Introduction
Pakistan has one of the highest infant mortality rates at 78 deaths per 1000 live births per year;
one in every eleven children born in Pakistan dies before his or her fifth birthday. Timely and
complete vaccination is a proven and highly cost effective public health intervention to reduce
death and disability, especially in children. Among other causes, vaccine preventable infectious
diseases are a significant contributor to child mortality in Pakistan.
Study methods
We assessed the association of contextual and individual level characteristics, using Pakistan
Demographic and Health Survey (PDHS) data, with the initiation, completion and up to date
(UTD) immunization for children 12 to 23 months old in Pakistan. The contextual level
characteristics included province of residence, rural or urban place of residence and whether the
child was issued a health card by the Expanded Program of Immunization (EPI) staff. The
individual level determinants included the household wealth index, gender of the head of the
household, maternal age, maternal education and ethnicity, gender and birth order of child and
number of children under five years of age in the household. Design based analysis was carried
out to account for the complex survey design of the Pakistan DHS 2006 and 1990 datasets. Using
logistic regression analysis, we assessed the independent association of individual, household and
contextual variables with initiation and completion of vaccination for the routine immunization
schedule of children aged 12 to 23 months. Following unadjusted analysis, the independent effect
estimates were assessed along with biological plausibility to select variables for adjusted
multivariable logistic regression analysis. Confounding of variables was measured through
stratified analysis. Interaction terms were included in the final model to assess their interactive
effect but there were no statistically significant interactive effects of the variables in the final
iv
models. Chi-square goodness of fit was used to compare appropriateness between different
models.
Results
The results of the analysis of the PDHS 2006 showed the household wealth index and maternal
education as statistically significant factors with vaccination initiation. Children in the richest
household wealth index were more likely to initiate vaccination compared to children in the
poorest wealth quintile. Similarly, children whose mothers had higher education were more likely
to initiate vaccination. In 1990, maternal education was significantly associated with vaccination
initiation but mothers with primary education did not behave differently from mothers with no
education. Also household wealth was statistically significant; children in the richest wealth
quintile had ten times higher odds of vaccination initiation compared to children in the poorest
quintile. In 1990 province of residence was a significant predictor of vaccination initiation only in
the province of Sindh.
In the PDHS 2006 data analysis, having a vaccination card and mothers’ education were
statistically significantly associated with vaccination schedule completion. In the PDHS 1990
analysis, the wealth index and maternal age were found to have a significant association with
vaccination schedule completion. Between 1990 and 2006 immunization coverage improved
significantly only among the middle wealth quintile with little improvement in households in the
poorest and poor wealth quintiles. Having a health/immunization card was significantly
associated with vaccination completion in both 1990 and 2006 and it was the only predictor of up
to date immunization in the PDHS 2006 analysis. The children for whom the card was seen at the
time of the interview were 21 times more likely to be UTD compared to children who were never
issued a health card. The interaction terms of possession of health/immunization card, wealth
index, and maternal education were not statistically significant.
v
Discussion
Although position in a higher wealth quintile and mothers’ higher education are individual level
characteristics, these associations may also suggest that the health care system is geared to cater
to the needs of the population with better access to sources of income and human development.
The government should increase its efforts to reach marginal populations with limited financial
resources and access to social services like education, better transport and mass media. Once the
health care system aims for equitable access of health care services, then even children in lower
wealth quintiles are more likely to complete their vaccination schedule.
Increasing female literacy is crucial for improving the health of the population and thus reducing
infant and under-five mortality. The government has strengthened its efforts to increase female
enrollment and keep girls in school beyond primary level but it also needs to focus on the quality
of education provided to these girls. The odds of immunization initiation and completion were
similar for mothers with up to primary education and those without a formal education; therefore
while girls are enrolled in primary school every effort should be made to educate them about
health and benefits of preventive health services. It is recommended that the primary school
curriculum needs to be updated to address the significance and effect of immunization.
Possession of a health card and seen at the time of the interview is a measure of maternal attitudes
and understanding towards the importance of immunization completion but it is also the single
most significant factor predicting vaccination completion. But in our sample only 13% of the
mothers in 2006 were able to show the household held immunization record at the time of the
interview and a quarter of children initiating the immunization schedule were never issued an
immunization card according to the mothers’ reports. The EPI program must invest in
digitalization of immunization and health records for children. The importance of a health card in
vaccination schedule completion has been documented in other parts of the world, including
Pakistan. Therefore until digital immunization records are widespread, the EPI program must
vi
make provision of waterproof, tear resistant and easily stored immunization cards to each and
every child initiating the schedule. Also cards that have a better system of reminding mothers of
the vaccine due dates are highly recommended to improve immunization completion, especially
among mothers who do not have access to electronic reminder systems like cell phones and those
who lack education. The EPI program needs to evaluate its performance on more stringent
criteria. Instead of just reporting DTP 1 and 3 among children 12-23 months of age, it should also
monitor Up to date (UTD) immunization status of children 52 weeks of age. Although
vaccination and immunization programs are in their fourth decade of global introduction, there
still is a need to create a social value for immunization to reap the maximum benefits of the
vaccines on the health of children and the general population