examinations in pulmonary tuberculosis:
How many sputum specimens are adequate?
Adnan YILMAZ, Ümmühan BAYRAM SELVİ, Ebru DAMADOĞLU, Sinem GÜNGÖR, Mualla PARTAL, Esen AKKAYA, Turan KARAGÖZ
Süreyyapaşa Göğüs Hastalıkları ve Göğüs Cerrahisi Eğitim ve Araştırma Hastanesi, İstanbul.
ÖZET
Akciğer tüberkülozunda tekrarlanan balgam incelemelerinin tanısal değeri:
Kaç tane balgam örneği yeterlidir?
Bu çalışmanın amacı; akciğer tüberkülozunun tanısında çok sayıda balgam örneğinin incelenmesinin değerini araştırmak- tır. 2002 yılı süresince kültür pozitif akciğer tüberkülozu tanılı hastaların balgam yayma ve kültür sonuçlarını analiz ettik.
Bin yirmi yedi hastada tanı balgam örneklerinde tüberküloz basilinin saptanmasıyla elde edildi. Laboratuvara gönderilen balgam sayısı 634 olguda bir, 167 olguda iki, 186 olguda üç ve 48 olguda üçten fazla idi. Yedi yüz altmış (%74) olgu pozi- tif yayma sonucuna sahipti. Yayma pozitif olguların %82.3’ünde birinci balgamın yayma incelemesi pozitif idi. Bu grup ol- guların %94.9’unda birinci veya ikinci balgam örneğinin yayma incelemesi pozitif bulundu. Üçüncü balgamın yayma in- celemesi %4.2 ek tanı değerine sahipti. Sekiz yüz altmış üç (%84) olguda birinci balgamın kültürü pozitifti. İkinci ve üçün- cü balgamların kültür incelemeleri sırasıyla %11 ve %4.5 ek tanı değeri sağladı. Kültür pozitif olguların %95’ine ilk iki bal- gam örneğinin kültür incelemesiyle tanı konuldu. Sonuç olarak; akciğer tüberkülozlu olguların büyük çoğunluğuna iki balgam örneği incelemesiyle tanı konulabilir. Üç veya daha fazla balgam incelemesi yapılması küçük bir ek tanı değerine sahiptir.
Anahtar Kelimeler: Akciğer tüberkülozu, balgam, tanı değeri, yayma, kültür.
Yazışma Adresi (Address for Correspondence):
Dr. Ebru DAMADOĞLU, Merkez Mahallesi Yüksel Sokak Ülfet Apartmanı No: 10/18 Küçükyalı 34840 İSTANBUL - TURKEY
e-mail: edamadoglu@yahoo.co.uk
Tuberculosis is a major world health problem.
According to World Health Organization (WHO) estimations, approximately one third of the world’s population is infected with Mycobacteri- um tuberculosis, and 9 million new cases of tu- berculosis and more than 2 million deaths occur every year (1). The most common clinical ma- nifestation of tuberculosis is pulmonary disease.
It accounts for about 80% of disease in HIV-sero- negative people (2).
The diagnosis of pulmonary tuberculosis (PTB) is made by identifying and subsequently cultu- ring of M. tuberculosis from respiratory speci- mens, most commonly expectorated or induced sputum (3,4). In diagnosing PTB, it is standard practice to collect three sputum specimens on consecutive days for acid-fast bacillus (AFB) staining and culture (3,5,6). Although most text- books and guidelines continue to recommend collection of three sputum specimens for diag- nosis of PTB, recent reports suggest that one or two specimens may be adequate for this purpo- se (7-10). The aim of this study was to assess
the diagnostic yield of repeated sputum exami- nations in diagnosis of PTB.
MATERIALS and METHODS
The present study was conducted at Süreyyapa- şa Center for Chest Diseases and Thoracic Sur- gery, which is located in İstanbul, Turkey. It is a tertiary care teaching hospital for chest diseases and tuberculosis. We retrospectively reviewed the microbiological laboratory records of pati- ents with a definite diagnosis of PTB during the year 2002. We detected 1095 cases with cultu- re-proven PTB. Sixty eight cases were excluded from the study because culture positive speci- mens were bronchial lavage or pleural fluid. The study included 1027 cases whose diagnosis we- re established by the detection of M. tuberculo- sis in sputum samples. None of these patients received prior therapy for tuberculosis. All spu- tum samples were obtained for diagnosis not for follow-up. We analyzed sputum smear and cul- ture results of these 1027 patients. The sputum smears were examined using Ziehl-Neelsen sta- ining method for detection of AFB. Lowenstein- SUMMARY
Diagnostic value of repeated sputum examinations in pulmonary tuberculosis:
How many sputum specimens are adequate?
