The Indian Journal of Tuberculosis

Vol. 47 Editorial BCG REVISITED

BCG vaccine is in the news again. Results of the 15-year follow-up of the Chingleput BCG efficacy trial conducted by ICMRs Tuberculosis Research Centre were published recently1. Some termed the results as the the great Madras failure while others called them the great BCG hoax. Editorials in newspapers questioned the continued inclusion of BCG in the immunization programme. Concerns were echoed in the Parliament too! The 15-year follow up results merely confirmed what was known since 1980, when results of the 71/2year follow-up of the same study were published2. BCG offered no overall protection against adult type bacillary tuberculosis and a low level of overall protection (27%, 95% CI - 8 to 50%) in children. Ever since BCG was first used (in 1921) as an anti-tuberculosis vaccine in humans, doubts concerning its efficacy have been raised repeatedly, but proper efficacy evaluation did not begin until 1930s. In the randomized controlled clinical trials of BCG vaccine carried out since then, its protective efficacy against tuberculosis has varied from 0 to 80%, in different populations3. In general, the protective efficacy was found to be greater in trials conducted in northern latitudes, where tuberculosis prevalence is low compared with studies elsewhere. Observational studies, including those in contacts and populations have confirmed the variability in the level of protection. Variability in protective efficacy has been observed not only in the different trials but also between sites of the disease. However, a meta-analysis has shown that summary BCG protective effect against miliary tuberculosis or tuberculous meningitis in randomized control trials was 86%, and in case control studies it was 75%4. Further, case control studies carried out in India (not included in the quoted meta-analysis) also show a protective efficacy of BCG against tuberculous meningitis of not less than 75%. It is precisely because of the beneficial effect that the continued use of BCG among infants in immunization programme has been recommended. The current strategy is for only one BCG vaccination at birth. To date, there is no evidence that repeat BCG vaccination or boosters increase the protection. A major issue affecting BCG vaccination is the duration of protection. A quantitative review of annual change in efficacy in ten randomized trials of BCG has shown that even when BCG gives good initial protection, it may wane with the time, at a rate corresponding to 5% to 14% annual decrease in the rate ratio of tuberculosis in unvaccinated compared to vaccinated individuals5. Published data do not provide evidence of the protective effect of BCG beyond ten years after vaccination. This observation is critical, when trying to assess the value of BCG vaccination as part of a tuberculosis control programmes because BCG is given in infancy while the major burden of the disease is in adulthood. There is, therefore, an urgent need for an improved vaccine which does what BCG fails to do. The availability of the whole genome sequence of M.tuberculosis, which provides access to the entire antigenic and epitopic repertoire, coupled with better understanding of cellular immunity, the availability of cytokine assays for epidemiologic studies and significant advances in mycobacterial research, using modern tools, provides an ideal platform to launch a new vaccine or an improved possibility of developing better vaccine candidates. As alternatives to BCG, there are several exciting experimental approaches, including rationally attenuated M.tuberculosis, recombinant BCG, over-expressing M.tuberculosis antigens, protein

New Delhi, January, 2000

No. 1

EDITORIAL

sub-unit vaccines (antigens of 39, 11, 31, 8.4 kDa and a fifth unspecified antigen have been studied extensively out of approximately 80 proteins which have been identified)6 and DNA vaccines (so far three types of antigens have been tested - Ag 85, Pst S-1, and hsp 65)7. These candidate vaccines are at various stages of development and it will be quite some time before any of these advance, from bench to bush. Even if an improved vaccine were available for human trials, the widespread use of BCG would present important ethical and methodological issues related to efficacy testing. It may be difficult to find young populations that are at an appreciable risk of getting tuberculosis, yet have not received BCG vaccination. Most of the high risk populations that have not received BCG vaccination are probably not suitable for long term follow-up required in a mycobacterial vaccine trial; the scarcity of BCG-free populations means that careful planning will be needed for any future trial. Till then, BCG should continue to be used for infants in the immunization programme in view of its beneficial effect in protecting children against the dreaded childhood forms of tuberculosis - meningeal and miliary tuberculosis. The results of Chingleput trial are a forceful reminder to the scientific community as well as the funding agencies to step up their research efforts for an improved vaccine against tuberculosis by pooling resources, both intellectual and financial.
Lalit Kant REFERENCES
1. Tuberculosis Research Centre, Fifteen year follow-up of trial of BCG vaccines in south India for tuberculosis prevention; Ind. J. Med. Res.; 1999, 110, 56 2. Tuberculosis Prevention Trial, Madras. Trial of BCG vaccines in south India for tuberculosis prevention; Ind. J. Med Res,; 1980, 72 Suppl.: 1 3. Fine PEM; BCG vaccination against tuberculosis and leprosy; Br Med. Bull.; 1988, 44, 704 4. Rodrigues LC, Diwan VK, Wheeler JG. Protective effect of BCG against tuberculosis a meta-analysis; Int. J. Epid .; 1993,22, 1154 5. Sterne JAC, Rodngues LC, Guedes IN. Does the efficacy of BCG decline with time since vaccination?; Int. J. Tuberc. Lung. Dis.; 1998, 2, 200 6. Jacobs GG, Johnson JL, Boom WH, Walis RS, Whalen CC, Ginsberg AM; Tuberculosis vaccines : how close to human testing?; Tubercle and Lung. Dis. 1997, 78, 159 7 Huygen K.; DNA vaccines: Application to Tuberculosis; Int. J. Tuberc. Lung Dis.; 1998, 2, 971

Leading Article

Ind. J. Tub., 2000. 47,3

TUBERCULOSIS AND DIABETES : AN APPRAISAL

Since ancient times, physicians have been aware of the association between tuberculosis and diabetes mellitus: perhaps the earliest to note it was the great Indian physician Susruta, in 600 A.D, while Avicenna (780-1027 A.D.) had commented that phthisis frequently complicated diabetes.1 The global burden of diabetes mellitus was estimated by the World Health Organisation2 in 1998. It has been projected that the prevalence of diabetes among adults world wide will more than double, from 135 million (4%) to 300 million (5.4%), by the year 2025. The major part of this tremendous increase will occur in developing countries, like India and China, wherein a 170% increase, from 84 million to 228 million is projected. With the revision of the criteria for the diagnosis of diabetes (Appendix), by the American Diabetes Association, in 19973, which are simpler to apply compared with those proposed by the National Diabetes Data Group of the National Institute of Health, in 19794, the prevalence rates of diabetes are expected to increase further. Tuberculosis has already been declared a global emergency by the WHO5 in 1992, with an estimated one third of the worlds population infected with Mycobacterium tuberculosis and tuberculosis recognised as the single biggest killer. Now, with diabetes assuming epidemic proportions in the first quarter of the 21st century, it is imperative to take measures for the prevention and control of this deadly duo. Diabetes mellitus is also recognised as an independent risk factor for developing lower respiratory tract infections6. Whereas infections with Staphylococcus aureus, Gram-negative bacteria, and fungi occur more frequently, those with organisms like Streptococcus, Legionella, and Influenza cause significantly more morbidity and mortality7.8. Tuberculosis occurs with an increased frequency in diabetics 9.10 and causes a significantly greater mortality11.12. Increased reactivation of tuberculosis lesions has also been recorded in diabetics13. At the same time, tuberculosis appears to aggravate diabetes, with patients requiring higher than before doses of insulin. The incidence of diabetes as such appears to be higher among tuberculosis patients14-16 as compared to the general population.
Tuberculosis Complicating Diabetes Mellitus

In 1883, Windle autopsied 333 known diabetic subjects and observed pulmonary tuberculosis in more than 50% of them17. In a classic study, Root reported that 2.8% of 1373 hospitalised diabetics had pulmonary tuberculosis. Of the 750 juvenile diabetics, 1.6% had tuberculosis as compared to 0.12% among school children. After studying the association between diabetes and tuberculosis, he made the following observations: (i) The development of tuberculosis occurred ten times more frequently in juvenile diabetics. (ii) In 85% of the patients, tuberculosis had developed after the onset of diabetes. (ii) The occurrence of pulmonary tuberculosis increased with the duration of diabetes. Hence, he concluded that a diabetic patient appeared doomed to die of pulmonary tuberculosis if he succeeded in escaping diabetic coma. Root, however, postulated that the association between the two diseases was one sided i.e., diabetic patients tended to contract tuberculosis but the reverse was rare. The Philadelphia survey revealed that 8.4% of the 3,106 diabetics had pulmonary tuberculosis as compared to 4.3% of the 71,767 presumably healthy industrial workers10. Tuberculosis was present in 17% of the diabetics who had had the disease for more than 10 years compared to 5% in the diabetics with less than 10 years of the disease. A higher prevalence of tuberculosis was found in diabetics requiring more than 40 units of insulin per day. Diabetes mellitus was present in 8.3% of the cases of reactivation tuberculosis in New York city1, the second most common association after alcoholism. A Korean study18 found that pulmonary tuberculosis had developed in 170 patients among 8,015 diabetics (2.1%) as compared to 4,935 patients among 806,698 control subjects (0.6%). Estimated annual incidence rates of bacteriologically confirmed cases of tuberculosis in diabetics and non-diabetics were 281 and 55 per 10,000 respectively. The study concluded that the relative risk of developing pulmonary tuberculosis, bacteriologically confirmed,

AMRIT GUPTAN AND ASHOK SHAH

was 5.15 times higher in diabetics than in matched controls. Some of the later studies, done in developed countries have failed to demonstrate an epidemiological association between tuberculosis and diabetes19, 20. Perhaps, this is largely due to the low prevalence of tuberculosis in these areas. However, in countries like India, diabetes remains one of the most important risk factors disposing towards tuberculosis, along with malnutrition, alcoholism and HIV infection. The prevalence of pulmonary tuberculosis in diabetics in India varies from 3.3% to 8.3%, about 4 times that of general population. Is Immune Dysfunction in Diabetics a Predisposing Factor ? A probable cause of increased incidence of pulmonary tuberculosis in diabetics could be defects in host defenses and immune cell functions (Table 1) 2 1 . 2 2 . The immune derangements predominantly involve the cell-mediated arm of the immune system. Also, the degree of hyperglycemia has been found to have a distinct influence on the microbicidal function of macrophages, with even brief exposures to blood sugar level of 200 mg% significantly depressing the respiratory burst of these cells.23.24 This is borne out by the observation that in poorly controlled diabetics, with high levels of
Table 1. List of defects in diabetics immunologic makeup and physiologic pulmonary functions
Immunologic abnormalities in diabetics Pulmonary physiologic dysfunctions in diabetics

glycated haemoglobin, tuberculosis follows a more destructive course and is associated with higher mortality. Multiple pulmonary physiologic abnormalities have also been documented in diabetics that contribute to delayed clearance of and spread of infection in the host.22 Infection with tubercle bacilli leads to further alterations in cytokines, monocyte-macrophages and CD4/CD8 T cell populations.25. 26 The balance of the T lymphocyte subsets CD4 and CD8 plays a central role in the modulation of host defenses against mycobacteria and has a profound influence on the rate of regression of active pulmonary tuberculosis.27 Glucose Intolerance in Tuberculosis In the early part of this century, the prevailing view, as suggested by Root9 was that tuberculous patients do not develop diabetes with any greater frequency than the non-tuberculous". However, Nichols14, in 1957, changed this view when he described 178 tuberculous patients of whom 5% had diabetes and a further 22% had an abnormal screening test. In India16, a multicentric study done in 1987 found the prevalence of unsuspected diabetes in tuberculous patients to be of the order of 9.7%: In males above 40 years, the rate was 17.8% compared with 5.1% in those below 40 years. And in females, the respective rates were 23.5% and 4.0%, the overall rates for males and females being 10% and 8.7% respectively. Various other Indian studies have estimated the prevalence of glucose intolerance in tuberculosis to be between 1.5% to 14%. Two recent studies conducted in Africa also had similar findings. In the Tanzanian study,28 done as per WHO criteria, in 506 consecutive patients admitted w i t h sputum positive pulmonary tuberculosis, 9 of whom were known diabetics, diabetes was diagnosed by the consecutive oral glucose tolerance tests (OGTTs), in 11 additional patients giving a crude diabetes prevalence rate of 4%. Impaired glucose tolerance (IGT) was present in 82 patients (16.2%). In comparison, a similar OGTT survey, carried out by the authors in a community, revealed prevalence of 0.9% in respect of diabetes and 8.8% for IGT. The Nigerian study29, done on 54 patients with active pulmonary tuberculosis found that 3 patients had OGTT values in the diabetic range and 20 had IGT.

Abnormal chemotaxis, adherence, phagocytosis and microbicidal function of polymorphonuclear Decreased peripheral monocytes with impaired phagocytosis. Poor blast transformation of lymphocytes. Defective C3 opsonic function

Diminished bronchial reactivity

Reduced elastic recoil and lung volumes. Reduced diffusion capacity Occult mucus plugging of airways Reduced ventilatory response to hypoxaemia and

Adapted from: Infections in Diabetes Mellitus21-22

TUBERCULOSIS AND DIABETES : AN APPRAISAL

An 8 year study, done in Japan30 from 1987 to 1994, revealed that in 2,659 patients the prevalence of diabetes in cases of active tuberculosis was 13.2%: In males above 40 and 50 years of age, the rates were 22% and 21.3% respectively. The prevalence in males was significantly higher than in females. Moreover, the prevalence of diabetes among patients with active pulmonary tuberculosis was significantly higher during 1991-1994 compared with during 1987-1990. Impaired glucose tolerance in tuberculosis is much higher than overt diabetes. Although IGT reverts to normal in a large number of cases with effective chemotherapy, the higher percentage with IGT is significant because, according to the National Diabetes Data Group of NIH, one to five per cent of patients with IGT may progress to overt diabetes, annually. Causes of Glucose Intolerance in Tuberculosis Acute severe stress is an important cause of the development of impaired glucose tolerance. Fever, protracted inactivity and malnutrition stimulate the stress hormones epinephrine, glucagon, cortisol and growth hormone, which acting synergistically raise the blood sugar level in excess of 200 mg%.31 Plasma levels of IL-1 and TNF alpha are also raised in severe illness which can stimulate the anti-insulin hormones32. Age, co-existent illnesses and alcoholism also influence the host response. Tempting though it may be to ascribe the metabolic derangements in tuberculosis to stress, the complex host parasite relationship suggests otherwise. Serum levels of adrenocortico-tropin hormone, cortisol and T3 have been found to be decreased in patients with tuberculosis.33 Clinical and sub-clinical hypoadrenalism has been described frequently in these patients.34 These abnormalities make the patients ability for a stress response doubtful. The endocrine function of pancreas has also been found to be adversely affected in severe tuberculosis, and a higher incidence of chronic calcific pancreatitis occurs in patients w i t h concomitant diabetes and tuberculosis35 leading to an absolute orrelative insulin deficiency state. A family of fatty-acid-transporter proteins in the tubercle bacillus may cause dysregulation of energy homeostasis in the disease.36 The fatty acid transporter protein gene of mycobacterium when expressed in mammalian

hepatocytes increases preferentially the uptake of long chain fatty acids (LCFAs). The LCFAs are an important source of energy for most organisms and also function as blood hormones regulating key functions such as hepatic glucose metabolism.36 Derangements of lipid metabolism have been described in patients with tuberculosis.37 Effect of Anti-tuberculosis Drugs on Blood Sugar Level Rifampicin is a powerful inducer of the hepatic microsomal enzyme system and frequently interacts with other drugs. It lowers the serum levels of sulphonyl ureas and biguanides.38 Hence, patients with co-existing diseases should have their doses of oyal anti-diabetic dnigs adjusted upwards according to plasma glucose concentration. Takayasu et al39 observed that Rifampicin induced an early phase hyperglycemia which he attributed to augmented intestinal absorption. However, no case of overt diabetes was observed and it was felt that Rifampicin was not diabetogenic. Rifabutin, a newer rifamycin, also induces hepatic metabolism but is not as potent an inducer as is Rifampicin.40 Although its interactions with Zidovudine, Fluconazole and Clarithromycin have been studied extensively, the effects of concomitant administration of oral hypoglycemics have still to be clarified. Other anti-tuberculosis drugs interfere very rarely with blood sugar level. An overdose of INH41 may cause hyperglycemia while in rare circumstances, diabetes may become difficult to control in patients on Pyrazinamide.42 Hypoglycemia may rarely be seen in patients on Ethionannde.43 Clinical Aspects of Concomitant Tuberculosis and Diabetes Symptoms of one disease often mimic those of the other. Loss of weight, loss of appetite and lassitude are common to both the diseases. The association is more common among those above 40 years of age and males appear to be at a somewhat greater risk compared with females. Holden and Hiltz43 described 106 patients with tuberculosis and diabetes in which diabetes appeared first in 48, while tuberculosis appeared first in 40 and the two conditions were diagnosed simultaneously in 18 patients, as did Sumrova come to similar findings.17

AMRIT GUPTAN AND ASHOK SHAH

Patients of tuberculosis who develop diabetes have greater clinical severity at the onset, a greater degree of lung involvement and residual changes. The diabetics who develop pulmonary tuberculosis have higher blood sugar levels and develop complications like coma and diabetic microangiopathies.17
Radiological Aspects of Concomitant Tuberculosis and Diabetes

2. 3. 4. 5.

The radiological aspects of concomitant tuberculosis and diabetes were first described by Sosman and Steidl 4 4 . They reported that diabetic tuberculosis has a special radiological pattern consisting of confluent, cavitary, wedge shaped lesions spreading from the hilum towards the periphery, predominantly in lower zones, to the extent of around 20%.I4-4S Marias46 observed lower lung field tuberculosis in 29% of patients with diabetes, as compared to 4.5% in the non-diabetic patients. However, in other studies 47 , cavitary disease and multi-lobe involvement was found to be more common in patients with pulmonary tuberculosis and diabetes. A study48 to evaluate the CT features of pulmonary tuberculosis in immunocompromised and diabetic patients compared with patients with no underlying disease was done in Japan. Diabetics and immunocompromised patients had a higherprevalence of nonsegmental distribution (30%) and multiple small cavities (44%). Unusual localisation of the disease including basal segment of lower lobe, anterior segment of upper lobe or right middle lobe occurred equally in both the groups (17% and 18%). Earlier studies17-43 had also noted that these patients have a greater extent of lesion and more frequent cavitation. Besides, there is more frequent tuberculous pleural effusion 49,51.52 Since the association of diabetes only and none of the other factors was found to be statislically significant, such radiographic presentations should be first investigated for the presence of diabetes.
Principles of Management of Co-existent Tuberculosis and Diabetes53,54

6.

7.

8. 9.

Patients with poor diabetic control should be hospitalised for stabilizing their blood sugar level. Ideally, insulin should be used to control blood sugar levels. Oral hypoglycemics should be used only in cases of mild diabetes. Drug interaction with Rifampicin should be kept in mind. Glycaemic equilibrium is essential for the success of anti-tuberculosis therapy and must be achieved in every patient with co-existent disease. The goals of therapy are: fasting plasma glucose < 120 mg% and glycated Hb <7%. Vigorous and good chemotherapy is essential. Monitoring for adverse effects, particularly of hepatic and nervous systems should be done. Use of potentially neuropathic agents (INH) in patients with peripheral neuropathy demands special consideration with mandatory administration of pyridoxine. Duration of chemotherapy is entirely dependent upon the control of diabetes and response of the patient to treatment. A longer treatment course may be needed. Supportive therapy for diabetes must be actively pursued. Management of co-existent illnesses, malnutrition and rehabilitation of the alcoholic diabetic remain of prime concern.

PROPHYLAXIS

1. Proper care.

All diabetics require regular medical examination and bi-annual chest radiograph. This should be followed more rigorously in patients who are more than 40 years of age or with weight less than 10% of the ideal body weight. Any diabetic- who suddenly develops cough, loss of weight, abnormal chest radiograph or needs increasing doses of insulin to control blood glucose should be investigated for presence of tuberculosis. The American Thoracic Society recommended, in 198656, that diabetics, particularly poorly controlled IDDM patients, should be given INH chemoprophylaxis. Although, primary chemoprophylaxis may be useful in certain communities with high prevalence rates of diabetes and tuberculosis, like the Oglala Sioux natives of North America, there seems to be no reason for using primary chemo-

TUBERCULOSIS AND DIABETES : AN APPRAISAL

prophylaxis in well controlled diabetic patients, in general. Secondary chemoprophylaxis in tuberculin positive diabetic patients is generally recommended, although some investigators have questioned the actual benefit to the diabetic patient.56 Amrit Guptan & Ashok Shah Department of Respiratory Medicine, Vallabhabhai Patel Chest Institute, University of Delhi, Delhi REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Barach JH. Historical facts in Diabetes. Ann Med Hist; 1928, 10,387 King H, Aubert RE, Herman WH: Global Burden of Diabetes 1995-2025: Prevalence, numerical estimates and projections. Diabetes Care 1998 21, 1414 Report of the Expert Committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 1997, 20, 1189 National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979, 29, 1039 Malin A S, McAdam K P W J: Escalating threat from tuberculosis: the third epidemic. Thorax 1995, 50, 37 W i n t e r b a u e r R, Bedon G, Ba l l W: Recurrent Pneumonia : Predisposing illness and clinical pattern of 158 patients. Ann Intern Med 1969, 70, 689 Isreal H, Weiss W, Eisenberg G. et. Al: Delayed resolution of pneumonia. Med Clin North Am 1956, 40, 1291 Kirtland S, Winterbauer R, Dreis D. et. al. A clinical profile of chronic bacterial pneumonia. Report of 115 cases. Chest 1994, 106, 15 Root HF. The association of Diabetes and Tuberculosis. NewEnglJMed 1934, 210, 1:78, 192 Boucot K, Cooper P, Dillon E, et. al: Tuberculosis among Diabetics. The Philadelphia survey. Am Rev Tuberc l952, 65(Suppl. 1): 1 Kondo A, Sakatani M, Tsuchiy a T, et. al: Multidisciplinary analysis of chronic excretors of M.tuberculosis bacilli, Kekkaku 1996, 71, 25 S l u m L V, A t a g u n M: Diabetes m e l l i t u s at t h e tuberculosis division of Baltimore City Hospitals Maryland Med J 1963, 12, 10 Edsall J, C o l l i n s J, Gray J. The reactivation of tuberculosis in New York city in 1967. Am Rev Respir Dis 1970, 102, 725 Nichols GP. Diabetes among young tuberculous patients. Am Rev Tuberc 1957, 10, 16 Zack MB, Fulkcrson LL, Stein E. Glucose intolerance in pulmonary tuberculosis. Am Rev Respir Dis 1973, 108, 1164 Research Committee of the Tuberculosis Association

of I n d i a : Prevalence of diabetes me l l i t u s among patients of pulmonary tuberculosis. Ind J Tub 1987, 34, 91
17. Smurova TF. Lung tuberculosis w i t h associated diabetes mellitus. Excerpta Medico Chest Dis Thorac Surg Tuberc 1980,37,660

18. 19.

20. 21. 22. 23.

24. 25.

Kirn SJ, H o n g YP, Lew WJ et a l . I n c i d e n c e of pulmonary tuberculosis among diabetics. Tubercle Lung Dis 1995, 76: 529 Hcndy M, Stableforth D. The effect of established diabetes mellitus on the presentation of infiltrative pulmonary tuberculosis in the i m m i g r a n t A s ia n community of an inner city area of the United Kingdom. Br J Dis Chest 1983, 77, 87 Masztalerx J, Miller M. The role of diabetes as a factor for increased risk of infection wi t h tuberculosis. Pneunionohgia Polska 1990, 58, 378 McMahon MM, Bistrian Bruce R. Host defenses and susceptibility to infection in patients with diabetes mellitus. Infect Dis Clin North Am 1995, 9, 1 Koziel H, Koziel MJ. Pulmonary complications of diabetes mellitus. Infect Dis Clin North Am 1995, 9, 67 Repine J, Clawson C, Goetz F. Bactericidal function of neutiophils from patients with acute bacterial infections and from diabetes. J Inject Dis 1980, 142, 869 M a c R u r y Sm, Gcmtnel C G, Patcrson K R et al Changes in phagocytic function with glycacmic control in diabetic patients. J Clin Pathol 1989, 42, 1 143 Tsukaguchi K, OkamuraH, Ikuno M ct al. The relation between diabetes mellitus and IFN-gamma, I L - 1 2 and I L - I O production by CD4+ T cells and monocytcs in patients with pulmonary tuberculosis. Kekkaku 1997. 72, 617
Karuchunskn MA, Gcrgcrt Via, lakovlcva OB. Specific features of cellular immunity of pulmonary tuberculosis in patients w i t h diabetes mellitus; Problem Tnberk 1997, 6, 59 Yu CT, Wang CH, H u a n g TJ et, a l . R e l a t i o n of bronchoalvcolar lavagc T lymphocyte subpopulations to rate of regression of active pulmonary tuberculosis Thorux 1995, 50, 869 M u g u s i F, Swai A B, A l b c r t i KG. M e l a r t y !)G Increased prevalence of diabetes mellitus in patients with pulmonary tuberculosis in Tan/ama Tubercle 1990,71: 271 Oluboya PO. Erasmus RT. The significance of glucose intolerance in pulmonary tuberculosis. Tubercle 1990. 71, 135 Yamagishi F, S u z u k i K, Sasaki Y ct al Prevalence of coexisting diabetes mellitus among patients with active p u lmo n ary tuberculosis. Kekkaku 1996, 7 1 , 569 Shamoon M, H c n d l c i R, S h e r w i n R. S y n erg istic i n t e r a c t i o n among a n t i - i n s u l i n h o r m o n e s in t h e pathogcncsis of stress hyperglyccmia in humans J CIni

26.

27.

28.

29

30.

31.

A M R I T GUPTAN AND ASHOK SHAH

32. 33.

34. 35. 36. 37. 38. 39. 40 41

42 43 44 45 46 47 48

Endocnnol Metab 1981,52, 1235 Fong Y, Marano MA, Moldawer LL ct. al. The acute splanchnic and peripheral tissue metabolic response to cndotoxm in humans. JClin Invest 1990, 85, 1896 Sinurova TF, Egorova IL. Endocrine disorders and principles of thejr correction in patients with pulmonary tuberculosis and concomitant diabetes mellitus. Klinichex Media, 1993, 71, 58 Prasad GA, Sharma SK Kochupillai. N. Adrenocortical reserve in patients with tuberculosis. Indian J Chest Ms Allied Sci 1996, 38, 25 Mollcntzc WF, Pansegrouw DR, Stcyn AF. Diabetes mellitus, pulmonary tuberculosis and chronic calcific pancreatitis revisited. South Afr Med J 1990, 78 : 235 Hirsch D, Stahl A, Lodish HF. A family of fatty acid transporters conserved from mycobactenum to man. Proc Nail Acail Sci USA 1998, 95, 8625 Geigorcva G Y. Scrum lipids in pulmonary tuberculosis patients with various types of diabetes mellitus. Problem Tnherk 1997, 3 : 56 Acocella G, Canti R. Interaction of rifampicin with other drugs. Tubercle 1980, 61, 171 Takayasu N, Yamada T, Muria H et. al. Rifampicin induced curly phase hyperglycemia in humans. Am Rev Rexpir Dis. 1982, 125, 23 Blaschkc TF. Skinner MH. The clinical pharmucokmcticsof rifabutin. Clinical Infect Dm 1996, 22 (Suppll), 515 Mandcll GL, Sande MA. Drugs used in the chemotherapy of tuberculosis and leprosy. In Goodman a n d O i l m a n s - The p h a r m a c o l o g i c a l b a si s of Therapeutics. Ed. Oilman AG. Goodman LS. Gilman A. 6th Edition Macmillan Publishing Co. Inc. New York, 1980. p1200 Girling DJ. Achersc effects of antitubcrculosis drugs. Dnigi 1982, 23:1, 56 Holden HM, Hiltz JE. The tuberculous diabetic. Can Med Asso J 1962. 87, 797 Sosman MC, Steidl JH. Diabetic tuberculosis. A J R 1927, 17, 625 Weaver R. U n u s u a l radiographic presentation of pulmonai) tuberculosis in diabetic patients. Am Rev Respir Dis 19974, 109, 162 Marias R M . Diabetes mellitus in black and coloured tuberculosis patients. South Afr Mcd J 1980, 57, 483 Umut S, Tosun GA, Yildirin N. Radiographic location of pulmonaiA tubeiculosis in diabetic patients. Chest 1994. 106. 326 Ikezoe J.Takeuchi N, Johkoh Tctc. al. CT appearances of p u l mo n a r y tuberculosis in diabetic and immunocompromiscd patients : Comparison w i t h

49. 50. 51.

