Survey with MMR remains as one of the most important methods available for measuring the size and extent of tuberculosis, specially in developing countries. Its value in case-finding programmes is well recognised. Nevertheless, mass miniature radiography with a single picture of the chest has a wide margin of error owing to the intra & inter-individual differences in X-ray reading. A study was undertaken to know the errors involved by repeating an X-ray picture after an interval of 3 to 4 months and judging the first picture in the light of a comparative reading of the two pictures. It is postulated that two pictures taken at an interval, may afford better judgement regarding the assessment of a case than a single picture only. A prevalence survey was carried out in Tumkur district in 1960-61, among 62 villages and 4 towns; 20 villages were selected for this study. A total of 8,000 persons were registered, 5,300 of them were X-rayed and re-read by two readers. Photofluorograms were repeated after three and a half months after the first picture. At the time of repeat X-ray, a spot sample of sputum was collected from persons with abnormal shadows.

Briefly the findings of the study were: (1) About 20% of bacillary cases were among those with inactive or non- tubercular shadows on the basis of a single X-ray film. (2) Inter-individual agreement for X-ray active cases was of the order of 50%. (3) Intra-individual agreement for X-ray active cases was 52% for one reader and 69% for the two readers. (4) Mass miniature radiography with a single film, in spite of its inherent limitations, is the best available method both for surveys as well as for case-finding programmes due to its ability to find cases as well as potential cases in a short time. (5) Even the agreement between two sputum samples collected within an interval of 1-3 days was 42% for positive results.

Limitations of a single X-ray picture for locating and interpreting shadows in the chest had been studied earlier. In order to reduce these limitations, it was suggested that two pictures of each person be taken where the second picture was to be taken after a vertical displacement of X-ray tube, up or down by about 4 to 5cms. The advantages of taking two pictures simultaneously as compared to a single picture have not been studied so far. Two mobile X-ray units each with an odelca camera were alternated for the single and double picture examinations. A total of about 2,000 persons were X-rayed and were read independently by 3 readers. A spot sample of sputum was collected 3-4 days later from persons with abnormal X-ray shadows and was examined by direct smear microscopy.

Comparison of the readings of the two sets of pictures did not show a better agreement between different (inter- individual) readers or between two different readings of the same reader (intra-individual) when the two picture technique was used. The X-ray cases detected by double picture only by any one reader were not confirmed, more often than those detected by single picture only. The X-ray pictures of the bacillary cases were also not interpreted more often as active tuberculosis by the two picture technique. It was concluded that the double picture technique does not offer any advantage over the single picture technique.

The purpose of this investigation was to compare the corresponding longitudinal and transverse diameters of tuberculin test and study the influence of reader variations, size of reaction and the age specific infection rates, in order to understand the effect of switching over from the transverse to longitudinal readings. The study was carried out in villages in south India where no BCG or tuberculin testing had been undertaken. A total of 1240 persons were given tuberculin test with 1 TU RT 23 in both longitudinal and horizontal diameters. The indurations were read after 2-4 days by two readers independently. The study showed that though the longitudinal diameters were bigger than the corresponding transverse diameters, these differences did not influence infection rates calculated at 10 mm or more induration level. In National Tuberculosis Institute (NTI), Bangalore the practice of reading transverse diameter was altered to longitudinal diameter in the epidemiological surveys as it was comparatively easier to read the longitudinal diameter. Obtaining almost similar infection rates at the 10 mm or more level of induration in this study, irrespective of readers and diameters has minimized the effect of the changeover from transverse to longitudinal diameter reading in the epidemiological surveys at NTI. This would also not pose any problem in comparing the results of NTI studies with other research studies by any national or international organisation where transverse diameters have been measured.