INFANT AND CHILD MORTALITY 8

This chapter presents estimates of childhood mortality, measured by perinatal, infant, and child
mortality rates. The chapter includes a description of the indicators, the current levels and trends, differen-
tials by selected background characteristics, and factors that contribute to elevating children’s mortality
risks.

Analyzing the levels and trends of childhood mortality is important in gauging the impact of ma-
ternal and child health programs improving the health of infants and children. The data can also be used
as input in population projections. Differentials in childhood mortality by selected characteristics are use-
ful in identifying groups in need for priority attention and in planning meaningful strategies to address
them.

8.1 DEFINITIONS AND ASSESSMENT OF DATA QUALITY

Six indicators are analyzed in this chapter: Infant mortality rate, neonatal mortality rate, postneo-
natal mortality rate, child mortality rate, under-five mortality rate, and perinatal mortality rate.

The infant mortality rate (IMR) is defined as the number of infant deaths per 1,000 live births
during the first 12 months of life. It is described as the probability of dying between birth and age one
year. Because the level of mortality is higher at the early ages of infancy than at the later ages of infancy,
it is useful to break up the rate into neonatal mortality (NN), the probability of dying within the first
month of life, and postneonatal mortality (PNN), the probability of dying after the first month of life but
before age one year.

The child mortality rate is the probability of dying between exact age one and age five, defined as
the number of deaths of children age 1-4 years per 1,000 children surviving to age 12 months. Another
useful summary indicator of infant and child mortality is the under-five mortality rate or the probability of
dying between birth and exact age five, defined as the number of deaths below age five per 1,000 live
births during the given period.

In this report, the five childhood mortality estimates pertain to periods of 0 to 4, 5 to 9, and 10 to
14 years preceding the survey.

Perinatal mortality is also presented in this chapter. It is defined as the number of stillbirths and
early neonatal deaths that occurred zero to four years preceding the survey per 1,000 pregnancies of seven
or more months’ duration.

The above mortality rates are estimated directly from the information derived from the questions
asked in the reproductive history section of the Women’s Questionnaire. There are two types of data col-
lected in this section. The first refers to a woman’s total number of pregnancies, classified as live births
and non-live births. To elicit complete reporting of all live births, interviewers asked the respondents to
report the number of children still living and those who died, each classified by sex. The second type of
data relates to detailed information on each of the woman’s pregnancies from the first to the last. The fol-
lowing information was collected: whether the pregnancy resulted in single or multiple births and the out-
come of the pregnancy (born alive, born dead, or lost before full term).

Infant and Child Mortality | 107

 For all live births, the name, sex, date of birth, and survival status of the child were recorded. For
surviving children, their age in completed years at last birthday was recorded. For dead children, the age
at death was noted. If the child was born dead (stillbirth) or the pregnancy was lost before term (miscar-
riage), the date of pregnancy termination and duration of pregnancy at the time of loss were also recorded.
For these pregnancies, the women were asked whether the pregnancy loss was induced.

The accuracy of these estimates depends on the respondent’s full recall about all of her births,
particularly those who have died, and her ability to accurately report the children’s date of birth and age at
death. Table C.4 in Appendix C shows that there are no substantial differentials in the distributions of
reported birth dates between living and dead children. The percentage of births in 1997 to 2002 with
complete birth dates for living and dead children is equal to 100 percent and close to 100 percent for the
earlier period. Moreover, a close examination of the pattern of reporting of age at death (Table C.6 in
Appendix C) reveals that for deaths reported to have occurred zero to four years preceding the survey,
there is no evidence of substantial heaping of age at death at any age. For deaths 5 to 9 and 10 to 14 years
before the survey, there is a heaping at age 12 months. However, it is too small to have an impact on the
estimated IMR. The reporting of date of birth and of age at death zero to four years before the survey is
reasonable. However, deaths reported in the earlier periods before the survey should be accepted with
some reservation.

