ASSESSMENT OF DIAGNOSIS OF PULMONARY TUBERCULOSIS BY SPUTUM MICROSCOPY IN A DISTRICT TUBERCULOSIS PROGRAMME pdf

Ind. J. Tub., 1971, 28/1, 10-21 Ind. J. Tub., Vol. XVIII, No. Reprinted from The Indian Journal of Tuberculosis,Vol.XVIII,No1 ASSESSMENT OF DIAGNOSIS OF PULMONARY TUBERCULOSIS BY SPUTUM MICROSCOPY IN A DISTRICT TUBERCULOSIS PROGRAMME K.P.RAO, S. S NAIR, N. NAGANATHAN AND R.RAJALAKSHMI (From National Tuberculosis Institute, Bangalore) Introduction An infectious case of pulmonary tuberculosis is diagnosed by demonstrating tubercle bacilli in sputum. The laboratory methods available for this purpose are sputum microscopy, isolation of tubercle bacilli in culture followed by identification of the bacilli and animal pathogenicity tests. Evidence based on all the three methods establishes bacteriological diagnosis beyond any doubt, but even economically advanced countries may not consider such an elaborate procedure essential for routine diagnosis. Very often sputum microscopy alone is considered sufficient for diagnosis, when combined with clinical findings. Therefore a realistic and economical approach for developing countries would be to provide only facilities for sputum microscopy. One of the basic principles of the Indian District Tuberculosis Programme (DTP) 1 is to offer diagnosis by sputum microscopy to persons presenting at any general health institution with complaints of symptoms such as cough for 2 weeks or more, pain in chest, fever and haemoptysis here-in-after called ‘symptomatics’. In the programme, non specialised staff of general health institutions perform diagnostic and treatment activities for tuberculosis, which are co- ordinated and supervised by a few specialised staff from District Tuberculosis Centre (DTC). These non-specialised staff of general health institutions have, therefore, to be trained in microscopy by the staff of the DTC. In actual practice the efficiency of diagnosis by microscopy will depend upon the skill, aptitude, experience and conscientiousness of the staff trained. Besides, their multiple responsibilities, training environment and the competence of the trainers, the distribution of work among trained technical persons and such other operational factors would indirectly influence the quality of their microscopy. To achieve and maintain a satisfactory standard in sputum examination, routine supervision of the general health institutions at regular intervals constitutes no less an important activity of the staff of DTC. At the same time, periodical technical assessment is essential to quantify the quality and reliability of microscopy, organised under such conditions, and to take timely corrective actions for improving technical performance. Eventually, such an assessment might help development of mathematical models in the field of tuberculosis. Objectives The primary objective of the present study was therefore to assess some aspects of the technical quality and reliability of sputum microscopy, as carried out in general health institutions under conditions of DTP. It was also envisaged to utilise the data, to formulate simpler assessment methodologies to measure the efficiency of the sputum diagnostic technique applied in DTP. Material and Methods The Bangalore district tuberculosis programme (the programme assessed). The Bangalore District has an area of 7798 square kilometers and had a population of about 1.3 million according to the census of 1961 (excluding the metropolitan area). The District has 13 towns and 2477 villages. The DTP was implemented in 1963, not by the staff of the DTC (as happens in other districts) but by trainees of two courses held in 1963 at the National Tuberculosis Institute (NTI) Bangalore. In all, 15 microscopy centres were organised in general health institutions of the District. Since, at that time, a mobile X-ray unit was also made to visit 6 centres, these centres were excluded from the study. Of the remaining 9 centres, 6 centres belonged to the period of initial implementation and 3 to the later period with programme running for 6 months only prior to initiation of the study. The microscopy centres were placed in different types of health institutions, with differing staffing pattern and functions. Method of data collection : Under the programme, at each microscopy centre, a single specimen of sputum is collected from ‘symptomatics’ and examined by Ziehl- Neelsen (ZN) method. For the study, these smears, along with the corresponding sputum specimens, were collected and brought to NTI laboratory, ASSESSMENT OF DIAGNOSIS OF