Measuring effective coverage of measles immunization in low-resource settings: A method and application from the baseline evaluation of Salud Mesoamérica 2015
Colson, Katherine Ellicott
MetadataShow full item record
Background: Estimating vaccination coverage is challenging in resource-poor settings where accurate records are sparse. Household surveys, a key source of coverage information, typically capture data from child health cards and rely on maternal recall when cards are unavailable. Little is known about the accuracy of these sources in low- and middle-income countries. Moreover, such metrics provide estimates of crude coverage (vaccination), but do not capture the gap between crude and effective coverage (seroconversion). In this thesis, I harness data from the baseline evaluation of Salud Mesoamérica 2015 to compare survey-based estimates of measles immunization coverage to the seroprevalence of anti-measles antibodies in dried blood spots (DBS) collected from children in poor areas of Mexico and Nicaragua. This thesis addresses gaps in knowledge regarding the feasibility and validity of using DBS methods in low-resource settings, the accuracy of survey-based measures of vaccination coverage, and the prevalence of biological protection against measles (effective coverage) in the study areas. Methods: Data for this study were collected as part of the baseline evaluation of Salud Mesoamérica 2015. Among the poorest 20% of the population in Mexico and Nicaragua, probability samples of 5,428 and 2,071 households (respectively) with resident women of reproductive age (15 to 49 years) or children under 5 years of age were selected for interviews. Surveys addressed maternal and child health topics such as antenatal and delivery care, immunization, child nutrition, and recent illness. During home-based interviews, consent was obtained to collect DBS, from children aged 12 to 23 months. ELISA methods for measuring the seroprevalence of measles immunoglobulin G (IgG) antibodies in dried blood spots were validated by comparing test results for matched serum and DBS samples. After establishing the validity of the method, I compare the levels and characteristics of measles vaccination coverage estimates based on maternal recall, health card documentation, and antibody presence. I assess how survey-based estimates of measles immunization coverage differ from serology-based estimates, and identify individual, maternal, household, community and health facility characteristics of children that have health card documentation of receiving measles immunization but who lack antibodies. Findings: In both countries, survey-based sources of measles vaccination coverage were significantly higher (83% in Mexico and 85% in Nicaragua) than the seroprevalence of protective antibodies (68% in Mexico and 50% in Nicaragua). A large proportion of children in both settings (19% in Mexico and 43% in Nicaragua) had health card documentation of having received measles vaccine, but lacked antibodies. These discrepancies were geographically concentrated in several specific areas, particularly the North Atlantic Region of Nicaragua, suggesting that these differences are driven by variations in health service delivery rather than individual or household characteristics. In multivariate regression analysis, few factors aside from geography significantly increased the likelihood that a health card-positive child would lack antibodies. Clear and consistent geographic patterns in both crude and effective coverage highlight that some surveyed areas are very high-performing while others are experiencing substantial challenges. Conclusions: This study shows that using health cards and/or caregiver recall to estimate population protection against measles is unwise, because these sources overestimate effective coverage of measles immunization. Results from this study are of great importance as they point the areas and groups where protection against measles may have been wrongly assumed. These findings will enable the health authorities in Chiapas and Mexico to examine and address the reasons for the differences between crude and effective coverage.
- Global health