Edward A. Allen

Intrathoracic paediatric tuberculosis: a report of 202 cases.

Until recently paediatric tuberculosis rates were in decline in developed countries, but more recently there has been a significant change in these trends. Tuberculosis in British Columbia (BC) is centralized and all paediatric cases diagnosed between 1979 and 1988 were reviewed. There were 252 notifications during the study period but 50 cases were excluded (31 extrapulmonary cases, 11 with missing records and 8 because of incorrect diagnoses). There were 109 (54%) female patients. The age range was 4 months-15 years (6.4 +/- 4.1 years). 52% were 5 years old or less. 75 (37.1%) were aboriginal Canadians, 61 (30.2%) were Asian, 50 (24.8%) were Caucasian and 16 (8%) were included in miscellaneous or unknown groups. Primary pulmonary disease occurred in 197 (97.5%); the remaining 5 had post-primary disease. A history of close contact was elicited in 158 children (78.2%). All but 8 (4%) had a positive PPD response to 5TU. 40% of children had symptoms, the most frequent being cough (51%) and fever (28.4%). Mycobacteriology was carried out in 169 (83.7%) children. Bacteriologic confirmation was achieved in 45 patients (22.2%). Chest X-rays were reviewed in 186 cases. Lymphadenopathy occurred in 93.5%, with hilar (34%) and combined hilar-paratracheal (26%) being the commonest findings. The majority of patients were treated with isoniazid and rifampin, with a minority receiving streptomycin or pyrazinamide in the initial intensive phase. 14 (6.9%) children had adverse reactions to their drugs but these were rapidly reversed with the withdrawal of the responsible drug. Response to therapy was excellent and there were no deaths.

Deaths in tuberculosis patients in British Columbia, 1980-1984

Records of all 1884 newly notified tuberculosis cases, over the 5-year period 1980-1984 in British Columbia, Canada, were reviewed and 201 deaths were identified, including 48 diagnosed only after death, and 153 who died while on treatment; 56 of unrelated causes, 67 in whom tuberculosis was a contributing cause, and 30 in whom it was the principal cause. Significant predictors of death while on treatment (specific to tuberculosis) were the extent of disease, history of previous disease and sputum smear-positive for acid-fast organisms. Significant predictors of failure of diagnosis, in patients who died, were the presence of disseminated disease and the absence of a history of previous disease. The presenting features were not different in those dying, in whom the diagnosis was made before, as compared with after, death. The most frequent mode of death due to tuberculosis was respiratory failure, followed by multiple organ system failure and haemoptysis. The case fatality rate was low (1.6%) and did not change over 10 years. One-half of patients whose death was due to tuberculosis were diagnosed only after death and this had not changed over 10 years. We conclude that death due to tuberculosis is uncommon in patients while on treatment and that the main reason for death due to tuberculosis is that some patients are not diagnosed, and therefore not treated, before they die of the disease.