A nation-wide cross-sectional study was conducted in India to assess the prevailing epidemiological situation of TB, by estimating ARTI. For the purpose of the study, the country was stratified into four zones viz., north, south, west and east, each having about one fourth of the country's population (Map-1). The study was designed to obtain estimates of average annual risk of tuberculous infection for each zone.

The study was conducted among children 1 to 9 years of age. The sample size for each zone was estimated considering the expected prevalence of tuberculous infection as 8% and to obtain the estimates within 10% of the true value at 5% level of significance. A design effect of 2.33 based on previous tuberculin surveys was applied for the estimation of the sample size. The estimated sample size was distributed into rural and urban strata in proportion to the rural-urban population in the zone. A two-stage sampling procedure was adopted for selection of clusters within the stratum. At the first stage, a sample of six districts was selected by systematic sampling, in each of the north, south and west zones. Within a stratum, the assignment of number of clusters to districts was done in proportion to the respective district population. In east zone, the sample size was distributed into eight similarly selected districts in order to have larger representation from northeastern states. The districts selected in the four zones are shown in Map-2.

The fieldwork in north, west and east zones was conducted by NTI and that in south zone by TRC. The necessary local support was provided by the district health authorities. The NDTC assisted in the supervision of fieldwork in three districts of north zone and MGIMS in three districts of west zone. The fieldwork was conducted by six teams of specially recruited contractual health workers. The teams were trained in various aspects of the survey at TRC & NTI.

In the selected clusters, the children were registered by house-to-house visits. They were subjected to tuberculin testing with 1TU PPD RT23 with tween 80 at a makeshift testing centre in the cluster. The tests were administered on mid-volar aspect of the left forearm after obtaining informed written consent from the parents/guardians. The date of testing along with presence or absence of BCG scar and identification particulars were recorded in individual child cards. The reactions were read about 72 hours later and maximum transverse diameter of induration was recorded. The presence of any unpleasant reactions like vesicles, bullae and necrosis was also recorded.

The proportion of children with BCG scar in different zones is given below :-

A significantly higher proportion of children elicited BCG scars in urban compared to rural areas, in all the zones.

The analysis was restricted to children without BCG scar numbering 25,816 in north zone, 17,811 in south zone, 22,259 in west zone and 19,322 in east zone. The number of test read children without BCG scar was larger than the estimated sample size, due to larger proportion of children without BCG scar than anticipated.

The frequency distributions of tuberculin reaction size were plotted separately for rural and urban strata of each zone. These were found to be bimodal with the second mode at 20 mm in case of north, west and east zones and 19 mm for south zone. The antimode (cut off point) for reactions attributable to infection with tubercle bacilli varied marginally as under :-

The proportion of infected children in the rural and urban areas of each district was estimated as weighted average of the proportion of children infected in individual clusters, the weight being the inverse of the initial probability of selection of the cluster. The proportion of infected children in the rural and urban strata of the zone was further estimated by combining the respective district estimates, using the proportion of district population (rural/urban) in the zone as the weight. The zonal estimate of prevalence of infection was obtained by similarly pooling the rural and urban estimates. ARTI was computed from the estimated prevalence of infection by using the appropriate equation.

The estimated prevalence of infection in each zone and the computed ARTI rates are given in the table below : -

The estimated prevalence of infection was similar by method I and method II, in respect of all zones except in west where it was marginally lower by method II. The estimated ARTI was highest in north zone at 1.9% followed by west zone - 1.8% by method I and 1.6% by method II. In east zone, the ARTI was estimated at 1.3%. It was lowest in the south zone at 1.0-1.1%.

The zonal estimates were pooled to estimate the national level ARTI, which worked out to 1.5%.

The children residing in urban areas were found to be at a significantly higher risk of infection than those residing in rural areas. For the purpose of comparison, the estimates of prevalence and ARTI in the rural and urban strata of four zones (by method I) are presented at Fig. 1.1 & 1.2.

The study provided useful information on the prevailing epidemiological situation of TB in the four zones of India. The results indicate a high rate of transmission of infection due to high load of infectious cases in the community. This is in consonance with unsatisfactory performance of NTP in terms of case detection and treatment during the last four decades. The high ARTI rates imply that the incidence of disease shall continue to be high in the coming years and therefore TB control measures need to be intensified on a prolonged basis. The RNTCP, which has covered approximately half of the country's population by 2002 (70% by June 2003), has shown remarkable improvement in cure rates of smear positive cases. However, the case detection rates need to be improved further in order to control TB. The RNTCP has to be extended to the remaining population and the private and non-governmental sectors have to be encouraged to follow national guidelines in case detection and treatment.