8.2 LEVELS AND TRENDS IN INFANT, CHILD, AND UNDER-FIVE MORTALITY

Table 8.1 shows early childhood mortality rates based on the 2003 National Demographic and
Health Survey (NDHS). The IMR during the five-year period before the survey, which centers in year
2000, is 29 deaths per 1,000 live births. The neonatal mortality rate for the same period is 17 deaths per
1,000 live births, and the postneonatal mortality rate is 12 deaths per 1,000 live births. Child mortality is
12 deaths per 1,000 and the under-five mortality rate is 40 deaths per 1,000 live births.

Table 8.1 Early childhood mortality rates

Neonatal, postneonatal, infant, child, and under-five mortality rates for five-year periods preceding the
survey, Philippines 2003

Years Neonatal Postneonatal Infant Child Under-five

preceding Approximate mortality mortality mortality mortality mortality
the survey calendar years (NN) (PNN)1 (1q0) (4q1) (5q0)
0-4 1998-2003 17 12 29 12 40
5-9 1993-1997 17 14 31 12 43
10-14 1988-1992 18 16 34 19 52
1
Computed as the difference between the infant and neonatal mortality rates

These estimates are associated with sampling errors. For example, at 95 percent confidence lim-
its, for the zero to four years preceding the survey, the actual IMR lies between 24 and 33 deaths per
1,000 live births (Appendix B).

The 2003 NDHS data confirm the pattern of decline in childhood mortality in the past 15 years.
Under-five mortality has declined from 54 to 48 to 40 per 1,000 as reported in the 1993 NDS, the 1998
NDHS, and the 2003 NDHS, respectively. Infant mortality rates from the three surveys also show a de-
cline, from 34 to 35 to 29.

108 | Infant and Child Mortality

8.3 SOCIOECONOMIC DIFFERENTIALS IN CHILDHOOD MORTALITY

Childhood mortality varies according to residence, education, and socioeconomic status. Data in
Table 8.2 show that mortality rates in urban areas are much lower than those in rural areas. For example,
the IMR in urban areas is 24 deaths per 1,000 live births, compared with 36 deaths per 1,000 live births in
rural areas. Childhood mortality is inversely related to the mother’s education level and wealth status.
The IMR for children whose mothers have no education is 65 deaths per 1,000 live births, compared with
15 deaths per 1,000 live births for children whose mothers have college or higher education. The IMR is
higher than the national average in seven regions: MIMAROPA, Western Visayas, Eastern Visayas,
Northern Mindanao, Davao, Caraga, and Autonomous Region in Muslim Mindanao (ARMM). While
there seems to be substantial differentials in childhood mortality by region, the large sampling errors (ex-
ceeding 20 points per 1,000 in some regions) suggest that the observed differences should be used with
caution.

Table 8.2 Early childhood mortality rates by socioeconomic characteristics and region
Neonatal, postneonatal, infant, child, and under-five mortality rates for the 10-year period preced-
ing the survey, by background characteristic, Philippines 2003

Neonatal Postneonatal Infant Child Under-five

Background mortality mortality mortality mortality mortality
characteristic (NN) (PNN)1 (1q0) (4q1) (5q0)
Residence
Urban 14 10 24 7 30
Rural 21 15 36 17 52

Region
National Capital Region 15 9 24 8 31
Cordillera Admin Region 6 8 14 20 34
I - Ilocos 19 10 29 11 39
II - Cagayan Valley 17 10 28 8 35
III - Central Luzon 15 10 25 6 31
IVA - CALABARZON 17 9 25 6 31
IVB - MIMAROPA 18 26 44 25 68
V - Bicol 19 9 28 15 43
VI - Western Visayas 22 18 39 11 50
VII - Central Visayas 18 10 28 11 39
VIII - Eastern Visayas 24 12 36 22 57
IX - Zamboanga Peninsula 6 21 27 17 43
X - Northern Mindanao 24 15 38 11 49
XI - Davao 18 20 38 10 47
XII - SOCCSKSARGEN 15 13 27 10 37
XIII -Caraga 21 14 35 14 49
ARMM 18 23 41 33 72

Education
No education (33) 32 65 42 105
Elementary 22 21 43 20 62
High school 18 9 26 9 35
College or higher 9 7 15 3 18

Wealth index quintile

Lowest 21 21 42 25 66
Second 19 13 32 15 47
Middle 15 10 26 6 32
Fourth 15 7 22 4 26
Highest 13 6 19 1 21

