Background: Achieving complete child immunization is a rare phenomenon in developing
countries, however, even in the face of abject poverty and extreme deprivations, two regions in
Ghana (PDRs) have thrived to do so. Effectively addressing variations in coverage among the
regions, to inform immunization strategies in Ghana and elsewhere, requires understanding of
such regions.
Objective: Comparing the PDRs with each other, and with the other eight regions, the study
investigated if there were regional-level characteristics that were unique to the PDRs, and that
differentiated them from the other eight regions.
Methods: With women response rate of 96.5%, a two-stage design was adopted. The first stage
was analysis of data on 542 children (12-23months), to calculate the dependent variable
“complete/incomplete immunization” and then to compute the rate of complete immunization for
each region. The second stage was cluster analysis of the regional-level factors, with the
dependent variable.
Results: Findings showed that besides complete child immunization rates of over 90%, the PDRs
were also unique on frequent use of maternal health facilities, and on high maternal health care
decision. However, exploratory hierarchical cluster analysis, applying squared Euclidean
Distance, as the similarity measure, revealed not homogeneity, but great heterogeneity, in the
socio-demographic composition and health care services availability-and-use characteristics of
the PDRs.
Conclusions: With evidence from the PDRs, it concluded that even with the poorest of all
regions, with very disparate characteristics, it is possible to achieve high immunization
coverage. Nevertheless, the GDHS data do not help to develop hypotheses to explain why the
PDRs are positive deviant.