52.

53. 54. 55.

56.

patients who had no underlying disease. A J R 1992. 159, 1175 Kuaban C, Fotsin JO K o u l l a Shiro S et. al. Lower l u n g field tuberculosis in Yaounde, Cameroon. Central Afr J Mcd 1996, 42, 62 Boucot K. Diabetes m e l l i t u s and pulmonarv tuberculosis. J Chronic Di.i 1957, 6, 256 Spencer D, Yagan R, Blinkhorn R, Spagnuolo P.I Anterior segment upper lobe tuberculosis in the adult Occurrence in primary and reactivation disease. Chest 1990, 97, 493. W ilcke JT, Askgaard DS, NyboJensen B, Dossing M. Radiographic spectrum of adult pulmonary tuberculosis in a developed country. Respir Mcd 1998, 92, 493 Shah A : Tuberculosis and diabetes. Ann Indian Med A.isoc Accid Med Specialities 1991, 5, 54 American Diabetes Association : Standards of medical care for patients with diabetes mellitus (Position Statement). Duiheie.i Care 1997, 20 (Suppl 1) s8 American Thoracic Society. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Re.ipir Dix 1986, 134, 355 Rose D, S i l v e r A, Schechter C. Tuberculosis chcmoprophylaxis for diabetics : Are the benefits of isoma/id worth the risk? Mount Sinai Med J 1985, 52, 253.

APPENDIX Criteria for the Diagnosis of Diabetes Mellitus*

1. Symptoms of diabetes plus casual2 plasma glucose of > 11.1 mmol/L(200mg/dL) or 2. Fasting 3 plasma glucose of > 7.0 mmol/L (126 mg/dL)
or

3. 2h plasma glucose of > 1 1.1 mmol/L (200 mg/dL) during an OGTT with 75 g anhydrous glucose * Adapted from the report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, l997. 1. The classic symptoms of diabetes include polyuna. polydipsia and unexplained weight loss. 2 Casual is defined asalany time of the day without rcgaid to time since last meal 3 Fasting is defined as no caloric intake for at least 8 hours. In the absence of unequivocal hypcrglycaemia with acute metabolic decompensation, these c r i l e n a should be confirmed In repeat testing on a di Herein day. The third measure is not lecommendcd for routine clinical use.

Original Article

Ind.J. Tub., 2000:47.9

ESTIMATING MORTALITY FROM TUBERCULOUS MENINGITIS IN A COMMUNITY : USE OF AVAILABLE EPIDEMIOLOGICAL PARAMETERS IN THE INDIAN CONTEXT
A.K. Chakraborty* (Received on 13.10.98; Accepted on 12.11.99)
Summary : Occurrence of tuberculous meningitis (IBM) and deaths from it are related to the epidcmiological situation of tuberculosis in a community. Since information from observational studies is lacking in India., a simple method of estimating these from available data is presented. Transmission of infection, expressed as Annual Risk of Infection (ARI) is known to have a parametric relation with TBM deaths, among the 0-4 year aged: 1% of the ARI could be expressed as estimated IBM deaths. Applying this parameter to available data on natural dynamics of tuberculosis, derived from a 23 year epidemiological survey conducted in Bangalore district, it is estimated that in an average Indian district of 1.5 million population, about 25 TBM deaths are likely to occur annually (1.5/100,000 population). The reduction in the number of transmitters and consequently the ARI, as observed in the NTI Longitudinal Study area, is estimated and its possible effect on the estimated TBM deaths is discussed. Key Words: Childhood tuberculosis, Meningeal tuberculosis. Epidemiology, Animal risk of tuberculosis infection.

INTRODUCTION

Transmission of tuberculosis infection is generally recognised as indicative of the tuberculosis situation in a community. It is measured by estimating the Annual Risk of Infection (ARI) i.e., the number of newly infected children in a years time.1 ARI is considered as the direct consequence of tuberculosis infectious cases prevalent in a community, the number infected per prevalence case giving the dimension of transmission. The latter could be expressed as contagion parameter1 or infectivity in a broader sensed It is important for paediatricians to observe that while incidence of infection is studied in children, the sources of transmission happen to be in the adult population. A working relationship between the two is described in the literature.3 The incidence of tuberculous meningitis (TBM) is of great interest and concern, to paediatricians in particular because of the urgency in its clinical presentation. Its occurrence could also be taken as a marker for the size of tuberculosis situation in a community, because of the relatively stable relationship it is reported to have with ARI4. In a recent study in South Africa, the incidence of TBM was observed to be higher among the comparatively

marginalised people, socially as well as geographically5, apart from predominantly occurring among those aged 0-4 years. The occurrence of TBM, and clinical tuberculosis of all forms in children have been mathematically related up in the above study, making it possible to estimate one from the other. Following widespread BCG vaccination among children, under the Universal Immunization Programme (UIP), a reduction in the incidence of TBM is expected to occur in India. However, no baseline data on incidence of TBM seem to be available, in the absence of which it may not be possible to discern changes, if any. It is difficult to diagnose TBM accurately in the peripheral health/ medical institutions in the country. The problem of confounders of diagnosis or the so called background illness-variables present in the hospital records, if available at all, would seriously come in the way of arriving at a reliable estimate. The disease is also not notifiable. It may be difficult to obtain an estimate of TBM incidence from the population based repeat surveys, in view of the acute nature of the illness and death supervening in a considerable proportion of children between surveys, besides the problems of diagnosing TBM. In this context, a mathematical model approach could be useful,

*Maj: (Dr.) A.K. Chakraborty, Epidemiology Analyst, BIKALPA, 557, 4th Block, 8lth Main, Koramangala, Bangalore 560 034 Fax: 080-5525395.

10

A.K.CHAKRABORTY Annual Tuberculosis Infection Rate (in %)

X 1000

wherein deaths from TBM could be estimated by using the parametric relationship derived from some of the published studies.4 Estimates of death from TBM are useful for programme planning as well as assessment of the tuberculosis situation. Therefore, an attempt has been made to estimate tuberculosis transmission, and deaths from TBM, in young children in a community, using some of the epidemiological parameters, currently available.
METHOD

Prevalence of culture + cases (in 100,000)

The present paper computes the contagion parameter, infectivity, mortality rate due to TBM and preventable mortality from TBM through District Tuberculosis Programme (DTP) intervention. The data used are from the following sources: 1. Longitudinal Survey conducted by the NTI6-9 2. Tuberculosis Research & Surveillance Unit, The Hague (TSRU) - Report NO. 14 3. Report on monitoring of DTP10 The hypothesis used in the paper is given under the broad heads A and B: A. Sources of Infection in a community and extent of transmission A. 1 Prevalence of the infected persons, culture positive and smear positive cases6 i) Population (Longitudinal Survey) 43,8896 ii) Infected 29.5%6 iii) Uninfected 70.5% iv) Culture positive cases (178) (405/100,000)6 v) Smear positive cases (83) 47% of culture positivity (189/100,000) vi) ARI - 1.1% in 1961-62 1%7 A.2 Two slightly different definitions are used: For computing infectivity2, all the culture positive cases, and for contagion parameter1, only the smear positive cases (confirmed on culture) are considered. i) Contagion Parameter Annual Tuberculosis Infection Rate (in %)
x 1000 Prevalence of sources of infection, i.e., Smear + cases (in 100,000)

B. TBM Mortality The following criteria were used for calculating the TBM mortality rate, and for estimating deaths in absolute numbers, including those preventable by intervention through DTP in an average Indian district: B.I TBM Mortality Rate i) TBM mortality is considered only in 0-4 year age group. ii) ARI is taken as 1 % (see A. 1) iii) The ratio of annual mortality from TBM in 0-4 year age group to ARI: 0.7% - 1% (say 1%)4 iv) Mortality rate due to pulmonary tuberculosis (NTI Survey area) - 90/ 100,009 v) Proportion of children aged 0-4 years in the total population - 15%6 B.2 TBM mortality (absolute number) in an average Indian district: i) Population - 1.5 million ii) Population aged 0-4 years (15%) - 2,25,000 iii) Newly infected (at ARI - 1%) - 2,250 B.3 Preventable TBM mortality through DTP intervention i) Efficiency of case-finding in an average DTP - 33% of the potential of 2,000 cases10 ii) Average number of cases actually diagnosed in a DTP 660 iii) Case per household 1.02 iv) Size of average household 10 v) 0-4 years aged among contacts at(iv) 15% vi) Infected among child contacts of case households 26%11 vii) Uninfected among (v) 74% viii) Incidence of infection in case households, being higher in children in general popn. say, ARI x 2.0 2.0%
RESULTS

ii) Infectivity

Sources of infection in a community and extent of transmission

ESTIMATING TUBERCULOUS MENINGITIS MORTALITY

11

Table 1 shows that annually, on the average, about five persons are infected by a smear positive case and three by a culture positive case of tuberculosis. Table 2 shows that the sources of infection would decline with time, as per the results of repeat surveys carried out in Bangalore rural district by the NTI, over a period of 23 years 7 . Simultaneously, with the ARI declining on the average by 2.3% per year, the smear positive case prevalence is observed to have declined by 2.8% per year. This gives the estimate of a likely correlation between the extent of transmission and the size of sources of infection in the community. IBM Mortality As per the hypothesis B.I (lii), annual TBM mortality among 0-4 years old children is taken as 1% of ARI. With ARI itself being 1% (see A.I), there would be one TBM death in 10,000 children (0.01 x 0.01) (Table 3). The mortality rate from TBM in general population is thus calculated to be 1.5/ 100,000. Comparing the data given in B.I (iv), the ratio of TBM mortality to pulmonary tuberculosis mortality in general population, could be expressed as 1.5 : 90.0 (or 1 : 60). Size of problem in an Indian district In an average Indian district of 1.5 million

population, the likely number of deaths due to TBM could be 23 (say 25), at the rate of 1.5/100,000 (Table 3). The likely occurrence of death from TBM among contacts of the cases in a DTP could be calculated as under: Population in case house holds = 660 x 5 = 3,300 (see B.3., I-iv). Uninfected among them in age group 0-4 years = 366 (see B.3., v and vi). Incidence of infection in 0-4 group (366 x 2%) = 7.0 (as per ARI given in B.3.viii) - Deaths from TBM in a year 7 x l % = 0.07, say 0.1 (see B.I, iii for the ratio of TBM deaths to ARI). - Therefore, approximately one death is estimated to occur once in 10 years due to TBM among the household contacts of the smear positive cases diagnosed by the DTP, with a case-finding efficiency of 33%. DISCUSSION Tuberculosis infection rates in children, converted into a set of ARIs over time may reflect the trend of tuberculosis in the community as a whole. ARI could be correlated with incidence of sputum smear positive cases in the community,3 or the size of the transmitters of infection in the community who

Table 1. Sources of Transmission of Infection Prevalence of cases (per 100,000) ARI % Smear 4 1 189 Culture + 405 Contagion parameter Col.l/2x 1000 5.3 (say 5) Infectivity Col. 1/3 x 1000 2.5 (say 3)

Table 2. Decline in Transmission by Programme Efficiency* Time of Survey Programme efficiency ARI (%) Prevalence of Cases (Per 100,000) SM Cultuie + + 189 68.0 405 438 Decline (%) (Annual) Sm + 64.0 (2.8)

# ARI 41.0 (2.3)

Culture + Nil

1962 1986

0@
33%** 0.61

1.1

* Only pertains to the aiea surveyed repeatedly for 23 yeais7 # Based on detailed calculation, given in Chakraborty et al7 @) The area studied did not have District Tuberculosis Piogiammc (DTP) upto 1971
** Aveiage efficiency observed for DTP implemented in 1972 - hypothetical lime of the efficiency calculation, middle of 1979

12 A.K.CHAKRABORTY are the target for any tuberculosis control programme. is true that BCG vaccination is highly effective in The average number of persons infected by a case preventing haematogenous dissemination). (smear positive and culture positive, separately) is However, those who recommend chemoalso calculated, in this paper, making it possible to prophylaxis and treatment of child contacts of estimate the damage and its public health smear positive index cases should note that by significance. The estimated decline in the prevalence carrying out surveillance and preventive of smear positive transmitters (by 2.8% per year) is treatment, only among the child contacts of the reported to be related to that of the ARI (by 2.3% tuberculosis cases registered under DTP, one per year) in a population in south India, surveyed by could possibly prevent just one death from TBM the NTI, for the last 23 years.8 The above trend could in 10 years with DTPs current efficiency, and be attributed to the natural dynamics, accentuated to an un-estimated morbidity in them provided some extent by socio-economic changes that are chemoprophylaxis is delivered at 100% taking place in the area, along with the effects of efficiency and is truely effective in preventing intervention through DTP, if any. disseminated disease). The operational feasibility TBM, being one of the serious consequences of of undertaking such a surveillance and preventive tuberculosis infection, is of major concern to treatment activity in the country, under the current paediatricians as well as programme planners. Its socio-economic constraints, can only make it an occurrence or TBM mortality is sometimes taken as illusory objective for any health programme to a marker of the tuberculosis situation in the area achieve. and effect of the intervention effort. Since notification TBM deaths are shown as directly proportional of TBM, or tuberculosis of any other form is note to ARJ, the latter showing a reduction at 2.3% per practised in India, it would be desirable to estimate year, as reported in the NTI Longitudinal Study area the size of the TBM from the ARI,4 to review the over a period of 23 years7. This reduction could have tuberculosis situation. brought down TBM deaths as well. Correlating the It is possible to estimate, from ARI of 1% (as reduction of smear positive transmitters to the decline observed in the south Indian district, data fromwhich in ARI and TBM deaths, it would appear reasonable are used in this model), that TBM mortality rate in to say that an efficient DTP could prove effective in the population could be 1.5/100,000 (cf: reducing TBM deaths in the community as well, over pulmonary tuberculosis mortality rate : 90/ the long term, especially when supported by BCG 100,000). For an average Indian district (population vaccination, as done at present. considered as 1.5 million), nearly 25 TBM deaths There is paucity of reliable information on could occur in a year. The estimated number would incidence of TBM and mortality from it in general vary, depending on the ARI (1.7% in Tamil Nadu12) population. The data available from a recent which is not likely to be uniform throughout the observational study from South Africa show that out country12a. The range of TBM deaths could be of the infected children, in 0-4 year age group, 15.7% appropriately calculated from the area-specific had suffered from clinical tuberculosis and 0.5% had ARI, as per the method shown in this paper. developed TBM (ARI - between 2% and 3%). Our It is believed that TBM deaths could largely be TBM mortality rate (at 10/100,000 children or 1.5/ prevented by BCG vaccination of the newborns and 100,000 population) compares favourably with that observed among the 0-4 year old coloured children the non-infected children (provided, the hypothesis
Table 3. Computation of Mortality due to Tuberculous Meningitis Distt. population all ages (in Millions)# 1.5
# Children aged 0-4 years 15%

Annual Risk of Infection % 1

*1%otARI

Death from TBM in children 0-4 years (in 10,000)* 1

**No estimated tbi 1.5 million (col.1)

TBM Mortality (per 100,000 population) 15(22-23)**

ESTIMATING TUBERCULOUS MENINGITIS MORTALITY

13

of the predominantly rural areas in South Africa (incidence rate of TBM : 29.0/100,000 in 0-4 year old children with 24% deaths) because black children below the age of five years, living away from the urban and peri-urban areas, expectedly had higher TBM deaths. It is also worth noting that only 44% of the TBM cases identified were found in the records of the South Africa health department. These finding indicate the seriousness of the problem among underprivileged sections of the community, apart from underlining the constraints in the measurement of the problem through health services data. In the Indian context, the proportion of TBM fatalities out of all tuberculosis deaths was reported to be as high as 18.6%, from the vital statistics records of Nagpur Municipal Corporation13. But city hospitals attract serious emergencies like TBM more than the insidiously dying tuberculosis cases, besides the problems encountered in diagnosing TBM. Hence, sentinel centres data have their own limitations in projecting these to the community at large. It may not be out of context to mention the process of arriving at ARI in Great Britain, from the notification data on TBM. Even though Vynnyckys model14 is, in a manner, converse application of the parameters compared to the use made in our analysis, both are essentially similar approaches as they illustrate the feasibility of arriving at a set of data, not readily found, from those which are available. Even though the data on TBM mortality presented in this paper are only preliminary estimates, which also do not permit cross classifications of any sort, these could still be used in planning for services for the paediatnc age group. Needless to say that the estimates require to be validated on the lines of the study carried out in South Africa.
REFERENCES
1 2 Styblo K., Recent advances in epidemiological research in tuberculosis Advance in Tuberculosis Research 1980, 20 : Pp 1-63, Kargcr, Basel Krishna Murthy MS, Channabasavaiah R, Nagaraj AV, Cliandrasckhar P., Incidence of tuberculosis infection in a south Indian village with a single sputum positive

case : An epidemiological case study. Ind J Tub 1991, 38, 123 3. Styblo K., The relationship between the risk of tuberculosis infection and the risk of developing infectious tuberculosis. Bull Int Union Tuberc, 1985, 60, 117 4. Styblo K, Meijer J, Sutherland L. The transmission of tubercle bacilli, its trend in a human population. Tuberculosis Surveillance Research Unit Report No. 1. Bull Int Union Tuberc, 1969, 49, 5 5. Berman S, Kibel MA, Fourie PB, Strebel PM., Childhood tuberculosis and tuberculous meningitis: Higher incidence rates in the Western Cape of South Africa. Tubercle & Lung Dis, 1992, 73, 349 6. National Tuberculosis Institute, Bangalore., A five year epidemiological study. Bull Wld Hlth Org, 1974, 51, 473 7. Chakraborty AK, Chaudhuri K, Sreenivas TR, Krishna M u r t h y MS, Channabasavaiah R., Tuberculosis infection in a rural population of south India : 23 year trend. Tubercle & Lung Dis, 1992, 73, 213 8. Chakraborty AK, Suryanarayana HV, Krishna Murthy VV, Krishna Murthy MS, Shashidhara AN; Prevalence of tuberculosis in a rural area by an alternative survey method without prior radiological screening of population; Tubercle & Lung Dis., 1995, 76, 20 9. Chakraborty AK, Gothi CD, Dwarakanath S, Smgh H., Tuberculosis mortality rates in a south Indian rural population. Ind J Tub 1978, 25, 181 10. National Tuberculosis Institute, Bangalore Performance of district tuberculosis programme during the year 1991. NTl Newsletter 1992, 81, 39 11. N a i r SS, Ramanatha Rao G, Chandrasekhar P. Distribution of tuberculous infection and disease in clusters of rural households, Ind J Tub 1971, 18, 3 12. Chakraborty A K . Prevalence and incidence of tuberculosis infection & disease in India: a comprehensive review, WHO SEARO, New Delhi, 1996. Pp 1-26 (Fig 1-17, Tables 1-14, App. l-vi) 12. Ma y u r n a t h S, Vallishayee RS, Radhamani MP, Prabhakar R. Prevalence study of tuberculous infection over fifteen years in a rural population in Chingleput district (south India). Ind J Med Res 1991, [A] 93, 74 13. Ramanatha Rao G. Tuberculosis mortality in an urban complex in central India: A study of long term trends. Tubercle, 1982, 63, 187 14. Vynnycky E, fine PEM. The ARI with M.tuberculosis in England & Wales since 1901. Int J Tuberc Lung Dis, 1997, 1, 389.

14

LEAVES FROM HISTORY - 3

Rene Theophile Hyacinthe Laennec (1781-1826) Laennec, the inventor of stethoscope - every physicians constant companion - was born in Quimper, France, to a poor family in 1781 Laennecs early ambitions vveie to become an engineer as well as a poet, but an uncle who was a physician, persuaded him to take to medicine. He joined the great Ecole de Medicine in Pans to become a doctoi There, he was a contemporary of Jean Nicolas Corvisart, who later became Napoleon Bonapartes favourite physician While at the medical school, he developed asthma, the deceptive guise which tuberculosis often takes Struggling with poverty as well as his weak lungs, he became acutely aware of strange happenings in his lungs and a fascination towards that part of the human anatomy Becoming interested in diseases of lungs and heart, he began deep investigations into them, dining which he invented the stethoscope, in 1816, for listening to the sounds in lungs and heart without having toapply his ear to patients chest His original stethoscope was a wooden tube with trumpet like ends He published his findings under the title, A Treatise on Mediate Auscultation, in 1819 That book changed the prevailing concepts, regarding lung diseases in general and phthisis in particular In 1819 itself, Laennec came to the conclusion that his asthma was in reality pulmonary tuberculosis for which bed rest was the treatment prescribed then With rest, his health improved and he began to work and teach, once again Physicians horn all over Europe thronged to hear his story of recovery from phthisis and teachings about how to diagnose the disease But the disease relapsed in 1826, and another pioneer fell victim to tuberculosis.

Original Article

Ind.J. Tub., 2000, 47,15

A COMPARATIVE STUDY OF TUBERCULIN REACTIONS TO 1 TU AND 2 TU OF PPD - RT23

V.K. Chadha1, P.S. Jagannath2, A.V. Nagaraj1, D. Narayana Prasad3 and N. Anantha3 (Received on 17.12.98; Accepted on 10.11.99)
Summary: Reactions to 2 TU of PPD - RT23 with Tween 80 tuberculin were compared with those of 1TU in order to determine a suitable dose for use in tuberculin surveys. The frequency distribution of tuberculin reactionsizes to I T U among unvaccmated children showed a bimodal distribution and a clear demarcation level for identification of the children probably infected with M.tuberculosis was observed at 17 mm. However, such a demarcation could not be observed with 2 TU dose. In addition, a significant proportion of the children with moderate reaction to 1 TU showed large reaction to 2 TU. Almost all the children with unpleasant skin reactions to the 2 TU dose had similar reactions to I TU as well. Comparison of the reaction-sizes, as read by the standard reader showed a high degree of correlation with those read by trained readers, between both the doses. The results do not support the thesis that a shiftover to 2 TU instead of 1 TU dose of PPD - RT23 for conducting tuberculin surveys would serve a useful purpose, at least not under situations similar to the present study area. Key Words: Tuberculin survey, PPD - RT23 with Tween; Epidemiology of tuberculosis.

INTRODUCTION

The tuberculin test is the only tool available for identification of tuberculosis infection. Hence, it is important for the study of epidemiology of tuberculosis. Of the various epidemiological indicators available for surveillance of tuberculosis (TB), the annual risk of tuberculosis infection (ARI) is generally considered to be the best single indicator for evaluating the tuberculosis problem, especially in the developing countries.1 ARI is estimated from the prevalence rate of infection obtained through scientifically planned tuberculin surveys done among children not vaccinated with BCG.1 Most of the tuberculin surveys done in India have been carried out by using 1 tuberculin unit (TU) of PPD-RT23 with Tween 80, as per the recommendations of WHO2. The National Tuberculosis Institute, Bangalore (NTI) has recently proposed a nation-wide survey for the estimation of ARI in different parts of the country. However, some experts have suggested the use of 2 TU of PPD -

RT23 with Tween 80 for the proposed survey. It had earlier been reported that the reactions to 1 TUdose were softer and newly trained inexperienced tuberculin readers committed errors in reading soft reactions. The World Health Organization4 and the International Union against Tuberculosis & Lung Diseases (IUATLD)5 have also suggested the use of 2 TU for tuberculin surveys. While 2TU is being used for tuberculin surveys in some countries,6,7 others continue to use the 1 TU dose,8 as does the NTI, in India. It was thought appropriate that a comparative study of the use of the two doses of tu b ercu lin be conducted with the following objectives: 1. To compare the tuberculin reaction-sizes to 1 TU and 2 TU doses of RT 23 with Tween 80; 2. To decide which of the two doses has a better correlation between the readings of the standard reader and of the newly trained readers; and 3. To compare the rates of unpleasant skin reactions to either dose.

1. Epidemiologist; 2. Statistical Assistant; 3. Team Leader

National Tuberculosis Institute, Ministry of Health & Family Welfare (Directorate General of Health Services), Bangalore. Correspondence: Dr. V.K. Chadha, National Tuberculosis Institute, No. 8, Bellary Road, Bangalore-560 003

16 MATERIAL AND METHODS

V.K.CHADHAETAL

The study was conducted using a double-blind design wherein each child was subjected to dual testing with the two doses of PPD - RT 23 with Tween 80. The pre-coded tuberculin vials containing 1 TU and 2 TU doses per 0.1 ml were procured from BCG Laboratory at Guindy, Chennai. Separate tuberculin syringes and sterilizers were used for the vials and labeled similarly. The individual child cards were also pre-coded for the dose to be given on a particular ami. Therefore, the sites of injection for the two tests in individual children were randomized. Each child, after ascertaining the identification particulars was examined for the presence of BCG scar by inspection of both the shoulders. Children having identifiable scar were considered to be vaccinated with BCG, while those without scar were identified as the other group. Children with doubtful scar were excluded from the study. Each child was administered dual intradermal injections with 1 TU and 2 TU doses of tuberculin on the volar aspect of the forearm. The secretary to the tuberculin tester made sure that the correct dose of tuberculin was injected on the assigned arm as per the code given on the individual child card. The same tester, having 30 years experience in tuberculin testing, gave all the tests during the study. The reactions to both the tests were read on the third day by palpating the induration and measuring its maximum transverse diameter. All the reactions were read independently by three readers: one standard reader (an experienced reader whose consistency has been tested and maintained through an assessment process9) and two trained readers (Reader I and Reader II), separately. The two test results for each individual child were read at half an hour interval of each other. The reaction sizes, as measured by each of the readers were dictated to individual secretaries. The presence of unpleasant skin reaction, seen at the test sites, like edema, bullae, vesicles and necrosis were also recorded at the same time.

For intra-reader and inter-reader comparison of tuberculin reaction-sizes, Pearsons correlation coefficient was calculated. The comparison of proportional data was made by X2 test with continuity correction. The t test was applied for comparing the mean values. For all these tests, a P-value of <0.05 was considered significant.

Study Population
The study was conducted among 5 to 9 year old school children in the selected villages of Anekal taluq of Bangalore district, from April to September, 1998. The selection of the study area was purposive so that children covered in any of the earlier tuberculin surveys carried out by the NTI were excluded. A total of 609 BCG-unvaccinated children were registered for the study. The children among whom injection of either of the two doses was unsatisfactory and the children whose reactions could not be read by all the three readers were excluded from analysis. (Fig. 1)
Fig. 1 Study P o p u l a t i o n (5 to 9 year aged BCG

unvaccmated rural children)

RESULTS

Statistical Analysis
The data analysis was performed using SPSS and Epi-Info software packages. Wilcoxon Signed Ranks test was performed for the comparison of frequency distributions of tuberculin reaction-sizes.

Frequency Distribution of Tuberculin Reactions The frequency distribution curve of tuberculin reactions to 1 TU, as read by the standard reader showed a distinct bimodal distribution (Fig. 2). The mode on the left was seen at 3-5 mm and on the right at 23-25 mm. The reactions around the left mode were distributed in the range of 0-9 mm. These reactions were termed as small reactions. The reactions around the right mode were distributed in the range of 17 mm and formed a distinct group.