Total 17 13 30 12 42
Note: Figures in parentheses are based on 250-499 unweighted exposed persons.
1
Computed as the difference between the infant and neonatal mortality rates

Infant and Child Mortality | 109

 8.4 BIODEMOGRAPHIC DIFFERENTIALS IN CHILDHOOD MORTALITY

Table 8.3 presents early childhood mortality by demographic characteristics. As expected, the
mortality rate is consistently higher for males than for females. For instance, the IMR for males is 35
deaths per 1,000 live births, compared with 25 deaths per 1,000 live births for females. Mother’s age at
birth can affect a child’s chances of survival. The table shows that early childhood mortality rates exhibit
the expected U-shaped relationship with the mother’s age: high at young ages, low at middle ages, and
high at old ages (Figure 8.1). The higher rates for younger and older women may be related to biological
factors that lead to complications during pregnancy and delivery.

Table 8.3 Early childhood mortality rates by demographic characteristics

Neonatal, postneonatal, infant, child, and under-five mortality rates for the ten-year period
preceding the survey, by demographic characteristics, Philippines 2003

Neonatal Postneonatal Infant Child Under-five

Demographic mortality mortality mortality mortality mortality
characteristic (NN) (PNN)1 (1q0) (4q1) (5q0)
Child's sex
Male 21 14 35 14 48
Female 13 12 25 9 34

Mother's age at birth

<20 28 14 42 15 56
20-29 16 11 26 9 36
30-39 15 14 28 15 43
40-49 32 34 66 24 89

Birth order
1 19 11 29 7 36
2-3 14 9 23 8 31
4-6 14 16 29 16 45
7+ 31 25 56 29 83

Previous birth interval2

<2 23 16 39 20 58
2 years 10 16 26 13 38
3 years 10 9 19 11 30
4+ years 15 10 25 6 31

Birth size3
Small/very small 29 22 52 na na
Average or larger 11 9 20 na na

na = Not applicable
1
Computed as the difference between the infant and neonatal mortality rates
2
Excludes first-order births
3
Rates for the five-year period before the survey

110 | Infant and Child Mortality

 Figure 8.1 Under Five Mortality Rates by Background
Characteristics

RESIDENCE
Urban 30
Rural 52

MOTHER'S EDUCATION
No education 105
Elemenatry 62
High school 35
College or higher 18

MOTHER'S AGE AT BIRTH

<20 56
20-29 36
30-39 43
40-49 89
0 20 40 60 80 100 120

NDHS 2003

The 2003 NDHS results show that there is no clear pattern for neonatal and infant mortality rates
by birth order: The IMR is lowest for second- and third-order births (Table 8.3). However, there is a clear
positive association between birth order and the probability of dying between ages one and five: Higher
order births have higher mortality risks. While the child mortality rate for first-order births is 7 deaths per
1,000, the corresponding rate for births of seventh order or higher is 29 deaths per 1,000.

In general, childhood mortality rates decline as the birth interval increases. For example, the IMR
for children born less than two years after a previous birth is 39 deaths per 1,000 live births, compared
with 25 deaths for children born after an interval of four or more years. Children born three years after a
preceding birth have the best chance of surviving infancy, with an IMR of 19 deaths per 1,000 live births.

A child’s size at birth has been shown to be associated with the risk of dying during infancy, par-
ticularly during the first months of life. For all children born during the five-year period before the sur-
vey, mothers were asked about their perception of the child’s size: whether the child was very small,
small, average size, large, or very large at birth. Although subjective, the mother’s judgment has been
shown to correlate closely with the actual birth weight. The 2003 NDHS results confirm that mortality
levels are higher for children perceived by the mother to have been small or very small at birth than
among other children. The neonatatal mortality rate for infants who were judged to be small or very small
at birth by their mothers are, for example, more than two times higher than that for infants who were re-
ported to be average or larger at birth (29 and 11 deaths per 1,000 live births, respectively).