Adnan YILMAZ, Ümmühan BAYRAM SELVİ, Ebru DAMADOĞLU, Sinem GÜNGÖR, Mualla PARTAL, Esen AKKAYA, Turan KARAGÖZ
Süreyyapaşa Chest Diseases and Thoracic Surgery Training and Investigation Hospital, İstanbul, Turkey.
The aim of this study was to assess the value of examining multiple sputum specimens in the diagnosis of pulmonary tu- berculosis (PTB). We analyzed sputum smear and culture results of patients diagnosed with culture-proven PTB during 2002. In 1027 patients, the diagnosis was established by detection of Mycobacterium tuberculosis bacilli in sputum samp- les. The number of sputum specimens submitted to laboratory was one in 634 cases, two in 167 cases, three in 186 cases and more than three in 48 cases. 760 (74%) cases had positive smear examination result. The first sputum smear examina- tion was positive in 82.3% of smear positive cases. Either the first or the second sputum was diagnostic in 94.9% of these cases. Smear examination of third sputum revealed 4.2% additional diagnostic yield. In 863 (84%) cases, culture examina- tion of the first sputum was positive. The second and the third sputum culture examinations revealed additional diagnos- tic yield of 11% and 4.5%, respectively. Percent 95 of culture-proven cases were diagnosed with the first two sputum cultu- res. In conclusion the majority of PTB cases can be diagnosed with the examination of two sputum specimens. Three or mo- re sputum specimens submitted obtain a small additional diagnostic yield.
Key Words: Pulmonary tuberculosis, sputum, diagnostic yield, smear, culture.
Jensen medium was used for culture examinati- on in all samples.
RESULTS
During one year period, 1027 patients with cul- ture-proven PTB were identified. Of these 1027 patients, 61.8% had only a single specimen sent for AFB smear and culture, 16.2% had two spu- tum samples sent for examination of M. tubercu- losis bacilli and 18.2% had three sputum speci- mens collected. The total number of sputum specimens submitted to the laboratory was 1703. Table 1 shows the distribution of patients with respect to sputum specimen number sub- mitted to laboratory.
Percent 74 of the patients (760 of 1027) had at least one positive smear examination for AFB.
Smear examinations were negative in 26% of patients (267 of 1027). Of 760 smear examina- tion-positive, 625 (82.3%) were identified on the first smear and 96 (12.6%) were identified on the second smear. Percent 94.9 of smear-positive patients (721 of 760) were diagnosed with the first two smears. Third smear examination pro- vided 4.2% additional diagnostic yield. Results were summarized in Table 2.
Culture examination of the first sputum sample was positive for M. tuberculosis in 84% of pati- ents (863 of 1027). Of 1027 culture-proven PTB, the diagnosis of 113 (11%) patients was established with the second sputum sample exa- mination. Culture examination of the first two sputum samples identified 94.9% of culture-pro-
ven patients (976 of 1027). Culture examination of the third samples provided 4.5% additional di- agnostic yield (Table 3).
Table 4 summarizes diagnostic yield and costs of 4 different approaches with respect to the num- ber of sputum samples submitted to the labora- tory for culture examinations. If the strategy of examining two samples was used, the number of smears examined and cost would be 16.6% less.
DISCUSSION
Diagnosis of active PTB is a challenge for clini- cians. In patients with signs and symptoms con- sistent with disease, examination of respiratory specimens is the most reliable approach to con- firm the diagnosis (8). Although smear micros- copy and clinical diagnosis play important roles in the diagnostic process, the gold standard for
Table 2. The diagnostic yield of sputum samples in diagnosis of patients with smear-positive PTB.