COMPARATIVE STUDY OF I TU & 2 TU TUBERCULIN TESTS

17

Fig. 2 Fiequency disttibution of tubeiculm ieact.ion-sizes to I TUand2TUofPPD-RT23(Standaid Reader)

Fig. 3b Frequency distributions of tuberculin reaction-sizes to 2 TU (three readers)

These reactions were termed as large reactions The reactions between these groups i.e 10-16 mm in size were termed as moderate reactions. Therefore, following the method generally used for identifying the demarcation level,10 all reactions 17 mm to 1 TU were considered to be due to infection with M tuberculosis. The frequency distribution of tuberculin reaction-sizes to 2 TU, as read by the standard reader was statistically different from that of 1 TU (Fig. 2). Unlike that for 1 TU, the right mode and demarcation level for infected group of children could not be identified in the case of 2 TU. Thus, there was the problem of separating out the infected from the unmfected with 2 TU dose, necessitating the use of same criteria of classifying the reactions for both the tests. The distributions of reactions to 1 TU, as read by readers I and II were statistically similar to that of the standard reader; the demarcation level of 17 mm for the identification of infected children was seen in the frequency distribution of each of the

reader (Fig 3a). Similarly, the frequency distributions of reaction-sizes to 2 TU, as read by the three readers were similar (Fig 3b). Comparison of tuberculin reaction to 1 TU and 2 TU The con-elation of reaction-sizes to 1 TU and 2 TU doses, as measured by the standard reader is presented in Fig. 4. Most of the reactions were concentrated around the diagonal. However, many of the children with small reaction to 1 TU were found to have moderate reaction to 2 TU. The proportions of small, moderate and large reactions to 1 TU and 2 TU are presented in Table 1. Of the children with small reactions to 1 TU, 9.2% had moderate reaction to 2 TU and 0.2% had large reaction Of the moderate reactors to lTU, 25% had large reactions to 2 TU None of the children with large reaction to 1TU had small or moderate reaction to 2 TU. Only 1% of small reactors to 2 TU had moderate leaction to 1 TU. About 73% of moderate reactors to 2 TU had small reaction to 1 TU. None of the moderate reactors to 2 TU had large reaction

Fig. 3a Fiequency distnbutions of tubcic u l m reaction-sizes to 1 TU (three readers)

18

V.K.CHADHAETAL

Table 1. Proportion of small, moderate and large reactions to 1 TU and 2 TU according to Standard Reader.

2TU Small 1TU Mod. Large Total Small 444 4 448 Mod. 41 15 56 Large 1 5 27 33 Total 486 24 27 537
Standard Reader

Table 2. Proportion of small, moderate and large reactions according to Standard Reader 1 a) 1 TU Reader I Small Small Std. Reader Mod. Large Total 480 5 445 Mod. 6 16 1 23 b) 2 T U Reader 1 Small Std. Reader Mod. Large Total Small 440 14 454 Mod. 8 41 49 Large 1 33 34 Total 448 56 33 537 Large 3 26 29 Total 486 24 27 537\

Fig. 5a Correlation between Standard Reader and Reader I for tuberculin reactions to I TU

Standard Reader

Fig. 5b Correlation between Standard Reader and Reader 1 for tuberculin reactions to 2 TU

to 1 TU. Of the large reactors to 2 TU, 15% had moderate reaction to 1 TU and 3% had small reaction to 1 TU. Thus, many of the children with small reaction to 1 TU had moderate reaction to 2 TU; the proportion of moderate reactors to 2 TU became significantly higher. A high proportion of moderate reactors to 1 TU was labeled as large reactors to 2 TU; none of the small or moderate reactors to 2 TU was labeled large reactor to 1 TU. Similar observations, as above, were made when reactions to 1 TU and 2 TU, as measured by readers I and II were compared (Table not given).

measured by the standard Reader and Reader I are shown in Fig. 5a and that of 2 TU in Fig. 5b. Most of the reactions were scattered around the diagonal and a very high degree of correlation was seen in the case of both the doses (0.94 for 1 TU and 0.96 for 2 TU). More than 95% of the readings by Reader I
Table 3. Mean reaction-sizes in mm according to reader and dose of tuberculin

1 TU Reader Reader 1 Reader II Mean of all reactions 46(4.8) 4.7(49)

Mean of large reactions

2TU

57(5.7) 5.8(5.7) 5.8 (5 8)

Std. Reader Reader 1 Reader II

23.2 (3.2) 22.1 (29) 22.5 (3.0)

23.5 (3.9) 23.9 (3.7) 23.2 (3 9)

Inter-reader Comparison of Tuberculin ReactionSizes The correlations of reaction-sizes to 1 TU, as

Standard deviations in parentheses

COMPARATIVE STUDY OF 1 TU & 2 TU TUBERCULIN TESTS

19

-9

-I

-7

-5

-4

-3

-2

-I

difference in reaction size (mm)

Fig. 6 Frequency distributions of inter-reader differences of reaction-sizes (Standard Reader and Reader I)

were within 3 mm of the Standard Reader (Fig. 6). The proportions of small, moderate and large reactions to 1 TU, as measured by the Standard Reader and Reader I are compared in Table 2a and those to 2 TU in Table 2b. Similar observations were made on comparing the reaction-sizes, as read by the Standard Reader and Reader 11; the value of the correlation coefficient to 1 TU was 0.85 and that to 2 TU was 0.86. About 90% of the readings by reader II were within 3 mm of the Standard Reader. Mean Reaction-Sizes The mean of tuberculin reaction-sizes to 2 TU was significantly high compared with 1TU (Table 3). However, no difference was found when the respective means of large reactions to 1 TU and 2 TU alone were compared. The means oi reaction-sizes by different readers were not statistically different from one another in case of either of the doses. Proportion of Complications In all, 4.6% of the children showed unpleasant skin reactions to 1 TU and 5.3% to 2 TU test. These rates were not statistically different. There was also no inter-reader difference in these rates.
DISCUSSION

The present study was undertaken to find the difference between tuberculin tests done with 1 TU & 2 TU dose of PPD-RT23 with Tween 80. Since the dual tests were given double blind and in a standard way, the reactions to the two tests were fully comparable. The freq u en cy distribution of tuberculin reactions with 1 TU showed a bimodal pattern; the

demarcation level for identification of the group presumably infected with M.tuberculosis was obtained at 17 mm. No such distribution and demarcation level could be obtained with the 2 TU test and the distinction between the infected and uninfected groups was totally obscured due to a higher proportion of moderate reactions to 2 TU, than to 1 TU. Many of the children with small reaction to 1 TU showed moderate reaction to 2 TU and presumably got merged with the group of the truly infected children. In view of the difficulty, estimates of the prevalence of M. tuberculosis infection may have to be based on arbitrary criteria rendering them highly erroneous. Regarding the significant proportion of children showing moderate reaction to 1 TU but large reaction to 2 TU, it is difficult to identify their epidemiological significance from a cross-sectional survey. Only a prospective study which compares the incidence of tuberculosis disease among the various groups (viz, children giving moderate reaction to 1 TU but large reaction to 2 TU, large reaction to both the doses and small reaction to both the doses) could help decide their epidemiological significance. However, such a study seems to be impracticable in the epidemiological situation in south India with high prevalence of infection with mycobactena other than tuberculosis. A very large number of children will have to be followed up over a number of years. Therefore, the choice may lie with the predictive values of the 1 TU or 2 TU tests i.e., whether a more sensitive test is desirable or a more specific one in the context of the more prevalent11.13 nonspecific sensitivity compared with infection with M.tuberculosis and a significant proportion of the BCC vaccinated children not having a discernible scar14, a more specific test shall be more desirable for epidemiological surveys even if it is slightly less sensitive. A significant proportion of children in the present study showing moderate reaction to 1 TU gave large reaction to 2 TU, thus losing on specificity with the use of the higher 2 TU dose, which may lead to errors in the estimation of prevalence of infection. Moreover, a high sensitivity has been observed even with the 1 TU test during the Longitudinal Study conducted in Bangalore.15 One might have expected that the use of higher dose, of 2 TU, would cause intolerably large troublesome reactions. However, the rate of unpleasant skin

20

V.K. CHADHAETAL in high prevalence countires, Tubercle & Lung Dis., 1996, 77(suppl) 6. Bosman MCJ, Swai OB, Kwamanga DO, Agwanda R, Idukitta G & Misgenovic O: National Tuberculin Survey of Kenya 1986-1990; International J. Of TB & Lung Dis., 1998, 2/4, 272. Liared R, Tazir M, Boulahbal F and Perdrizet S: Use of two methods of analysis to estimate the annual rate of tuberculosis infection in southern Algeria; Tubercle & Lung Disease; 1996, 77, 207 Hong YP, Kirn SJ, Lew WJ, Lee EK and Han YC: The seventh nationwide tuberculosis prevalence survey in Korea, 1995; InternationalJ of TB & LungDis.; 1998, 2/1, 27 Shashidhara AN: National Tuberculosis Institute, Bangalore. An introduction to tuberculin testing and BCG vaccination, 1980, IBM Prakashana, Gandhmagar, Bangalore

reactions to 1 TU and 2 TU in this study were more or less similar. The data obtained in this study does not support the hypothesis favouring a shift-over to the use of 2 TU dose of PPD RT23 for tuberculin surveys in India.
ACKNOWLEDGMENTS

7.

The authors are indebted to Dr. (Mrs.) Prabha Jagota, Director, NTI, Bangalore for her suggestion that this study be conducted, and to all the members of the Technical Co-ordination Committee for their valuable suggestions. The authors are thankful to Mr. MS Krishnamurty and Mr. AN Shashidhara for their contributions during formulation of the study protocol. The hard work of all other members of Epidemiology Section, in data collection, that of Ms. G Sumathi & Mr. V. Magesh, in data processing and Ms Vijayalakshmi for secretarial assistance is also acknowledged.
REFERENCES
1. Styblo K, Meijer J, Sutherland I. The transmission of tuberculosis, its trend in a human population, TSRU Report No. I, Bull hit UN Tub, 1969, XLII-I 2 3. World Health Organization : The WHO standard tuberculin test, WHO/TB/Techn Cinde/3, 1963 National Tuberculosis Institute. Tuberculosis in a rural population of south India : A five year epidemiological study. Bull World Health Org.; 1974, 5 1, 473 Ten Dam HG Surveillance of tuberculosis by means of tuberculin surveys, 1985, WHO/TB/85. 1 4 5 Annadottir, Rieder HL, Trbucq A and Waaler HT: Guidelines foi conducting tuberculin skin test surveys

8.

9.

1 0 . Bleiker MA, Sutherland I, Styblo K, Ten Dam HG and Misljenovic; Guidelines for estimating the risk of tuberculous infection from tuberculin test results in a representative sample of children; Bull IUAT & LD; 1989, 64, 7 11. Chakraborty AK, Ganapathy KT and Kul Bhushan: Prevalence of non-specific sensitivity to tuberculin in a south Indian rural population, Indian J Mecl Res; 1976, 64, 649 1 2 . Edwards LB, Meijer J and Benjamin PV- Specific and non-specific tuberculin sensitivity in India; IndJ. Tub; 2; 1955, 1 2 , 6 6 1 3 . Raj Narain, Krishnamurthy MS and Anantharaman DS: Prevalence of non-specific tuberculin sensitivity in some parts of India; fnd J Med Res; 1975,63, 1098 14. Chadha VK, Krishnamurthy MS, Shashidhara AN, Magesh V: Findings of BCG scar survey in Bangalore city; Indian J Prev Soc Med; 1997, 2 8 , 8 1 1 5 . Raj Narain Nair SS, Ramanath Rao G and Chandrashekar P: Distribution of tuberculin infection and disease among households in a rural community; Bull Will Illth Org; 1966, 34, 639

4. 5

Original Article

Ind. J. Tub., 2000, 47,21

PREVALENCE OF HIV INFECTION AMONG TUBERCULOSIS PATIENTS IN DELHI - A SENTINEL SURVEILLANCE STUDY* S.K. Jain1, J.K. Aggarwal2, S. Rajpal1 and U. Baveja4
(Received on.28.5.99; Accepted on 11.11.99)
Summary: To determine the prevalence of H I V infection among newly diagnosed untreated tuberculosis patients in Delhi, serum specimens of about 400 patients between 15 and 45 years of age were collected from each of six randomly selected tuberculosis clinics of Delhi and screened by the ELISA test. Out of the total 2,361 specimens thus collected from six clinics, 16 (0.68%) were found to be positive for HIV-I antibodies. A l l the 16 H I V positives were among 1,409 males examined (1.14%), suggesting a higher risk of H I V co-infection present among males suffering from tuberculosis. Among them, illiterates and those educated upto primary level were significantly more compared to those with middle and higher education. The other risk factors for H I V and tuberculosis co-infection were (i) heterosexual relations with multiple partners, (ii) intravenous drug abuse and ( i i i ) concomitantly present sexually transmitted disease. Key Words: HIV/TR Co-infection. HIV stains of TB patients. Sentinel Surveillance of HIV/TB

INTRODUCTION

Tuberculosis and HIV infection together form a very grave public health hazard. WHO estimates that there are 2.5 million HIV seropositive persons in India 1 . There are reports that prevalence of tuberculosis infection is higher in seropositive individuals2. It is also hypothesized that persons infected with the tubercle bacillus are at an increased risk of developing clinical disease if they become infected with HIV3. Recently, WHO has given alarming estimates of HIV related tuberculosis in developing countries4 where tuberculosis is now recognised as one of the most common opportunistic diseases among persons seropositive for HIV1. Tuberculosis prevalence and incidence are increasing in some parts of the world5-6. There are about 16-20 million tuberculosis cases in the world and nearly 8 million cases are added each year7. About 22 million pepple are currently suffering from HIV/AIDS1. Nearly 85% of HIV infections are now taking place in the developing world4 where one third of population is already infected with Mycobacterium tuberculosis, exposing these people to a

higher risk of developing tuberculosis. In 1990, it was estimated that HIV contributed 4% to the total tuberculosis cases (i.e. 3,00,000) in the world and this estimate may rise to 14% (i.e. 1.4 million cases per year) by 2000 AD, if both HIV and tuberculosis are not tackled effectively8. In India, by December 31, 1997, a total of 32,27,557 persons from high risk groups were screened for HIV antibodies by National AIDS Control Organisation (NACO), from the inception of the National AIDS Control Programme in 1986, and 71,400 (2.21%) were found to be seropositive. In Delhi, 0.4% of the 3,17,103 persons similarly screened were found HIV positive. The major high risk groups included heterosexually promiscuous, intravenous drug abusers and blood donors9. But, there is very little information on HIV seropositivity among tuberculosis patients in India. It has been suggested that tuberculosis and gastro-intestinal diseases might increase in seropositive individuals and vice versa10. Tuberculosis often appears before other opportunistic infections occur in persons infected with HIV11.

* The study was funded by the National A ID S Control Orgam/ation (NACO), New Delhi 1 Epidemiologist; 2. Director, 3. Medical O f f i c e r - New Delhi Tuberculosis Centre, 4. Consultant (Micro), National Institute of Communicable Diseases, New Delhi Correspondence: Dr. S.K. J a m , E p i d e m i o l o g i s t , N e w D e l h i T u b e r c u l o s i s Centre, lawaharlal N e h r u Marg, New Delhi-1 10002.

22

S.K.JAIN E T A L

Since varying HIV seropositivity rates among tuberculosis patients have been reported in India11-l7, it was considered desirable to determine the magnitude of the problem in the state of Delhi. OBJECTIVE The objective of the study was to determine the prevalence of HIV infection among newly diagnosed, untreated tuberculosis patients in Delhi state. MATERIAL AND METHODS Sampling Procedure: The guidelines issued by NACO for sentinel surveillance were followed. The population of Delhi is served by 14 TB clinics/ hospitals. Each clinic is responsible for tuberculosis control services in a well defined territory. All these clinicsfulfilledthecriterialaiddownforbeingselected as sentinel site/hospital. It was decided to limit the study to six tuberculosis clinics,** selected in a random manner (Fig. 1). Their areas included urban, rural and slum segments of the Delhi population. As per the NACO guidelines for sentinel surveillance, about 400 patients from each sentinel clinic were to be inducted during 2 months. In case of slower rate of intake, the period of intake was extended to four months in that particular clinic. The total duration of field work being fixed at one year, serum samples from two clinics were collected at one time. Selection of Patients: All the newly diagnosed patients during the period of the study in each clinic satisfying the following criteria were included: a) Either smear positive or smear negative and/ or having extra-pulmonary tuberculosis. b) Aged between 15 and 45 years. c) Untreated or taken anti-tuberculosis drugs for less than 30 days. To minimize participation bias, an unlinked anonymous HIV testing approach was adopted. The ethical aspect was addressed by maintaining absolute confidentiality and fulfilling all the requirements of unlinked anonymous testing. Study Procedure: Serum specimen collection was begun in September, 1997. Every day, two

technicians visited two chest clinics to collect specimens and connected relevant information. Patients newly diagnosed on that day and fulfilling criteria for inclusion in the study were listed without further selection. The basic socio-demograplnc and laboratory data (namely sex, age, etc.) were recorded on a pretested form (Proforma). Treatment to all patients was started without delay. Blood collected from each patient was stored in a labeled tube, for routine laboratory examinations, given a study number and linked to the individuals form, without personal identifiers. The blood specimens thus collected were stored at 40 Celsius at the New Delhi Tuberculosis Centre. Blood specimens from 400 consecutive patients were collected from each clinic. Once the specimen collection work at a clinic was over, the New Delhi Tuberculosis Centre laboratory technician was shifted to the next clinic
PROFORMA I IDENTIFICATION 1. 1 Serial Number _______________ 1 2 Date ________________ _ _ I .2 Name of Clinic __________________ _ _ _ _____________________

II Background Information 2. I Age___ 2.2 Sex (M 1 F 2)___2.3 Marital Status 2. 4 Religion ___2.5 Education ___ 2.6 Stay in Delhi
(Yrs) 7 7 Place of Stay Rural 1 Urba ..2 Slum ..3 7 8 Present Occupation __________ 2.9 Duration __ _ (Yrs ) 2 10 Presently living with spouse7 Yes ..1 No .. 2 NA . 3 2 1 1 Number of children Son(s) ____ Daugher(s) ______ 2 12 Monthly Income (Rs ) _______________ _________________ III 1 2 GENERAL MEDICAL HISTORY 3 2 Medical History No/Yes No/Yes 2. Persistent skin diseases Pulmonary - Bacillary )+) 1 Recipient of bl & bl Prod Pulmonary - Bacillary (-)

3 1 Diagnosis

Extra-pulmonary (specify)

3. History of STDs No/Yes 4 . Undergone major surgery No/Yes 5 Addiction, Drug Abuse No/Yes
6 Extra-marital sex No/Yes 7 Number of sex partners One/More 8 Contact with sex workers No/Yes

Education Status Illiterate Primary Unmarned2 Middle Graduate 1

Religion Hindu

If Yes - Frequent 1 Sometimes 2 Occupation M a r i t a l 1 Prof-Tech 2 Semi-professional 1 Married 1 2 3 4 5 6 7

2 Muslim

3 Christiana 3 Clerk. Shop/Farm owner 4 Skilled Worker 5 Semi-skilled worker Unskilled worker No work 5 Others

High School 4 Sikhs PG & Above 6

**Shahdara Chest Clinic, Patparganj Chest Clinic, Motinagar Chest Clinic, New Delhi Tuberculosis Centre, Nehru Nagai Chest Clinic and Gulabi Bagh Chest Clinic

PREVALENCE OF HI V INFECTION AMONG TUBERCULOSIS PATIENTS

23

and the same procedure was repeated. In this way, six TB clinics were covered from September, 1997 to August, 1998. At the New Delhi Tuberculosis Centre, serum was separated out and forwarded to the National AIDS Reference Centre at National Institute of Communicable Diseases for ELISA testing without any other information or identification provided to them. The samples were first screened by ELISA method using Innotest HIV-l/HIV-2 Ab. s.p. method. The samples found to be non-reactive on first testing were declared as negative. All samples found to be reactive on ELISA testing were subjected to additional E/R/S tests, based on different antigens/ different principles. For the second test Capillus HIV, Spot/Immunocomb, HIV-l/HIV-2 kits were used. These are rapid tests and the results are available within a few minutes. The samples found to be reactive with these tests were again tested by one of the E/R/S. All the samples found reactive by these tests were further subjected to Line Immunoassay for confirmation, using the highly specific INNOLIA HIV-l/HIV-2 Ab kits.
RESULTS

In all, 2,361 serum specimens from as many patients were collected during the September, 1997 to August, 1998 period from the 6 centres (Fig. 1).

Adequate number of specimens was collected from all the clinics except Gulabi Bagh chest clime (numbers shown in Fig. 1). Of the 2,361 specimens, 36 (1.52%) were found to be reactive by the initial ELISA test. But, only 16 of these were found positive by the subsequent two tests (Capillus and Immunocomb). All the 16 were also positive for HIV-1 antibodies by the highly specific INNO-LIA test. Thus, seroprevalance of HI V infection among newly diagnosed, untreated, tuberculosis patients of Delhi was 0.68% (95% CI = 0.35%-1.01%). All the 16 HIV positive cases were found among the 1409 males examined (1.14%; 95% CI = 0.59% - 1.69%); (Table 1) which suggests a significantly higher risk (p=0.0023) of co-infection among males suffering from tuberculosis. Further detailed analysis was, therefore, limited to males only. The age distribution suggests (Table 1) an increase in seroprevalance of HIV with rising age: prevalence was 0.51 % in 15-24 years group, 1.55% in 25-34 years group and 1.66% in the 35 years and more group. Seroprevalance of HIV was 1.61% (Table 2) among unmarried males and those living singly, while only 0.88% of male tuberculosis patients living with their spouses had HIV infection. A highly significant association (p=0.020) was observed between literacy level and HIV (Table 3).

Fig. 1 Map of Delhi showing the location and number of samples collected from six chest clinics

24

S.K. JAIN ET AL

Table I H I V scroprevalence among tuberculosis patients according to age and sex MALE Age (yrs) Total Tested 590 517 302 1 ,409 HIV Positive No. (%) 3 (0.51) 8 (1.55) 5 (1.66) 16 (1.14) FEMALE Total Tested 490 308 154 952 HIV Positive No. (%)
__

15-24 25-34 35 & above Total

(-) (-) (-)

_

(-)

Table 2. HIV scroprevalence among tuberculosis patients according to manta! status and sex MALE Total Tested Living with 913 HIV Positive No. (%) 8 (0.88) 8 (I 61) 16 (1.14) FEMALE Total Tested 704 248 952 HIV Positive No. (%) (-) .(-) _ (-)

Prevalence of HIV infection among different categories of tuberculosis patients is given in Table 4. Prevalence was 1.86% in the bacillary positives, 0.71% in abacillary and 1.52% in extra-pulmonary tuberculosis patients. Information about the extent of initial radiological disease was available for only 1,168 (82.90%) out of 1,409 male patients (Table 5). Only in 6 HIV positive cases was one lung involved and the other 10 cases had bilateral disease. Multiple cavities were seen in 5 HIV positive cases, parenchyma! infiltration, pleural effusion and mediastinal lymph nodes were present in 14, 2 and 1 HIV positive cases respectively. In 6 (37.50%) out of 16 HIV positive patients, history taking failed to elicit any risk factor (Table 6). The recognised risk factors for H I V and tuberculosis co-infection in 10 cases were, a) heterosexual relations with multiple partners, b) intravenous drug abuse and c) presence of concomitant sexually transmitted diseases.
DISCUSSION

spouse Unmamed or 496 l i v i n g singly 1 ,409 Total

Table 3. HIV seroprevalence among tuberculosis patients according to educational status and sex
Educational Status Illiterate Primary School Middle School High School Graduate MALE Total Tested 696 187 228 261 30 HIV Positive No. (%) 11 (1.58) 4 (2.14) 1 (0.44) 0 (0.00) 0 (0.00) Post-graduate 7 & above Total 1,409 0 (0 00) 16 (1.14) 2 952 122 142 109 34 FEMALE Total Tested 543 HIV Positive No (%) _ (-) (-) (-) .(-) (-) (-) (-)

The published reports about seroprevalance of HIV among tuberculosis patients give highly variable rates world wide. Enki et al18 found that 66% newly diagnosed tuberculosis patients in Kampala (Uganda) were HIV seropositive. Eilhot et al19 reported 60% seroprevalence among tuberculosis patients in Zambia. But, Onorato and McCray 11 had reported that 3.4% of the 3,077 tuberculosis patients had HIV co-infection in U.S.A.

Table 4. H I V sciopiexalcnce among tubeiculosis patients, accoidmg to diagnostic catcgorics

Catcgoiy MALE Total Tested Bacillaiy positive Bacillaiy negative Extiapulmonaiy Total 430 847 132 1 ,409 HIV Positive No. (%) 8 (1. 86) 6 (071) 2 (1.52) 16 (1. 1 4 ) FEMALE Total Tested 236 574 142 952 HIV Positive No. (%) (-) (-) (-) (- )

PREVALENCEOF H I V INFECTION AMONG TUBERCULOSIS PATIENTS

25

In India too, wide variations in HIV seroprevalence among tuberculosis patients have been observed. Solomon et al16 found 0.77% of tuberculosis patients HIV positive, in 1991 and a higher seroprevalence, in 1993 (3.35%). Banavaliker et al20 found 0.5% HIV seropositive in hospitalized tuberculosis cases while Jayaswal et al21 reported 4.0% seroprevalence in Military Hospital, Pune. Mohanty & Basheer12 had reported an alarming increase of HIV infection, from 2.56% in .1988 to 10.15% in 1993-94, among hospitalised tuberculosis patients in Mumbai, while Anuradha et al22 reported 12 of 3,071 (0.4%) tuberculosis cases positive for HIV from Tuberculosis Research Centre, Chennai. In the current study confined to Delhi state, covering 2,361 tuberculosis patients reporting for the first time at 6 randomly selected tuberculosis clinics, the seroprevalence found (0.68%) was no different from the observation made by Jain et al14. However,
Table 5. H I V seroprevalence among tuberculosis patients according to radiological status
MALE Radiological Status* Chest X-ray i) Normal 58 0 (0.00) 6 (148) 10 (1.42) 5 (2. 02) 14 (1.27) 0 (0. 00) 2 (1.83) 1 (1.89) 0 (0.00) 16 85 298 395 (-) (-) (-) (-) (-) (-) (-) (-) (-) Total Tested HIV Positive No. (%) FEMALE Total Tested HIV Positive No. (%)

Table 6. Distribution of H I V infected tubeiculosis patients according to recognised risk factors

MALE Risk factois Total tested HIV positive No. Heterosexual multiple partncis 10 Homosexuals Recipients of blood/blood pioducts Intravenous drug users Sexually transmitted diseases History of suigery No i isk factors Total Patients 8 1,345 1,409 0 6 16 2 947 952 5 3 1 19 2 0 26 0 4 1 7 0 0 0 FEMALE Total tested HIV Positive No

* More than one factor n some patients.

