PNAS – Proceedings of the National Academy of Sciences of the United States
(Accessed 22 February 2014)
Distance to health services affects local-level vaccine efficacy for pneumococcal conjugate vaccine (PCV) among rural Filipino children
Elisabeth Dowling Roota,1, Marilla Lucerob, Hanna Nohynekc, Peter Anthamattend, Deborah S. K. Thomasd, Veronica Tallob, Antti Tanskanenc,e, Beatriz P. Quiambaob, Taneli Puumalainenf,
Socorro P. Lupisanb, Petri Ruutug, Erma Ladesmab, Gail M. Williamsh, Ian Rileyh, and Eric A. F. Simõesi
Edited by Burton H. Singer, University of Florida, Gainesville, FL, and approved January 21, 2014 (received for review July 22, 2013)
Although pneumococcal conjugate vaccines (PCVs) are widely available in industrialized nations, the cost of these vaccines and the strategy of universal vaccination of infants, as endorsed by the World Health Organization, are daunting obstacles to the adoption of these vaccines in developing countries. Using spatial epidemiological methods to examine the spatial variation in vaccine efficacy (VE) in an 11-valent PCV trial in Bohol, Philippines, we suggest an alternative strategy to universal vaccination. Our main finding suggests that areas with poor access to healthcare have the highest VE. An alternative vaccination strategy could target vaccination to areas where children are most likely to benefit, rather than focus on nationwide immunization.
Pneumococcal conjugate vaccines (PCVs) have demonstrated efficacy against childhood pneumococcal disease in several regions globally. We demonstrate how spatial epidemiological analysis of a PCV trial can assist in developing vaccination strategies that target specific geographic subpopulations at greater risk for pneumococcal pneumonia. We conducted a secondary analysis of a randomized, placebo-controlled, double-blind vaccine trial that examined the efficacy of an 11-valent PCV among children less than 2 y of age in Bohol, Philippines. Trial data were linked to the residential location of each participant using a geographic information system. We use spatial interpolation methods to create smoothed surface maps of vaccination rates and local-level vaccine efficacy across the study area. We then measure the relationship between distance to the main study hospital and local-level vaccine efficacy, controlling for ecological factors, using spatial autoregressive models with spatial autoregressive disturbances. We find a significant amount of spatial variation in vaccination rates across the study area. For the primary study endpoint vaccine efficacy increased with distance from the main study hospital from −14% for children living less than 1.5 km from Bohol Regional Hospital (BRH) to 55% for children living greater than 8.5 km from BRH. Spatial regression models indicated that after adjustment for ecological factors, distance to the main study hospital was positively related to vaccine efficacy, increasing at a rate of 4.5% per kilometer distance. Because areas with poor access to care have significantly higher VE, targeted vaccination of children in these areas might allow for a more effective implementation of global programs.