Canadians, 15, and aboriginal Canadians, 13. Since 1982, a new high-risk group has emerged with the highest relative risk for tuberculosis ever recorded, namely, patients with the acquired immunodeficiency syndrome (AIDS). All high-risk groups together accounted for 80% of all the cases diagnosed at that time: 33% in Asian-born Canadians; 17% in previous cases; 12% in aboriginal Canadians; 8% in contacts; 7% in residents of urban slums and 3% in patients with AIDS.
Urban centres have traditionally experienced higher rates than rural areas [25]. In British Columbia, Canada, from 1970 to 1985 [34] notification rates were about twice as high in the large urban centre of Vancouver as compared with the predominantly rural remainder of the province. Although immigrants accounted for some of the difference, when only non-aboriginal Canadian-born persons were considered, the urban/rural difference remained. To study this further, five separate sections of the city, chosen on the basis of socioeconomic characteristics, were compared; median annual income varied from $6000 to $20 000 per year; rate of unemployment from 18% of employable men to 3%; completion of secondary school education from 59% to 93%. The annual rate of notification of active tuberculosis between 1980 and 1982 also varied across the sections of the city from the lowest socioeconomic level (with an annual notification rate of 242 per 100 000) to the highest level (with a notification rate of 2). A similar relationship was observed for two other Canadian cities, Edmonton and Calgary (Fig. 2.4). The difference in rates typically found between single and married males was entirely due to the concentration of single males in the lowest income census tracts of the city, and, when this was taken into consideration, there were no continuing difference according to marital status. Unemployment was the single most important predictor of level of notification rate in the city. Moreover, in comparing the notification rates with similar reports in poor areas of Buffalo, New York in 1939 [35], the rates were quite similar for the two cities, even though there was a much lower likelihood of previous infection with tuberculosis in Vancouver in the later period. These findings indicated that low socioeconomic conditions led to an increase in current transmission of tuberculosis.
Over the two decades since 1970, the proportion of notified cases who had been born outside Canada, rose from 20 % of all cases to nearly 50% [36, 37], even though some groups had lived in Canada, on average, for more than 40 years, indicating the continuing importance of infection in early life in determining subsequent rates of tuberculosis in low-prevalence countries.
Some investigators [38] have observed that the notification rate of tuberculosis in immigrants is highest in the years immediately following immigration. This higher rate, however, is due to a bias related to incorrect identification of individuals with active tuberculosis as 'immigrants' when, in fact, they were not [36].
Immigrants from high prevalence countries have a higher prevalence of fibrotic lesions than those born in low prevalence countries and those with fibrotic lesions who have never taken chemotherapy have an increased risk of active tuberculosis. A study of immigrants to British Columbia during 1982-1985 from five countries of Asia (Japan, Korea, the Philippines, China and India) [39] found a greater prevalence of fibrotic lesions than in residents of British Columbia (6% compared with 1%). In the years following immigration, a larger number of cases occurred in the group of immigrants who had fibrotic lesions than in those who did not (33 cases compared with 30). In those with fibrotic lesions, the greatest number of the cases were discovered on the initial examination after entry into Canada, in spite of having been investigated by chest radiograph and bacteriology prior to immigration. Moreover, almost all the cases in those with fibrotic lesions among the immigrants occurred in those who indicated no previous chemotherapy for tuberculosis, with a notification rate for bacillary cases of 1% per annum. As these individuals had been previously identified and were required to have an examination upon arrival, the disease might have been prevented by chemotherapy with isoniazid upon their arrival in the country.
2.3 THE IMPACT OF CHEMOTHERAPY ON TUBERCULOSIS
The principal aim of programmer for control of tuberculosis is to reduce the transmission of tuberculosis and, in this way to eliminate the disease. This is accomplished by the reduction in the number of sources of infection (smear-positive pulmonary cases) by permanently rendering them bacteriologically negative-through treatment or, where no treatment is given, unfortunately only by death of the patient. In evaluating the results of programmes, therefore, it is most important to determine the ability of such programmes to reduce the residual pool of smear-positive cases in the community. The results of treatment programmes under various conditions have been reviewed [40]: when 'no chemotherapy' is given; [41,42] under 'ideal' programme conditions, where chemotherapy is individualized to the patient based upon results of culture and sensitivity of the organism; and under 'mass chemotherapy' conditions prior to the introduction of rifampicin [43, 44]. With 'no chemotherapy' one-quarter of the patients die within 2 years and 50% die within 5 years. Another 25% cure themselves (their immunological defence is capable of overcoming the micro-organisms within their bodies and render them inactive). Another 25%, however,