404 n) Unilateral involvement 706 m)Bilateral involvement Type of Lesion* 248 i) Cavities ii ) Parenchymal infiltration ni)Miliary i v ) Pl e m a l effusion v) Mediastinal lymph nodes v i ) Hydropneumothorax Total 1 ,098 1 109 53 3 1,168

during an earlier study, Jain et al14 reported no HIV positive case at New Delhi Tuberculosis Centre in 1988, and 0.4% seroprevalence in 1992. Banavaliker et al20 found 0.5% seroprevalence in hospitalized tuberculosis patients of Delhi in 1994-95. The entire information available for Delhi is shown in Fig. 2. The distribution of HIV cases in the higher ages and the association between literacy level and HIV seroprevalence in the present study is similar to the findings of Dey et al23. This study suggests that prevalence of HIV among tuberculosis patients

134 612 2 2 66 29 778

Ycar of Study

* Radiological status, ot 415 tubeiculosis patients was not available * * One in mine lesion pel patient

Fig. 2. Picvlcncc of H I V among Tuberculosis patients m Delhi

26 S.K. J A I N E T A L Programme, Journal oj Indian Medical Association. in Delhi is lower than that reported from other 8. 1996, 94, 372 parts of the country. The need for continuous Dolin P.J., Raviglione M.C. and Kochi A : Global 9. surveillance of HIV in tuberculosis patients of tuberculosis: incidence & mortality during 1990Delhi remains, as does the need for preventive 2000. Hull World Health Org. 1994, 72, 213 intervention among high risk groups identified, 10. National AIDS Control Organization. Ministry of in this study and elsewhere. Health & Family Welfare, Govt. of India, Monthly
ACKNOWLEDGEMENTS
11. 12.
Update, Dec. 1997 ICMR, H I V Infection: on going studies & future Research Plans, ICMR Bulletin Nov. 1986, 19( 11), 115 Onorato I.M. and McCray E. : Prevalence of Human Immunodeficiency Virus infection among patients attending tuberculosis clinics in the United States. Journal of Infectious Diseases. 1992, 165, 87 Mohanty K.C. and Basheer D.M.M. : Changing trend of H I V infection and Tuberculosis in a Bombay area since 1988. Ind. J. Tub. 1995, 42, 117 Deo S. : Prevalence of H I V infection in patients with tuberculosis. Ind J Tub. 1995, 42, 183 Jain N.K., Aggarwal J.K., Chopra K.K. and Khanna S.P.: Prevalence of HIV infection among tuberculosis patients. Ind J Tub. 1996, 43, 105 Solomon S. Kumarasvvamy N., Anuradhu S.. Vennila R. and Pal JaykarS.A.: Tuberculosis and HIV infection - an association, Indian J Med Microbiol. 1994, 12(4), 313 Solomon S., Anuradha S. and Rajasekaran S. : Trend of H I V i n f e c t i o n in patients w i t h p u l m o n a r y tuberculosis. Tubercle & Lung Disease. 1995, 76, 17 R a n a d i v a S.N., Tan k ar M.R., K u l k a r n i S.S., Sukumaran K., Gokhale N.T., Rodngues J.J and Bancrjee K.: Prevalence of H I V - I antibodies in patients attending tuberculosis clinics in Punc .Jassoc Physicians India. 1995, 89, 5 Eriki P.P., Okwera A. and Aisu T. : The influence of human immunodeficiency virus infection on tuberculosis in Kampala, Uganda. Am Rev Respir Dis. 1991.42, 128 ElliottA.M. LuoN. andTemboG.: Impact of HIV on tuberculosis in Zambia : a cross-sectional study, fir Med.I. 1990, 301, 412 Banavaliker J.N., Gupta R., Sharma D.C., Goel M.K. and Kumari S. : H IV seropositivity in hospitalised pulmonary tuberculosis patients in Delhi. Ind J Tub. 1997, 44, 17 Jayaswal R., Arora P.N. and Panda B.N. : H I V in tuberculosis, Medical Journal Armed Forces. India 1995, 51, 259 Anuradha S., Solomon S. and Rajasekaran S. : HIV seropositivity in patients with respiratory diseases. Ind J Tub. 1993,40, 13 Dey S.K., Pal N.K. and Chakrabarty M.S.: Cases of HIV and tuberculosis : early experience of different aspects. IIIWA1DS Research in India, National AIDS Control Organisation, Ministry of Health & Family Welfare. Govt. of India, 1997, 610.

The authors are grateful to the National AIDS Control Organization, Ministry of Health & Family Welfare for having given financial support for carrying out the study. The study was initiated by Dr. S.P. Khanna, the then Director, New Delhi Tuberculosis Centre, who did the initial planning. Dr. D.R. Nagpaul, Technical Advisor to the Tuberculosis Association of India provided guidance throughout the study as well as during presentation of the present report and his contribution is gratefully acknowledged. Thanks are also due to Mr. G.P. Mathtir for his help in the preparation of this report. The field work could not have been completed without the active cooperation of Dr. J.N. Banavaliker, State TB Control Officer (Delhi), Dr. V.K. Arora, Dr. R.K. Joshi, Dr. (Mrs) M.M. Sethi, Dr. R.K. Mehra, Dr. (Mrs) U.C. Dhawan and their contribution is gratefully acknowledged.

13. 14. 15. 16.

17.

18.

REFERENCES
1.
National AIDS Control Organization, National AIDS Control Programme, Ministry of Health & Family Welfare, Govt. of India, Country Scenario - An Update. Dec. 1996, pages l , 3 & 8 World Health Organization : Groups at risk : WHO report on tuberculosis epidemic, 1996, 12 Colebunders R.L., Rymer R.W. and Nzilambi N.: HIV infection in patients with tuberculosis in Kinshasa, Zaire. Am Rev Respir Dis. 1989, 139, 1082 Narain J.P., Raviglione M.C. and Kochi A.: H I V associated tuberculosis in developing countries, epidemiology and strategies for prevention. Tubercle & Lung Disease. 1992, 73, 311 Meeran K.: Prevalence of H I V infection among patients with leprosy and tuberculosis in rural Zambia. BrMedJ. 1988, 298, 364 Slutin G, Leowski J and Man J.: Tuberculosis and AIDS: The effects of the AIDS epidemic on tuberculosis problems and tuberculosis programme. Bull Int Union Tuberc. 1986, 63, 21 K h a t r i G . R . : N a t i o n a l T u b e r c u l o s i s C o n tro l

19.

2. 3.

20. 21.

4.

22. 23. 24.

5. 6.

7.

Original Article

Ind. J. Tub., 2000, 47,27

SURVEILLANCE OF DRUG RESISTANCE IN TUBERCULOSIS IN THE STATE OF TAMIL NADU

C.N. Paramasivan1, K. Bhaskarair, P. Venkataraman1, V. Chandrasekaran4 and P.R. Narayanan5 (Received on 23.8.99; Accepted on 26.10.99)
Summary: Surveillance of drug resistance was carried out at State level to obtain data which are standardised and compaiable using guidelines prescribed by the WHO/IUATLD Working Group on Anti-tuberculosis Drug Resistance Surveillance. Objective: To determine the proportion of initial and acquired drug resistance in cases of pulmonary tuberculosis in Tamil Nadu, m order to use the level of drug resistance as a performance indicator of the National Tuberculosis Programme. Methods: Two specimens of sputum from each of a total of 713 patients attending 145 participating centres all over the state were tested by smear and culture examination and drug susceptibility tests of Isoniazid, Rifampicin, Ethambutol and Streptomycin. Results: Out of 400 patients for whom drug susceptibility results were available, 384 (96%) had no history of previous anti-tuberculosis treatment. Of these, 312 (81%) were susceptible to all the drugs tested. Resistance to Isoniazid was seen in 15.4% of patients and to Rifampicin in 4.4%, including resistance to Isoniazid and Rifampicin in 3.4%. Conclusion: There has been a gradual increase in initial drug resistance over the years in this part of the
country.

Key Words: Pulmonary tuberculosis; Initial drug resistance, Acquired drug resistance

INTRODUCTION

The level of drug resistance is known to provide an epidemiological indicator to assess the extent of resistant bacterial transmission in the community as well as success or otherwise of the National Tuberculosis Programme (NTP). High levels of resistance have been reported in certain regions of the world, particularly in Asia and parts of Africa1-7. Potential causes of drug resistance include inadequate treatment provided by health services, poor case-holding, poor drug supply, poor quality of drugs, non adherence of patients to the prescribed drug regimens and indiscriminate use of antituberculosis drugs in the private sector. However, the most important cause of drug resistance may be the error of health care workers in not providing correct regimens8. High levels of Rifampicm

resistance, for instance, require adjustment of the control progrmme to guarantee high cure rate. The recommendation to use drug susceptibility tests for monitoring and guiding tuberculosis treatment programmes was made many years ago9. In view of the practical difficulties in collecting comparable data, the World Health Organization (WHO) proposed a programme of global surveillance of drug resistance in tuberculosis through its collaborating centres for bacteriology of tuberculosis which would function as Supranational Reference Laboratories (SRL) for the respective regions9. The proposed programme was based on random sampling of patients reporting to clinics for tuberculosis treatment. Susceptibility testing was to be performed by the reference laboratories based on a common protocol including uniform laboratory methods. As a first step, a regional survey was carried out in 10

1. Deputy Director 2. Director of Medical Services, Govt. of Tamil Nadu 3. Senior Technical Officer 4. Technical Assistant 5 Director Tuberculosis Research Centre (ICM R), Chennai - 600 031 and Directorate otMedical Services, Government of Tamil Nadu Correspondence. Dr C.N. Paramasivan, Deputy Director, Tuberculosis Research Centre, Mayor V.R Ramanathan Road, Chetput, Chennai-600031

28
10

CN PARAMASIVANETAL

Latin American countries . The overall experience gained in Latin America suggested that a sample survey of drug resistance with large failure rates of more than 5% may indicate inadequate routine treatment and high levels of initial resistance, which made survey of drug resistance a priority8. Several countries in Asia and Africa undertook national surveys in accordance with the protocol, but regional multi-centre surveys were not completed. Meanwhile, several countries, including Tanzania11 and South Africa12 established systematic national surveillance programmes. In India, there was an urgent need to establish surveillance of drug resistance at country level to obtain standardized and comparable data within and among countries. In view of the large size of the country, it was proposed to undertake surveillance of drug resistance in one state, Tamil Nadu to begin with, based entirely on the guidelines of the protocol prescribed by the WHO8. The study was conducted by the Tuberculosis Research Centre, a WHO collaborating centre.
MATERIAL & METHODS 1. Organizational and Survey Outline Survey area The entire State of Tamil Nadu in south India formed the survey area: Central Laboratory The Tuberculosis Research Centre (Indian Council of Medical Research), Chennai (TRC) was the National Central Laboratory. A l l bacteriological investigations namely, smear, culture, identification and susceptibility testing were earned out there. Supranational Reference Laboratory (SRL) External quality control assessment was earned out at the Centre for Public Health Sciences, Brisbane, Australia which validated the results of susceptibility tests done by the Central Laboratory (TRC). The Brisbane laboratory had already been identified by the WHO as a Supranational Reference Laboratory for the region. The method of quality control, according to internationally accepted standards, was developed by the SRL before starting the survey. Such a system included, (a) testing at TRC of a sample of strains sent by the SRL, and (b) re-

testing at SRL of a sample of strains tested at TRC. Diagnostic Centres All the District TB Centres (DTC) at the district headquarters and X-ray centres (XC) in taluks of the entire state of Tamil Nadu formed the diagnostic centres. In addition, hospitals attached to the medical colleges in the state were also in clu d ed . In all, a total of 145 centres participated in the study. They comprised 23 DTCs, 117 XCs and 5 medical college hospitals. Supervision A Central Coordinator and a team of medical officers closely supervised all diagnostic centres to ensure good cooperation. They promptly i d e n t i f i e d and corrected any operational problems. It was also verified that the procedure for sputum smear microscopy followed at the diagnostic centres was of an acceptable standard13. Preparatory phase In the preparatory phase of the survey, the Coordinate! and the cential laboiatoiy jointly assessed all the relevant indicators of the programme, the infrastructure and piocedures Furthei, the Coordinator arranged a training programme at Chennai for t h e personnel managing the District Tuberculosis Programmes During naming, the personnel of the Central Laboiatoiy gave a detailed day-long explanation of all aspects of surveillance activities such as case selection, collection of specimens of sputum, entries to be made in the relevant forms and despatch of the specimens to TRC through the postal services. The district personnel, in turn, conducted training programmes for medical officers and laboratory technicians at the XCs of then respective districts. Data Analaysis The data analysis was done by the TRC using the EpiInfo software
2. Sampling It was estimated that a sample size of at least 500 was needed for the overall sun ey to be able to detect resistance level of about 30% expected in this area with a confidence interval of 26% to 34% and a level of confidence of 95%14.15. I lence, fioni each of the 145 diagnostic centers,

DRUG RESISTANCE IN TAMIL NADU

29

2 specimens of sputum from 4 to 5 consecutive, newly diagnosed smear-positive cases were collected over a two-month period, about 700 cases in all (allowing for contamination rate of about 5%), and sent to TRC. The diagnostic centres also filled up a form for each specimen giving information about previous treatment the patient had received which was of help in distinguishing between initial and acquired resistance. 3. Intake of patients Inclusion criteria Patients were eligible for inclusion if: 1. they had been newly registered (reregistration of the same patient in different centres was to be avoided); 2. they had been found to be sputum-smear positive, at least once during the intake period. 3. Children under 15 years of age fulfilling the above criteria were also eligible. Transport of sputum specimens Before the start of the study, the TRC supplied, through the Coordinator, sufficient numbers of sputum collection bottles (sterile universal containers), each containing 5 ml of sterile 1% cetyl pyridinium chloride solution. At the centres, sputum specimens were collected from patients, smears made, stained by Z-N method and examined as per the programme procedure. If a smear was positive, and the patient fulfilled the other eligibility criteria, two more specimens (5 ml each) were collected from the patient in the sputum collection bottles containing CPC. A total of 1375 sputum specimens was received from 713 patients during February-March 1997. Of these, 51 patients provided only one specimen each while the rest gave two specimens each. 4. Bacteriological investigations at the Central Laboratory Smear examination At TRC, smears were made from all specimens, stained by the auramine-phenol method and examined using fluorescence microscopy. Positive smears were graded as 1+, 2+ and 3+.16

Culture The specimens were processed by modified Petroff s method17, using 4% sodium hydroxide and inoculated on to two slopes of L-J medium. The slopes were incubated at 37 C and examined weekly for evidence of growth till 8 weeks. The growth was graded as 1 + (20-200 cols.), 2+ (more than 100 discrete cols.) and 3+ (confluent growth). Actual colony counts were reported if growth was less than 20 colonies. Identification Culture isolates were identified as M. tuberculosis based on niacin test, catalase at 68C/pH 7 and growth on L-J medium with p-nitrobenzoic acid.18, 19 Susceptibility testing Susceptibility tests were performed for Streptomycin by the resistance ratio (RR) method and for Isoniazid, Rifampicin and Ethambutol using the absolute concentration (MIC) method on L-J medium. A 3mm loopful of suspension of the growth, containing approximately 4 mg/ml, was inoculated on to different concentrations of the drugs and read after 28 days of incubation. Resistance to Streptomycin was defined as RR of 8 or more. An MIC of 1 mg/1 or more for Isoniazid, 128 mg/1 or more for Rifampicin and 8 mg/1 or more for Ethambutol was interpreted as resistant.20-21 RESULTS Smear On smear examination at TRC, 6.5% of the 1375 specimens were found to be smear-negative (Table 1). Duplicate smears from the same patient showed a fair agreement with 91% concordance even in grades (Table 2).
Table 1. Smear results of all specimens tested
Smear Grade Neg 1+ 2+ 3+ TOTAL Total No. 90 1,054 224 7 1,375 Specimens % 6.5 76.7 16.3 0.5 100.0

30

C.N. PARAMASIVAN ET AL

Table 2. Distribution of smear results of duplicate specimens from same patient according to grade of smear
Smear II Smear I Neg Neg 1+ 2+ 3+ NA 16 24 1 0 0 1+ 30 425 55 1 42 553 77.6 2+ 3 52 52 0 8 115 16.1 3+ 0 0 1 2 1 4 0.6 49 501 109 3 51 713 6.9 70.3 15.3 0.4 7.2 Total No. %

quality assurance studies were undertaken by the Brisbane laboratory, the results of which are presented in Table 4. Of the 20 cultures received from the SRL in Round I, there was 100% agreement in the case of Isoniazid and Ethambutol, while it was 95% for Rifampicin and 90% with Streptomycin. However,
Table 3. Summary of anti-tuberculosis drug resistance. No history of Previously treated/acquired resistance No. % 16 100.0 50.0 50.0 12.5 0.0 0.0 0.0 6.2 12.5 0.0 6.2 6.2 0.0 6.2 0.0 0.0 0.0 0.0 50.0 25.0 25.0 treatment/initial resistance No. % Total tested 384 312 72 29 2 2 7 2 4 0 7 7 5 5 0 0 2 0 59 100.0 Fully sensitive Xny resistance Resistance to H R E S HR HRE HRS HRSE HE HS HES RE RS RES ES Any H Resistance

Total No. 41 5.8

Culture A little more than half of all the specimens were culture negative and more than half of all positive cultures yielded 100 colonies or less. An identical pattern was obtained when duplicate specimens from the same patients were examined (not tabulated). The contamination rate was also extremely low (0.6%). All the strains isolated were identified as Mycobacterium tuberculosis. Drug susceptibility test results were available for 400 patients, including 384 patients without any history of previous treatment and 16 patients with previous treatment. The pattern of resistance observed in these patients is shown in Table 3. Of the 384 patients without any history of previous treatment, the isolates from 312 (81.2%) patients were fully susceptible to all the drugs tested, while isolates from 72 patients had resistance to one or more drugs. Resistance to Isoniazid, alone or in combination with other daigs was seen in 59 (15.4%) patients. Similarly, resistance to Rifampicin, alone or with other drugs was seen in 17 (4.4%) patients. Resistance to Isoniazid and Rifampicin together was observed in 13 (3.4%). Of the 16 patients with history of previous treatment, isolates from 8 patients were fu lly susceptible. Resistance to Isoniazid, alone or with other drugs was seen in 8 cases, Rifampicin resistance was observed in 4 patients, all of whom were also resistant to Isoniazid. External Quality Control Studies During the study period, two rounds of external

81.2 8 18.8 8 7.6 0.5 0.5 1.8 0.5 1.0 0.0 1.8 1.8 1.3 1.3 0.0 0.0 0.5 0.0 4.4 3.4 2 0 0 0
1

2 0
1 1

0
1

0 0 0 0 4 4

15.4 8

Any R Resistance 17 Any HR Resistance 13

Table 4. Agreement in drug susceptibility results between TRC, Chennai and SRL, Brisbane* Reproducibility (%) Retesting at Round 1: TRC SRL Round II TRC SRL S 90 100 100 100 H 100 90 100 100 R 95 100 100 100 E 100 100 100 100

*SRL, Brisbane used BACTEC radiometric method hi le TRC used conventional (RR/MIC) methods

DRUG RESISTANCE IN TAMIL NADU

31

Table 5. Comparison of direct culture method with Petroff s method in specimens containing CPC Petroffs Method Neg. Col. 1+ 2+/3+ Total Neg. 6 1 2 0 9 Direct Culture Col. 7 0 0 0 7 1+ 3 0 0 0 3 . Total 2+/3+ 22 38 7 2 4 35 8 4 4 54

in Round II, the reproducibility was 100% for all the drugs tested. Similarly, of the 20 strains re-tested at the SRL in Round I, there was total agreement between the two laboratories in the case of Streptomycin, Rifampicin and Ethambutol. In round II, there was 100% agreement with all the four drugs tested.
DISCUSSION

The WHO/IUATLD Global Project on Antituberculosis Drug Resistance Surveillance recorded considerable variation in the prevalence of drug resistance among 35 countries in 5 continents. The median prevalence of drug resistance among patients with no history of previous treatment was 10% with the range from 2%-40%. Although the prevalence of MDR-TB is generally low, there are several countries where the situation demands immediate intervention. Overall, the median prevalence of primary MDR-TB was 1.4%, ranging fromO-14%10. Among the South East Asian Region (SEAR) countries, the prevalence of primary resistance is readily available only for Nepal and Thailand since they participated in the WHO supported Global Project on Anti-tuberculosis Drug Resistance Surveillance in 1994-1997. The median prevalence of primary resistance to any drug was recorded as 23.2% with range of 9.8% to 36.6%. The median prevalence of primary MDR-TB was 2.5%, significantly higher than the global mean of 1.4%14. However, such infonnation for other SEAR countries, based on standardized protocols and methods, is not available. Although drug resistant tuberculosis has frequently been encountered in India and its presence has been known from the time drugs were introduced for the treatment of tuberculosis, there is no comprehensive report mainly due to limited facilities

available for culture and susceptibility tests across the country. The present report on drug resistance in the entire state of Tamil Nadu, using internationally acceptable guidelines and a standardized methodology gives reliable information. Considering smear results of all specimens tested, it was observed that 6.5% of the specimens were negative by microscopy at this Centre. It should be noted that one of the criteria for inclusion of patients in this study was that they should have been smear positive at least once during the study period. The specimens that were sent to TRC were not examined for smear at the diagnostic centres. Hence this discrepancy could have been due to differences between specimens from the same patient. Another explanation could be that the sputum specimens got diluted by the addition of 5 ml of CPC solution which could have resulted in a small proportion of smears being reported as negative. Nevertheless, the agreement of 93.5% shows that sputum microscopy at the diagnostic centres, with minimal infrastructure, was up to an acceptable standard. That a very large number of smear-positives were culture negative in this study is of utmost concern. This, it was realized, was due to defective methodology provided in the-study protocol which laid down that the CPC containing specimens should be processed by Petroffs method using 4% sodium hydroxide (NaOH). This combination of CPC-NaOH was perhaps responsible for culture negativity23. However, by the time this was realized, the intake was almost completed. However, towards the end of the study, an alternative method of processing was employed. The CPC-containing specimen was directly centrifuged as such and the deposit was inoculated on to two LJ slopes (direct method). Subsequently, the deposit was re-suspended in water to the original volume and processed by PetrotTs method. A comparison of the results of this study on 61 specimens is shown in Table 5. It was observed that the direct method yielded 45 positive cultures as against only 16 by Petroffs method, a highly significant difference. Further, 35 out of 54 cultures (64.8%) yielded 2+ / 3+ grades as against only 4 (7.4%) by Petroffs method. The CPC-NaOH combination had not only inhibited the growth of M.tuberculosis but perhaps also that of the contaminants, which accounted for

32

C.N.PARAMASIVANETAL

the very low (0.6%) contamination observed in this stu dy. This observation has been repeated in a survey conducted by us in North Arcot district (Tamil Nadu), in 1999. A total of 635 sputum specimens from 320 patients were collected in transport medium containing 5 ml of 1% CPC and 2% NaCI solution and were processed directly. An interim analysis has revealed culture positivity of 95%, negativity of 3% and contamination of 2%. The results of this study confirm our earlier observation once again (unpublished data). It might be observed that the 1988-89 level of 2% MDR-TB in the North Arcot district has gone up to about 4% in the succeeding decade. Considering the level of drug resistance observed in this study, the proportion of resistance to H, R and HR was of the order of 15.4%, 4.4% and 3.4%, respectively, in previously untreated cases. These estimates are similar to other surveys conducted by the TRC in the recent past. Thus, in a study conducted in the North Arcot district of Tamil Nadu during the period 1985-89 on 2,779 cases, with no history of previous anti-tuberculosis treatment or with previous treatment of less than 2 months duration, initial resistance to H was 13% and HR resistance was 1.6%14. In the district of Raichur in Karnataka State, initial resistance to H was of the order of 13.2%24. Data from TRC on primary resistance among 6,742 patients, from 1956 onwards for 18 different chemotherapeutics, had shown a gradual increase in the prevalence of primary resistance to antituberculosis drugs. For Isoniazid and Streptomycin, the resistance rates were similar and ranged from 314% with the highest level of 14% observed during 1990 for Isoniazid. Initial resistance to Rifampicin started appearing during the 1990s and is 1.2% at present. Resistance to HR or SHR was observed to be less than 1% in 1990-957. However, observations on acquired resistance in India vary in different surveys. A retrospective analysis on a cohort of 3,357 smear-positive patients initiated on anti-tuberculosis treatment, between April 1986 and March 1988 in North Arcot District had shown acquired resistance of 67% to H and 12% for R, either alone or in combination with other anti-tuberculosis drugs25. Elsewhere in India, an alarmingly h i g h proportion of acquired resistance has been reported

from the state of Gujarat26, resistance to H being 3560% and 3-37% for R. Other recent reports on acquired resistance (from previously treated cases) in centres such as Sewagram, Wardha (1982-89), New Delhi (1990-91) and Tamil Nadu (1995) showed a high level of resistance to Isoniazid (20.9%, 50.7% and 23.6% respectively) and MDRTB (9.6%, 33.7% and 23.3% respectively)27-28 are not surprising considering their specialized status where chronically ill patients with a history of mismanagement and bad treatment are referred. On the other hand, drug resistance reported from the Military Hospital at Pune showed a very low level of initial resistance to H (0.6%) which could be due to the minimal chance of indiscriminate treatment received prior to reporting at the hospital.30 There has perhaps been a gradual increase in primary drug resistance in this part of India, over the years. However, there appears to be a marginally higher acquired resistance, which contributes to the problem of rise in primary drug resistance. This could be overcome by a strong control programme which can reduce the creation of drug resistance in the community such as directly observed therapy with short course regimens (DOTS)4. Since no newer drugs for tuberculosis are likely to become available in the near future, the only options left for the prevention of drug resistance are effective casefinding, prompt and correct diagnosis and successful treatment of patients. Apart from a strong control programme, continuous surveillance of drug resistance will provide information which will serve as a useful parameter in the evaluation of control programmes.
ACKNOWLEDGEMENTS

This research project was funded by the World Health Organization. The assistance rendered by Dr. Marcos Espinal, Dr. F. Bustreo a n d Dr. M. Raviglione of WHO/TB, Geneva, Dr. T.R. Fneden, WHO/SEARO, New Delhi, is g r a t e f u l l y acknowledged. We also wish to thank Dr. David Dawson, Cen tr e for Pu b lic Health Sciences, Brisbane, Australia for his assistance in the external quality assessment. We are grateful to the Directorate of Medical Services, Government of Tamil Nadu, the District TB Officers of the various districts and th e M e d i c a l O f f i c e r s as well as lab o r ato r y technicians of all the participating Centres for their

DRUG RESISTANCE IN TAMIL NADU

33

cooperation in the conduct of the study. We also thank the technical staff of the Bacteriology Laboratory for undertaking meticulously all the bacteriological investigations. The secretarial assistance of Smt. Jothi Segaran is also gratefully acknowledged.
REFERENCES
1. 2. 3. Kochi A, Vareldzis B, Styblo K. Multi-drug resistant tuberculosis and its control. Rex Microbiol 1993, 144: 104 K i m SJ, Hong YF. Drug resistance to Mycobacterium tuberculosis in Korea. Tubercle & Lung Dis 1992, 73: 219 Chandrasekaran S, Jagota P, Chaudhury K. I ni t i a l drug resistance to anti-tuberculosis drugs in urban and rural district tuberculosis programme, Ind J Tub 1992, 39: 171 Frieden TR, Sterling T, Pablos-Mcndes A, et al. The emergence of drug-resistant tuberculosis in New York City. New Eng J Med 1993, 328: 521 Van der Werf TS. H i g h i n i t i a l drug resistance in pulmonary tuberculosis in Ghana. Tubercle 1989, 70: 249 Braun MM, Kilburn JO, Smithwick RW, ct al. H I V infection and primary resistance to anti-tuberculosis drugs in Abidjan, Cote dIvoire. AIDS, 1992, 6: 1327 Paramasivan CN. An overview of drug resistant tuberculosis in India. Ind. J Tub, 1998, 45, 73 World Health Organisation Guidelines for surveillance of drug resistance in tuberculosis. WHO/TB/96, 216 World Health Organization. Surveillance of drug resistance in tuberculosis : a global random sample survey of i n i t i a l and acquired resistance; WHO/TB, 1984, 143, 1 World Health Organization. Anti-tuberculosis drug resistance in the world (the WHO/IUATLD Global Project on A n t i - t u b e r c u l o s i s Drug Resistance Surveillance 1994-1997) WHO/TB/97-229 Chodc TM. The role of bacteriological surveys in the N a t i o n a l Tuberculosis and Leprosy Programme in Tanzania. Bull Int Union Tuberc Lung Dis; 1989 Weyer K, Kleeberg HH. Primary and acquired drug resistance in adult black patients with tuberculosis in South Africa : results of continuous drug resistance surveillance programme involvement. Tuberc Lung Dis, 73, 106 IU ATLD. Technical guide for sputum examination for tuberculosis by direct microscopy; Bull Inl Union Tuberc 1978, suppl 2 P a r a m a s i v a n CN, C h a n d r a s e k a r V, S a n t h a T, Sudarsanam NM and Prabhakar R. Bacteriological investigations for short course chemotherapy under the tuberculosis programme in 2 district in India. Tuberc Lung Dis; 1993, 74, 23 Lwanga SK, Lcmeshow S. Sample si/cdeterminations

in health studies. A practical manual. World Health Organization, Geneva, 1991 16. Holst E, Mitchison DA, Radhakrishna S. Examination of smears for tubercle b a c i l l i by fluorescence microscopy. Ind J Med Res, 1959, 47, 495 17. Petroff SA. A new rapid method for isolation and cultivation of tubercle bacilli from sputum and faeces. JExpMed; 1915, 21,38
18. A l l e n B, Baker RJ. In: Mycobacleria: Isolation, identification and sensitivity testing. P. 17, 1968. Butterworth, London

4. 5. 6. 7. 8. 9.

