Mantoux and contact positivity in tuberculosis
http://www.100md.com
《美国医学杂志》
Department of Pulmonology, Institute of Child Health and Hospital for Children, Egmore, Chennai, India
Abstract
Objective: To study the role of Mantoux and contact history in various forms of Childhood tuberculosis. Methods. 605 children registered with TB clinic of Institute of Child Health and Hospital for Children, Chennai over a 5 year period from January 2000 to October 2005 with various forms of tuberculosis were recruited in the study. Clinical examination findings, basic investigations, chest skiagrams, computerized tomography (CT) wherever warranted, sputum or gastric aspirates for AFB smear, histopathology wherever possible were analyzed. Results. The study showed that Mantoux positivity in various forms of tuberculosis studied is 34.7%. The positivity of Mantoux was highest in lymph node tuberculosis (53%) and the lowest with CNS tuberculosis (21.2%). Among the other forms, Mantoux positivity was 36.4% in TB abdomen, 44.4% in Skeletal TB, 30.3% in pulmonary tuberculosis. The contact positivity was 30.4% in the sample studied. Conclusion. The study also reflects that the extra pulmonary forms of tuberculosis seems to be more common in the pediatric population which constituted 79.8% of the cases included in the study.
Keywords: Mantoux; Contact history; Childhood tuberculosis
Despite the availability of effective preventive measures and chemotherapy, the prevalence of tuberculosis (TB) is increasing in the developing world and in much of the industrialized world as well.[1] In developing countries the annual risk of tuberculosis infection in children is 2- 5 per cent. Nearly 8-20 per cent of the deaths caused by tuberculosis occur in children.[2] Early diagnosis plays a vital role in control of TB. Although acid fast bacilli (AFB) microscopy, and conventional Lowenstein Jensen (L-J) culture remain the cornerstone for the diagnosis of TB, these traditional methods are either slow or their sensitivity is quite low.[3]
Among all the available investigations, Mantoux test plays an important role in the diagnosis of tuberculosis as it denotes the particular subject is sensitized to tuberculous antigen. Tuberculous infection is first signaled by a reactive Mantoux test. Its operating characteristics are superior to many other tests commonly used in the diagnosis of tuberculosis, despite the fact that this test is performed with a multiple antigenic mixture and its result is dependent on the variability in immunologic reactivity.[4] The estimated lifetime risk of developing tuberculosis disease for a young child infected with Mycobacterium tuberculosis as indicated by positive tuberculin test is about 10 per cent.[5]
Approximately 10% of otherwise normal children with culture-proven TB do not react to tuberculin initially.[6] Studying Mantoux positivity in various forms of pulmonary and extra pulmonary tuberculosis may give useful information on the present status of Mantoux test. The present study objectives were to estimate the role of Mantoux test (tuberculin skin test) and to determine the magnitude of contact positivity in children suspected and/or confirmed as tuberculosis.
Materials and methods
Six hundred and five children less than 12 years of age diagnosed with various forms of tuberculosis like pulmonary tuberculosis, lymph node tuberculosis, abdominal tuberculosis, skeletal tuberculosis, tuberculous meningitis and tuberculoma, registered at the TB Clinic, Department of Pulmonology, I.C.H & H.C during the period of January 2000 to October 2005 who satisfied our inclusion criteria were retrospectively studied. Emphasis was given for age, Mantoux status and contact history details. The subjects were labeled as confirmed cases when they had either bacteriological or histopathological evidence and highly probable tuberculosis cases that satisfied all the epidemiological criteria but for bacteriological or histopathological support.
Inclusion criteria for various forms of tuberculosis
Tuberculous lymphadenitis: Cervical and axillary lymph nodes (>1 cm), inguinal lymph nodes (> 1.5 cm) with / without matting, not responding to antibiotic therapy for 2 weeks, with histopathological evidence by FNAC or by excision biopsy.
Abdominal tuberculosis: Children presenting with ascites, sub acute obstruction or pyrexia of unknown origin with or without hepato-splenomegaly supported by barium meal follow through showing pulled up caecum with or without peritoneal or lymph node biopsy.
Pulmonary tuberculosis: Pulmonary tuberculosis may be classified as primary, progressive primary and post primary tuberculosis.
Children presenting with persistent fever and/or cough >3 weeks, unexplained recent loss of weight or appetite with persistent radiological lesions (unilateral exudative pleural effusion; pneumonia with an enlarged mediastinal lymph node and miliary lesion highly suggestive of tuberculosis) beyond 4 weeks inspite of antibiotic therapy supported by AFB positivity in gastric lavage, bronchoalveolar lavage or sputum were diagnosed as progressive primary complex.
Children presenting with the cavitory lesions, supported by AFB in gastric lavage, bronchoalveolar lavage or sputum were diagnosed as post primary tuberculosis (adult type tuberculosis).
Tuberculous meningitis: Tuberculous meningitis was diagnosed in children with the following clinical criteria and investigations as suggested by fever for more than 2 weeks and/or altered behavior / change in personality of recent origin, headache, vomiting, movement disorder, focal deficit, seizures supported by CSF studies and/or CT scan.