8.5 DIFFERENTIALS IN CHILDHOOD MORTALITY BY WOMEN’S STATUS

Although there is no direct association, women’s status has been found to influence infant and
child mortality rates through women’s ability to control resources and make decisions. In the 2003
NDHS, women were asked about certain aspects of their autonomy, including the number of household
decisions in which the woman participates, the number of reasons for which a woman feels a wife is justi-
fied in refusing sexual relations with her husband, and the number of reasons that justify wife beating.
With regard to participation in household decisions, the following question was asked: “Who in your
family usually has the final say on the following decisions?” Decisions about which women were asked
include deciding about her own health care, making large household purchases, making household pur-

Infant and Child Mortality | 111

 chases for daily needs, visits to family or relatives, and what food should be cooked each day. A woman
is considered more independent if she participates in a larger number of household decisions.

The second measure is the number of reasons to refuse sex with her husband. The following was
presented to the respondents: “Husbands and wives do not always agree on everything. Please tell me if
you think a wife is justified in refusing to have sex with her husband when: she knows her husband has a
sexually transmitted disease, she knows her husband has sex with other women, she has recently given
birth, and she is tired or not in the mood.” A woman is considered more independent if she agrees with a
greater number of reasons for a woman to refuse sex.

The third indicator is the number of reasons wife beating is justified. The respondent was asked
the following: “Sometimes a husband is annoyed or angered by things that his wife does. In your opin-
ion, is a husband justified in hitting or beating his wife in the following situations: if she goes out without
telling him, if she neglects the children, if she argues with him, if she refuses to have sex with him, and if
she burns the food.” A woman is considered less independent if she accepts more reasons for justifying
wife beating.

Table 8.4 presents childhood mortality rates by women’s status indicators. On the basis of the
three indicators, there is no clear relationship between women’s status and childhood mortality. However,
the decisionmaking indicator has a weak relationship to under-five mortality, such that the children of
women at the extremes of decisionmaking—that is, women who make no decisions and women who par-
ticipate in all decisions—are at higher risk for under-five mortality. Women who cannot make decisions
are likely to be less empowered to make independent childcare decisions.

Table 8.4 Early childhood mortality rates by women's status

Neonatal, postneonatal, infant, child, and under-five mortality rates for the ten-year period preceding
the survey, by women's status indicators, Philippines 2003

Neonatal Postneonatal Infant Child Under-five

Women's status mortality mortality mortality mortality mortality
indicators (NN) (PNN)1 (1q0) (4q1) (5q0)
Number of decisions in which
woman has final say2
0 * 13 13 29 42
1-2 15 11 25 9 35
3-4 17 14 31 8 38
5 18 13 31 13 44

Number of reasons to refuse sex

with husband
0 (16) 9 25 15 40
1-2 36 20 56 9 64
3-4 17 13 29 12 41

Number of reasons wife beating

is justified
0 16 10 27 11 38
1-2 21 20 41 13 54
3-4 20 9 29 20 48
5 * 20 24 8 32

Note: Figures in parentheses are based on 250-499 unweighted exposed persons. An asterisk indicates
that an estimate is based on fewer than 250 unweighted exposed persons and has been suppressed.
1
Computed as the difference between the infant and neonatal mortality rates
2
Either by herself or jointly with others

