Before-after (1998 and 2008) trend analyses on regional clustering of clinical dentist-to-population ratio in all 1,976 municipalities of Japan

Purpose: The main purpose of this study was to obtain geographic clustering information in order to identify shortages (“cold spots”) and surpluses (“hot spots”) of dentists in all municipalities of Japan. Methods: Pretreatment steps were conducted to recover the lost comparability between pre-1998 and post-2008 data due to the large-scale merging of municipalities (42.1% reduction) in the Heisei era. Moran’s I, LISA and spatial multiple regression analyses with AIC were performed to verify regional clustering. Dependent variables of the regression analyses were the clinical dentist-to-population ratio in 2008 (Model 1) and the difference between 1998 and 2008 (Model 2). Results: The R2 was 0.8379 (p<0.0001) for Model 1 and 0.5832 (p<0.0001) for Model 2. The initial dentist-to-population ratio in 1998 showed the highest significance in both models. However, the coefficient of Model 2 was negative, which was exactly the opposite of that of Model 1. Furthermore, indices relating to urbanization and hospital dentist-to-population ratio in 1998 were also highly significant (p