Introduction Atomic bombs were dropped on Hiroshima and Nagasaki in 1945, and then A-bomb survivors' health handbooks (shortly ‘handbooks’) were issued by the Japanese government to help A-bomb survivors in 1957. They have been able to receive free medical checkup twice for a year and free medical care for designated disorders. The purpose of the study is to evaluate effects of A-bomb survivors' health handbooks focusing on the relationship between a mortality risk and a length of having handbook.

Methods Objects for analysis were selected from the ABS database of RIRBM Hiroshima University. The number of over-all deaths is 58 599 and the number of censored data is 101 244. Cox's proportional hazard model was applied for analysing the data. The observation period is from 1970 to 1997 and the time variable is a time from 1^st January 1970 to an occurrence of death. Length of having a handbook was defined as the period from registration year as an A-bomb survivor to 1970. Sex, age at A-bomb exposure, radiation dose and a length of having a handbook are used as covariates.

Results and Conclusion In men, there was significant negative relationship between a length of having a handbook and a relative mortality risk after being adjusted for sex, radiation dose and age at bombing, but in women such a relationship was not found. A man who got a handbook at a young age had a lower mortality risk compared to a man at an old age.