PULMONARY T.B BY SPUTUM MICROSCOPY IN A DIRECT T.B. PROGRAMME Ind. J. Tub., Vol. XVIII, No.1 11 three times a week for independent evaluation. There, in addition to the re-examination of the original smears, a fresh duplicate smear was prepared from the material and examined by NZ; microscopy at NTI was performed by an experienced technician without knowledge of the results of centres’ examination. Futher, sputum specimens were cultured by swab method and all positive cultures were further examined for growth at room temperature, rate of growth at 37 °C, pigment production in the dark and after exposure to light, for catalase and peroxidase reactions, niacin production, and drug sensitivity to Isoniazid, Streptomycin and P-amino-salicylic acid (PAS). At NTI, smears were scanned for 5 minutes, bacilli were counted and the smears were graded as follows: Neg … If no bacilli were seen; <50 … If the actual number of bacilli is less than 50; + … If 50 or more bacilli were recorded; ++ … If masses of bacilli were seen in most field ; Positive cultures were classified as follows : <20 … If actual number of colonies was less than twenty ; + … If 20-100 colonies were counted ; ++ … If more than 100 colonies were seen and +++ … If confluent growth was observed. Material was collected from the field over a period of 18 months commencing from September, 1963. The number of specimens collected at each centre and the proportion of positive among them are shown in Table 1. From the 9 centres, 1857 specimens were collected for further examination at NTI of these 4 specimens could not be examined by culture for want of adequate quantity of sputum and 106 or 5.7 percent cultures were contaminated. For another 66 specimens one smear result (either of centre examination or of re-examination at NTI) was not available. The analysis is therefore based on 1681 specimens for which all the result, viz., centre examination, re-examination at NTI, duplicate smear and culture were available. TABLE 1 Number of sputum specimens collected at each centre and the proportion of positives among them. Culture Positives Centre smear Positives Code No* Centres No of speci- mens examined All the four** results available No. % to Col.4 No. % to Col.4 1 2 3 4 5 6 7 8 1. Devanahalli 439 407 101 24.8 79 19.4 2. Doddahejjaji 278 254 21 8.3 22 8.7 3. Sulebele 199 182 23 12.6 23 12.6 4. Bidadi 193 178 22 12.4 5 2.8 5. Vijayapura 133 123 15 12.2 21 17.1 6. K.R.Pura 144 127 16 12.6 14 11.0 7. Dommasandra 131 115 15 13.0 9 7.8 8. Hesaraghatta 105 94 10 10.6 2 2.1 9. K.D.Halli 235 201 5 2.5 4 2.0 Total: 1857 1681 228 13.6 179 10.6 *These code numbers only are used hereafter (both in the text and the tables). **(1) Direct microscopy by the centre, (2) Re-examination of the centre smears by NTI technician, (3) Duplicate smear examination by NTI technician and (4) Culture examination of specimen. K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No.1 12 Analysis and Presentation: As the main objective of the study was technical assessment of sputum microscopy, these findings have been considered first. For this purpose culture results have been taken as standard for comparison (Method1). Since assessment using culture as the yard-stick can be carried out only on a limited scale, the material has been analysed further to provide information on the following two simpler methods of assessment also: (i) Assessment based upon the results of examination of duplicate and that of the re- examination of the centre smears (Method 2). (ii) Assessment based upon the result of the re- examination of centre smears only (Method 3). For each assessment method, under diagnosis* (missed cases) and over-diagnosis* (false cases) respectively have been estimated. The definition of a ‘case’, for such estimation differs under each method. A case is defined as culture positive in method1 and duplicate smear positive in method2, and re-examination smear positive, in method3. Under diagnosis and over- diagnosis have also been studied fore each individual centre. As cases were too few in some centres, the conclusions drawn should be considered suggestive only and illustrative of the methodology of analysis of such assessment data. Findings Technical assessment based on culture results (Method1): Under-diagnosis : Varying proportions of culture positive cases have been missed by the 9 centres(Table2). The last column of this table shows that microscopist at NTI had missed 15.8 per cent of the cases while on the average, centre microscopists missed 38.2 percent. The proportions missed by centres 