The question of carrying out a repeat epidemiological survey in India has been engaging the attention of many for quite some time. The first nationwide tuberculosis prevalence survey was conducted in India during 1955-58. It served as an eye opener and produced data which were profitably used by the planners to decide about the form and state of national control programme. Doing a repeat survey will be useful only if it would be capable of yielding epidemiological information on the future course of action. At the time of formulation of the District Tuberculosis Programme (DTP), it was perhaps presumed that programme would work with optimum efficiency as in the operational studies and as such the real performance was not envisaged. Secondly, due to low prevalence rates of tuberculosis as shown in all the surveys could reflect a small rate of change or no change at all, thus these longitudinal surveys with inadequate samples, did not have enough discriminatory power to observe a statistically valid change with time.

It is now globally realised that instead of looking at mortality rates or small changes in the prevalence rates of cases, it is the Annual Risk of Infection (ARI) which holds the key to epidemiological trend in a community. However, through a model recently constructed at the National Tuberculosis Institute, it is possible to extrapolate the findings of well planned small surveys in certain areas. It gives an idea what to expect over a period of 50 years - a slow decline. Therefore, when the present efficiency of Case-finding programme is about 33%, treatment efficiency also of the same order or even worse and with persistent rise in the population, it is futile to talk of epidemiological assessment through repeat surveys. Instead, we should concentrate on raising the efficiency of the DTP as near to the level which could be called the critical level of efficiency. Till then nation wide surveillance through the calculation of ARI is the only choice.