Survey of deaths in Hong Kong attributed to tuberculosis during a five-year period.

For the 12-month period from September 1st 1975 to August 31st 1976, and for the years 1979 and 1980, records were obtained for patients certified on part I of the death certificate as having died from tuberculosis in Hong Kong. In addition, records were obtained for patients in whom tuberculosis was a contributory cause of death (part II certifications) in 1979 and in 1980. The records for each survey were reviewed by an independent assessor. In the assessor's opinion, the diagnosis of tuberculosis was adequately established in 93%, 92%, and 92% of the patients in the 3 periods respectively, but among the part I certifications tuberculosis had in fact been the direct cause of death in only 74%, 78%, and 76%, active disease being the cause in 53%, 48% and 42%, and the late results of previous disease, inactive at the time of death, in 21%, 30% and 35%, respectively. Among the part II certifications, 39% in 1979 and 45% in 1980 should have been part I certifications, and only 32% and 24% respectively had been correctly certified in part II. If all the patients certified in part I had really died from tuberculosis, the death rates would have been 13.1 per 100 000 population in 1975/76, 9.7 in 1979, and 10.3 in 1980. According to the assessor's classification these should have been 9.7, 7.5, and 7.9 respectively, the rates from active disease declining from 6.9 in 1975/76, to 4.6 in 1979, and 4.3 in 1980. Among both part 1 and part 11 certifications there was a marked preponderance of males, and the death rates were particularly high in patients of both sexes aged 55 years or over.