remain smear-positive sources of infection within the community. Under 'ideal' programmes, 8% of patients die while on treatment, and 84% are cured, leaving 2% smear-positive at the end of treatment (in 6%, results are unknown because they moved or absconded from treatment). With 'mass chemotherapy', the fatality rate is dramatically reduced to only 10%, and the proportion of patients cured is increased to about 60%. This, however, leaves as many as 30% smear-positive or unaccounted for. The irregular treatment of many patients (who leave treatment when feeling better but return when symptoms return) allows them to survive but fails to cure them. Thus, the net effect of such programmes may be to increase the number of sources of infection within the community. Under mass treatment conditions what is normally good chemotherapy becomes, in fact, poor chemotherapy. Poor compliance (which is the normal state of affairs with patients) is the most important reason. Directly observed chemotherapy is the only means to ensure compliance and direct observation of drug-taking must be the practice, at a minimum, in that portion of therapy that is most important (the initial intensive phase in smear-positive pulmonary patients).
The second reason is the presence of bacteriological resistance to chemotherapeutic agents. It is very uncommon in countries like Canada [38, 45], Algeria [46] or Tanzania [47], where initial resistance to one or more drugs occurs in less than 10% and acquired resistance in more than 35%. In some other countries, however, the problem is very much greater. In Korea, for example, in a national prevalence survey, initial resistance was present in 24% of cases and acquired resistance in 74% [43], in China, the corresponding figures were 27% and 72% [48]. In Korea, 53% and in China, 62% of all cases discovered in the prevalence surveys had previously been treated; these 'residual' cases (representing a failure of the national programme) were drug-resistant sources of infection in the community. A mathematical model [49] based upon the data derived from the periodic national prevalence surveys of Korea and Taiwan, estimated the relative infectiousness of the two types of cases, new and chronic. A chronic case was estimated to be more important as a source of infection in the community (produced more new cases) than a new case.
The introduction of rifampicin-containing regimens of chemotherapy into some national programmes has had a great impact. The most important part of the short-course regimen is the initial intensive phase, continued in smear-positive cases until the sputum smear has become negative (in most cases, after 60 days of treatment). If this period of drug-taking can be assured, overall results can be good enough to have a positive impact on the epidemiological situation through reducing the number of sources of infection in the community, as demonstrated in the National Tuberculosis Programmes (NTP) assisted by the International Union Against Tuberculosis and Lung Disease [50]. In Tanzania, the NTP has been in operation since 1978 (see Chapter 15b). At the outset, the regimen of treatment was 12 months of isoniazid and thioacetazone, supplemented with streptomycin in the first 2 months. Treatment results were poor: less than one-third of patients were cured, one-half of patients abandoned their treatment and one-fifth remained positive. With the same regimen, but improved programme structure, the results were considerably improved, but not yet satisfactory: slightly less than 60% were cured, approximately 30% absconded and 1% remained positive. With the introduction of rifampicin-containing regimens, a further improvement was achieved: nearly 80% were cured, 12% absconded, 6% died and 1% remained positive (results close to those achieved under 'ideal' programmes).
Another example of the impact of rifampicin can be seen by looking at the results of treatment of tuberculosis within the Social Security System [51] in Mexico. This branch of the government is responsible for providing health care for the poorest segment of Mexican society, serving over one-third of the entire population. Rifampicin-containing regimens were introduced into the system in the middle 1980s and gradually replaced 12-month chemotherapy, which was either self-administered or supervised. Results of treatment by cohort analysis for the three types of treatment regimen between 1983 and 1987 indicated successful completion of the regimen in 62% of the self-administered, 68% of the supervised 12-month and 82% of the supervised 6-month treatment groups. The fatality rate was similar in the three groups, at about 5%. The remainder were absconders with highest rates in the self-administered group and lowest in the supervised 6-month group.
Good results with rifampicin-containing regimens are obtained under a precise set of prerequisites [47], including a political commitment to ensure correct operation of the programme (including designated personnel at a central level for supervision, coordination and training and a policy that tuberculosis patients are not discriminated against in obtaining necessary treatment), a regular supply of materials, diagnosis based upon bacteriological examination, correct recording and reporting of diagnosis and treatment results, and direct observation of medication swallowing during the initial, intensive phase. Failure to follow these prerequisites often leads to very poor results, such as those observed in the Harlem district of New York City [52] where the results of treatment are worse than in the poorest developing countries with a consequent and dramatic increase in reported cases (Fig. 2.5), including an alarming increase in the rate of multiple drug resistance.
2.4 HUMAN IMMUNODEFICIENCY VIRUS (HIV) AND TUBERCULOSIS
Infection with HIV has disrupted the balance that previously existed between the human host and the tubercle bacillus and has thereby increased the rate of development of disease in those who have been infected with tuberculosis. Its action is initially noted in those individuals who are dually infected with both HIV and tuberculosis. Thus, instead of only one in six infected individuals developing tuberculosis during their lifetime, at least one in three who are dually infected can be expected to develop active tuberculosis [53]. The impact of HIV on the tuberculosis situation in the community, however, will be exerted through the increase in the number of infectious cases of tuberculosis in the community caused by HIV infection and the subsequent increase in the rate of transmission within the community.
The impact of HIV on tuberculosis incidence was first noted in the USA [54], where the previous steady decline in tuberculosis incidence was halted and then reversed in the mid 1980s. This effect was, in part, due to the coexistence of the two infections in certain sub populations within the USA,
