Spatiotemporal Analysis of Infant Measles Using Population Attributable Risk in Shandong Province, 1999–2008

BACKGROUND Reduction of measles incidence and mortality has been encouraging in China. However, it remains an important public health concern among infants. This study aimed to examine the space-time distribution pattern of infant measles occurrence for the period of 1999-2008 in Shandong, China. METHODS AND FINDINGS Measles cases among infants aged younger than 1 year were obtained from the national infectious diseases reporting information system. A spatiotemporal analysis using population attributable risk percent (PAR%) was used to distinguish between multiple geographic clusters of potential interest. The analysis detected 29 statistically significant space-time clusters with the most likely cluster in Zaozhuang City from 2006 to 2008. Of the 28 secondary clusters, 22 were found in 2008. The map of PAR%, relative risk (RR) and space-time cluster analysis indicated that the clusters were generally unchanged, and were found south-west and north-west of Shandong. The Lanshan District in Linyi had the highest PAR%, while highest RR was in the Yicheng District in Zaozhuang. CONCLUSION There were significant space-time clusters of infant measles in Shandong over the study period. PAR% is an effective way to analyze multiple clusters from their application like RR. Interrupting measles circulation and maintaining routine coverage over 95% may be the only effective strategy to achieve measles elimination.