The five year longitudinal epidemiological study in south India (1961-68) showed that a considerable proportion of bacteriologically proven cases found in a survey got cured naturally without the facility of organised treatment in the survey area. This "natural cure" could be an epidemiologically significant phenomenon- depending on the stability of such a cure or in other words, the frequency of relapses among the naturally cured. In all, 108 naturally cured cases of tuberculosis out of a total of 269 cases, from among about 62,000 persons surveyed twice, were followed up for varying periods of 1 to 3½ years.

It was observed that the average relapse rate was 85.4 per 1000 person years of observation, there being no difference between the two sexes. Relapse rates were however higher in persons aged 20 and more compared to those 5-10 years old. Relapses were not dependent on the bacteriological status at initial diagnosis i.e., whether positive by culture alone or positive by smear and culture. The death rate among the naturally cured was 42.7 per 1000 person years and together with relapse constituted the unfavourable fate after natural cure. It has been calculated that as an input, adding to the pool of bacillary cases in the community, the ratio of relapse cases to cases arising afresh from the general population in a year would roughly be in the order of 1:16. It is concluded that the naturally cured status could be considered as an epidemiologically favourable situation, though much less so when compared to the chemotherapeutically achieved cure.