1,4,7 and 8 (particularly centres 4 and 8) were significantly higher than 15.8 percent. Excluding *Under-diagnosis and over-diagnosis are complements of sensitivity and specificity respectively. centres for 4 and 8, under-diagnosis in the other 7 centres was reduced from 38.2 to 30.6 percent. Some idea of the reasons for under-diagnosis at the centres could be obtained by comparing percentages of culture positives missed by the centres with similar percentages for re- examinations and duplicate smear examinations. As re-examination of the centre smears reduced under-diagnosis from 38.2 to 29.4 per cent, reading variations may be one factor. The reduced percentage further dropped down to 15.8 per cent when a duplicate smear was prepared and read by NTI technician. Defective smear preparation and/or intra reader differences appear to be the other possible reason (s). As intra-reader variation of NTI technician, in classifying a certain proportion of smears as positive, may not be high between re-examination and duplicate smear examination (and is also likely to occur in both directions) it may be inferred that the smears prepared and stained by the centres were somewhat inferior. In centres 4 and 8, reading error might have been a more prominent cause than defective smear preparation and staining. Excluding these two centres, the percentage missed by re-examination (28.1 per cent)was not much different from the centre results (30.6 per cent), but under diagnosis due to defective smear preparation and staining probably persisted as the duplicate smear was able to reduce the same to 14.8 per cent  a statistically significant reduction (P<0.001) Defects in smear preparation and staining might have occurred in centres 4,5 and 7 in which the percentage of culture positives missed by re- examination was considerably higher than the percentage missed by duplicate smear examination. But, the differences were not significant, most probably because of the small numbers involved. However, improvements in smear preparation and staining are likely to further reduce under-diagnosis in these centres. Over-diagnosis : Table3 shows that 2.6 per cent of the culture negatives were declared smear positive by the centres. Duplicate smear examination at NTI had shown over-diagnosis of 1.2 per cent. Compared to this, over-diagnosis was significantly higher ASSESSMENT OF DIAGNOSIS OF PULMONARY T.B BY SPUTUM MICROSCOPY IN A DIRECT T.B. PROGRAMME Ind. J. Tub., Vol. XVIII, No.1 13 TABLE 2 Extent of under-diagnosis by different smear examinations taking culture positives as Standard. Cult. pos. missed by centre’s exam Cult. pos. missed by re-exam Cult. pos. missed by duplicate smear examination Centre No. Total cult. Pos. No. percent No. Percent No. Percent 1 2 3 4 5 6 7 8 1. 101 27 26.7 26 25.7 12 11.9 2. 21 7 33.3 8 38.1 7 33.3 3. 23 7 30.4 5 21.7 4 17.4 4. 22 19 86.4 9 40.9 5 22.7 5. 15 6 40.0 6 40.0 1 6.7 6. 16 5 31.2 4 25.0 2 12.5 7. 15 7 46.7 5 33.3 2 13.3 8. 10 8 80.0 3 30.0 2 20.0 9. 5 1 20.0 1 20.0 1 0.0 Total 228 87 38.02 67 29.4 36 15.8 TABLE 3 Extent of over-diagnosis by different smear examinations taking culture negatives as the standard Smear positive by Centre’s exam Re-exam Duplicate smear exam Centre Culture negatives No. % No. % No. % 1 2 3 4 5 6 7 8 1. 306 5 1.6 4 1.3 7 2.3 2. 233 8 3.4 1 0.4 2 0.9 3. 159 7 4.4 7 4.4 2 1.3 4. 156 2 1.3 2 1.3 0 0.0 5. 108 12 11.1 2 1.9 1 0.9 6. 111 3 2.7 1 0.9 4 3.6 7. 100 1 1.0 1 1.0 1 1.0 8. 84 0 0.0 1 1.2 1 1.2 9. 196 0 0.0 0 0.0 0 0.0 Total 1453 38 2.6 19 1.3 18 1.2 K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No.1 14 only at centre 5. Excluding this centre, this proportion came down from 2.6 per cent to 1.9 per cent for all the other 8 centres together. The significantly smaller percentage of smear positives on re-examination (1.3 per cent as compared to 2.6 per cent) shows that reading as differences contributed mainly to over diagnosis. The percentage of smear positives among culture negatives did not vary between re-examination and duplicate smear examination by NTI technician. Defective smear preparation and staining, therefore, might not have influenced over-diagnosis. The differences between re- examination and duplicate smear examination, for different centres, could be attributed mainly to chance fluctuations in sampling of specimens for smear preparation and intra-reading differences of NTI technician. Thus, neither under-diagnosis nor over- diagnosis was a ‘serious problem’ in 6 out of the 9 centres. Of the remaining 3 centres, under- diagnosis was substantial in