Tuberculoma: Children presenting with focal neurological deficit (seizures) with CT evidence of granulomas larger in size (>20mm), with significant edema persisting more than 3 months as evidenced by repeat CT. Granulomas at the gray-white junction, smaller in size were excluded. Since histopathology is not routinely implicated in the diagnosis of tuberculomas, we have included only highly probable cases.
Skeletal tuberculosis: Spinal tuberculosis and joint tuberculosis diagnosed by skiagram supported by CT and MRI with/without histopathology evidence.
Exclusion criteria : Tuberculosis in children with immunosuppressive states including HIV.
Mantoux test
1 TU PPD with RT 23 Tween 80 was injected intra-dermally and the reaction read between 48-72 hours. Induration of 10 mm or more in largest diameter was suggestive of natural infection, irrespective of prior BCG vaccination. In case a patient presented late but within 7 days of the test, then any reaction above 10 mm is was considered as a positive test.
Contact: Contact was defined as any child who lived in a household with an adult taking anti-TB therapy or had taken such a therapy in the past 2 years.[7]
Statistical analysis
Proportions of Mantoux and contact positives in childhood tuberculosis as a whole and in various forms were arrived. To compare these among confirmed against probable cases, chi-square test wsa used. P< 0.05 was considered statistical significance. Sensitivity, specificity, positive and negative predictive value for Mantoux test and contact history were arrived. For this, histopathology or bacteriological proof was considered as gold standard.
Results
Of 605 children diagnosed for tuberculosis, 309 (51.1%) children were less than 6 years and 296 (48.9%) were of 7-12 year age group. Of the total study group, 198 (32.7%) were confirmed by bacteriology or by histopathology. The remaining 407 (67.3%) cases were included as highly probable cases of various forms of tuberculosis. The number of confirmed and highly probable cases in various forms of tuberculosis is shown in table1 Mantoux was positive in a total of 210 (34.7%) cases which included 101 (51%) confirmed cases and 109 (26.8%) probable cases (p < 0.05). Mantoux positivity in less than 6 year age group was in 103 (33.3%) in comparison to 107 (36.15%) in 7-12 year age group (p = 0.47).
Of 122 (20.2%) children with pulmonary tuberculosis, only 46 (37.7%) had bacteriological evidence (38 Progressive primary and 8 Post primary), the rest were diagnosed on epidemiological factors. Mantoux was positive in 44 (36.1%) children with pulmonary tuberculosis with progressive primary complex having 37 (36.6%) and post primary tuberculosis having 7 (33.3%) cases of Mantoux positive children. In progressive primary complex, the Mantoux positivity was 20 (52.6%) in confirmed cases in contrast to 17 (27%) in probable cases (p = 0.01). Similarly in post primary tuberculosis Mantoux was positive in 4 (50%) of confirmed cases and in 3 (23.1%) of probable cases (p = 0.20). Totally among pulmonary tuberculosis, Mantoux was positive in 24 (52.2%) confirmed cases in comparison to 20 (26.3%) probable cases (p = 0.004).
Extra pulmonary tuberculosis constituted 483 (79.8%) cases. This included lymph node tuberculosis 117 (19.3%), skeletal tuberculosis 99 (16.4%), abdominal tuberculosis 22 (3.6%). CNS tuberculosis which included tuberculous meningitis 160 (26.4%) and tuberculoma 85 (14%) constituted 245 (40.5%) cases.
Mantoux positivity was observed in 62 (53%) of lymph node TB of which 55 (56.1%) were confirmed cases and 7 (36.8%) probable cases (p = 0.123). Eight (36.4%) of TB abdomen cases were Mantoux positive which included 7 (43.8%) confirmed cases and 1 (16.7%) probable case (p= 0.24). In skeletal tuberculosis, 44 (44.4%) cases were Mantoux positive of which 9 (52.9%) were confirmed and 35 (42.7%) probable cases (p = 0.439). Six (28.6%) confirmed cases and 28 (20.1%) probable cases of tuberculous meningitis were Mantoux positive (p=0.379). The overall Mantoux positive in tuberculous meningitis cases was 34 (21.3%). Mantoux was positive in 18 (21.2%) cases of tuberculoma. Contact tracing revealed 184 (30.4%) of children had contact with adult tuberculosis that included 61 (30.8%) confirmed and 123 (30.2%) probable cases (p = 0.883).
Outcome
All the children who were included in the study received anti tuberculous drugs from the hospital. Of the 407 children with probable TB 28[6.8%] were lost for follow up. The remaining 379 completed their treatment and had recovered from their illness. Of the 198 confirmed cases 14 [7.1%] children were lost for follow up. And the rest 184 had recovered.
Discussion
Most recently developed sensitive and specific diagnostic tests for tuberculosis like Bactec, PCR, have not found a place in the routine evaluation of children with suspected tuberculosis as they are not widely available and too costly to be included in the routine screening. In spite of extensive research it is yet to evolve a simple diagnostic test in place of Mantoux. Even after hundred years, Mantoux stands to be a simple, easily available test, making it very useful for the diagnosis of TB though it has its own limitations.