112 | Infant and Child Mortality

8.6 PERINATAL MORTALITY
Table 8.5 Perinatal mortality
In the 2003 NDHS, women were Number of stillbirths and early neonatal deaths, and the perinatal mortality
asked to report all pregnancy losses in the rate for the five-year period preceding the survey, by background character-
istics, Philippines 2003
five years before the survey. For each such ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
pregnancy, the duration of pregnancy was Number of
Number pregnancies
recorded. In this report, perinatal deaths Number of early Perinatal of 7+
include pregnancy losses occurring after Background of neonatal mortality months’
1 2 3
characteristic stillbiirths deaths rate duration
seven completed months of gestation (still- ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
births) and deaths to live births within the Age at birth
<20 9 15 38 649
first seven days of life (early neonatal 20-29 30 39 18 3,840
deaths). Thus, the perinatal mortality rate 30-39 24 27 23 2,187
40-49 13 11 68 355
is the sum of the number of stillbirths and
early neonatal deaths divided by the num- Previous pregnancy
interval in months
ber of pregnancies of seven or more First pregnancy 16 26 23 1,837
months’ duration. The distinction between <15 9 16 35 724
15-26 15 15 18 1,740
a stillbirth and an early neonatal death may 27-38 18 13 28 1,102
be a fine one, depending often on the ob- 39+ 17 22 24 1,627
served presence or absence of some faint Residence
signs of life after delivery. The causes of Urban 33 39 21 3,494
Rural 43 53 27 3,536
stillbirths and early neonatal deaths often
overlap, and examining just one or the Region
National Capital Region 8 13 19 1,058
other can understate the true level of mor- Cordillera Admin Region 2 1 23 117
tality around delivery. For this reason, in I - Ilocos 4 7 36 314
II - Cagayan Valley 3 0 14 228
this report, both event types are combined III - Central Luzon 6 8 20 689
and examined together. As with other IVA - CALABARZON 11 10 25 827
IVB - MIMAROPA 2 3 21 243
childhood mortality measures, sampling V - Bicol 7 5 26 439
errors associated with perinatal mortality VI - Western Visayas 11 10 43 493
VII - Central Visayas 2 4 13 537
rates are very large. VIII - Eastern Visayas 2 7 24 359
IX - Zamboanga Peninsula 3 3 21 279
X - Northern Mindanao 3 5 24 304
Data in Table 8.5 show that, over- XI - Davao 3 4 24 298
all, 76 stillbirths and 92 early neonatal XII - SOCCSKSARGEN 3 4 21 326
XIII - Caraga 4 3 35 207
deaths were recorded in the survey, ARMM 3 6 28 313
resulting in a perinatal mortality rate in the Education
Philippines of 24 deaths per 1,000 preg- No education 2 4 45 135
nancies. Perinatal mortality is slightly Elementary 25 35 29 2,065
High school 32 38 23 2,984
higher in rural than in urban areas (27 College or higher 17 15 17 1,847
compared with 21 deaths per 1,000 preg- Wealth index quintile
nancies). Perinatal mortality rates are Lowest 18 28 25 1,876
negatively and monotonically associated Second 17 23 25 1,607
Middle 22 13 26 1,374
with the mother’s education, ranging from Fourth 14 20 29 1,176
45 deaths for women with no education to Highest 4 7 11 997
17 deaths per 1,000 pregnancies for those Total 76 92 24 7,030
with college or higher education. Varia- ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
1
Stillbirths are fetal deaths in pregnancies lasting seven or more months.
tions across mother’s age at birth are the 2
Early neonatal deaths are deaths at age zero to six days among live-born
same as those for childhood mortality, children.
3
The sum of the number of stillbirths and early neonatal deaths divided by
with the lowest levels of mortality at a ma- the number of pregnancies of seven or more months' duration, multiplied by
ternal age of 20-29. 1,000.

Perinatal mortality is highest among pregnancies with a preceding birth interval of less than 15
months. Apart from the wealthiest group demonstrating the least perinatal mortality, no substantial differ-
entials are evident across wealth status.

Infant and Child Mortality | 113

 8.7 HIGH-RISK FERTILITY BEHAVIOR

Maternal fertility patterns and children’s survival risks have been known to have a strong rela-
tionship. Generally, infants and children have a greater probability of dying if they are born to mothers
who are too young or too old, if they are born after a short birth interval, or if they are of high birth order.
These factors are of particular interest since they are easily avoidable at low or no cost.

For purposes of the analysis of high-risk fertility behavior presented in Table 8.6, a mother is
classified as too young if she is less than 18 years of age and too old if she is over 34 years of age at the
time of delivery. A short birth interval is defined as a birth occurring less than 24 months after the previ-
ous birth, and a child is of high birth order if the mother had previously given birth to three or more chil-
dren (i.e., if the child is of birth order four or higher). Although first births are commonly associated with
high mortality risk, even if they occur when the mother is between 18 and 34 years old, they are not in-
cluded in the high-risk category unless they occurred too early or late; instead, they are considered un-
avoidable.