two and over- diagnosis in one. Resultant over-diagnosis: The proportion of false positives (as judged by culture) among the total cases diagnosed as positive by microscopy in a centre “resultant over-diagnosis”  may be of particular interest. The resultant over-diagnosis at general health institutions (Table not shown) was 21.2 per cent, for re-examination of smears by NTI technician 10.6 per cent, and for duplicate smear examination by NTI technician 8.6 per cent. The resultant over diagnosis is influenced by both under and over- diagnosis. For instance, centres 1 and 4 had nearly equal over-diagnosis (1.6 percent and 1.3 per cent) but showed 6.3 per cent and 40 per cent resultant over-diagnosis. This way largely due to under- diagnosis at centre 4 (86.4 per cent as compared to 26.7 per cent in center 1). Comparison with some other tuberculosis laboratories: The quality of diagnosis by sputum microscopy in general health institutions (in the present study) was compared with that observed in some other tuberculosis laboratories in India. Data from some published studies of Else Holst, Sikand, and Raj Narain were used to calculate under and over-diagnosis based on culture. Under- diagnosis in the general health institutions of the present study (38.2 percent) falls in the range observed in tuberculosis laboratories involved in the above mentioned studies, which varied from 26 percent to 41 percent. Similarly, making an over-diagnosis of 2.6 per cent in the general health institutions falls in the range of 1 percent to 7.2 percent reported by these tuberculosis laboratories. Thus, there is no reason to believe that microscopy at all general health institutions would be inferior to that in specialised laboratory services. On the other hand, the wide variability observed may due to differences in the material e.g. survey population, clinic patients etc., 5 and in the techniques of microscopy and culture examination ,and emphasises that these factors may be even more dominant. Methods of assessment without employing culture examination: Assessment based upon examination of duplicate smears (Method 2) : Under-diagnosis : Duplicate smears prepared and examined at NTI from 1681 specimens showed 210 positives. Percentages of these cases missed microscopy at the 9 centres are shown in Table 4. The extent of under-diagnosis was 31.9 per cent for all the centres together. Taking the duplicate smear results of NTI as the standard, all centres (except the last with only 4 positives) had missed fairly high proportions of cases (i.e. significantly higher than zero which would have been the proportion missed if centres examination had been equally efficient as duplicate smear examination at NTI). Centres 4 and 8 performed poorly in this respect. The duplicate smear positives which were missed by the centres included 13 specimens unconfirmed by culture. If these were excluded, under- diagnosis by the centres would be reduced, but this would not affect centre 4 and 8. Comprason of the percentage missed by centre examination (31.9 per cent)with that missed by re-examination (21.9 per cent) shows that reading errors may be a cause of unser-diagnosis, the difference being statistically significant : 21.9 per cent missed by re-examination itself might be partly attributed to defects in smear preparation and staining. The percentage of duplicate smear positives missed in re-examination of centre smears was quite high for centres 4,5 and 6 (significantly higher than zero). 15 K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No. Over-diagnosis: Percentages of negative duplicate smears read positive by centres and by re-examination, separately, are given in Table 5. Centre 5 showed considerable over-diagnosis and Centres 2 and 3 TABLE 4 Extent of under-diagnosis by centre’s examination and re-examination taking duplicate smear positives as the standard Duplicate smear positives Culture result of missed cases Duplicate smear positives Culture result of missed cases Missed by centre Missed by re- exam Centre No. Total No percent Pos Neg. No percent Pos. Neg. 1 2 3 4 5 6 7 8 9 10 1. 96 21 21.9 15 6 19 19.8 15 4 2. 16 5 31.2 3 2 4 25.0 3 1 3. 21 4 19.0 3 1 2 9.5 1 1 4. 17 14 82.4 14  6 35.3 6  5. 15 5 33.3 5  5 33.3 5  6. 18 6 33.3 3 3 6 33.3 3 3 7. 14 5 35.7 5  3 21.4 3  8. 9 7 77.8 6 1 1 11.1 1  9. 4 0 0.0   0 0.0   Total 210 67 31.9 54 13 46 21.9 37 9 TABLE 5 Extent of over-diagnosis by centre’s examination and re-examination taking duplicate smear negatives as the standard Smear positive by Centre Re- exam Centre No. Duplicate smear negative No . % No. % 1 2 3 4 5 6 1. 