Spatial autocorrelation analysis of measles in China, 2005-2014

Abstract

To detect the spatial features of measles in China by means of spatial statistical analysis. Data of prefecture-level measles cases and incidence from 2005 to 2014 were collected from the China Information System for Disease Control and Prevention. Information collected from the system included demographic characteristics, spatial distribution information, and diagnostic reports. Cases of unconfirmed measles and those with unknown address were ruled out. Cases from Hong Kong, Macao, Taiwan, and foreign countries were not included in this study. Maps were obtained from geographical boundary data at prefecture level from the Chinese Center for Disease Control and Prevention and demographic data from the National Bureau of Statistics. Based on different measures of measles elimination, we divided the data from 2005 to 2014 into three stages: stage 1 (2005-2008), stage 2 (2009-2012), and stage 3 (2013-2014). ArcGIS software was used to describe the spatial distribution and for global and local spatial autocorrelation analysis. The total number of confirmed measles cases reported in the system was 650 222, with average incidence 0.46/100 000. The highest reported incidence was in 2008 (9.95/100 000) and the lowest in 2012 (0.46/100 000). Average incidences for stages 1, 2, and 3 were 8.87/100 000, 1.99/100 000 and 2.96/100 000, respectively. Global Moran's I coefficients from 2005 to 2014 were 0.31, 0.08, 0.36, 0.56, 0.26, 0.48, 0.34, 0.20, 0.29 and 0.52, respectively; all were significant (P<0.05). Average incidences for high-high (H-H) clusters in 2005-2008, 2009-2012, and 2013-2014 were 33.02/100 000, 7.06/100 000, 11.91/100 000, respectively. Western China had high-value clustering consistently throughout all three periods; however, the number of prefectures covered by high-value clusters and discrepancy in the incidence between western and eastern regions were low. Northeast and northern China had H-H clustering in stages 2 and 3. In this study, measles incidence was spatially autocorrelated at the prefecture level from 2005 to 2014. Although China has made great progress in the elimination of measles, H-H clusters were consistently present. A need remains in China for targeted measles prevention and control measures.