Table 8.6 presents the percentage of births in the five years preceding the survey and the percent-
age of currently married women according to the level of risk category in relation to the fertility behavior
of the mother. The table also presents the risk ratio or the relative risk of dying estimated by comparing
the proportion dead among births in a specific high-risk category with the proportion dead among births
not in any high-risk category. This information is useful for designing and monitoring programs both to
avoid high-risk behavior and to cope with elevated risks.

Data in Table 8.6 show the percentage of births occurring in the five years before the survey that
fall into these various risk categories. Twenty-one percent of children in the Philippines are not in any
high-risk category, and 25 percent have an elevated mortality risk that is unavoidable (first births for
which any risk is considered unavoidable). Among those who are at risk, 34 percent of births are in only
one of the high-risk categories, while 20 percent are in multiple high-risk categories (because of a combi-
nation of mother’s age, birth interval, and birth order).

The single avoidable high-risk categories with the highest percentage of births are birth order
higher than three (15 percent) and births with intervals of less than 24 months (14 percent). Compared
with births with no elevated mortality risk, the mortality risk increase associated with these categories
(1.97 and 1.34, respectively) is less than that for births to mothers who were under 18 years at birth
(3.41). Mortality risks are most elevated for the single-risk categories of too young mothers and too old
mothers: 3 percent of births fell in the latter category, while 2 percent of births fell into the former.

The multiple high-risk category with the largest proportion of births is high-order births to older
mothers: 10 percent of children fall in this category. Compared with births with no elevated risk, these
births are 2.5 times more likely to die in early childhood. The multiple high-risk category with the high-
est risk ratio is the combination of mothers giving birth after age 34, birth interval less than 24 months,
and birth order higher than three; the 3 percent of children in this category are more than five times as
likely to die as children with no elevated mortality risk.

114 | Infant and Child Mortality

 Table 8.6 also presents the distribution of currently married women according to category of in-
creased risk if they were to conceive at the time of the survey. Although many women are protected from
conception through the use of family planning, postpartum insusceptibility, and prolonged abstinence, for
simplicity, only those who have been sterilized are considered to be in the no-risk category solely on the
basis of their contraceptive method. Two in three currently married women (66 percent) are at risk of
conceiving a child with an elevated risk of dying; 30 percent of women are at risk because of a single
high-risk factor, while 36 percent of women have multiple high-risk factors. The most common risk is
late childbearing combined with high birth order (27 percent of currently married women).

Table 8.6 High-risk fertility behavior

Percent distribution of children born in the five years preceding the survey by
category of elevated risk of mortality and the risk ratio, and percent distribu-
tion of currently married women by category of risk if they were to conceive a
child at the time of the survey, Philippines 2003
Births in the 5 years Percentage of
preceding the survey currently
Percentage married
Risk category of births Risk ratio women1
Not in any high-risk category 21.4 1.00 28.12

Unavoidable risk category

First order births between ages 18
and 34 years 25.1 1.60 5.7

Single high-risk category

Mother’s age <18 2.2 3.41 0.3
Mother’s age >34 3.1 2.18 10.9
Birth interval <24 months 13.8 1.34 8.7
Birth order >3 14.5 1.97 9.9

Subtotal 33.7 1.83 29.7

Multiple high-risk category

Age <18 and birth interval <24
months 3 0.2 4.06 0.1
Age >34 and birth interval <24
months 0.4 0.00 0.6
Age >34 and birth order >3 10.1 2.48 27.2
Age >34 and birth interval <24
months and birth order >3 2.5 5.10 3.1
Birth interval <24 months and
birth order >3 6.6 3.47 5.5

Subtotal 19.8 3.11 36.4

In any avoidable high-risk category 53.5 2.30 66.1

Total 100.0 - 100.0

Number of births 6,954 - 8,671

Note: Risk ratio is the ratio of the proportion dead among births in a specific
high-risk category to the proportion dead among births not in any high-risk
category.
1
Women are assigned to risk categories according to the status they would
have at the birth of a child if they were to conceive at the time of the survey:
current age less than 17 years and 3 months or older than 34 years and 2
months, latest birth less than 15 months ago, or latest birth being of order 3 or
higher.
2
Includes sterilized women
3
Includes the category age <18 and birth order >3

Infant and Child Mortality | 115