10.

11. 12.

13. 14.

15.

19. KubicaGP; Differential identification of mycobacteria. V I I . Key to features for identification of clinically significantmycobacteria.Am Rev Resp, 1973,107,9 20. Canetti G., Fox W, Khomenko A, et al. Advances in techniques of testing mycobacterial drug sensitivity and the use of sensitivity tests in tuberculosis control programme. Hull WHO; 1969, 4 1 ,2 1 2 1 . Tuberculosis Research Centre, Madras. Study of chemotherapy regimens of 5 and 7 months duration and the role of corticosteroids in the treatment of sputum-positive patients with pulmonary tuberculosis in South India. Tubercle, 1983, 64, 73 22. Tuberculosis Research Centre, Madras. Ethambutol p l u s isoniazid for the treatment of pulmonary tuberculosis - a controlled trial of four regimens. Tubercle; 1981,61, 13 23. Smithwick RW, Stratigos CB, David HI. Use of cetyl p y r i d mi u m chloride and sodium chloride for the d eco n t ami n at i o n of sputum specimens that are transported to the laboratory for the isolation of M.luberculosis. J. Clin Microbiol; 1975, 1,411 24. G o p i PC, V a l l i s h a y e e RS, Appe Gowda B N , Paramasivan CN, Ranganatha S, Venkataramu KV, P h a n i a r a j BS, K r i s h n a m a c h a r y a L, Devan J, Ponnuswamy R, Komaleeswaran G and Prabhakar R. A tuberculosis prevalence survey based on symptoms questioning and sputum examination. Ind.I Tub: 1997, 44, 171 25. Datta M, Radhamani MP, Sclvaraj R, Paramasivan CN, Gopalan BN, Sudeendra CR, Prabhakar R. Critical assessment of smear-positive pulmonary tuberculosis patients after chemotherapy under district tuberculosis programme. Tub Lung Dis; 1993,74, 180 26. Tnvedi SS, Desai SC. Primary anti-tuberculosis drug resistance and acquired rifampicin resistance in Gujarat, India. Tubercle; 1988, 69, 37 27. Narang P ct al. Personal communication 28. Jain NK, Chopra KK, Prasad G. Initial and acquired r i f a m p i c i n resistance to M.tuberculosis an d its implications for treatment. Ind J Tub; 1992, 39, 121 29. Vasanthakumari R, Jagannath K. Multi-drug resistant tuberculosis - A Tamil Nadu study. Lung India; 1997, 15, 178 30. Jena J, Pande BN, Nema SK et al. Drug resistance pattern of Mycobacterium tuberculosis in a chest diseases hospital of armed forces. Lung India; 1995, 13, 56.

34

Smoking Withdrawal Syndrome

Inhaled nicotine bonds with some receptors in the body which provide perceived pleasurable feelings. When the effect of nicotine dissipates, the receptorssend signals to the brain which are perceived as a displeasurable feeling - the withdrawal syndrome - compelling the smoker to light up. The act of smoking does not, however, happen in a vacuum: It is almost always attended with particular "settings", situations or emotions peculiar to each individuals desires. These unfulfilled desires make the withdrawal symptoms all the more painful. It is mainly the psychological associations which urge a smoker to resume smoking. Or, when the nicotine level in the body is low enough to trigger the receptors. Smokers associate cigarettes and smoking with many aspects of their daily life. Whenever these associations recur, there is a craving to smoke. After stopping smoking, the physical effects of the withdrawal persist for between 7 to 30 days. However, the craving for a cigarette continues for months, despite the withering away of receptors after having been deprived of nicotine. The craving persists because of the associated settings and human emotions of the smoker.

YOU HAVE TO QUIT

Published by the Tuberculosis Association of India in the mteiest of public health

Original Article

Ind. J. Tub.. 2000, 47,35

CONTROL OF TUBERCULOSIS AMONGST THE TIBETAN REFUGEE COMMUNITY IN NORTHERN INDIA

D.F. Wares1, T.D. Sadutshang2, N.J. Beeching1 and P.D.O. Davies4 (Received on 10.3.99; Accepted on 22.7.99)
Summary:

Objective: To review the Tuberculosis Control Programme based on the refugee-run Tibetan Delek Hospital, Dharamsala (north India), during 1985-1992. Design: Retrospective review of hospital In-Patient Register and Files, Laboratory Registers, TB Register and available individual TB Patient Record Cards. Results: Case-holding of Tibetan pulmonary tuberculosis (PTB) patients (n= 973) was 86.4% and among the smear positive PTB patients (n = 403), 80.1% got cured, 1% were treatment failures; 8.2% died and 10.7% defaulted. Case-holding for the Indian PTB patients (n = 144) was 43.1%; only 43.2% of smear positive PTB cases (n = 74) were cured, 2.7% died and 47.3% defaulted. Conclusion: Overall performance of the Tibetan refugee-run Tuberculosis Programme is good. Important success factors were: a relatively stable refugee population, existence of a programme manual, regular and reliable drug supplies, motivated health workers based close to the patients, regular supervision and on-going training of the health workers, programme monitoring, health education to patients and general refugee population. It has to be noted that the programme is funded by overseas donors. Present concerns of the programme are: increasing case load (new and drug resistant cases), and poor treatment completion rates amongst Indian patients. Innovative methods need to be considered to overcome these problems.
Key Words: Refugees; Pulmonary tuberculosis; Case-holding; Tuberculosis control programme

INTRODUCTION

Never before have there been so many refugees and internally displaced persons (IDPs) in the world as today. The 1994 data provided by the United Nations High Commissioner for Refugees show more than 23 million refugees and 27 million IDPs.1 The majority of refugees originate from, and remain within, the areas of high tuberculosis (TB) endemicity. Up to 50% of all refugees in the world may be infected with tuberculosis.2 The cramped and over-crowded living conditions typical of refugee populations provide the ideal environment for the transmission of TB infection. Refugees, therefore, form a particularly vulnerable and susceptible population, and disease management is difficult due to the unstable nature of many refugee communities

and mobility of the populations concerned. It is well recognized that TB emerges as an important health problem in many refugee settings, once the emergency phase is under control.3,4 Since 1959, approximately 120,000 Tibetans have fled into exile in India and live in settlements, ranging in population from < 100 to > 10,000, scattered throughout the country. This community has become an established refugee population in India. However, between 1,000 - 5,000 new refugees still arrive annually. TB is a major health problem within the refugee community and remainsamajorcauseof death (Table 1). Since 1980, a TB Control Programme (TBCP) based on the Tibetan Delek Hospital (TDH) in Dharamsala, has been run for the Tibetan refugees living in the northern Indian State of Himachal

1. Director, The Britain-Nepal Medical Trust, Biratnagar, Nepal. 2. Chief Medical Officer/Administratoi, Tibetan Delek Hospital, Dharamsala, Himachal Pradesh. 3. The School of Tropical Medicine, Liverpool, England, UK. 4. Director, TB Research Unit, Cardiothoracic Centre, Liverpool, England. Correspondence: Dr. D.F. Wares, The Britain-Nepal Medical Trust, PO Box 9, Biratnagar, Morang 56601, Nepal. Fax: 00977 21 25232. Email: tbphnmt@mos. Com.np

36
5

D.F. WARES ETAL

Pradesh, on lines similar to that of the Indian National Tuberculosis Programme (NTP). The TBCP is primarily based on passive casefinding and sputum testing, serving both Tibetan refugee settlements and Indian communities living close to these settlements. A Settlement Health Worker (SHW) is responsible for investigating suspects, supervising treatment of existing cases and retrieving defaulters, both in the allotted refugee settlement and any nearby Indian community from where patients may have been enrolled into theprogramme. There are also comprehensive guilines in regard to BCG vaccination. The programme is administered and monitored by a TB Medical Officer and 2 TB Health Workers based at the TDH. Whilst the majority of Tibetan patients live in Himachal Pradesh (HP), quite a few come from other parts of India and even from Nepal. Most Indian patients reside in Dharamsala itself or the surrounding hills. A team from the TDH visits all the refugee settlements within HP three times in a year, primarily to monitor the TB programme. Treatment regimens are based on WHO guidelines6 and the TBCP Manual is continually being updated.7 The antituberculosis drugs are not provided free but at a highly subsidized price, which is considerably lower than the market cost of the drugs. An unusual feature for a TB programme in a developing country is that reserve drug regimens are also available to patients who have failed on standard regimens and/or are documented drug resistant cases who are not responding to treatment.* MATERIAL AND METHODS Reported here is a retrospective review of the TBCP as observed from the Tibetan Delek Hospital records, from 1985 to 1992. Data were extracted from the following records: 1) Admission Register and individual In-

Patient files (1987-1992), Laboratory Registers (1986-92), TB Register and individual TB Patient Record Cards (19851992), as obtained from the TDH. 2) Census data for Tibetan refugees in Himachal Pradesh, the Home Affairs Office of the Tibetan Administration in Exile, Dharamsala, Himachal Pradesh. As this was a retrospective review, problems surrounding patient classification and treatment outcome definitions were encountered.** Therefore, WHO reporting system definitions were adapted.8 The term Episode was defined as the first presentation of a patient with the diagnosis of TB to the Delek Hospital TBCP. Although comprising mainly new cases, this group may have included default and relapse cases from outside the programme and possibly even cases who had failed treatment elsewhere. Even if a patient was a true new case, he/ she may have consulted either a private medical practitioner or an Indian Government hospital prior to presentation at the Delek Hospital. Up to 70% of TB patients in India may commence their treatment in the private sector, and possibly 50% remain there9. With 40 different Intensive Phase (IP) regimens identified in this review, many of these started by private practitioners, it is likely that the Tibetan refugees had done the same, although possibly to a lesser extent. The majority of these cases, diagnosed elsewhere, entered the TBCP with their initial sputum smear status unknown to the programme. Relapse was defined as clinical re-presentation with the diagnosis of active PTB*** one year or more after completion of a successful course of antituberculosis therapy (ATT). Relapse, within the review period was coded as 2nd Episode. However, if less than one year had elapsed since the previous treatment, then their data was included in the 1st Episode.

*Betbre 1990, new cases of PTB were treated with 2SHT/1 OHT. Regimens containing R+Z were reserved for treatment failures Short Course Chemotherapy [2S(E)HRZ/6HT] was introduced in 1992. From 1995 onwards, this was replaced by 2S(F:)HRZ/4HR Reserve drugs used included: K.anamycin,Capreomycin,Cyclosenne,Thioaceta/.one, PAS, hthionamideand Fluoroqumolones(Ciprotlo\acm or Ofloxacin) Regimens contain at least 4 drugs, of which at least 2 would be drugs that the patient had never received before Duiation of treatment is fora mini mum of 18 months. ** Although most patients registered as new cases had no previous history of TB, it was seen that a number may have had treatment for TB before. Relevant information may be withheld at the time of presentation, sometimes due to ignorance and other times deliberately Also, data were sometimes incomplete, especially from the earlier years of the review period ***A relapse patient with active pulmonary TB could be either smear positive or negative. If negative, the patient had to ha\ c radiological abnormalities on chest X-iay, determined by a medical officer to be consistent with TB disease, prior to the decision to start tieatment

TUBERCULOSIS AMONGST TIBETAN REFUGEES

37

WHO definitions of treatment outcomes in sputum smear positive (sm+ve) TB patients,8 were also adapted, as follows: Cure a sputum sm+ve patient who completes treatment, clinically improves and has, including the continuation phase, a negative smear at the end of the treatment course. Treatment a sputum sm+ve patient who failure remains, or again becomes, sm+ve at the time of completion of the treatment course. Default lost from treatment for > 1 month, and no subsequent return for further treatment. The data collected were stored and analysed using a version of Epilnfo.10 Statistical analysis was performed using 2 x 2 2 tables. For case-holding and treatment outcome results, only 1st Episode data were analysed. Outcomes of reserve drug regimen treatment are included in the overall totals.
RESULTS

Table 1. New TB cases detected among Tibetan refugees living in Himachal Pradesh New cases Case detection Tibetan Year registered rate per refugee population under the Delek 1 00,000 Hospital TBCP population 1987 1991 1992 15,994 19,105 19,354 217 250 369 1,357 1,309 1,907

Table 2. Sputum smear results, 1986 - 1992 Year 1986 1987 1988 1 989 1 990 1991 1992 Total New cases Positive 91 (7.2%) 71 (7.3%) 65 (8.3%) 79 (6.3%) 64 (7.5%) 86 (9.4%) 75 (7.3%) 531 (7.5%) Negative 1,179 (92.8%) 904 (92.78%) 718 (91.7%) 1,182 (93.7%) 794 (92.5%) 829 (90.6%) 951 (92.7%) 6,557 (92.5%)

A) Overall Review 1. Hospital Admissions In the six year period (1987-1992), 23.2% (1166/5020) of hospital admission were forpatients with tuberculosis. Overall mortality in the hospital was 2.7% (136/5020), of which 30.1% (34 Tibetan + 7 Indian) were due to TB which caused two thirds (14/21) of all deaths in the age group 15-34 years. 2. Hospital Laboratory Work-load Almost 14,600 smear results were recorded during the review period. In all, 48.6% (7088/14598) were for new patients. Of the sputa examined from new patients, 7.5% (531/7088) were AFB positive (Table 2). 3. TB Control Programme Between 1985 and 1992, a total of 2487 cases were enrolled under the TBCP. Individual Patient Treatment Cards were available for 87.2% (2169) of these cases. Of the 2113 patients reviewed, 89.8% were Tibetan (T) and 10.2% Indian (I) : 50 patients (45 T+5 I) had two episodes of TB and 3

Tibetans had three episodes during the review period. (Fig. 1) The composition of Tibetan and Indian patients was somewhat different. There was a significant difference between the male : female ratios (Mantel Haensel/>value <0.01), with Tibetan cases being predominantly male; 57% of Tibetan and 46% of Indian patients were within the age group 15-34 years (no statistical difference) (Table 3). In all, 85.5% of Tibetan cases were pulmonary TB (PTB). Of Tibetan PTB patients resident in Himachal Pradesh, 46.9% were smear positive (sm+ve). Among the Indian patients, 87% were PTB and 44.5% were sputum smear positive. The HIV status of all patients was unknown. B) Status of cases from Himachal Pradesh: For all Tibetan PTB cases residing in HP, caseholding* was 76.6% (841/1096 x 100%). For the sm+ve Tibetans, case-holding was 80% (327/409 x 100%). However, after excluding those cases with a

*Case-holding is defined as the percentage of patients who, having commenced treatment, complete the f u l l course of treatment. Deaths and transfers out were not included in the denominator.

38

D.F. WARES ETAL

50 patients (45 I + 5 I) had 2 episodes and 31 ibetans had 3 episodes recorded during it-view pei tod **Includes 3 patients with wrong diagnosis and 1 who stopped due to side effects of AT I" NK = not known

Fig. 1 Flow chart of cases registered under the TBCP not known outcome, case-holding becomes 56.4% (841/973 x 100%) and 88.4% (327/370 x 100%) for these two groups respectively. For all Indian PTB patients residing in HP, comparative figures were 41% (62/151 x 100%) and 43.1% (62/144 x 100%) respectively. For Tibetan sm+ve PTB cases, who had a known treatment outcome (n = 403), 80.1% were cured, 1 % failed treatment, 8.2% died and 10.7% defaulted (Table 4) while 4% (n = 16) of the Tibetan cases relapsed. For the Indian sm+ve cases (n = 74), the
Table 3. Characteristics of Tibetan and Indian tuberculosis cases Tibetan Indian Male : Female Mean age, in years Pulmonary TB cases 2: 1 28.2 85.5% [I622/ 1898] 1.36: 1 33,6 87% [187/215]

respective percentages were 43.2%, 0%, 2.7% and 47.3%. Although based on small numbers, there was a highly significant difference (MHp-value <0.01) in cure and default rates between the male and female Indian patients. Male Indian patients had a cure rate of just 32% and a default rate of 59.6%, compared to 70.8% and 25% respectively for female Indian patients. No significant difference was present between male and female Tibetan patients. Treatment outcome for respective drug regimens was not studied. Thus, a number of patients received standard therapy, then followed by SCC. In 161 cases (154 T + 7 I), a reserve drug regimen was also used. With 40 different Intensive Phase (IP) drug regimens, analysis of treatment outcome by regimen would have been extremely complex to perform and numbers in some groups were few (for sm+ve Tibetan PTB cases, 15 different IP regimens - 10 more than the possible maximum advised by the programme - were identified). From the data available, it was impossible to quantify exactly how many patients on correct

TUBERCULOSIS AMONGSTTIBETAN REFUGEES

39

Table 4. Treatment outcome in smear positive PTB csaes (residents of Himachal Pradesh only)
Treatment outcome Cure Treatment failure Default Died Transfer out Not known* Total Tibetan 323 4 43 33 0 39 442 Indian 32 0 35 2 1 4 74

Tibetan case-holding (CH) versus Indian CH : MHpValue <0.001. *Not known = patient treatment card missing or incompletely filled in.

regimens had started these on the advice of private practitioners, but almost 8% were prescribed inappropriate regimens for sm+ve PTB. The situation was worse for those patients whose initial sputum status was unknown, when they started treatment, with 25% being prescribed inappropriate regimens.
DISCUSSION

The good case-holding and cure rates seen amongst the Tibetan patients are positive findings. However, as the definitions used in the review were different from those of the WHO reporting system,8 one has to be cautious in making comparisons. Besides, a significant number of PTB cases-entered the programme with an unknown smear result and their treatment outcomes were not analysed. Also, no attempt was made to analyse results by the lespective treatment regimens. Although low, compared to other reports from India11, the number of defaulters and deaths remain significant amongst the Tibetan patients. With the widening use of SCC and the gradual introduction of directly observed treatment (DOT), it is hoped that mortality and default will reduce as long as the present high levels of supervision and programme monitoring arc maintained. An important finding of the review is that the good results were obtained from within a system that is generally considered as unsupervised treatment. Cases were not routinely admitted to the hospital, exceptions being a sm+ve case from a school, hostel or monastery/

nunnery, and seriously ill patients. Treatment was not directly observed, with patients being issued with weekly medicines. Obviously, patients on Streptomycin (S) needed to attend the hospital or clinic daily for their injection. With almost 90% of the 1st Episode Tibetan sm+ve PTB cases having (S) in their IP regimen, the intake of just one drug was monitored daily in the majority of patients. But intake of oral medicines, including Rifampicin, was not observed or monitored. This lack of direct observation has to be placed against the central importance of TB nurses and settlement health workers, role in the programme and needs to be recognised. In a number of recent studies and commentaries, the pivotal importance of this type of worker, living close to the patient and being a member of the same community, to the success of TB programmes is prominently mentioned.11-13 The favourable case-holding and cure rates found amongst the Tibetan patients in this review further supports this hypothesis and calls for greater attention. The follow-up of the Tibetan patients was easier because there is a trained health worker present in each refugee settlement. If a patient misses coming for his medicines, action can be taken by the SHW, or T B nurse, quickly. Increase in the number of welltrained SHWs, w i t h continued monitoring, supervision and on-going training of those health workers already in position, must be encouraged. Although their numbers are small, there is a serious problem with the care of Indian patients enrolled under the programme (default rate of >50%), which seems to be a more serious for male patients as female patients had a cure rate approaching that of the Tibetan patients. The followup of Indian patients by the Tibetan refugee TBCP is difficult, as they often live far from either the hospital or a settlement clinic. This contrasts with the situation of the Tibetan patients and must be a contributory factor to the poor results seen. The situation needs to be urgently addressed. Various suggestions such as the employment of an "Outreach Worker" from the hospital itself or the recruitment of local people e.g. shop-keepers14 to function as treatment supervisors must be considered. The reasons why Indian patients, especially the males, and Tibetan patients default in the first place need to be explored.

40

D.F. WARES ETAL

The Tibetan refugee Tuberculosis Control Programme in Himachal Pradesh has a number of features, which may explain, at least partly, its relative effectiveness: i) It is relatively well funded, mainly by overseas donors. ii) Due to the regular adequate funding, the drug supply is virtually always sufficient for its needs. iii) The population (i.e. Tibetan refugees), lives largely in a small number of well-defined settlements. iv) All refugee settlements have a health worker available, a functioning health clinic exists in a l l th e settlements and there is a designated referral centre (TDH). v) Clear programme guidelines exist, set out in the TBCP Manual, which health workers must adhere to for dealing with patients enrolled under the TBCP. Since 1995, the Manual provides forjust 5 regimens, which are based on WHO guidelines, vi) Constant supervision of health workers, with on-going training for them, and monitoring of the programme, are seen as essential components. vii) An active health education programme operates within the refugee community. The advantages of early diagnosis and treatment are generally well understood by the community. Despite a good programme infrastructure and improvement in the overall socio-economic condition of the refugee community, tuberculosis remains a major health problem with 23% of admissions and 30% of deaths at the TDH being due to TB. Although comparisons are inadvisable, notification rates amongst Tibetan refugees in Himachal Pradesh are 10 times higher than that of the 1990 estimate for India (153.2/100,000).15 Notifications continue to rise, though only in the number of cases. With no annual risk of tuberculosis infection rate being available, it is difficult to state whether the disease situation is worsening within the refugee community or not. The observed increase cannot be explained solely by an increase in cases amongst new comers (refugees who have lived <1 year in India), even though there has been a small increase in their numbers amongst the TB cases (increasing from 6%

of annual total in 1991 to 9% in 1992). Socioeconomic conditions may be improving for the refugee community, overall, but many live in less than satisfactory conditions. Schools, hostels and monasteries/nunneries, having limited living space, are facing over-crowding as many of the new refugees arriving each year are young children, monks or nuns. In a recent study of adult Tibetan immigrants to Minnesota, USA, 98% were found to be tuberculin positive and 8.4% developed disease within 2-3 years of immigration.16 A major problem throughout South East Asia is the widespread inappropriate prescribing of drugs, both for tuberculosis and for non-tuberculosis illnesses. In this review, 40 different ATT regimens were noted, the majority of which were started by private practitioners prior to the patients enrolment in the programme. This has many consequences for both TB control programmes and the patient: which is the correct regimen to be given to the patient, how to correctly classify the patient and, of course, the greaterpossibility of drug resistance. Poor knowledge of the disease and poor management of tuberculosis patients by private practitioners have both been well documented.17-19 However, the private sector constitutes a large segment of the health care services in India, and up to 70% of TB patients may initiate their TB careers with private practitioners.9 A partnership is needed between the different sectors of the health care services in developing countries,9.18 to ensure good TB control programmes and prevent the development of drug resistance. Whether the refugeerun TBCP could contribute to such a partnership (e.g. by offering a forum for continuing education of local private practitioners) needs to be explored.
ACKNOWLEDGEMENTS

Our thanks go to all staff of the Delek Hospital, especially to Mrs. Sangay Choedon (TB Nurse) and Mr Karma (TB I lealth Worker). Thanks also to Dr. U.K. Sood, Kangra District TB Officer, NTP, for his cooperation and assistance. The review study was supported by a grant from the Liverpool School of Tropical Medicine.
REFERENCES
I Un i t e d Nations High Commissioner foi Refugees ( U N H C R ) Populations of concern to UNHCR: a

TUBERCULOSIS AMONGST TIBETAN REFUGEES

41

statistical overview, Geneva; UNHCR, 1994 World Health Organisation (WHO); Groups at risk: WHO report on the tuberculosis epidemic, Geneva: WHO, 1996 3. WHO. Tuberculosis control in refugees and displaced persons. Geneva; WHO, 1996 4. Centers for Disease Control and Prevention. Famine affected refugee and displaced populations: recommendations for Public Health issues. MMWR; 1992,41, 1 5. Kramer AC. Tibetans in Exile: a health perspective; Pmc R Coll Physicians Edinb; 1996; 26: 250 6. WHO. Tuberculosis Treatment Guidelines for National TB Programmes. Geneva; WHO, 1992 7. Tibetan Dclek Hospital. Delek Hospital TB Control Programme M a n u a l , 5th E d i t i o n ; Tibetan Delek Hospital; 1996 8. WHO. Treatment of Tuberculosis: G u i d e l i n e s for National Programmes. Geneva; WHO, 1997 9. WHO. TB patients and private for-profit health care providers in India. Geneva; WHO, 1997 10. Dean AC, Dean JA, Coulombier D. Epi Info, version 6: a word processing, database and statistics program for epidemiology on microcomputers. Atlanta; Center for Disease Control and Prevention, 1994 11 Datta M, Radhmani MP, Selvaraj R, Paramasivan CN, Gopalan BN, Sudeendra CR, Prabhakar R. Critical assessment of smear-positive pulmonary tuberculosis p a t i e n t s a f t e r chemotherapy u n d e r t h e d i s t r i c t 2.

12.

13.

14.

15. 16.

tuberculosis programme; Tuberc & Lung Dis, 1993; 74:1 8 0 Gangadharam PRJ. Editorial - Chemotherapy of tuberculosis under programme conditions with special reference to India; Tuberc & Lung Dis. 1994; 75: 241 Chowdhury AMR, Chowdhury S, Islam MN, Islam V, Vaughan .IP. Control of tuberculosis by community health workers in Bangladesh; Lancet. 1997, 350, 169 Wilkmson D. High compliance tuberculosis treatment programme in a rural community; Lancet: 1994, 343: 647 WHO. T u b e r c u l o s i s Programme R e v i e w I n d i a . Geneva: WHO, 1992 Truong DM, Hedcmark LL, Mickman JJ, Moshcr LB, Dictrich SE, Lowry PW. Tuberculosis among Tibetan immigrants from India and Nepal in Minnesota, 1992\995;JAMA; 1997,277, 735

I 7. Uplckar MW, Shepard DS. Treatment of tuberculosis by private general practitioners in India; Tubercle: 1991; 72: 284 18. Marsh D, Hashim R, Hassainy F, ct ul. Front-line management of pulmonary tuberculosis: an analysis of tuberculosis and treatment practices in urban Sindh, Pakistan; Tuberc & Lung Dis, 1996, 77, 86 19. Singla N, Sharma PP, Singla R, Jain RC. Survey of knowledge, attitudes and practices for tuberculosis among general practitioners in Delh i, India; Int J Tuberc & Lung Dis; 1998, 2(5): 384.