Smear-positive cases
Sputum n %
First 625 82.3
Second 96 12.6
Third 32 4.2
Fourth 5 0.7
Fifth 1 0.1
Sixth 1 0.1
Total 760 100
Table 3. The diagnostic yield of sputum samples in diagnosis of patients with culture-proven PTB.
Culture-positive cases
Sputum n %
First 863 84
Second 113 11
Third 46 4.5
Fourth 3 0.3
Fifth 1 0.1
Sixth 1 0.1
Total 1027 100
Table 1. The distribution of patients with respect to sputum specimen number submitted.
Number of Patients Total number sample submitted n (%) of samples
One 634 (61.8) 634
Two 167 (16.2) 334
Three 186 (18.2) 558
Four 28 (2.7) 112
Five 7 (0.7) 35
Six 5 (0.4) 30
Total 1027 (100) 1703
the diagnosis of PTB is still a positive culture for M. tuberculosis (11). It was reported that sensi- tivity, specificity and positive predictive value were 55.3%, 99.9% and 98% for smear examina- tion, respectively (12). Levy et al. found that sensitivity and specificity were 53.1 and 99.8 percent for smear examination, 81.5 and 98.4 percent for culture, respectively (13).
The current WHO and International Union Aga- inst Tuberculosis and Lung Disease (IUATLD) recommendations for diagnosis of PTB are ba- sed on examination of three sputum specimens collected from each suspect (14,15). Similarly, the technical guidelines of the Centers for Dise- ase Control and Prevention (CDC) and Ameri- can Thoracic Society (ATS) recommend obta- ining three sputum samples for diagnosis of PTB (5). However, many studies suggest that one or two sputum specimens may be as sensitive three sputum specimens (3,8-10,16-20). Nel- son et al. reported that for 95% of the patients, either the first or the second specimen collected proved to be diagnostic (3). They found that for only 5% of the patients were a third specimen re- quired to make a definitive diagnosis of PTB. A previous report confirmed that the first two samples provided diagnosis of PTB in the majo- rity of patients (92%) (17). Walker et al. repor- ted that 77.1% of smear-positive patients were found on the first smear, a further 15% on the se- cond smear and 7.9% additional cases were identified on the third smear (18). The additional value by the third specimen was 0.8% in Gopi’s report and was 3.2% in Wu’s report (19,20). Go- pi and Wu concluded that there was no need for a fourth or fifth sputum specimen and that for maximum efficiency, two sputum smears should be examined (19,20). Our findings substantially agree with those shown by previous reports (3,17-20). In our study, the first two sputum spe- cimens identified 94.9% of smear-positive pati- ents and 95% of all patients. The third specimen revealed a small additional diagnostic value.
It is a well-known fact that the smear positivity value increases as the number of sputum speci- mens collected per PTB suspect increases. The large number of smears prepared however, incre- ases the workload of technician and compromi- Table 4. The diagnostic yield and costs of different approachs. Number (%) ofCosts (USD) Specimens SpecimensCasesIncremental*Per each ApproachesPatientsexaminedreduceddetectedcasesspecimenTotalIncremental* Only one sample 10271027676 (39.6)8638631313351-- Two samples10271420283 (16.6)97611313184605109 Three samples1027164657 (3.4)10224613213982938 Present 10271703--102751322139741 * An incremental value considers changes from one strategy to another.
se the quality of microscopy because of less time for processing and reading (16,20). It is known that the number of smear-positive specimens processed within a laboratory is associated with the risk of specimen cross-contamination (7).
Reducing the number of sputum specimens co- uld reduce the workload of laboratory staff, allo- wing more time for examining each slide (20).
According to Harries et al., when a policy of exa- mining two samples was used, the number of smears examined would be 29.9% less (21).
Crampin et al. reported that using two smears rather than three, the sensitivity for detecting cul- ture-proven patients was only very marginally reduced, and the specificity was marginally imp- roved (9). A strategy of examining two samples reduces the service’s and patient’s costs (16).
According to Walker et al., this policy is a cost- effective strategy (18). If the strategy of exami- ning two samples was used in our center, the number of smears examined and cost would be 16.6% less.
In conclusion, the majority of pulmonary tuber- culosis cases can be diagnosed with examinati- on of two sputum specimens in our center. Three or more sputum specimens submitted to the la- boratory obtain a small additional diagnostic yi- eld. This strategy will reduce the workload of la- boratory staff and costs.
