大阪府下における膀胱腫瘍患者の地域集積性

Abstract

Chemical carcinogens as causative factor of the occupational bladder tumor were found nearly a century ago. However, spontaneously occurring bladder tumor would not cease to appear despite many efforts of thinkable preventive measures. Geographic epidemiology play the most important role in such situation to detect more precise causative factor in environment. But conventional statistical method is not good enough to ensure any clue for the problem when epidemiologic study was done in a highly industrialized and densely populated large city. There are millions of people in a small plane, and bladder tumor patient can move one urological clinic to another freely under present medicare system. These circumstances make it difficult to continue a long term follow up study for some specified case for years. On the basis of recent report of the Bladder Tumor Study Group in Osaka, 1977 to 1978, a new test for significance of geographic aggregation of bladder tumor patient in restricted area was examined. The test employed a technique of Monte Carlo simulation and cluster analysis of events. A new computer program for this purpose was developed in our laboratory and compiled in FORTRAN language. Then tabulated data were processed by FACOM 230-OS2/VS system. In thirty six areas in suburbs and 26 in city zone, incidence of bladder tumor on both sex were categorized in five hierachies as mean value of the whole area represents median of the distribution. A successive calculation which compare real number of patients in identical area with expected integer value of each concordant category pairs, can reveal aggregation of the bladder tumor patient in Osaka Pref. At the results, significant concentration(p<0.001) of the patients was found in city zone with high incidence, and moderate clusters were observed in adjacent suburbs at the same time. The test will give not only distinctive interpretation of geographic pattern of the disease but a tool to approve statistical significance of distribution.