The present definition of the objectives of the National Tuberculosis Programme is too vague. A proper definition of the objectives, both longterm and intermediate, is needed. It should clearly state the index to be used for measurement of the problem and the expected values of this index at specific points of time. Another serious problem in assessment is to find out how much of the observed problem reduction is due to the impact of the programme and how much due to (or in spite of) the natural trend (downward or upward). Repeated surveys cannot provide this information and keeping of control groups is not feasible. Epidemetric models help in choosing the index for measuring the problem and fixing intermediate and long term objectives in terms of this index. They also help to take the natural trend into account, while assessing the programme.

Prevalence of infection is the least sensitive index. Prevalence of disease and incidence of infection may lead to over optimism. Incidence of disease is most suitable but difficult to get in developing countries. Hence, prevalence of infection or disease has to be chosen. Difficulties of the former are interference by BCG vaccination and non- specific sensitivity. The use of BCG induration to estimate prevalence of infection has some advantages and it is worthwhile to investigate further this possibility. Using epidemetric models, two methods of assessment of effectiveness are suggested. One to carry out prevalence surveys but need not be attempted unless programme efficiency has been quite high for at least 10 years. The other is the assessment of efficiency of the programme which can be easily carried out.