REFERENCES
1. Raviglione MC, Snider DEJ, Kochi A. Global epidemi- ology of tuberculosis: Morbidity and mortality of a worldwide epidemic. JAMA 1995; 273: 220-6.
2. Shafer RW, Edlin BR. Tuberculosis in patients infected with human immunodeficiency virus: Perspective on the past decade. Clin Infect Dis 1996; 22: 683-704.
3. Nelson SM, Deike MA, Cartwright CP. Value of examining multiple sputum specimens in the diagnosis of pulmo- nary tuberculosis. J Clin Microbiol 1998; 36: 467-9.
4. Kim TC, Blackman RS, Heatwole KM, et al. Acid-fast ba- cilli in sputum smears of patients with tuberculosis. Am Rev Respir Dis 1984; 129: 264-8.
5. ATS/CDC. Diagnostic standards and classification of tu- berculosis in adults and children. Am J Respir Crit Care Med 2000; 16: 1376-95.
6. Shinnick TM, Good RC. Diagnostic micobacteriology: La- boratory practices. Clin Infect Dis 1995; 21: 291-9.
7. Stone BL, Burman WJ, Hildred MV, et al. The diagnostic yield of acid-fast-bacillus smear-positive sputum speci- mens. J Clin Microbiol 1997; 35: 1030-1.
8. Finch D, Beaty CD. The utility of single sputum speci- men in the diagnosis of tuberculosis. Comparison betwe- en HIV-infected and non-HIV-infected patients. Chest 1997; 111: 1174-9.
9. Crampin AC, Floyd S, Mwaungulu F, et al. Comparison of two versus three smears in identifying culture-positive tuberculosis patients in a rural African setting with high HIV prevalence. Int J Tuberc Lung Dis 2001; 5: 994-9.
10. Leonard MK, Osterholt D, Kourbatova EV, et al. How many sputum specimens are necessary to diagnose pulmonary tuberculosis? Am J Infect Control 2005; 33: 58-61.
11. Dam T, Bose M. Culture examination-an essential step to increase diagnostic yield for pulmonary tuberculosis in developing countries. Int J Tuberc Lung Dis 2000; 4: 282.
12. Gordin F, Slutkin G. The validity of acid-fast smears in the diagnosis of pulmonary tuberculosis. Arch Pathol Lab Med 1990; 114: 1025-7.
13. Levy H, Feldman C, Sacho H, et al. A reevaluation of sputum microscopy and culture in the diagnosis of pul- monary tuberculosis. Chest 1989; 95: 1193-7.
14. World Health Organization. Guidelines for national prog- rammes. WHO/TB/97.220. Geneva: WHO, 1997.
15. Enarson DA, Rieder HL, Arnadottir T, Trebucq A. Mana- gement of tuberculosis: A guide for low income countri- es. 5thed. Paris: IUATLD, 2000.
16. Yassin MA, Cuevas LE. How many sputum smears are necessary for case finding in pulmonary tuberculosis?
Trop Med Intern Health 2003; 8: 927-32.
17. Cascina A, Fietta A, Casali L. Is a large number of spu- tum specimens necessary for the bacteriological diagno- sis of tuberculosis? J Clin Microbiol 2000; 38: 466.
18. Walker D, McNerney R, Mwembo MK, et al. An incre- mental cost-effectiveness analysis of the first, second and third sputum examination in the diagnosis of pulmonary tuberculosis. Int J Tuberc Lung Dis 2000; 4: 246-51.
19. Wu ZL, Wang AQ. Diagnostic yield of repeated smear microscopy examinations among patients suspected of pulmonary TB in Shandong province of China. Int J Tu- berc Lung Dis 2000; 4: 1086-7.
20. Gopi PG, Subramani R, Selvakumar N, et al. Smear exa- mination of two specimens for diagnosis of pulmonar tu- berculosis in Tiruvallur District, South India. Int J Tuberc Lung Dis 2004; 8: 824-8.
21. Harries AD, Mphasa NB, Mundy C, et al. Screening tuberculosis suspects using two sputum smears. Int J Tuberc Lung Dis 2000; 4: 36-40.