42

Dr. S.P. PAMRA (May 1911-January 2000) OBITUARY

With grief and deep sorrow we have to inform that Dr. S.P. Pamra passed away peacefully on the 29th January, 2000 after a brief illness. A renowned clinician and a doyen of tuberculosis workers, Dr. Pamra had been associated with the TB Association of India for almost five decades. Born in May, 1911 at Phagwara (Punjab), Dr. Pamra passed his M.B.BS. from King hdward Medical College, Lahore in 1933 with honours and did TDD fiom Madras Medical College in 1946. Di. Pamra started his tuberculosis career in the Silver Jubilee TB Hospital, Delhi and subsequently he joined K.E. Sanatorium, Dharampur and TB Clinic, Shimla. Dr. Pamra joined the New Delhi TB Centre in 1948 and retired as its Director in 1980. Dr. Pamia had wide teaching experience and had been a teacher for undergraduates and postgraduate students. He was awarded the TAI Gold Medal in 1978 in recognition of his meritorious service in the field of tuberculosis After retirement from the New Delhi TB Centre, Dr. Pamra became the Editor of the Indian Journal of Tuberculosis and Hoi orary Technical Adviser of the Association. During the peuod when Dr. Pamra was associated with the TAI, he was the Eastern regional lepresentative to the Executive Committee of the International Union Against 1 ubeiculosis and Lung Disease, Pans and served on the Treatment Committee of the IUAT-LD. Dr. Pamra was also the Editor of the Bulletin oi the Eastern Region of the IUAT-LD. Dr Pamra had been a member of the Scientific Advisory Committee of the Tuberculosis Research Centie, Madias for many years. Dr. Pamra had not only been associated with the TB Association of India but also with ICMR, Government of India, TB Hospitals, etc. He was associated with various institutions in various important and responsible capacities, specially with the Tuberculosis Association of India. A noble and endearing person, he was humane and liked by all His sad demise on 29th January, 2000 came as a shock to the Tuberculosis Association of India and those associated with him. The void created by Dr. Pamras demise is difficult to fill His demise is a great loss not only to the Tuberculosis Association of India but also the entire country . The President of the Association, its Chairman, Office-bearers and functionaries, staff and the functional ics of the New Delhi TB Centre, its officers and staff, convey their heartfelt condolences to his near ones, and pray to God fro eternal peace to the departed soul. to his near ones, and pray to God for eternal peace to the departed soul.

Case Report

Ind.J. Tub , 2000. 47,43

TUBERCULOSIS OF GALL BLADDER PRESENTING AS MALIGNANCY : A CASE REPORT

K. Alam,1 S. Ahmad,2 G. Mehdi,1 S.M. Ashraf3 and V. Maheshwari2 (Received on J8.1.99: Accepted on 8.6.99)
Summary: A 40 year old female with hcpatoblllary symptoms suggesting stone in gall bladder was offered cholecystectomy. At surgery, a nodular growth in liver and thickened gall bladder were discovered. Histopathologic examination of the gall bladder and liver nodule revealed tuberculous pathology. The case is presented because of its rarity.

INTRODUCTION

Tuberculosis of gall bladder is extremely rare. Tuberculomata may be multiple, white, irregular or caseous abscesses surrounded by fibrous capsule. Their naked eye distinction from Hodgkins disease, secondary carcinoma or actinomycosis may be difficult. 1
CASE REPORT

A 40 years old female presented with intermittent pain in the right upper abdomen for the past 5 months, associated with vomiting and fever with chill and rigours. The pain aggravated on taking fatty meals and was relieved by anti-spasmodic drugs. There was no associatedjaundice or blood in vomitus. There was no significant past history, especially regarding tuberculosis. General examination revealed a middle aged female of average built and nutrition with normal vitals and no significant findings. There was no lymphadenopathy. Systemic examination of the cardio-vascular, central nervous and respiratory systems was normal. Abdominal examination showed a normal scaphoid abdomen with no visible peristalsis or lump. Organomegaly was not present. Tenderness was present in the right hypochondrium with raised local temperature. Bowel sounds were also present. No free fluid was present in the abdomen.

Haemoglobin was 11.8 g%, TLC-20, 550/ cu.mm., DLC - P 70%, L 30%, Blood sugar - 70 mg%(R), Blood urea - 34 mg%, S. creatinine1.0 mg%, 1:CG and X-ray chest were normal. Ultrasonography revealed cholelithiasis. It was decided to remove the gall bladder surgically. Cholecystectomy was done, along with the stones in the gall bladder. Common bile duct was normal in caliber and no stones were felt in it. A nodular growth was seen and felt in the inferior border of liver. Presumptive diagnosis of malignancy with secondaries in liver was made. The nodule was resected and sent forhistopathological examination. Gross examination of the resected gall bladder: measurements 4cm x 2 cm x 1 cm, wall thickness 1 cm, neck diameter 0.8cm, fundus showed a solid whitish yellow, homogenous growth like area, which on cross section was haemorrhagic and necrotic, mucosa was velvety and bile stained. No sentinel lymphnode was present in the specimen. Microscopic histopathologic examination showed a non-caseating granuloma in the mucosa of the gall bladder and dense infiltration of lymphocytes in its muscle layer(Fig. 1). The section from the liver nodule showed non-cirrhotic periportal fibrosis and granulomas consistent with tuberculosis (Fig. 2). Short course anti-tuberculosis therapy was started with a four drug regimen and the patient responded to treatment clinically.

1. Lecturer, 2. Reader, 3. Professor Departments of Pathology and General Surgery, J.N. Medical College, Aligarh Muslim University, Aligarh Correspondence: Dr. Kiran Alam, Alam Man/il, Zohra Bagh, Dodhpur, Aligarh - 202 001

44

K ALAM ET AL

F ig 1 Photomiciogiaph of gall bladdu showing chionic granulomas inf i l t i d t i n g the muscle (H & E x 50)

Fig 2 P h o t o n n c i o g i a p h of l i v e : n o d u l e s h o w i n g cpiththoid gianuloma with giant cells and dense lymphocylic infiltiation (H & E, x 50)

DISC USSION I lepatic biliai y tubeiculosis piesenting as a gall bladdci tumoui isiaie No pathognomonic diagnostic chaiactenstics can be relied upon C onfirmation of diagnosis is made on histopathology, though possibly PolymeiaseC ham Reaction (PC R) may make eaily diagnosis and t h e i a p y possible Ben et aF have icpoiled a 64 )eai old female patient similarto 0111 case while Alimacl et al 3 have iepoi ted a similai case in a male patient Pseudo tu mo ial hepatic tubciculomas are a l s o r a r e . There may be no evidence of e x t r a hepatic tubeieulosis The distinctive features ol tubeieulosishistopathologically,arethepiesence of AFB easeation and destitution ol the icticulin f i a m e w o i k cpithclioid cell gianulomas having irregularity oi the eontoui with apaiticulaily dense cu f f ol l y mp h o c y t e s . 4 lubeiculous eholangitis is exti cinch i ai e and the caseous centi al ai eas become bilestamed 1 icatmentisthatfortubeiculosisandno specific tieatmentfoi h\ei is indicated REFERLNCES
1. Ach em SR, Kotts BE, G r i s n i k J e t a l . : Pseudo tumorai hepatic tubeiculosis Atypical picscntation and u i m p i c l u n s i v e i c v i c w of t h e l i t c i a t u i e J Cl i n . G a stero en tero l; 1992, 14, 72 2 B c n R I Young T LccHS Hepatobihaiy tubeiculosis presenting as a gall bladder tumour; Scand. J. Infect., D is . ; 1995, 27 , 4 1 5 3. Ahmad MN, aigai HU Shahdab NA, S a p r u A , Kaur S. Trberculosis of g a l l bladder A ease rc p o rt ; J P o s t g r a d . Med; 1983, 29, 258 4 S h i e l d S h u l o c k In Diseases of the h \ e i and b i l i a i y sy stem 10 t h Edn Blackwcll Scientific 1 td London 1997 MM

Case Report

Ind.J. Tub., 2000. 47,45

TUBERCULOSIS OF MIDDLE EAR - A CASE REPORT K.B. Gupta, Sanjeev Tandon, S.K. Mathur* and Rajnish Kalra*
(Received on 18.1.99; Accepted on 10.6.99)
Summary: A case of middle ear tuberculosis in an apparently healthy male who presented with painless otorrhoea without pulmonary lesion is presented because of its rarity.

INTRODUCTION

Tuberculosis affects the middle ear rarely. It is difficult to assess its true incidence as the large reported series have been selected from hospitalised sub-groups with established tuberculosis.1-3 These patients usually present with unexpectedly severe hearing loss, facial paralysis, past family history of tuberculosis, presence of abundant pale granulations along with the presence of normal mastoid cellular development.1 However, these patients usually have the initial symptom of painless otorrhoea which is not often considered as tuberculous in the differential diagnosis till it leads to severe damage to middle ear. Early diagnosis and prompt treatment may prevent ear damage and the central nervous system complications. We are reporting an apparently healthy young man who had otorrhoea but was found to have middle ear tuberculosis, without lung involvement.
CASE REPORT

Laboratory studies revealed haemoglobin of 9.2 g%, TLC - 9500\mm3 with 75% polymorphonuclear leukocytes. Tuberculin test was positive (20mm). Skiagram chest did not show any p leu ral or parenchyma! abnormality. Specimens of ear discharge were negative forpyogenic and acid fast bacilli on smear and culture. Tympanomastoidectomy, type II was done. The mastoid antrum was found to be full of granulation tissue, Incus was found to be floating, and malleus was absent. Biopsy from the middle ear revealed fibrocollagenous tissue, blood clot, a strip of stratified squamous epithelium along with caseating epithelioid cell granulomas with Langhans giant cells - a picture compatible with the diagnosis of tuberculous otitis media (Fig. 1). The patient was treated with antituberculosis treatment (2RHZE\4HR). Subsequent follow up a f t e r three months showed marked improvement with complete resolution of symptoms and healing of perforations.
DISCUSSION

A 17-year old male, non-smoker, presented with painless otorrhoea from left ear, varying from scanty to profuse discharge for 4 months and partial hearing loss for 2 months. There was no history of previous illness or hospitali/ation and no known exposure to tuberculosis. He had been treated with antibiotics by a local practitioner without any relief. Otoscopy revealed the external auditory canal to be full of thick pus. Clearing the same revealed two small attic perforations.

Tuberculosis of middle ear is known to occur in all age groups but occurs specially in children (50%). About h a l f of these cases usually present w i t h painless otorrhoea, often of long duration. The presence of pale, often abundant, granulation tissue is usually mistaken for Cholesteatoma, and demands detailed mastoid exploration. The route of spread of tuberculosis to middle ear has been argued for many years; the most logical

Dcpaitments of Chest and Tuberculosis and Pathology*, Pt. B.D Sharma Post-Graduute I n s t i t u t e of M e d i c a l Sciences.
Rohtak- 1 2 4 0 0 1

C o r r e s p o n d a n c e : Dr. K.B. Gupta, 9.J/37, Medical Enclave, Rohtak-124 001

46

K B GUPTA ET AL

Fig. 1 Microphotograph of middle ear cuicttmgs showing epithehoid cell gianulomd with Langhans giant cells ( H & E Stain X 400)

route of entry of organisms being via the pharyngotympanic tube Adams in a study of tuberculosis patients undergoing thoracoplasty showed abnormal pharyngo-tympanic tube patency in all patients who developed otitis media 2 However, haematogenous spread has also been described in cases with mihary tuberculosis In one series, half the patients had no evidence of tubeiculosis elsewhere, as happened in the present case .4 The results of treatment are good The cases treated by anti-tuberculosis chemotherapy recover well Recently, the role of surgery has been revised In the past, it was done to provide drainage, to control spiead to central nervous system and to relieve facial paralysis The advent of specific chemotherapy has changed all this, and today surgery should be reserved foi decompi ession of the facial nerve and for removal of neciotic mateiial which might piovide a nidus foi the organisms to lemain out of teach of antituberculosis drugs

Today, tubeiculous otitis media is an uncommon disease But, if left untreated, it causes gieat damage to the middle ear and the suirounding stuictuies It is not easy to diagnose The clinician must, therefore, maintain a high index of suspicion and must be awaie of features which tend to distinguish the tuberculous ear from other conditions haily diagnosis and tieatment offers every prospect of eventless healing
REFERENCES
1 2 3 4 Windle Tayloi PC Bailey CM Tubueulous otitis media Laryngoscope; 1980, 90, 1039 Adams JG Tubciculoiib otitis media A complication of thoiacoplasty Ann Otol 1942, 51, 209 Pioctoi B I mdsay JR Tubciculosis of the. EAR;Arch Otol 1942, 15, 221 Mahajan M Ag g a i wa l DS Smgh NP Gadic DJ Fubuuilosis of the middle ear A c a s e R e p o r t ; I n d . j Tub 1995, 42, 55

Case Report

Ind J Tub 2000 47 47

HERNIATION OF LUNG FOLLOWING LUNG ABSCESS IN A CHILD : A CASE REPORT

R. Prasad1 and Surya Kant2

(Received on 19.11 98, Accepted on 10 6 99)

Summary: A pathological thoracic type of acquned heimation of l u n g f o l l o wi n g l u n g abscess in a six year old child is piesented The patient lesponded to conscivative management

INTRODUCTION

Lung hernia is defined as protrusion of lung tissue beyond normal thoracic boundaries through an anomalous opening in the chest wall, diaphragm or mediastinum along the parietal and visceral pleura.1 Lung hernias are classified according to anatomical location and etiology. Anatomically, they may be cervical, thoracic, diaphragmatic or mediastmal. Etiologically they are of two types, congenital and acquired. Acquired hernias are further classified as traumatic, spontaneous and pathological Pathological lung hernias are pare. We report a case of thoracic, pathological, acquired lung hernia following a lung abscess. To the best of our knowledge, no such case has been reported in the literature.
CLINICAL RECORD

size, crepitant, tender baloomng out on coughing and going back when reduced manually. Left inter- and infra-scapulai areas showed decreased respiratory movement, resonant note on percussion and decreased an entry with a few crepitations on auscultation. His haemoglobin was 8.5 g%, TLC - 17,900/ cumm, DLC - p78L16E2M2 with normal blood sugar, liver functions and urine analysis, Mantoux test showed an induration of 4 mm. in transverse diameter at 72 hours. Sputum for acid fast bacilli was repeatedly negative. Staphylococcm aurein grew on culture, sensitive to Amoxycilhn and Gentamicin Chest X-ray, PA view, revaled a lung abscess in the left lower zone. Left lateral view clearly visualised the abscess with hermation oi lung posteriorly into the subcutaneous plane(Figure 1). On flouroscopy,

RK, a thin-built male child aged 6 years was admitted with complaints of moderate to high grade fever, cough with minimal expectoration and left side chest pain for one month and the appearance of a swelling, on left side chest-back, ten days prior to admission His parents had noticed that the swelling got bigger on coughing and crying. There was no history of trauma, nor family history of tuberculosis On physical examination, the patient was of thin built, anaemic, pulse rate 120/minute, respiratory rate 22/mmute and 101F body temperature. Chest examination revealed an ill defined, smooth, soft swelling in the left inter-scapular area, 1 0 x 6 cm.

Fig. I X-iay Chest left lateial view showing lung abscess postei o-mfci 101 ly with hei mation of lung postei lorly

1. Head of Department, 2 Lecturer Department of Tuberculosis & Chest Diseases, K i n g Gcoigcs Medical College, Lucknow Couespondeine Dr. R. Prasad, Head, Department of T u b e r c u l o s i s & Chest Diseases, K i n g Georges Medical College, Lucknow-226 001

48

R PRASAD AND SURYA KANT

the lung tissue was protruding outside posteriorly. This protrusion increased on coughing and on Valsalvas manoeuvre. His CT thorax showed lung abscess in left mid and lower zones with irregular walls and a few septae and hernation of lung parenchyma posteriorly and subcutaneously with compression of surrounding subcutaneous tissue (Fig 2).

weakness precipitating lung hernia due to persistent coughing. The diagnosis of lung hernia is usually made by the clinical presentation of a soft swelling which becomes prominent on coughing, sneezing and straining and can be reduced manually Roentgenograms and fluoroscopy during Valsalvas manoeuvre show a radiolucent area protruding outside the confines of thoracic cage. CT scan of thorax is a useful diagnostic procedure8. Usually, conservative treatment is sufficient Surgical repair is indicated only when there is constant pain, recurrent infection, respnatory distress, progressive increase in size of hernia, cosmetic concern or when there are chances of excessive increase in intrathoracic pressure and impending lung rupture due to heavy physical activity. Surgical repair of the defect is done by means of peritoneal flaps pulled from the adjacent ribs.9 Marlex mesh or teflon bands have also been sued for repair.10-11
REFERENCES
1. 2. 3. 4. 5. 6. 7. 8. 9 10 11 Adkins, PC; The chest wall. In: Biam Blades ed. Surgical diseases of Chest St. Louis: CV Mosby Company; 1961,87 Bidstrup P, Noidentoft M, Peterson B, Hernia of the l u n g Brief survey and report of two cases; Acta. Radial. ; 1966, 4, 490 Hiscol B, Digman J; Types and incidence of lung hernias; J. Thor Surg; 1955, 30, 335 Donate AT, Hipona FA, Navani S, Spontaneous lung hernia; Chest; 1973, 64, 254 Prasad R, Muke r j i PK, Gupta H, Hermation of the lung, Indian J. Chest Dis. Allied Sci.; 1990, 32, 129 Bagga AS, Kakodkar U, Lawande D, Chatterji R, Hermation of the Lung : A case report, Ind J. Tub.; 1995, 42, 47 Ghanchi FD, Patel KR; Herniation of lung, Ind. J. Chest Dis. Allied Sic.; 1996, 38, 49 Sadler MA, Shapiro RS, Wagieich J, Halton K, Hecht A; CT diagnosis of acquired intercostal lung heriniation; Clinical Imaging; 1997, 21, 104 Goodmann HL; Hernation of l u n g ; .J. Thorax. Cardiovast Sing, 1933, 2, 368 Giunwald RP, Knox WG, Chest wall lung hernia method of icpair using marlex mesh; Am. J. Surg.; 1964, 108, 730 Munnell ER, Hermation of Lung, Ann Tlioiac S u r g . ; 1968, 5, 204.

Fig. 2 CT Thorax showing lung abscess with hermation of lung posteriorly

The patient was managed with Amoxycihn and gentamicin with supportive treatment like bed rest, analgesics, antipyretics, cough suppressants and haematmics The patient responded within a week: fever came down, the blood counts became normal and cough disappeared After two weeks, the superficial swelling was reduced in size and radiologically the abscess cavity was much smaller. The patient was discharged after one month but got lost to follow up.
DISCUSSION

A majority of lung hernias (82%) are acquired2; as many as 52% of them are secondary to penetrating or non-penetrating trauma to chest and just 1% are the pathological type3-5. About 65% of the lung hernias are located in the thoracic region and 35% in the cervical region.6 The pathological variety of the reported lung hennas were mainly following tuberculosis 3-7 The present case is of special interest as it developed following lung abscess. The child was thin-built and had weak muscle mass in the chest Probably, the inflammatory process of lung abscess extended to the mtei costal muscles leading to further

Short Communication

Iml.J. Tuh . 2(11)0, 47,49

TUBERCULOSIS TREATMENT MANAGEMENT UNDER A PRIVATE

MEDICAL PRACTITIONER
KaliCharan Mathur1
(Received on 20.8.98; Accepted on 21.9.99)

INTRODUCTION

Enlisting tuberculosis patients cooperation in adhering to the prescribed drug regimen, dosages, regularity of drug intake and completion of treatment under the conditions of a private medical practice in India is of topical interest. Since a busy private medical practitioner has less time, he may find it hard to enlist such cooperation. It is a common impression that private medical practice does not make adequate efforts to offer organised medical care to tuberculosis cases, which may be the main cause of the emerging MDR-TB in the country.
MATERIAL AND METHODS

All untreated and previously treated patients of tuberculosis who registered at the authors private clinic, from 1 October 1991 to 31 December 1995 constituted the study population. Patients registered for treatment at the authors private clinic are routinely made aware of and fully understand that: i) They have to pay for the consultations, investigations ordered, and the medicines prescribed (to be purchased from the market). ii) Two alphabetical (general and village-wise) indices are maintained at the clinic to monitor their progress, for which they must visit the clinic regularly, mostly every month, to get complete cure. iii) A chest X-ray will be required initially, and later at the end of 2, 6 and 9 months of treatment, as well as a sputum smear examination for AFB at the start and end of treatment ( if sputum is available). Sputum may be tested at the end of second month of

treatment as well. iv) Personal motivation is given by the author to the patient and/or family members at each visit. v) In the event of default in treatment a letter (maximum 2, if necessary) is dropped at the patients village address. The simple system of tuberculosis care detailed above has been found to be practicable and has worked well at the private clinic which is manned by the author and one helper. Neither do the patients wait long, or does the 2-member team work overtime. The following drug regimen is followed: a) First 2 months Rifampicin 10 mg/kg (RHEZ) b) Next 4 months (RHE) Isoniazid 6-8 mg/kg e) Last 3 months (RH) Ethambutol 25 mg/kg d) Self administered at Pyrazmamide 25-30 home mg/kg The chosen SCC regimen, the duration of treatment and the frequency of monitoring check up are somewhat different from those recommended under the governments programme. The changes were introduced keeping in view the best interest of the cases and what is practicable in a private practice. The clinics clientele is 20-25% from BikanerCity, another 25-30% from Bikaner district and the rest from neighbouring districts. They have to be treated at home by self administering the drugs. Around 20% expected irregularity in drug intake was sought to be covered by prolonging the treatment period from six months to nine months so that each case has the best chance of completing at least 7 months of treatment (in 9 months). Accordingly, the definition of completed treatment adopted is intake of drugs for at least 7 months during a period of 9 months.

1. Retired Professor of Tuberculosis and Chest Diseases, Medical College, Bikaner. Correspondence: Dr. K a l i Charan M a t h u r , C-56, Sadul Ganj, Bikaner-334 003 ( R a j a s t h a n )

50

KALI CHARAN MATHUR

Great care is taken that patients take at least 3 drugs in the initial phase of 2 months. Keeping in view the advanced stage of the disease in most of the rural patients and frequently delayed check up visit (upto 4-7 days), it is considered worthwhile to eliminate the costly Pyrazinamide and continue with RHE during the next 4 months and then RH in the last 3 months. More frequent X-ray examination is used as a tool to convince the patient that they are making steady progress. Rural patients are easier to convince after taking a chest X-ray. The same could not be done for sputum examination, because usually cough disappears and often there is no sputum. Monthly visits help the author to maintain a good level of health education and establishing motivation rapport with patients.
RESULTS

difference between the previously treated and the newly diagnosed cases regarding regularity of visits. From the information elicited from patients during the follow up visits regarding regularity of treatment taken, the following picture emerged (Table 3). As told by patients at the time of follow up visit, upto 80% of the patients had taken their treatment regularly for 7 months or more in 9 months. There was hardly any difference between the newly diagnosed and previously treated cases in this regard. The distance between patients residence, urban or rural residence, and whether coming from Bikaner city, Bikaner district or other districts did not interfere much with the cooperation extended for reporting at the clinic or in terms of regularity of drug intake.
Table 2. Distribution of patients according to visits to the c l i n i c and previous treatment status
Number who made avciagc visits <6 7-8 9+ Total 62(29%) 32(33%)
94(30%)

The total number of tuberculosis patients in the Treatment study were 307, of whom 96 had been treated Status previously and 211 were newly diagnosed (Table 1). Untrea.ed
Treated

92(44%) 57(27%) 34(36%) 3 0 ( 3 1 % )

126(41%)

211 96
307

Table 1. Distribution of study population by age, sex and previous treatment status Age Sex (Years) Male 79 25 26 - 50 >51 All 108 35 222 Treatment status Untreated Treated 76 39 101 34 211 46 11 96

Total

87(29%)

Female 36 39 10 85

Total 115 147 45 307

Table3. D i s t r i b u t i o n of patients according to previous tieatmenl status and present treatment adherence
Drugs taken during present treatment Treatment Status Untreated Treated Total Period < 6 months 43 20 63 7 (in months) -8 9Total 211 68 219 96 307

17
8 25

151

Male : Female ratio among the study population was 2.5:1, nearly half (48%) were 26 to 50 years old and 68% of the patients were previously untreated. Patients Cooperation According to the clinics schedule, each patient was expected to make a maximum of 9 visits for monitoring the progress for the duration of treatment period, irrespective of urban/rural residence, distance, etc. The extent of cooperation extended by patients in this respect can be seen in Table 2. More than two thirds of the patients were quite regular in coming to the clinic and there was no

Table4. Distribution of patients assessed foi efficacy of treatment according to different assessment cntei la used and previous treatment status Previous treatment status Criterion Relief in symptoms Radiological improvement Bacteriological conversion (smear negative) (n = 146 ) Untreated ( n = 168) 164 (98%) 157(94%,) 88 (90%.) 88/90 Treated (n = 76) 70 (92%) 60 (80%) 48 (85%,) 48/56

TUBERCULOSIS TREATMENT UNDER PRIVATE PRACTICE

51

Only 244 (80%) of the 307 patients in the cohort were available for assessing the efficacy of treatment at the end of 9 months period (Table 4). CONCLUSION In most tuberculosis control programmes, organisational failures are reported to be higher than any technical short comings1-2. Treatment completion rates have also improved, from about 30% to 55% with the adoption of Short Course Chemotherapy3. Treatment adherence can be further improved if repeated and systematic motivation is undertaken4.5. This study demonstrates that a private medical practitioner too can provide the basic organisational set up and regular motivation during each visit without difficulty. Prescribing drugs for purchase from the market and consultation for a fee are mostly acceptable to private clinic clientele. Therefore, satisfactory results can be obtained which are comparable to any good public sector control programme. The extent of patients cooperation extended to a private medical practitioner in the present study was about the same as reported by the author five years earlier, among freshly diagnosed cases only6.7, usmu 7 months treatment as the criterion for

completion of treatment. Adherence to treatment was about the same in both previously treated (elsewhere) and untreated cases, which suggests that if a reasonably good care is provided, the previous poor experience is no bar to enlist cooperation and get good results.

REFERENCES
1. 2 3. 4 5. M a t h u r K C Review of Tuberculosis Control work of District Clinic; NTI Newsletter, 1967, 4, 52 Leading Article - Result of Treatment, Tubercle, 1968, 49, 235 Nagpaul D.R Indias National Tuberculosis Programme - An Overview; Ind J Tub ; 1989, 36, 205 Ancja K.S., Seetha M.A., Hardhan Smgh and Leela V. Influence of Initial Motivation on treatment of Tuberculosis Patients; Ind J Tub, 1980, 27, 123 Arora V.K. and Bedi R.S. Motivation Assessment Scoring Scale - Its impact on case-holding u n d e r National Tuberculosis Programme, Ind. J Tub;- 1988, 35, 133 Mathur K.C. Problem of drug defaulting and response to treatment taken short-term chemotherapy under advice of a private TB practitioner Paper presented at the 47th National Conference on Tuberculosis and Chest Diseases held at Bombay. Mathur K.C, Singh K.D., Chakravorty A., Bash N. and Khandelwal M.L. Result of Domiciliaiy Treatment in an urban TB Clinic; Ind J Tub, 1964, 11, 68

6.

7.

