In the District Tuberculosis Programme (DTP) the diagnosis is based on sputum microscopy. Majority of health institutions in the district are provided with microscopes for this purpose. In the Peripheral Health Institutions, the programme activities have to be carried out by its staff after a short period of training given by District TB Centre personnel on the spot. So the microscopy work in the PHIs is likely to be carried out by any paramedical personnel and not necessarily by a qualified laboratory technician. It is therefore, necessary to know whether the standard of microscopy carried out by these paramedical personnel after a short training will be upto the mark. To assess the efficiency of smear examination done by these individuals, a study was conducted in Bangalore district covering nine microscopy centres in various types of health institutions, a few months after the implementation of the programme. Under the DTP a spot specimen is collected from every chest symptomatic attending the health institutions and a smear is made and examined for the presence of AFB and all positive cases are put under treatment. The sputum specimens and the smears examined in these nine centres were brought to National TB Institute laboratory. The smears were examined by an experienced laboratory technician. Duplicate smears were also prepared from these specimens and their results compared with results of re examination and centre's examination. All specimens were cultured by swab method and all positive cultures were subjected to sensitivity and identification tests.

Analysis of the results based on culture showed that barring a few centres where the performance was poor, the standard of examination was fairly good. The under and over diagnosis based on culture were 38.2% and 2.6% respectively, and these were within the limits observed generally. Comparison of results on re examination of centre smears and duplicate smears indicated that both reading variation and defective smear preparations and staining could have influenced under diagnosis in these centres. The study has also thrown some light on methodology of assessment of sputum examination that could be adopted wherever a tuberculosis control programme is functioning.

Of all the available methods for the diagnosis of pulmonary tuberculosis, bacteriological examination is the most reliable. Diagnosis of pulmonary tuberculosis is chiefly done by sputum microscopy and culture. This paper discusses sputum microscopy from various points of view. Sputum, which forms the basis of bacteriological diagnosis, is a variable source material. Type of specimen, its quality, quantity, bacterial content and viability of organism considerably influence the sensitivity and the specificity of the methods; and these in turn would vary under different diagnostic situations. One of the reasons for the observed variations could be the different criteria adopted for examination; another might be due to the observed range of diagnostic situations varying from an epidemiological survey situation at the one extreme to the other where cases seek treatment in a comparatively backward community with poor tuberculosis diagnostic services.

In epidemiological community survey (ICMR 1968), it has been found that culture positives that were also smear positives varied from 73% to 87%, whereas among patients attending rural general health institutions for diagnosis, about 82% of the infectious cases found by culture could also be discovered by microscopy of single spot specimens (Rao, 1966). Sikand (1965) from New Delhi Tuberculosis Centre, could get 67% of culture positives by microscopy, whereas Mitchison (1967) found that 35% were smear positive among the sputum positive patients reporting for the first time. In the longitudinal epidemiological study carried out in the Bangalore rural area, it was found that about 40%-48% were positive by both direct smear and culture and the rest by culture only. Reasons for these variations could be (i) different criteria adopted for examination (ii) different situations from where the sputum specimens were collected (iii) sensitivity and specificity of sputum microscopy technique adopted and (iv)the experience of the trained technician. It was observed that over diagnosis by the trained auxiliary staff in the general health institutions (1.9%) compares favourably with the over diagnosis of 1.3% by experienced technicians indicating simplicity of smear examination. Besides these aspects, other factors like the quality of sputum smear, time spent on examination, type of sputum specimen, the use of multiple smears, etc., also influence the results. The cost of bacteriological examination have also been studied, and the cost ratio between microscopy and culture have been worked out to be 1:6.6. Cost can become an important factor in deciding the suitability of bacteriological methods for diagnosis of pulmonary tuberculosis in various countries and in different diagnostic situations.

The findings of two studies, (i) one on comparison of Ziehl-Neelsen method of staining of acid fast bacilli with and without alcohol decolourisation and use of Gabbet's Methylene blue (in place of decolourisation and counter staining) and (ii) comparison of two different types of Basic Fuchsin dye used in the preparation of Carbol Fuchsin, have been presented. The first study has shown that omission of alcohol decolourisation or the use of Gabbet's Methylene Blue has not influenced the detection of positives, though the latter has more often produced a non- satisfactory background. The second study has brought out the fact that two types of Basic Fuchsin are similar in every respect. However, the findings does not rule out the possibility of a bad dye giving rise to poor results. Need for conducting studies for simplifying the staining procedure has been stressed.

Under the National Tuberculosis Programme, Ziehl-Neelsen method (ZN) is the recommended procedure for AFB staining. However, at the Peripheral Health Institutions, the applicability of this procedure appears to be adversely affected due to non- availability of alcohol for heat fixing, heating and decolourisation steps. Hence, it has become necessary to look for procedures when the use of alcohol could be avoided completely. In this study, three different methods of cold staining, wherein the use of alcohol is avoided, have been tried in 489 sputum specimens. One set of smears was fixed with acetone (CSB method), the second set with candle flame (CSD) and third set was not fixed (CSC). All were stained by a modified cold staining procedure. The (a) sensitivity (b) specificity (c) degree of agreement with culture and the standard ZN method and also (d) applicability of these methods, were studied. All the specimens were subjected for culture and ZN smear examinations. ZN method was significantly superior to other methods. As far as the agreement with culture is concerned, ZN method was found to be superior to CSB method by 29.26%, to CSC method by 24.3% and to CSD method by 31.7%.

The sensitivity and specificity of various methods were: ZN-76.1%, 98.4%, CSB-72%, 96.6%, CSD-72.7%, 96% and CSC-76.6%, 96.9% respectively.