 presents childhood mortality rates by demographic characteristics (sex of the child, mother’s age at birth, birth order, previous birth interval, and the child’s size at birth). Table 8.3 shows that childhood mortality rates are consistently higher among male children than among their female counterparts. For example, the infant and under-5 mortality rates for males are 70 deaths and 114 deaths per 1,000 live births, respectively, compared with 59 deaths and 98 deaths per 1,000 live births, respectively, for females. Although there is no clear pattern in the variation of childhood mortality rates by mother’s age at birth, these rates are lowest among children whose mother’s age at birth was 20-29. Childhood mortality rates are highest among children of first and seventh or higher birth order. For example, under-5 mortality is 120 deaths and 134 deaths per 1,000 live births for children of the first and seventh or higher birth order compared with 90 to 98 deaths per 1,000 live births for other children. Short birth intervals (those less than two years) substantially reduce children’s chances of survival. For example, the infant mortality rate is 95 deaths per 1,000 live births for children born less than two years following a preceding birth compared with 46 to 49 deaths per 1,000 live births for children born after longer intervals. Children’s weight at birth is an important determinant of their survival. Because many births in Uganda occur at home and, as a result, children often are not weighed at birth, data on birth weight are available for only a few children. However, in the 2011 UDHS mothers were asked whether their child was very large, larger than average, average, smaller than average, or very small at birth, and the answer was used as a proxy for a child’s weight. Babies who were reported as smaller than average or very small at birth had higher mortality rates than those who were reported as average or larger at birth. The data show that 66 in 1,000 children who were reported as small or very small at birth died before reaching their first birthday compared with 50 deaths per 1,000 children who were reported as average or large. This differential is most pronounced for neonatal mortality (38 deaths per 1,000 live births for children born small or very small compared with 23 deaths per 1,000 live births for those born average or larger). Table 8.3 Early childhood mortality rates by demographic characteristics Neonatal, postneonatal, infant, child, and under-5 mortality rates for the 10-year period preceding the survey, by demographic characteristics, Uganda 2011 Demographic characteristic Neonatal mortality (NN) Postneonatal mortality (PNN)1 Infant mortality (1q0) Child mortality (4q1) Under-5 mortality (5q0) Child's sex Male 34 36 70 48 114 Female 27 33 59 41 98 Mother's age at birth