The Tuberculosis Association of India

3, Red Cross Road, New Delhi-110001 TB Seal Campaign
Tuberculosis patients should take their medicines regularly, in prescribed doses and as long as the doctor advises. They should visit the doctor for check-up and assessment as and when called. The medicines are supplied to the patients free in primary health centres, Government hospitals and TB Clinics. Patients who take medicines regularly do not pose danger to their family and, therefore, isolation is not necessary. Do remember the tuberculosis stricken brethren during the festivals by buying TB Seals A TB Seal Costs Rs. Two Only
Donate generously : Donations to the Tuberculosis Association of India are exempt from payment of Income Tax under Section 80G of Income Tax Act, 1961.
Fight TB

Help

WE SOLICIT YOUR PARTICIPATION IN ALLEVIATING HUMAN SUFFERING Educate people about causation and prevention of TUBERCULOSIS
Courtesy space: M/s. Tata Hydro Electric Power Supply Company Limited, Mumbai M/s. Tata Power Company Limited, Mumbai M/s. Andhra Valley Power Supply Company Limited, Mumbai

Special Article

Ind. J. Tub., 2000, 47,53

A COMPREHENSIVE, MULTIPURPOSE, NATIONAL SAMPLE SURVEY ON TUBERCULOSIS -- A CHALLENGE AND A GOLDEN OPPORTUNITY
S.S. Nair* Till the end of 1940s, tuberculosis in India was considered to be an urban problem. Tuberculosis patients were diagnosed and treated in urban tuberculosis clinics, hospitals and sanatoria. Tuberculin tests carried out in rural and urban areas during the initial stages of the Mass BCG Campaign, in early 1950s, questioned this belief. The resulting doubts and confusion were convincingly cleared by the Tuberculosis Prevalence Survey carried out by the Indian Council of Medical Research (ICMR) in 195558. Systematic research (particularly, operations research)** and planning activities which followed the nation-wide survey resulted in the formulation of the National Tuberculosis Programme (NTP), in 1962. The NTP has been implemented in almost all the districts of the country. During the last 37 years (since the formulation of NTP), the need to ascertain the extent of reduction in the prevalence of Tuberculosis has been raised from time to time to justify the resources utilised for NTP and to take corrective actions, if necessary. NTP has been revised recently - but the above important question remains unanswered, even unattended to. Several limited tuberculosis surveys have been carried out by interested workers and institutions to ascertain current prevalence rates**. Since these could not provide a satisfactory answer to the basic question of effectiveness ofNTP, there have been repeated demands for repetition of national tuberculosis survey, using uniform design and techniques. But, the demands have been turned down because besides being very expensive, it was felt it was too early to conduct such a survey. The programme actually implemented could not have made any dent on the problem since it did not diagnose even a majority of the prevalence cases and the number of cases diagnosed was less than the number of new cases added every year. It was also too early to expect any measurable improvement in the tuberculosis problem because of limited general socio-economic improvement. More than forty years after the 1C MR survey, however, there is a need to ascertain current prevalence rates in different parts of the country. The fact that annual risk of infection has shown a declining trend: without an e ffecti ve control programme, indicates that the prevalence has come down, to some extent, most probably due to socio-economic improvement. Confirming this reduction is very important. It is possible that a fairly satisfactory answer to the trend of tuberculosis can be had by calculating Annual Risk of Infection, if this information can be systematically obtained on a national scale. However, this possibility has not succeeded and most probably will not succeed in curbing the demand for a national tuberculosis survey, which seems to be still looming large in the minds of a si/eable number of tuberculosis workers in the country. Moreover, it is significant that knowledge of the trend of tuberculosis alone will not suffice. It is equally (or even more) important to have adequate information on epidemiological, social, economic and cultural factors and their interactions which not only influence the spread of tuberculosis but also ought to influence choice of methods of tuberculosis control and their effective implementation. Our thinking seems to have been predominantly influenced by technical considerations (like epidemiology and chemotherapy) alone. While provider systems (both governmental and non-governmental) have often been blamed for faulty implementation of NTP, hardly any attempt has been made to provide information on all the above aspects (most of them non-technical) to them so that they can act with a fuller understanding of the situation. For these and some other reasons which are discussed in the following paragraphs a compre-

* Retired Director (Evaluation), Ministry of Health and Family Welfare. Government of India **Individual references arc not given because t h e i r number is larse. Correspondence: Shri S.S. Nair, 618, 3rd Cross, 3rd Block, Koiannmyaki Bangalore - ^60 034

54

S.S. N A I R

hensive, multipurpose, national sample survey on tuberculosis (CMNSST) will be the best and timely approach now, since it alone can satisfy all the information needs of tuberculosis workers, planners, programme implementors and evaluators. Looking back at the ICMR Survey (1955-58), its many limitations are now obvious. Due to some practical considerations, the sample population was selected from six convenient zones, which covered only 40% of the total population of the country. Further, the geographical distribution of these zones was such that some distinct regions of the country like western, central and northeastern regions were completely left out. It is difficult to convince tuberculosis workers in these regions that the problem was not different in their areas. Another important lacuna was that only epidemiological picture was provided, that too incomplete (i.e., rates of prevalence of disease only), even in the areas covered. While an attempt was made to study economic factors, by obtaining the prevalence rates for some economic groups, social and cultural factors were left out. Neither did the survey collect information about health services provider system, its rapport with community in general and chest symptomatics and tuberculosis cases in particular, and factors which influence the same. The interactions between all the above factors were not thought of. Evidently, the total picture, which provides various facets of the complex problem and type of control measures needed, was sadly lacking. Faced with the diversity of the problem, a large number of isolated surveys and studies** have been carried out in different parts of the country by individual research workers and institutions to fill up some of the gaps, which they visualised. These do provide information of sorts on a variety of topics, such as prevalence of tuberculin sensitivity, prevalence of non-specific sensitivity, incidence of infection and disease, effectiveness of BCG vaccination, awareness of tuberculosis, behaviour pattern of chest symptomatics, fate of treated cases, regularity of antituberculosis treatment, annual risk of infection, some bacteriological aspects, influence of H1V infection , etc. These were, no doubt, of topical interest then and were considered as important landmarks. But, w ith hindsight, it seems unfortunate that these sur**I bid, page 53

veys were done without central planning and without developing any concept about the total picture of the tuberculosis problem and its control, which ought to have been and should be the basis for research. They now look like isolated studies, with a limited purpose. Thus, what we got are pictures, like those in the story of blind men and an elephant, where each man got a partial picture and each insisted that he was right. No doubt, such isolated studies failed to meet the information needs for proper planning, implementation and evaluation of a scientific, humane, pragmatic and contextual tuberculosis programme. The situation, naturally, does not satisfy all tuberculosis workers (including social scientists and programme planners, implementors and evaluators) because not only do many questions remain unanswered but the partial answers provided have led to lack of clarity and direction. Moreover, the near total emphasis on technical aspects alone may have led to the failure of NTP. Success of the programme depends more on getting a clearer picture of the social, economic and cultural factors which influence behaviour of tuberculosis cases, chest symptomatics and health care provider systems (governmental and non-governmental) so that all these could be taken into account. Evidently, in the present context, knowledge of factors which could influence the programme is more important for success of the programme than technical matters alone. The only solution is to have a comprehensive, multipurpose, national sample survey of tuberculosis (CiMNSST) which alone can provide a complete picture. Are we prepared to face this reality squarely or shall we continue to sweep it under the carpet, due to lack of concept, courage or otherwise? The various aspects to be covered by CMNSST will have to be discussed by a multidisciplinary team of experts and placed before wider fora leading to their acceptance. Some of the aspects that may be considered for discussions are: 1. Prevalence of infection and disease (to ensure uniformity in diagnosis by x-ray, the possibility of utilising national x-ray readers and computer technology may be explored). 2. Incidence of infection and disease 3. Annual risk of infection.

COMPREHENSIVE, MULTIPURPOSE NATIONAL TUBERCULOSIS SAMPLE SURVEY

55

4.

Prevalence and incidence of non-specific sensitivity, how it affects prevalence and incidence of tuberculosis and possibly spread oftuberculosis. 5. Prevalence and incidence of chest symptomatics and their relief seeking behaviour pattern, at initial survey and subsequently. 6. Identification of the actual illnesses which afflict the vast majority (about 90%) of the non-tuberculous chest symptomatics in different regions and seasons, so that a proper programme to reduce their suffering can be started. This is essential not only from ethical considerations but also for building up confidence of chest symptomatics in health care provider systems, so crucial to the success of NTP. 7. Social, economic and cultural factors which influence behaviour of chest symptomatics, tuberculosis cases and health care provider systems with regard to awareness, actiontaking, regularity of treatment and hygienic measures to restrict spread of infection. 8. Influence of health care provider systems (governmental and non-governmental) on behaviour of chest symptomatics and tuberculosis cases. 9. Extent of community involvement in organising and utilising tuberculosis control services and in preventing spread of infection. 10. Mortality rates among (a) tuberculosis cases with different durations of treatment (b) chest symptomatics with different intervals between occurrence of symptoms and action taking and (c) non-chest symptomatics. 11. Cure rates for different durations and types of treatment among regional, demographic, social, economic and cultural groups. The assumption that all tuberculosis cases require the same duration of treatment is questionable. Studies by National Tuberculosis Institute, Bangalore (NTI)3 have shown that one-third of sputum positive tuberculosis cases, living in rural areas and mostly of lower economic status were cured without any proper treatment. This rate may be

higher for better economic classes.Even among cases not completing full treatment, substantial percentages achieved sputum conversion with different durations of treatment4. It is relevant that many countries controlled tuberculosis without chemotherapy. The declining trend in annual risk of infection, despite a large proportion of tuberculosis cases not being diagnosed and with inadequate treatment of diagnosed cases, if confirmed, also supports this view. It is, indeed, surprising that this very important aspect continues to be overlooked and no further research has been carried out despite such clear indications. 12. Number, type and geographic distribution of private medical practitioners/institutions contacted by tuberculosis cases and chest symptomatics for diagnosis and treatment. 13. Prevalence and incidence of drug resistance, respective contribution to it by different treatment agencies and the extent to which drug resistance affects the tuberculosis problem and its control. 14. Economic burden oftuberculosis on patients, community and the country. 15. Influence of HIV infection on epidemiology of tuberculosis and on behaviour pattern of chest symptomatics and tuberculosis cases. 16. Inter- linkages between the above factors. 17. Determining, to the extent possible, the factors which contribute to the natural decline in prevalence oftuberculosis. All these have to be studied for regional, demographic, social and cultural groups so that flexible and contextual programmes can be evolved. For a better understanding of the full picture, all these investigations have necessarily to be carried out among the same populations in different regions, through actual chest sympytomatics and tuberculosis cases becoming known from the survey. However, many of these investigations can be done on smaller samples (sub-samples) out of the total sample. A CMNSST will, no doubt, be very expensive. A closer scrutiny of cost will reveal that a major part of the cost of tuberculosis surveys is due to the inclusion of mass x-ray examination. If alternative methods, which could provide fairly reliable esti-

56

S.S. N A I R diagnosed in prevalence surveys. Since no tuberculosis survey was conducted in the same population, it was not possible to see what proportion of actual tuberculosis cases in the community were self-reported cases. In other words, a correction factor which could be applied to self-reported prevalence rate to obtain the true prevalence rate could not be calculated because a conventional prevalence survey was not conducted in the same population. On further reflection, it seems that the similarity in the observed distributions by age and sex and actual prevalence may be due to a certain (probably constant) proportion of cases becoming aware that they are tuberculosis cases which is not influenced by age and sex. Most of the tuberculosis cases, except the recent incidence cases, ought to be aware that they may be suffering from tuberculosis and could be selfreporting cases, when method (3) is used. If reliable correction factors can be calculated, any of these methods could be used to provide an estimate of tuberculosis prevalence, at a lower cost. Besides, cases found by methods (1) (a), (1) (b), (2) (a) and (2) (b) could be treated and/or followed-up to obtain estimates of cure and mortality rales, after different durations of treatment. Those found by method (3) could be referred for diagnosis and treatment and similarly followed-up. This will also provide an estimate of the proportion of actual tuberculosis cases among the self-reported cases. As stated earlier, CMNSST will provide baseline information on current status of various indicators of epidemiological, social and cultural aspects of the tuberculosis problem and contextual control measures required in different regional, demographic, social, economic and cultural groups. Further, a series of follow-up investigations could be carried out using this baseline without much additional cost because repeated mass x-ray examination would not be necessary. There is also a distinct possibility that the simpler methods could also be used for obtaining estimates of incidence of tuberculosis, at a much lower cost. All these together will throw light on the complete dynamics of the tuberculosis problem and how to effectively plan, implement and evaluate control measures. The gold mine of information which will become available will far outweigh the cost involved and will make CMNSST unique and cost e fleet n e. Needless to say, this picture of the total dynamics of the tuberculosis problem and control measures is a

mates of prevalence and incidence of disease, could be developed, subsequent surveys for studying incidence and time trend of tuberculosis would become manageable. One of the main aims of and, at the same time, justification of CMNSST should be evolution of simpler methods of survey. There are distinct possibilities that this is feasible. However, to test the suitability of the simpler methods, a onetime survey using mass x-ray examination and the simpler methods, on the same population groups is absolutely necessary. The one-time heavy expenditure on CMNSST will be definitely worthwhile because of two reasons. First, it will provide, for the first time, a total picture of the tuberculosis problem and an essential baseline for evaluation and for planning a series of studies on tuberculosis cases, chest symptomatics and provider systems for health services among regional, demographic, social, economic and cultural groups. Second, it is likely to provide a cheaper methodology for future surveys and studies. One or more of the following possibilities could be tried out for evolving cheaper methods of survey: 1. Examination of multiple specimens of sputum from chest symptomatics, by (a) smear and cultuie and (b) smear alone; 2. examination of multiple specimens of sputum from persons in the community who state on enquiry that they have suggestive symptoms and may be suffering from tuberculosis, by (a) smear and culture and (b) smear alone; 3. enumeration of persons in the community who have suggestive symptoms and may be suffering from tuberculosis; 4. repetition of all the three above after a short interval, instead of a one-time effort. Some more methods could be added through wider discussions and sharing of experiences because it will be very costly to try out other methods later, after the baseline survey with x-ray examination under CMNSST is over. Possibilities (2) and (3) above, particularly (3), may appear to be too naive or even absurd. But, a surprising finding from the baseline survey for a health care project conducted by NTP5 in Tamil Nadu was thai distribution by age and sex of self-reported tuberculosis cases, obtained by method (3), showed a pattern similar to that of cases generally

COMPREHENSIVE, MULTIPURPOSE NATIONAL TUBERCULOSIS SAMPLE SURVEY

57

pre-requisite for proper planning, implementation and evaluation of a scientific and pragmatic tuberculosis programme with enough flexibility and emphasis on contextuality to fully meet all local situations. May be, with this knowledge of the total dynamics of the tuberculosis problem and control measures required, a programme to eradicate tuberculosis, with active participation and full co-operation of all concerned could be evolved, in due course. The cost of CMNSST will be so high that it is unlikely to be funded from the budget of any single organisation. The best way will be to pool resources of different organisations such as ICMR, Government of India, WHO, World Bank, SIDA, DFID, DAN1DA, NORAD, Swiss Tropical Institute, Ross Institute of Tropical Medicine, European Commission, HIV funds etc. Keeping in mind the importance of such a survey for the whole world, in a country where this is possible, serious attempts deserve to be made to organise multiple funding for it with a sense of priority. It is hoped that this question will be suitably debated and a mechanism of multiple funding evolved using the ingenuity and committed sense of purpose of these organisations. It is relevant to mention that substantial resources are being spent even now by different organisations in various parts of the country and by international agencies for ad hoc studies conducted by them, which, as pointed out earlier, lead only to partial answers and lack of clarity and direction. It should be possible to pool at least a major part of these resources for CMNSST. These organisations could be encouraged to take up some, out of the large number of special studies which will have to be conducted in different regions of the country using the baseline provided by CMNSST, after adopting uniform designs and techniques to ensure comparability of findings. This will also ensure that most of the tuberculosis research conducted in the country ( i f not all) will be within the basic concept of the tuberculosis problem and its control emerging from CMNSST. If the need for CMNSST is agreed to, and mul-

tiple funding is feasible, the next steps will be to prepare a broad outline of the survey, identify a central nodal agency and study centres in different regions of the country to carry out the survey, assess available resources, which can be pooled and further resources needed. All these could preferably be entrusted to a multidisciplinary expert team consisting of at least one epidemiologist, statistician, social scientist, bacteriologist, management specialist and accountant. On the basis of the report of the expert team, it will be preferable to set up an autonomous nodal agency with adequate expertise to take full responsibility of planning and conducting CMNSST and preparation for a series of reports, all of which will take some years to be completed. This nodal agency could set up multidisciplinary committees to assist it on various aspects of the survey. The proposed CMNSST is, simultaneously, a challenge and a golden opportunity for taking a unique and bold step forward. The question is: are we wise enough, farsighted enough and bold enough to join together and take this unique step forward, now? REFERENCES
1. Indian Council of Medical Research. Tuberculosis in India - A Sample Survey (1955-58): Special Report Scries No. 34, New Delhi 2. C h a k r a b o r t y A K , C h o u d h a n K, S r e e n i v a s TR, Knshnamurthy MS, Shushidhara AN, Channabasavaiah R. Tuberculosis Infection in a Rural Population of south India; 23-year trend; Tubercle and Lung Dis.: 1992, 73. 213
3. National Tuberculosis Institute, Bangalore; Tuberculosis in a Rural Population of South India; A Five Year Epidcmiological Study; Hull WHO. 1974,51.473 4. Baily GV.I, Samuel GER and Nagpaul DR, A Concurrent Comparison of an Unsuperviscd Self-administered Daily Regimen and a Fully Supervised Twice Weekly Regimen of Chemotherapy in a Routine Out-patient Treatment Pogramme; Ind. J Tub.: 1974, 2 1 , 152 5. National Tuberculosis Institute, Bangalore. Vol 1 , 1 88. Report on the Baseline Survey DAN I D A Health Care Project, Tamil Nudu (1988)

To Philanthropists and Well Wishers

Donate Generously
Donations to the Tuberculosis Association of India are exempt from payment of income tax under Section 80-G of the Indian Income Tax Act, 1961. Donations for specific research work in tuberculosis and allied diseases of the chest given to the Tuberculosis Association of India, are exempt from Income Tax under Rule 6 Clause (ii) of sub section (1) of section 35 of the Income Tax Act, 1961 under the category "Institution" vide notification No. 1149 (F. No. DG/ IT (E)/ND-49/35(l) (ii)/90-91 dated 9th June, 94 published in part II section 3 (ii) of Ga/ette of India, Ministry of Finance (Department of Revenue), validated upto 31.3.1996 (and applied for revalidation).

Courtesy space: M/s. Tata Hydro Electric Power Supply Company Limited, Mumbai M/s. Tata Power Company Limited, Mumbai M/s. Andhra Valley Power Supply Company Limited, Mumbai

Special Article

1nd. J. Tub.. 2000, 47.59

RECOMMENDATIONS OF THE NATIONAL CONSENSUS CONFERENCE ON TUBERCULOSIS CONTROL - NEW DELHI - NOVEMBER, 1997*
DIAGNOSIS OF TUBERCULOSIS

1. Sputum microscopy is the primary diagnostic tool for pulmonary tuberculosis. Diagnosis of tuberculosis should be done systematically in patients attending health facilities. Three sputum smears, for AFB examination, should be obtained from all patients with cough for 3 weeks or more. Topmost priority should be given to diagnosing and curing patients with sputum positive tuberculosis as these patients are the source of infection in their community and because they have the highest risk of dying of pulmonary tuberculosis. 2. Unlike strategies in other disease control programmes, efforts to find tuberculosis patients in the community through active search or diagnostic camps should never be undertaken. Studies in India conclusively show that the vast majority of tuberculosis patients present to medical facilities soon after the onset of illness. Efforts at education should be directed towards the health staff to ensure that all patients with cough for 3 weeks or more undergo sputum examination in a competent laboratory. Active case finding in the community will have little yield, as compared to the inputs, and will substantially increase the number of patients who are likely to develop drug resistant tuberculosis, which they will spread to their family and community. 3. X-ray should not be used as the primary diagnostic tool for tuberculosis because it is nonspecific and will result in over-diagnosis of active tuberculosis. However, X-ray is an important complementary tool in the diagnosis of smear negative cases and for those who have only a single positive smear. 4. Intensive information should be disseminated about the free availability of diagnosistic facilities and location of diagnostic centres. Sputum microscopy services with systematic

quality control should be established by the government and used as a diagnostic resource by the medical community. 5. Sputum microscopy centres for diagnosis should be located in all medical colleges in RNTCP districts. These can also be used as training centres. 6. The expertise available in the medical colleges can be used for training purposes for which the core faculty at medical colleges may be trained at central institutes. 7. Mantoux test has no role in the diagnosis of tuberculosis above the age of 5 years. ELISA, PCR, rapid culture methods and other sophisticated tools are also not recommended for diagnosis, under programme conditions.
TREATMENT OF TUBERCULOSIS

1. Regimens, modes of administration and duration recommended by the National Tuberculosis Programme and the Revised National Tuberculosis Control Programme (RNTCP) are effective and are recommended to be used by all medical practitioners. The number of pills in the combipack should be reduced, if possible, maintaining the same dosage. 2. Experience with current dosage in terms of efficacy and toxicity in different body weight groups (especially less than 30 kg and more than 60 kg) should be systematically evaluated in India, and based on this information, policy is to be modified, if necessary. 3. A regular and reliable supply of good quality antituberculosis drugs must be ensured. Facilities for treating patient under the RNTCP should have patient-wise boxes in place to ensure that patients who are started on treatment will have the full course of medications to complete treatment. 4. Anti-tuberculosis drugs should not be available without a doctors prescription.

*Central TB Division, Directorate General of Health Services, M i n i s t ry of Health and Family Welfare, Nirman Bhawan, New D e l h i - 1 1 0 01 1.

60 5.

CONSENSUS CONFERENCE

Follow-up sputum examinations are essential for determining patient progress and treatment outcome. 6. Priority must be to implement the NTP and RNTCP in order to prevent MDR-TB, because a poorly performing programme will create MDR-TB at a rate faster than these cases can be cured, even if unlimited resources were available. A good tuberculosis control programme, such as the RNTCP can prevent the emergence of MDRTB, and can reverse the trend of MDR-TB in communities in which it has emerged. Treatment of MDR-TB is costly, toxic and often unsuccessful, and should only be undertaken in specialized centres. 7. Before any patient is labelled as a chronic case and considered for MDR-TB treatment, he should have received Category II treatment under direct observation. ADMINISTRATION AND ORGANIZATION OF ANTI-TUBERCULOSIS TREATMENT 1. Patients non-adherence to treatment, even when medications are available, is a serious problem which reduces the success of tuberculosis control measures and facilitates the spread of drug resistance. There are many causes for patient nonadherence, including poor interpersonal communication skills and inaccessibility of treatment facilities. The only reliable and proven method of ensuring adherence to treatment is in a well-functioning programme of directly observed treatment (DOT), in which a trained individual who is not a family member watches and assists the patient swallow his pills. DOT should be arranged as conveniently as possible to the patient, relying on strengths and resources which exist in each community (e.g. multipurpose workers, Anganwadi workers, trained Dais, Panchayat leaders, religious leaders, etc.). While taking due care against programme expansion without full preparation and ground work, the RNTCP, using DOT and short-course chemotherapy, should be implemented throughout India as rapidly as possible, while maintaining a high quality of patient services and supervision. 2. There should be systematic monitoring of and accountability for the outcome of all patients

started on treatment for tuberculosis, particularly of those who receive regimens co n tain in g Rifampicin. Standardized outcome measures, using follow up sputum smears for patients with smear positive tuberculosis, should be used to evaluate and improve performance in each health facility, treatment unit, district, state, and country as a whole. Sputum conversion rates and rates of cure, treatment completion, failure, transfer out, default and death should be calculated according to definitions of the RNTCP and should meet the standards laid down, especially those for cure. 3. When the RNTCP is implemented in a district, the whole area of the district should be covered at the same time. All the central, state and public sector, viz. Armed Forces, Railways, CGHS, ESI and medical colleges, etc. should be involved and should participate actively. They should establish treatment units and microscopy centres as per the population norms, and follow DOTS strategy and RNTCP guidelines. These agencies should submit quarterly reports as per schedule on prescribed formats. 4. All RNTCP centres should be provided with the particulars of location, etc. of other RNTCP districts/sub-districts in the country so as to facilitate transferring of the patients. 5. Non-RNTCP districts should also be strengthened by providing vehicles/drugs and required logistics at the earliest. Coordination ofTuberculosis Control Services 1. To achieve greater coordination of tuberculosis control services, there is need for continued political support and commitment. There is also a need to increase p u b l i c awareness, IEC activities and advocacy. 2. Governmental and non-governmental facilities providing tuberculosis services must co-ordinate better. Ways must be found by Central and State Governments to involve Railway s, Armed Forces, Medical Colleges, ESI and NGOs in the RNTCP. These agencies should be able to ensure over 85% cure rates. In a d d i t i o n , p r i v a t e practitioners, general hospitals and District 1 B Centres s h o u l d h a v e a p r o p e r s y s t e m to coordinate referrals for diagnosis and treatment. 3. NGOs and corporate bodies should share (heir

RECOMMENDATIONS FORTUBERCULOSIS CONTROL

61

experiences with the government in order to improve the programme. 4. Physicians have a responsibility towards their patients, their profession and their community. In the case of tuberculosis, their responsibility is to ensure t h a t patients are cured and stop spreading infection. This responsibility should be the central theme of teaching in medical colleges. As per the Medical Council of Indias recommendations, undergraduate students and interns should have postings in tuberculosis departments. 5. Leaders in the medical community, such as professors in medical colleges, are crucial to success of tuberculosis control efforts. They should set the standard of care, including using anti-tuberculosis regimens recommended by the Government of India. All doses of medicines given in hospital should be directly observed. To patients living in RNTCP areas, the RNTCP regimens should be given, and follow-up care including direct observation of treatment should be closely co-ordinated with the District T uberculosis Centre/Chest Clinic in the area. 6. Professional updates on tuberculosis should be

sent to all medical colleges for use in teaching, to the Indian Medical Association (IMA) for dissemination of information among its members, and to Directors of Health Services of States for dissemination to the hospitals and other health care outlets. 7. The IMA, Public Sector Undertakings and NGOs may be identified as individual units so that training and IEC activities, etc. can be planned by them with regular budgeting and support. 8. The medical education curriculum should made fully consistent with national policies on control of tuberculosis. State medical education authorities and programme implementing authorities should meet at regular intervals to discuss consistency of the curriculum with national policies and to ensure effective coordination so as to achieve the goals of the RNTCP. 9. Funds should be committed to organise CME programmes in each district on a regular basis. The IMA should be recognized as a nodal agency for this purpose. 10. Medical colleges which adopt RNTCP and open a DOTS centre should receive the supply of drugs from the national programme.

Ind J Tub 2000 4762

CONTEMPORARY ISSUES
INSTANT CROSS MATCHING OF MDR STRAINS

At the US Department of Energys Argonne National Laboratory, Chicago and the Russian Academy of Sciences Engelhardt Institute of Molecular Biology, Moscow, a bio-chip has been developed, almost simultaneously The biochip, like any other microchip, can carry out thousands of biochemical reactions instantly Mounted on a glass slide, there are about 10,000 micro-gel pads, each carrying a labelled DN A strand got from a drug resistant strain of M tuberculosis A DNA strand from any suspected drug resistant strain, when placed on the slide is instantly recognised and cross-matched with the known MDR strains.
A UNANI ANTI ASTHAMATIC DRUG

way into new regimens The adoption of new combinations and newer uses of the currently prescribed drugs is a distinct possibility Search tor new derivatives of Rifamycm, Fluoroqumolones and Nitroimidazole continues apace Immune amplifiers such as Intelferon-gamma, Inteileukui 2 and Intel leukm 12 also hold bright piospects With the complete mapping of M. tuberculosis genome, enzymes unique to the bacillus ai e likely to be used to interrupt the metabolic pathways and transcription factors that bind DNA but do not promote RNA Phages may deliver antisense nucleic acid tor inhibition of mycobacterial gene expression It is likely that the cm rent distinction between di ugs, vaccine and immunotherapeutic agents may disappear

HIV AND ETHICS

A reaction appears to be brewing in India regarding the limits to which confidentiality about the HIV status of an individual should be maintained Reportedly, the Delhi AIDS Control Society has written to the Indian Medical Association on the merits of keeping the HIV status a secret, even between spouses/would be spouses Confidentiality, it is claimed could not be placed above the right to health and life ol the conjugal partner. The move is apparently in line with the recent Superme Court judgement wh i c h held that the H I V positive appellants light to privacy had to yield to the fiancees right to life Some NGOs may be planning to move the court to review the said judgement because, it is thought, it may force some people not to get HIV tested in fear of jeopardising their marriage or happy married life prospects Others have started doubting if confidentiality is a proven key factor in controlling spread of the AIDS epidemic In fact, an informed spouse, it is argued, may be better equipped to arrange for protection and reduce the onslaught of the epidemic. The developing debate may highlight the need for an HIV positive person being given the information fiist along with the right to inform the spouse/children about the situation, before being persuaded and encouraged to divulge the information to the family personally, failing which the diagnostic centre should do so.