311 4 1.3 2 0.6 2. 238 11 4.6 2 0.8 3. 161 6 3.7 6 3.7 4. 161 2 1.2 4 2.5 5. 108 11 10.2 1 0.9 6. 109 2 1.8 1 0.9 7. 101 0 0.0 0 0.0 8. 85 0 0.0 0 0.0 9. 197 0 0.0 0 0.0 Total 1471 36 2.4 16 1.1 K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No.1 16 also showed somewhat showed considerable over-diagnosis and Centres 2 and 3 also showed somewhat higher than average figures, which were statistically significant ( different from zero, which is the expected proportion had centre- examination been as good as duplicate smear examination). The proportion of such ‘false’ positives was 2.4 per cent only for centre examination, but was even significantly less (1.1 per cent) for re-examination. This reduction indicates that reading errors had contributed to over-diagnosis. As only 1.1 per cent of negative duplicate smears were considered positive on re-examination of the centre smears by the same NTI technician, defective smear preparation and staining might not have contributed much towards over-diagnosis. Assessment based on re-examination of centre smears (Method3): Under-diagnosis : Table 6 shows that 23.9 per cent of positives at re-examination had not been considered positive by the centres. Proportions missed were significantly high (different from zero) for all the centres, except 2, 7 and 9. The percentage missed in centres 4 and 8, was significantly different from the other centers. Excluding these two centres, the percentage missed was reduced to 16.6. Of the remaining centres, 5 and 6 had missed fairly high proportions of re-examination positives (36.4 per cent and 30.8 per cent respectively), but not significantly higher than the remaining 5 centres. Of the 43 positives missed by the centres, 33 only were positive on culture and represent real under-diagnosis. But, 10 of these were negative on culture and might be a limitation of considering re-examination result as the standard. It is interesting that the number of culture positives missed was reduced from 33 to 18 by excluding the two centres (4 and 8) showing considerable under-diagnosis. Over-diagnosis : The extent to which re-examination negatives were considered positive by the centres is shown in Table7. It averaged 2.8 per cent for all centres, together. However, this percentage was much higher in centre 5 and higher than average in TABLE 6 Extent of under-diagnosis by centre’s examination taking re-examination smear positives as t he standard Re- examination smear positives Culture result of missed cases Missed by centre Centre No. Total No. Percent Pos. Neg. 1 2 3 4 5 6 1 79 9 11.4 6 3 2 14 3 21.4 2 1 3 25 4 16.0 2 2 4 15 11 73.3 10 1 5 11 4 36.4 3 1 6 13 4 30.8 3 1 7 11 2 18.2 2  8 8 6 75.0 5 1 9 4     Total 180 43 23.9 33 10 ASSESSMENT OF DIAGNOSIS OF PULMONARY T.B BY SPUTUM MICROSCOPY IN A DIRECT T.B. PROGRAMME Ind. J. Tub., Vol. XVIII, No.1 17 centres 1,2 and 6 as well; all differences being statistically significant (different from zero). Excluding centre 5, the proportion of ‘false’ cases in the other centres would fall to 2.0 per cent. TABLE 7 Extent of over-diagnosis by centre’s examination taking re-examination negatives as the standard Smear positives by centre Centre No. Re-exam. Negatives No. % 1 2 3 4 1 328 9 2.7 2 240 11 4.6 3 157 2 1.3 4 163 1 0.6 5 112 14 12.5 6 114 5 4.4 7 104 0 0.0 8 86 0 0.0 9 197 0 0.0 Total 1501 42 2.8 The above findings differ from those under methods 1 and 2. This is due to the limitation of considering re-examination results. For example, out of 42 smear positives by centers among re- examination negatives, 13 were culture positive. Limitation of culture as standard in assessment: Table 8 shows that 9 sputa which were considered positive both by center examination and re-examination, were actually negative on culture. Of these, 4 were also positive on duplicate smear examination i.e., positive by all the 3 types of smear examination. Similarly, there were 5 culture negatives which were read positive by the centers and duplicate smear examination. Further, there was one sputum, which was considered positive both by re-examination and duplicate smear examination but was negative on culture (not shown in table). Thus, in all, there were 15 sputa which were positive by at least two of the three smear examinations but did not grow either typical or other acid-fast bacilli on culture. Absence of growth may be due to non-viability of bacilli in sputum specimens. Thus, even assessment based on culture has its limitation. Out of the over-diagnosis reported under method 1 (2.6 per cent for centre smears and 1.2 per cent for duplicate smears), up to 1 per cent (i.e., 15 out of 1453 negative cultures) may have to be allowed on account of this limitation. This is on the assumption that culture negatives which are positive on two or more of smear examination are TABLE 8 Correlation of culture results with results of the three types of smear examination Cult. Pos. Cult Neg. Centre Smear Centre Smear Pos. Neg. Pos. Neg. Re-examination Pos. 128 33 9 10 Neg. 13 54 29 1405 Total 141 87 38 1415 Pos. 138 54 5 13 Neg. 3 33 33 1402 Duplicate smear examination Total 141 87 38 1415 K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No.1 18 all true positives. It may be emphasized here that the absence of growth on culture medium, which is highly selective for tubercle bacilli, cannot rule out presence of certain other acid-fast organisms in specimens, which could repeatedly give positive result on microscopy. Thus the actual extent of error in using culture to judge over- diagnosis by microscopy can be estimated only after further investigations. Even if all the15 sputa mentioned above are considered true positive, in addition to the 228 otherwise culture positive, the proportion of cases missed by culture works out to 6.2 per cent (15 out of 243) only and does not influence the utility of culture as the yardstick for judging under-diagnosis by microscopy. Discussion Assessment of DTP could have four major aims : 1. To measure or evaluate achievements of the institutions involved. 2. To indicate lines of corrective actions in order to improve the programme. 3. To indicate further studies needed to understand the reasons for observed shortfalls or deviations from expectations, and 4. To indicate improvements in methodology of assessment itself. In a technical assessment of the type described in this report, the relative importance of these aims would depend on the stage of development of the programme. In the initial stage, as in the present study, all the aims are likely to be equally important. It is very essential to be assured that the type of staff at general health institutions entrusted with the work of sputum collection and examination are capable of doing a technically sound job. Once this is established, and thus the modus operandi of integration of diagnostic examination for tuberculosis in general health institutions is justified, the second aim. viz., lines of corrective actions become more important, because both under and over-diagnosis are undesirable for the welfare of the individuals concerned; they affect also the assessment of treatment efficiency and the accuracy of district tuberculosis case index. The third and forth aims are likely to be relatively more important in the initial stages only. For a proper appreciation of the present study, it would be helpful to consider the result keeping these four aims in view. This assessment had demonstrated beyond doubt that non-specialised staff of general health institutions, after being trained by DTC staff are technically capable of carrying out satisfactory microscopy for diagnosis of tuberculosis. This assessment could also pick up certain centres which were not upto the mark, and indicate type of defects (under or over- diagnosis) and probable reasons for the same, viz., reading variation or defects in preparation and staining of smears. The part played by reading variation in both under and over-diagnosis emphasizes the need for a better appreciation of the extent of such errors in examination of smears. It is possible that defective staining may enhance reading variation, unless the technician is thoroughly trained and has considerable experience. These aspects deserve to be studied further. Possible defects in preparation and staining of smears in some centres emphasized the need for re-orientation of personnel concerned on proper techniques and/or a check on quality of chemicals used. As the study did not include a duplicate collection of sputum by trained technicians of the NTI, it has not been possible to study the influence of a sputum collection on under or over-diagnosis at different Assessment based on Centres with under diagnosis Centres with over diagnosis Remarks Culture Centres 1,4,7 and 8 Centre 5 Provides some idea of the factors responsible for under and over diagnosis Duplicate smear All centres expect centre 9 Centres 2,3 and 5 - Do - Re-examination of centre smear All centres expect centres 2,7 and 9 Centres 1,2,5 and 6 Does not provide any idea of the factors responsible for under & over diagnosis 19 