Cost analysis of paediatric tuberculosis treatment in Japan.

OBJECTIVE: To estimate the cost of treating a tuberculosis (TB) case and to analyse TB-related medical service utilisation, a cost-of-illness study was conducted for all patients with a primary diagnosis of TB admitted to a public hospital in Japan. METHODS: Retrospective analysis by abstracting in- and out-patient medical records of 57 paediatric patients diagnosed with TB during 1993-1998 at a public hospital in Osaka prefecture. Costs were estimated based on third party's payer perspectives according to the service utilisation pattern. In addition to cost data, sociodemographic information and service utilisation pattern were also extracted from the medical records. Cost of preventing a case of TB was abstracted from the published literature. RESULTS: The average cost of treatment was 8384 US dollars (95%CI 5667-11,099), while the average length of hospitalisation was 63 days (95%CI 43-84). Based on 20-80% vaccine efficacy, the cost of preventing a case of TB was 35,950-175,862 US dollars. In univariate analysis, site of TB (P = 0.04) was significantly associated with TB treatment cost, while case-finding method (contact tracing, symptoms, etc.) was associated with length of hospitalisation (P = 0.03). Multivariate regression analysis, however, showed none of the factors to be significant predictors of TB treatment cost and length of hospital stay. CONCLUSION: The cost of treating a case of paediatric TB is much lower than that of preventing one. Japan's universal BCG vaccination policy should be re-examined in the light of economic, social and political issues.

Economic evaluation of universal BCG vaccination of Japanese infants.

BACKGROUND: The international controversy surrounding the use and effectiveness of the Bacillus Calmette-Guérin (BCG) vaccine and the low incidence of tuberculosis (TB) among Japanese children prompted this study. METHODS: We compared 'universal BCG vaccination' with 'no vaccination at all' using a cost-effectiveness analysis. The study population was a hypothetical cohort comprising a total of 1.2 million infants born in 1996 at locations all over Japan. A model was developed to calculate the number of TB cases prevented by the vaccination programme. Assuming 40-80% overall vaccine efficacy (64-86% for TB-meningitis) and 10 years of protection, we calculated the cost and number of immunizations required to prevent one child from developing TB, the total number of TB cases averted by vaccination and total costs required for the programme. RESULTS: Based on an assumption of flexible vaccine efficacy (40-80%), we estimated that 111-542 TB cases including 10-27 of TB-meningitis would be prevented during the 10 years after BCG vaccination among the cohort of infants born in 1996. About US$35 950-175 862 or 2125-10 399 immunizations would be required to prevent one child from developing TB. Sensitivity analyses covering a wide duration of protection, incidence of TB, vaccine coverage and discount rate, revealed that other than vaccine efficacy, the cost of preventing a single case of TB is highly sensitive to the duration of BCG protection and TB incidence. CONCLUSION: The cost per case of TB prevented is heavily dependent on vaccine efficacy and the duration of protection, and is high compared with the cost of treating one child who has developed TB.

[Assessment of effectiveness of BCG vaccination against childhood tuberculosis].

We performed a retrospective analysis of the history of BCG vaccination of 367 childhood patients who were treated for active tuberculosis (TB) in our hospital from 1976 to 1994. Fifty-eight percent of 367 cases, 83.1% of cases under 5 years of age and 92% of tuberculous meningitis had not received BCG. To investigate the effectiveness of BCG vaccination against TB, we carried out case-control studies using 59 patients with TB and 118 controls without TB. The cases were patients treated for TB in our hospital from 1988 through Nov. 1994. Two controls were chosen for every case with matching for sex, age at admission, year of admission and place of residence. Based on whole 59 pairs, BCG vaccination was shown to have protective efficacy of 78% (95% confidence interval [CI], 57-89%). For 34 pairs of under 5 years of age, estimated efficacy was 92% (95%CI, 80-97%), for primary pulmonary disease (27 pairs) it was 92% (95%CI, 78-97%). For bacteriologically confirmed TB (27 pairs), it was 84% (95%CI, 58-94%), while for bacterilogically negative TB (32 pairs) 71% (95%CI, 27-89%). Our results indicate that BCG vaccination with multipuncture methold protected considerably against TB in infants including primary pulmonary TB.