The Central Council for Research in Unani Medicine has applied for an Indian patent for two herbsbased formulations for treating bronchial asthma The National Research Development Corporation will provide help in patenting as well as commercialisation of the Unami medicines Traditionally, Unani medications also take into consideration an individual patient s biological make up as well as temperament Human trials have shown that the said formulations are effective and cost much less than similarmedications available m the market These formulations reduce cough, wheezing, bronchial irritability leading to paroxysmal cough and viscosity of sputum-secietions Other effective Unani medicines are also being made available for treating malaria, infective hepatitis and rheumatoid arthritis
TUBERCULOSIS TREATMENT AT THE DRAWN OF THE NEW CENTURY

It is being anticipated that new anti-tuberculosis regimens will be introduced in the near future from (I) new uses of the known drugs, (ii) new ways of delivery of the known drugs, (iii) new derivatives obtained from within the known groups of drugs and (iv) entirely new drugs Some of the old drugs, such as Clofaznnine, its analogue Rifabutm,the macrolides, ammoglycosides, quinolones and perhaps Vitamin D may find their

Iml. J. Tub , 21)1)0, 47.63

FORUM
Hepatic Cryptococcosis

1. I have gone through with interest the paper Hepatic Cryptococcosis in Association with Pulmonary Tuberculosis and Amoebic Liver Abscess by M.P.S. Menonetal published recently (hid. J. Tub.; 1999, 46, 137). The case report deals prominently with the unusual detection of Cryptococcus in an amoebic liver abscess. However, the diagnosis of Cryptococcosis does not seem to be very convincing. There is no culture of detection antigen for C. neoformam. Only on the basis of histopathological study which revealed the presence of budding yeast forms in the aspirate after PAS and India In k staining teclirdques,was diagnosis of Cryptococcosis made. I presume that there was no biopsy study. The smear stained by PAS shown in the photographs Fig.2 and Fig. 3, specially in Fig. 3, does not confirm the presence of C.neoformans. For diagnosis of Cryptococcosis, only on microscopy, demonstration of capsule or direct fluorescent antibody positivity is essential. The capsule can even be seen as a negative stained area on PAS staining. Thus, the slides may be reviewed by an expert medical mycologist before confirming the case to be an unusual case of Cryptococcosis in an amoebic liver abscess. A. Chakrabarti Post-graduate Institute of Medical Educat. & Research, Chandigarh 2. Your attention is invited to the paper entitled, Hepatic Cryptococcosis in Association with Tuberculosis and Amoebic Liver Abscess by M.P.S. Menon et al (Ind J Tub, 1999,46, 137). It is disappointing to note that the authors have diagnosed Cryptococcosis without proper evidence. The diagnosis is claimed on the basis of the following perfunctory observations: Further histopathological study revealed the presence of budding yeast forms in the aspirate which after PAS and India Ink staining techniques were diagnosed to be cryptococci (Figs. 2,3). In this context, the following comments are warranted:

1. It is important to point out that the presence of budding yeast forms in a histopathological preparation is never diagnostic of Cryptococcosis. Apart from species of the genus Cryptococcus, yeast-like fungi belonging to diverse genera such as Candida, Rliodotontla, Malassezid, Saccharomyces. Sporoholonivcas, Trichosporoit, etc. all have budding yeast forms. 2. One of the distinguishing features Cryptococcus neoformans, the etiologic agent of Cryptococcosis, is the presence of a polysaccharide capsule around its cells, which is demonstrable in culture as well as in clinic specimens 1-4. Notably, no mention is made of this characteristic by the authors nor is this feature evident in the coloured photomicrographs which are otherwise of high quality. 3. In the absence of culture evidence, an unequivocal diagnosis requires that the liver aspirate be examined after mucicarmine staining by the authors which would have been specific for demonstrating C. neofornunu capsule 1-4. Moreover, no attempt was made to get the latex agglutination test done, which is highly specific for the diagnosis of Cryptococcosis. 1.5 4. Fig. 3 is purported to illustrate the budding yeast forms of the etiologic agent. The fact is that not a single budding yeast is seen in this figure. On the contrary, it illustrates profusely branched hyphae, and a number of deeplystained, thick-walled, rectangular to cylindrical spores which are known as arthroconidia (arthrospores) in mycological jargon. The presence of hyphae as well as these spores leaves no doubt that the etiologic agent in this case could be anything but C. neoformans. Actually, the illustrated morphology is highly suggestive of an arthroconidial fungus belonging to any of the genera, such as Trichosporon, Geotricliiini etc . My comments on the above paper are in the i n terest of the readers of IJT, lest they get misled reg a r d i n g the c o r r e c t d i a g n o s t i c c r i t e r i a for Cryptococcosis - a pulmonary and systemic mycosis

64

FORUM

of rapidly growing importance, particularly in patients with AIDS. (Prof.) Z.U. Khan, Dept. of Microbiology, Kuwait University, Kuwait References
1. Kwon-Chung, K.J., and Bennett, J.E.: Cryptococcosis. In Medical Mycology, Lea Fcbigcr, Philadelphia; 1992, 397 2. Littman, M.L. and Zimmcrman, L.E.: Cryptococcosis. Grunc and Stratton, New York, 1956 3. Cox, G.M. and Perfect, J.R.: Cryplococcux neoformans var neoformans and gatti and Trichosporson species. In: Toplcy and Wilsons Microbiology and Microbial Infections, 9th edition, Volume 4: Medical Mycology, Editors: L. Ajcllo and R.J. Hay, Arnold, London, 1998, 461 4. Chandler, F.W.: Histopathological diagnosis of mycotic diseases. Ibid, 1998, 1 1 1 5. Matthews, R., and Burnic, J.: Mycoscrology. Ibid, 1998, 89

Cryptococcosis should be considered in the histological diagnosis of virtually any yeast infection. 3. Regarding the comment there is no culture or antigen detection etc., attention is once again drawn to the case report which clearly states: A clinical diagnosis of fibrocaseous tuberculosis of lung with amoebic liver abscess was made ... and the patient expired even before the laboratory reports were received... Since Cryptococcosis was not suspected, no further mycological workup could be planned as the patient died in Hindu Rao Hospitals surgical unit. 4. The reasons and purpose of publishing this case report are: (a) the serendipitous way the diagnosis was brought to our notice by Dr. Sudhir Jain, (b) when we see an already well known pattern, why take the trouble of publishing it and (c) look for the unusual, then we may see something new. Public Health and Ethics That Pro lessors Porter and Ogden had taken my comments on their paper (Ind J. Tub; 1999,46,3) in the spirit in which these were written re fleets their academic stature. Even more encouraging is their contenlion that their paper was meant to stimulate a debate on ethical and public health aspects of the RNTCP. I am making this small comment in response to their seeking my opinion on the question of the role of the so-called outsiders in the debate on the RNTCP. Let me stress at the outset that I had used the term, outsiders in a positive sense, appreciating the authors giving a new perspective to the scientific debate. Interestingly, the RNTCP has outside genesis and it is being implemented by the authorities in India strictly on the lines set up by the outsiders. It reminds me of a communication to me from the British High Commission in New Delhi, inviting research proposals on the condition that they be within the confines of the framework of the RNTCP. 1 reminded them about my detailed monograph, entitled. Serious Implications of the Revised National Tuberculosis Control Programme and pointed out the incongruity of confining research within prescribed limits. Indeed, the question of outsiders and insiders does not arise at all when we discuss issues conceniinu science and scientific methods. Il also does

The authors reply: We are grateful to Dr. A. Chakrabarti and Professor Z.U. Khan for having gone through the case report so carefully (Ind J Tub; 1999, 46, 137). The comments are well received. Ordinarily, demonstration of capsule, is considered essential for the diagnosis of Cryptococcosis on microscopy, especially when culture, serology or fluorescent antibody confirmation are not possible. Dr. Chakrabarti is right in pointing this out, however: 1. When one deals with tissue specimens (direct microscopy) not all C. neofonnans exhibit a capsule (see Diagnostic Microbiology by Baily & Scott under Characteristic features of Cryptococcus in direct examination of clinical specimens). 2. Regarding Dr. Chakrabartis comments on Fig. 3 (he is sure that it is not C. neofonnans but is equally unsure what it is) we would like to quote from the Textbook of Pathology by Anderson regarding tissue sections containing C. neoformans: Budding cells are numerous in lesions that are attached to parent cells by narrow bases and multiple buds are occasionally seen. Pseudohyphae and rarely true hyphae may be formed in tissues; Because C.neofonnans is usually plcomorphic and its encapsulated forms are not always conspicuous,

FORUM

65

not permit taking sides. One has to take the side of scientifically established facts. If there are differences of opinion, then concepts and data from both the sides have to be presented to provide what the authors call a balance. This had not been done in the authors article or in their comments. I do not have anything to say on the authors response to my comments. It is, however, noteworthy that they chose not to respond to some other key issues I had raised, questioning the scientific validity of the RNTCP, for instance. Can one think of building a good, cost-effective, felt-need oriented tuberculosis programme on a health services infrastructure that is grievously wounded by the imposition of big vertical programmes from outside? D. Banerji Professor Emeritus Javvaharlal Nehru University, New Delhi
Agenda for Research

ernment officials should sit together and find a via media where patients can be benefited and Tuberculosis Control Programme can be pushed forward, successfully. I also agree that inadequately trained staff, poor supervision and poor corrective actions need to be looked into. It is also suggested that technical audit of the above programme be essentially envisaged as early as possible. R.P. Bhagi, Delhi
Mantoux Test: Longitudinal/Transverse Diameter Measurement

This is in reference to the special article An Agenda for Research in Tuberculosis Control (Ind J. Tub; 1999,46, 161). I congratulate Dr. Nagpaul for his bold and timely article on RNTCP. I entirely agree that strategies borrowed from successful programmes under different sociological and operational conditions cannot be depended upon to replicate success. And the position till now, of the revised NTP is not without concern as no results have been made available regarding the progress made by this programme. There have been frequent expressions of scepticism regarding one or the other component of the revised strategy from senior and experienced tuberculosis workers and that involvement of voluntary organisations which is worth mentioning actively in control programme has not been noticed. Dr. Nagpaul has also mentioned that surprisingly, no efforts are apparently being made to closely study the new HI V epidemic or the twin AIDS/TB epidemic m epidemiological terms. This is not only a serious matter but one of great and grave concern. In this connection, I would like to refer to my presidential address at Ahmedabad under the heading RNTCP in winch I suggested that our programme managers should take a healthy view, positive attitude and try to plug the lacunae sincerely. And the everything can be clone by us attitude may not suit the control programme. Also that NGOs and Gov-

Chadha et al (Ind. J. Tub; 1999,46,105) deserve appreciation for their original article Estimation of Annual Risk of Tuberculosis Infection among BCG Vaccinated. In almost all epidemiological studies, it has been the convention to measure the transverse diameter of induration in millimeters for the reading of Mantoux test. It may, however, be more specific to express the reading in both the diameters, if there is a difference between the transverse and longitudinal diameters of an induration, e.g. (10 x 14 mm). Would the authors like to comment why the measurement of longitudinal diameter alone was selected for their study? V.K. Dhingra New Delhi TB Centre New Delhi The authors reply: The technique of measuring the longitudinal diameter of tuberculin test indurations was developed by WHO consultants at NTI, Bangalore during its formative years (personal communication. M.S. Knshnamurthy, Senior Investigator (retired). It was felt by them that it would be comparatively easier to read the induration in this way instead of on the transeverse curvature. This procedure has been followed during all the subsequent surveys. Therefore, it was deemed appropriate not to change the practice to which the tuberculin readers were accustomed, since it is the frequency distribution of tuberculin test indurations that helps to decide the proportion of persons who could be considered as infected and not the results in individuals. However, the practice at this Institute lias been changed to the measurement of transverse diameter, since 1995.

Ind. J. Tub., 2000, 47,66

NEWS & NOTES

INAUGURATION OF CAMPAIGN IN STATES 50TH TB SEAL 3. Effectiveness of RNTCP diagnosis and treatment policies in HIV infected tuberculosis patients System must be appropriate in the Indian context, without involving complicated technology and/ or h i g h cost. Cost effectiveness m u s t be evaluated. For details, please contact Dr. G.R. Khatri, DDG (TB), Directorate General of Health Services, Nirnian Bhawan, New Delhi-110 001. Telephone andFaxNo.3018126

In Bihar, the campaign was inaugurated by Dr. Vijyeshwar Sliarma, Professor of Medicine, Patna Medical College Hospital, on 2nd October, 1999 in the office of the Bihar TB Association. The function was presided over by Shri U.N. Vidyarthi, Chairman of the Association. In Gujarat, the campaign was inaugurated by Dr. Kirtibhai M. Patel, President-elect, IMA Gujarat State branch on 3 rd October, 1999. Dr. Nitin S. Vora, Director, ESIS, Ahmedabad was the Guest of Honour. Dr. S.B. Trivedi, Chairman of the Gujarat State TB Association presided over the function.
OPERATIONAL RESEARCH ON RNTCP

HEALTH VISITORS COURSE

The Central Tuberculosis Division of the Directorate General of Health Services, Ministry of Health, Government of India is inviting proposals in the field of operational research on RNTCP to provide grant-in-aid. The following priority areas have been chosen for receiving proposals and brief guidelines are given below: 1. Pi l ot projects to improve coordination between p u b l i c and private sectors in implementing RNTCP Must be implemented in close coordination with the district (municipal) tuberculosis centres; inputs must be sustainable over a long period; medications must not be handed over to private practitioners but provided for those individual patients who are under their treatment; may involve individual practitioners, trust-hospitals, nursing homes, NGOs, etc.; record-keeping and defaulter retrieval must be ensured. 2. Effectiveness of different types of DOT providers at RNTCP sites, with the strengths and weaknesses of each type Different types of DOT providers to be identified i n d i v i d u a l l y , or through an NGO b u t not i n c l u d i n g f a m i l y members; either no cash incentive or upto Rs. 175/- paid per patient cured; strict adherence to the principles of RNTCP.

The 2000-2001 TB Health Visitors Course will commence in July, 2000. The course will be of nine months duration and will be held at the New Delhi TB Centre. The minimum qualification for admission to this course is 10+2 with Science and/or Hygiene. Science education upto 10th class is essential. Application forms for admission to the course can be had from the Secretary General, Tuberculosis Association of India, 3, Red Cross Road, New Delhi110 001. The last date for receipt of applications is 1st May, 2000.

CHANCHAL SINGH MEMORIAL AWARD - 2000

The Tuberculosis Association of India awards every year a cash prize of Rs.1000/- to a medical graduate (non-medical scientists working as bacteriologists, biochemists, etc. in the field of tuberculosis are also eligible) who is below 45 years of age and is working in tuberculosis, for an original article, not exceeding 30 double spaced foolscap size pages (approximately 6,000 words, excluding charts and diagrams) on a subject relating to tuberculosis. Articles already published or based on work of more than one author will not be considered. Papers may be sent, in quadruplicate, to reach the Secretary General, Tuberculosis Association of India, 3, Red Cross Road, New Delhi-110 001, before 1st July, 2000.

NEWS & NOTES

67

ESSAY COMPETITION- 2000

The Tuberculosis Association of India awards every year a cash prize of Rs.500/- to a final year medical student in India for an original essay on tuberculosis. The subject selected for the year 2000 competition is Status of HIV/Tuberculosis Epidemic in India. The essay should be written in English, typed double spaced, on foolscap size paper and should not exceed 15 pages (approximately 3,000 words including tables, diagrams, etc.). Four copies of the typescript should be forwarded through the Dean or Principal of College/University to reach the Secretary General, Tuberculosis Association of India, 3, Red Cross Road, New Delhi-110 001, before 1st July, 2000 with a certificate that the author is a final year medical student.
NEWS FROM KIYOSE (JAPAN)

Group Training Courses (i) Eleven National Tuberculosis Programme Management officers from 10 countries participated in a six week training course, from May 11 to June 19, 1998. Her Imperial Highness, Princess Akishino met with the participants. (ii) Twenty one nominees from all over the world attended the Tuberculosis Control Course II, from July 6 to October, 1998. (iii) Nine students from different countries attended the Tuberculosis Laboratory Services Course, from October 26 1998 to February, 1999. (iv) Twenty two participants from 16 countries attended the International Training Course on HIV/AIDS Prevention and Care in Asia course, from November 4 to December 11, 1998. In addition, a workshop titled DOTS: Experiences of Different Countries was held under the auspices of the Research Institute of Tuberculosis in kiyosechi and the Royal Dutch Anti-Tuberculosis Association
ANNUAL GENERAL MEETING OF TUBERCULOSIS ASSOCIATION OF INDIA The Annual General Meeting of the Tuberculosis Association of India was held on Thursday, the 2nd December, 1999 in the Associations Conference Hall.Before commencing the proceedings, all present

stood in silence for two minutes on the passing away of Dr. P.K. Sen, former President of the Association and Emeritus Editor of the Indian Journal of Tuberculosis, as a mark of respect to the departed soul. Dr. D.R. Nagpaul, Vice-Chairman , welcomed the dignitaries present at the meeting. Dr. S.P. Agarwal, Director General of Health Services and Chairman of the Association, presented the Annual Report of the Association for the years 1995-96 to 1998-99, and Shri M.P. Gupta, the Honorary Treasurer, presented the accounts of the Association for the years 1995-96 to 1998-99. Dr.S.P. Agarwal gave away awards for outstanding activities undertaken by the State Associations and also for the highest collections made from the TB Seal Campaign. Drs. M.M. Singh, Chander Mohini, I. Ranga Rao, B.M. Soni, N.R. Patel, P.L. Jindal and D.P. Mangla were elected members of the Central Committee for the year 1999-2000.
MEETING OF CENTRAL COMMITTEE

The Central Committee met under the Chairmanship of Dr. S.P. Agarwal, Director General of Health Services. The Committee adopted the recommendations made by the Annual General Meeting of the Association as well as recorded the various minutes of the Finance and Executive Committees. The accounts of the Association presented by the Honorary Treasurer were also adopted. The Central Committee co-opted Drs. V. K. Arora, P.R. Narayanan, R.C. Jain, P. Jagota and M.M. Singh as members on the recommendations of the Finance and Executive Committees. The Central Committee also re-constituted these Committees for the next year. The Central Committee reconstituted the New Delhi TB Centres Managing Committee for 19992000 and also appointed M/s. Ravi Verma & Company as auditors. The Central Committee unanimously elected Dr. D.R. Nagpaul and Dr.M.M. Singh for the offices of President and Vice-Chairman of the Association respectively. Meetings of Secretaries of State TB Associations as well as of the Standing Technical Committee were also held.

Ind.J. Tub, 2000, 47,68

ABSTRACTS
Fifteen Year follow-up of Trial of BCG Vaccines in south India for Tuberculosis Prevention
Tuberculosis Research Centre (ICMR), Chennai; Iml. J. Mad. Res.; 1999,110,66

A large scale community-based double blind randomised controlled trial was carried out in Chingleput district of south India to evaluate the protective effect of BCG against bacillary forms of pulmonary tuberculosis. From among 366,625 individuals registered, 281, 161 persons were vaccinated with BCG or placebo by random allocation. Two strains of BCG were used, the French and Danish, with a high dose (0.1 mg/0.1 mi) and a low dose (0.01 mg/0.1 ml) for each strain. The entire population was followed up for 15 years by means of resurveys every 30 months, and selective follow up for every 10 months, and continuous passive case-finding. There were 560 cases (189, 191 and 180 from the high dose, low dose and placebo groups respectively) arising over 15 years, among 109,873 persons who were tuberculin negative and had a normal chest x-ray at intake. This represents a small fraction of the total incidence of 2.6 per 1000 person years, most of which came from those who were initially tuberculin positive. The incidence rates in the three vaccination groups were similar, confirming the complete lack of protective efficacy, seen at the end of 7 1/2 years. BCG offered no overall protection in adults and a low level of overall protection (27%, 95% C.L 8 to 50%) in children. This lack of protection could not be explained by methodological flaws, or the influence of prior sensitization by non-specific sensitivity or because most of the cases arose as a result of exogenous re-infection. The findings after 15 years show that in this population with high infection rates and high non-specific sensitivity, BCG did not offer any protection against adult forms of bacillary pulmonary tuberculosis. Effectiveness of BCG Vaccination against Tuberculous Meningitis
Shally Awasllu and Sonfia Moin; Indian Pediatrics: 199, 36. 455

The protective effect of BCG vaccination against tuberculous meningitis was assessed in children 1 month to 12 years of age, after controlling for age, nutrition and socio-economic status. This casecontrol study was conducted in a hospital in Lucknow. Cases were those conforming to the definition of tuberculous meningitis and controls were patients admitted after every third consecutive case included in the study, from September 1995 till the end of August 1997, who did not suffer from any central nervous system disorder. Among the 192 cases and 70 controls, BCG scar was present in 57.8% and 75.7%, respectively. The crude odds ratio (OR) for tuberculous meningitis with a BCG scar was 0.44 (95% CI, 0.24-0.81; P=0.008), while the adjusted OR was 0.53 (95% CI, O.26-1.06; P value = 007) after controlling for weight, age, sex and place of residence. Higher weight for age and urban residence were associated with a decreased risk of tuberculous meningitis in the logistic model. It is concluded that BCG vaccination offers protection against tuberculous meningitis. Since improvement in weight for age was associated with a decreased risk of disease, further studies are needed to evaluate the association, if any, between nutritional status and vaccine efficacy. Prevalence of primary and acquired resistance of Mycobacterium tuberculosis to antituberculosis drugs in Benin after 12 years of short course chemotherapy
A. Trebucq et al; Int. J. Tub. & Lung Dis.; 1993,3,466

The objective was to measure the prevalence of primary and acquired resistance of M. tuberculosis to anti-tuberculosis drugs, i.e. Isomazid, Rifampicin, Ethambutol and Streptomycin in Benin, during 19941995, after 12 years use of short-course chemotherapy regimens. A prospective study, by cluster sampling according to the methodology recommended by the International Union Against Tuberculosis and Lung Disease (IUATLD) and the World Health Organisation (WHO) was conducted. The survey of primary resistance included 333 strains of which 28 (8.4%) were drug-resi.stanl. one

ABSTRACTS

69

to both Rifampicin and Isoniazid (multidrugresistant). For acquired resistance, out of 57 strains tested 26 (45.6%) were resistant, 6 of which (11%) were multidrug-resistant. Despite the considerable increase in the number of tuberculosis cases observed in recent years (52% between 1987 and 1995), direct observation of patients taking their anti-tuberculosis drugs during the intensive phase of treatment has limited the development of drug resistance in Benin. The 1997 nationwide tuberculosis prevalence survey in the Philippines T.E. Tupasietal; Int. J. Tub. & LungDis.; 1999, 3, 471 The objective was to determine the prevalence of tuberculosis as a basis for setting the targets for the National Tuberculosis Control Programme. A multistage cluster survey of a random sample of 21,960 subjects from 36 clusters nationwide was undertaken from April to July, 1997. BCG scar verification and tuberculin testing were performed for subjects aged 2 months and over, and chest radiography screening was done on subjects 10 years and older. Sputum samples were collected from individuals who were initially assessed to have abnormal chest radiographs to determine the prevalence of bacillary tuberculosis. Acid-fast smears by modified Kinyouns technique and culture on Lowenstein Jensen medium were done to demonstrate Mycobacterium tuberculosis. The prevalence of active pulmonary TB was 42/1000 population. The prevalence of culture-positive and smear-positive cases was 8.1 and 3.1/1000, respectively. The prevalence was similar in urban and rural areas. Morbidity from tuberculosis remains high. Allowing for methodological differences from the survey in 1981-1983, the prevalence of active pulmonary tuberculosis was regarded as unchanged. There was only a minimal decrease of 37% for smearpositive cases and 25% for culture-positive cases in the 14-year interval. Prevalence of sputum-positive pulmonary tuberculosis in tribal and non-tribal populations of the Ashti and Karanja tehsils in Wardha district of Maharashtra State P. Naranget al; Int. J. Tub. & LungDis.; 1999, 3, 478 To find and compare the prevalence of bacillary

positive pulmonary tuberculosis amongst the different tribes and in the non-tribal population, a prevalence study of pulmonary tuberculosis was conducted by house-to-house survey of symptoms among tribal (n=20,596) and non-tribal (n=93,670) populations aged 5 years and over, between September 1989 and November 1990. The prevalence of smear and/or culture-positive tuberculosis/100 000 population was 133 in the tribal and 144 in the non-tribal population. The difference in prevalence of symptomatic individuals and sputum-positive cases among the tribal and non-tribal populations was statistically significant only in the symptomatic individuals (P=0.01). The prevalence of cases in both groups was higher in males than females; however this difference was significant only in the tribal group (P = 0.05). Only 2 of the 46 tribes encountered, the Mana and Pawara tribes, showed a h ig h p rev alen ce of 730 and 6 1 2 / 1 0 0 000, respectively. The 3 other tribes with positive cases (the Gond group) had prevalence rates comparable to that of the non-tribal population. The prevalence of tuberculosis in tribal people was comparable to that of the non-tribal population. The importance of quality control of sputum smear microscopy: the effect of reading errors on treatment decisions and outcomes. T.N.L. Ngiiycn et al: Int. J. Tub. & Lung Dis.; 1999,3.483 The impact of slide reading errors at peripheral level on case-finding and treatment decisions was evaluated in Ho Chi Minh City, Vietnam. Over a 6month period in 1997, information on date, type of slide, results of other slides from the patient, and treatment status was collected for all slides, from District TB Centres, detected as having reading errors during smear microscopy quality control rereading. Reading errors were detected in 117 slides: 115 (98.3%) were incorrectly read as negative. In all 75 (65.2%) of these errors occurred in casefinding slides. In the 75 falsely negative case-finding slides, re-reading resulted in initiation of treatment in 38 patients (50.7%). The remaining 37 (49.3%) had only one positive slide and were told to return for follow-up after 6 months; the 2 (5.4%) who did return were both diagnosed with active tuberculosis. Detection of errors in the 40 false-negative followup slides resulted in treatment changes in 4 patients

70

ABSTRACTS

(10%). Quality control plays a crucial role in ensuring the timely diagnosis and treatment of new cases and appropriate management of patients currently on treatment. The usefulness of quality control could be enhanced by focusing greater efforts on case-finding slides initially read as negative. Cytochrome P-450 in Mycobacterium tuberculosis drug-resistant

corresponding sensitive strains. It, therefore, appears that cytochrome P-450 might play a role in causing drug resistance in tuberculosis. Early results from indirect drug susceptibility test for tubercle bacilli Sara Mathew et al; Ind. J. Med. Res.; 1999. 109, 167. Indirect susceptibility test results on LowensteinJensen medium for tubercle bacilli against Streptomycin, Isoniazid and Rifampicin were read at the end of 2 weeks and compared with the results at 4 weeks. Drug resistance could be correctly predicted in >70% of cultures including multi-drug resistant tuberculosis strains at the end of 2 weeks. The para-nitrobenzoic acid (PNB) susceptibility read at 2 weeks was able to distinguish nontuberculosis mycobacteria from Mycobacterium tuberculosis cultures. The early detection of resistance by this procedure requires only minimum inputs, and can benefit the majority of patients harbouring drug resistant tubercle bacilli.

Geetha Ramachandran et al; Current Science; 1999, 76, 1231 Methods for the isolation and spectrophotometric determination of cytochrome P-450 in mycobacteria were standardized. Cytochrome P-450 levels were estimated in Mycobacterium tuberculosis organisms sensitive to both Isoniazid and Rifampicin and resistant to either. Cytochrome P-450 was isolated and its presence was shown in M.smegmatis, M, fortuitwn, M.chelonae and M. tuberculosis H17Rv. The cytochrome P-450 content was significantly elevated m M. tuberculosis, resistant to both Isoniazid and Rifampicin when compared with the