Prevalence surveys are useful for estimating the tuberculosis problem in different countries. Three techniques are commonly used in surveys, tuberculin test, mass miniature radiography and sputum examination. Each has its own limitations. A limitation of sputum examination is that all the sputum positive cases in the community cannot be diagnosed when only one sample of sputum is examined from each eligible person. Multiple sputum examinations are not often possible under field conditions of surveys covering the whole community. It would be worthwhile to have some idea of the extent of under-diagnosis in sputum examination. For this purpose, during an epidemiological survey, four specimens of sputum were collected within seven days of X-ray examination from each person with an abnormal chest X-ray in 30 villages of a district of south India. Each specimen was examined by Fluorescent Microscopy (FM), Ziehl Neelson (ZN) technique and culture.

There were 34 culture positive cases among 2,164 persons for whom all the four culture examination results were available. Of them, 21 (62%) were found positive on one specimen. The second specimen increased the positivity to 32 (95%). Thus, for detecting both smear and culture positive cases two specimens are adequate. A third specimen is helpful for detecting cases positive by culture alone. An estimate of prevalence obtained from one sputum specimen can be estimated for the prevalence obtained from many specimens by applying correction factor of 1.67 and estimates based on two specimens by applying 1.26. Of the remaining 37 smear positive cases detected by one specimen, 20 were smear positive and culture negative. Of the remaining 17 smear positive and culture positive, 14(82%) were detected by one smear examination only.

ZN positives not confirmed by culture (mostly with less than four bacilli reported in the smear) increased from 7 from the first specimen to 18 from all four specimens, while positives confirmed by culture method showed only a marginal increase from 13 to 15. FM did not have this disadvantage as only two were culture negative among the 18 smear positive results by FM method. Examination of two specimens by FM detected about 95% of cases demonstrable by this method. But with the ZN technique additional specimens may add more “false positives”. Thus, for detecting cases both smear and culture-positive two specimens appear adequate. A third specimen is helpful for detecting cases positive on culture only.

The difficulty in staining tubercle bacilli is believed to be related to the complex surface structure containing a large amount of unsaponifiable wax. Any staining technique which can counteract the influence of this wax could therefore be expected to give better results. The standard method in vogue is the application of heat which renders the bacilli permeable to aqueous dyes. Several attempts have been made to develop a cold staining method for tubercle bacilli as for other organisms. Since this wax is soluble in chloroform, a cold staining method using carbol fuchsin containing chloroform was developed and the results of staining by this new method have been compared with the conventional Ziehl-Neelsen (ZN) method in the present study. Triplicate smears were made from 186 specimens and these were stained by ZN, Cold Staining (CS) and Fluorescent Microscopy (FM) methods. In addition, single smears of 343 specimens previously examined by FM were randomly divided into two subgroups and restained by ZN and CS methods respectively.

The results of examination of duplicate smears by ZN and CS methods showed a high degree of correlation with 75%(140/186) showing identical grading and only 8 were positive by one and negative by the other method. Of the 8 smears positive by CS alone, 7 were confirmed by culture, whereas 3 were positive by culture out of the 8 positive by ZN method. This indicates that those positive by CS alone are likely to be real cases, whereas those positive by ZN alone may include some false positive cases. As far as false negatives are concerned, there was no difference between ZN and CS methods. The reliability of these methods was judged on the basis of culture results and agreement among themselves. The cold staining method was found to be as efficient as ZN method in detecting different gradings of culture positives. In addition, preparation of stain, training of personnel for CS was also found to be as simple as ZN method.

A study was done to evaluate inter and intrareader differences in reading of smears stained by Fluorescent Method (FM) and Modified ZN Staining (CS) methods and variation in multiple smears made from the same specimens in order to find out to what extent sensitivity and specificity are influenced by repeated sputum smears from same specimens, by change of readers or by repeated reading.

Eighty sputum specimens with known results, 35 negative and 45 positive with different grades were selected. Ten smears were prepared from each specimen. All the smears were first examined by FM and later by CS method. Culture was also done for these specimens. The findings were: (i) FM was more sensitive than CS method. The specificity appeared to be equal in both. (ii) Change of readers influenced the relative sensitivity of both methods, but repeated examination by the same reader had no effect. (iii) Both reader as well as reading influenced the specificity of FM method but not that of CS method. (iv) Repeated sampling from the same specimen had no effect on the sensitivity of both the methods, while it had some effect on the specificity of FM method. (v) Consistency among readers and readings was found to be poor in smears with low grade positivity. (vi) The relative sensitivity of any method was influenced by the proportion of low grade positive cultures in the total pool. (vii) Examination of multiple smears from the same specimen by more than one reader and more than one reading by the same reader was more rewarding in CS method.