K.P. RAO, S.S. NAIR, N. NAGANATHAN AND R. RAJALAKSHMI Ind. J. Tub., Vol. XVIII, No. centres. A further assessment could also take this into account. The study has also provided some insight into 3 different methods of assessment based on (1) culture, (2) duplicate smear examination and (3) re-examination of centre-smear. The main conclusions can be summarised as follows: It would be observed that duplicate smear examination and re-examination of centre smears led to somewhat different conclusions from the assessment based on culture. It is, however, reassuring that if only those centres, which show very large discrepancies between centre smears and re-examination or the duplicate smear results, are considered to be below standard, then assessment based on direct microscopy may serve useful purpose. One serious limitation of assessment based on re-examination is that it does not provide any insight into the factors which may be responsible for the defects in each centre, and therefore, the type of corrective actions required. However, the great advantage of re-examination of centre-smears is its applicability on a large scale, thereby ensuring assessment of all districts on a regular basis. The present study has clearly shown that in such an assessment both positive and negative smears of each centre should be re- examined, because both under and over-diagnosis could occur in some centres. It may be emphasised that errors in direct microscopy limit to some extent, the utility of this method alone for the purpose of assessment. These errors may also vary according to efficiency of the assessment technician (in preparation, staining and reading of smears). Such subjectiveness may not arise in assessment based on culture which could also be extended to provide information on drug sensitivity of strains isolated. Further, results of culture could be used to have a check on quality of examination by technicians who are assigned such work, for assessments based on duplicate smear or re- examination of centre smears. On the other hand, taking culture as the standard could possibly lead to an ex-aggerated estimate of over-diagnosis (ref. Para 2,). Even so, technical assessment based on culture should be considered as the most reliable method, followed by assessment based on duplicate smear examination and then re- examination of centre-smears. However, from practical stand point, it may become necessary to use all three methods. The DTC may re-examine smears prepared and stained by general health institutions on routine basis and prepare and examine duplicate smears on a sample basis. It can take immediate corrective actions when large discrepancies are found. Institutions with culture facilities could take up, on sample basis, assessment based on culture to the extent facilities permit and supplement these by examination of duplicate smears by a well-trained technician. As stated earlier, it would also be necessary to have sputum collection made by well trained technicians, so that the effect of wrong techniques of collection can be studied. This will have to be restricted to an even smaller sample. Working methods could be so designed to assess not only technical performance of general health institutions, but also standards of district level routine assessment. Sputum examination standards are closely connected with many operational factors such as period of inservice training, types and number of staff doing sputum examination at a centre, efficiency of supervision, type of corrective actions taken etc. Efficiency is likely to be adversely affected by transfer of trained personnel and their replacement by untrained ones, inadequate supply of stains and other requirements, and quality of supplies particularly stains and slides. Some of these aspects could be studied only subjectively in the present study. Transfer of medical officers in three centres did not influence quality of smear examination, suggesting that supervision by these medical officers was either not there (or was ineffective), or that the standard of microscopy was so good that there was no room for effecting further improvement. The visit of DTC staff to these centres was reported not to have led to technical guidance to an extent to improve quality of diagnosis appreciably. Therefore, the initial training imparted to the staff for 2 to 4 weeks appears to have been satisfactory. Supplies were fairly regular and their quality was satisfactory and these might not have influenced results adversely. For a number of reasons, one of which is the influence of the training programme of NTI, the findings of this study should be considered illustrative of methodology of assessment of integrated diagnostic examination for tuberculosis. [...]