Estimation of cancer incidences in Aichi prefecture: use of a model area with good quality registry data.

In Japan, local government is responsible for organization of population-based cancer registries and the quality of the registration remains modest, mainly due to dependence on voluntary-based operations without legal obligations. Aichi Prefecture cancer registry covers a large population, estimated at 7 million, and its quality has yet to reach the level required internationally. The derived cancer incidences for Aichi Prefecture therefore tend to be underestimated. In the present study we set up a model area, located in the central part of Aichi Prefecture, with a good quality of registry data, covering a reasonable population, including both urban and rural areas. Our model area has typical demographic features of Aichi Prefecture. The materials were data on cancer incidence and deaths during the period of 1996-2000 in this model area of Aichi prefecture, with a population of approximately one million, under the jurisdiction of three public health centers, covering nine municipalities. The percentage of death certificated notified (DCN) cases for all sites was around 14% and the incidence/death ratio was around 1.9. Estimated age-adjusted incidence rates were found to be 256.0 (per 100,000) for males and 177.6 for females, these values being 10-15 % higher than those generated using data for the whole prefecture, and quite close to incidence rates in Japan estimated from the highest quality of data available. It is suggested that the cancer incidence in the Aichi prefecture is indeed being underestimated and that the actual figures may be closer to the estimates provided here.