... K.P RAO, S.S NAIR, N.NAGANATHAN, AND R RAJALAKSHMI Ind J Tub., Vol XVIII, No.1 ASSESSMENT OF DIAGNOSIS OF PULMONARY T.B BY SPUTUM MICROSCOPY IN A DIRECT T.B PROGRAMME 21 The success of DTP depends largely on the training imparted to staff of general health services and periodic supervision followed by corrective actions, to ensure proper standards of work In actual practice, the type of training imparted... the Institute to whom we are grateful We thank Miss C.S.Indira Bai and Miss.T.J.Alamelu for their sceretarial help REFERENCES 1 2 3 Among the 9 centres, under -diagnosis was substantial only in 2 centres and over -diagnosis in one, indicating the need for supervision and corrective training Comparisons of results on reexamination of centre smears and duplicate smears, indicated that both reading variations... bacilli by Fluorescence Microscopy , Ind Jour Med Res 47, 495-499 Sikand B.K and Ranga Rao (1958), A simple pretreatment technique for sputum comp rising the use of a combination of Pancreatin and Cetavlon for the routine cultivation of tubercle bacilli: Indian Jour Tuberc., 5 76-86 Raj Narain, A. Gaser, M.V.Jambunathan and M.Subramanyan (1963) Tuberculosis prevalence survey in Tumkur district Indian... present report is an illustration, though by no means complete, for developing such a methodology assessment based upon duplicate smears as well as re-examination of centre smears were also considered Results and limitations of these assessments are presented Re-examination of centre smears is applicable on a large scale and could ensure an assessment of all districts on a regular basis This method... laboratory for re-examination of the smears and examination of a fresh duplicate smear by ZN and culture In all, 1681 specimens were analysed (for which all four results viz., centre examination, re-examination of centre smear, duplicate smear and culture examination were available) The study showed that the nine centres had missed (under -diagnosis) 38.2 per cent of the culture positives on direct microscopy. .. variations and ‘defective smear preparation and staining’ could have influenced under -diagnosis by the centres As culture examination can be carried out only on a limited scale, some simpler methods of 4 5 Nagpaul, D.R., (1967), District Tuberculosis Control Programme in concept and outline’, Indian Jour Tuberc.,14 186-198 Else Holst, D .A. Mitehison and S.Radhakrishna (1959) Examination of smears for... imparted and the thoroughness of supervision and corrective actions are bound to vary from one district to another, and this necessitates periodical evaluation Such an appraisal based on a simple, purposeful, and systematic assessment methodology, should be considered as a series of efforts which are built into the system at appropriate levels of the organisation and should cover all aspects of the programme. .. Summary ACKNOWLEDGEMENT A study was conducted in 9 such microscopy centres in Bangalore district to find out whether quality of diagnosis was technically sound In each institution a single spot specimen of sputum had been collected from each symptomatic and examined by Ziehl-Neelsen method These smears along with corresponding sputum specimens were brought to National Tuberculosis Institute (NTI), laboratory... by assessment based upon culture in a sample of districts To pick out centres which follow defective techniques, duplicate examination could prove quite useful, on a much larger scale than is possible through assessment by culture All the three methods would be useful to develop a builtin assessment at different levels for assessment of case-finding by microscopy in the district tuberculosis programme. .. negatives as direct smear positives (overdiagnosis) The authors wish to thank the members of the Technical Co-ordination Committee of the National Tuberculosis Institute, Bangalore, in particular Dr.D.R.Nagpaul, Director, Dr.Savic, Senior WHO Officer and Dr.Gothi, Epidemiologist for their helpful criticism The study was conducted with the active participation of the staff of bacteriology department of . great advantage of re-examination of centre-smears is its applicability on a large scale, thereby ensuring assessment of all districts on a regular basis over- diagnosis) and probable reasons for the same, viz., reading variation or defects in preparation and staining of smears. The part played by reading