Even in developed countries, the diagnosis of tuberculosis in children is largely based on indirect epidemiological evidences where Mantoux test still plays a significant role. Clinical criteria tuberculin skin test results, radiographic changes, and documented exposure to an infectious adult, remains the standard diagnostic methods.[8] No single ancillary test is 100 percent accurate in making or excluding the diagnosis of tuberculosis. Although a Mantoux test should always be performed, it may be negative in 10 to 25 percent of patients with active disease.[9] The relatively low sensitivity and specificity of this test makes the test very useful for persons at high risk for TB infection or disease but undesirable for use in persons at low risk.[10]
Mantoux positivity in various forms of childhood tuberculosis studied was found to be 34.7%. Mantoux positivity in confirmed cases was 51% in contrast to 26.8% in probable cases which is statistically significant.
In pulmonary tuberculosis, the overall Mantoux positivity was 36.1%. Among pulmonary tuberculosis, Mantoux test was significantly positive in higher proportion in confirmed cases (52.2%) in contrast to 26.3% in probable cases which is statistically significant (p = 0.004).
Of 117 children diagnosed as lymph node tuberculosis, Mantoux positivity was observed in 62(53%) of lymph node tuberculosis with 56.1% in histopathology proven cases. The present study results are comparable with the study conducted by Ahmed et al which showed 65% positivity in lymph node tuberculosis.[11]
Skeletal TB included spinal TB (76) and various other bone and joint TB (23).44.4% of Skeletal TB was Mantoux positive. Though bacteriological support confirms TB, it is not possible in bone TB where imaging techniques play an important role. The high prevalence of tuberculosis in India precludes the need of histopathological diagnosis especially in skeletal TB as other alternative lesions including malignancy are rare in pediatric population. Radiographs are the first line of investigations to substantiate or refute a clinical diagnosis of tuberculosis of the spine. In spinal tuberculosis, the commonest site of involvement is the dorsolumbar vertebra in 62 cases (81.6%). The diagnosis of spinal tuberculosis was substantiated by CT evidence. In 5 children, the initial spine X-rays were normal inspite of highly suggestive clinical presentation and in these, CT confirmed the early tuberculous changes. Though the wide availability of CT scan and MR imaging has increased the use of these modalities in supporting the diagnosis of tuberculosis of the spine, the significance of Mantoux positivity is important in the absence of confirmatory evidence like histopathology.
Only 8 of the 22 cases of TB abdomen were Mantoux positive which constituted 36.4%. The contact positivity in confirmed cases of abdominal tuberculosis was found to be 22.7%. This is very similar to the study done earlier which shows 25.6%.[12]
21.3% of tuberculous meningitis was Mantoux positive which are comparable with studies done by Ahmet Yarami et al and Reider et al.[13], [14] Cases of tuberculomas included in the study were based on strong criteria which exclude the possibility of neurocysticercosis as suggested by Garg et al .[15] The inclusion criteria adopted was based on the consensus adopted by Amdekar et al.[7]
From the present study, it is evident that Mantoux positivity in childhood tuberculosis varies from 21.2 % to 53%. Also positivity of Mantoux is higher in lymph node tuberculosis 62 (53%) than other forms of tuberculosis. The significance of Mantoux is least in CNS tuberculosis 52 (21.2%) which includes tuberculous meningitis 34 (21.3%) and tuberculoma 18 (21.2%). The results obtained from the study are similar to the study done by Mahadevan et al.[4]
Contact investigation remains an important mechanism to identify potentially infected and infective patients. Although the most efficient way of identifying infected children is through contact investigations of adults with pulmonary tuberculosis, the reverse i.e., associate investigation, which includes tracking contacts of infected children, is also important.[16] So the authors studied the contact details of the children included in the study. It was found that only 184 (30.4%) children had contact history positive. This included 61(30.8%) children with confirmed tuberculosis and 123 (30.2%) children with highly probable tuberculosis (p =.883). The present study results are comparable to the studies done by Nair et al[17] and Somu et al[18] which shows the contact positivity of 35.5% in 1000 children and 61.2% in 227 children respectively.
It was found that the sensitivity of Mantoux was 51% and the specificity was 71.7%. The positive predictive value of the test was 52.6% and negative predictive value was 70.4%. The percentage of false negatives was 49% and percentage of false positives was 28.3%. Thus with such low sensitivity and specificity, the reliability of diagnosing tuberculosis with Mantoux alone is limited. With a higher false negativity rate, it cannot be reliable as a very good screening test for tuberculosis. The sensitivity of contact positives was found to be 30.8% and the specificity was 69.8%. The positive predictive value was 33.2% and the negative predictive value was 67.5%. With contact history many a times underreported, the reliability of contact history alone in diagnosing a case as tuberculosis is very low. Considering the above facts and figures, there arises definitely a need of an alternative test for Mantoux both for screening and for diagnosing tuberculosis.
Conclusion
False positive and false negative test make significant variation in Mantoux reactivity during the diagnostic work up in children with suspected tuberculosis. In the present study, the Mantoux positivity was in 210 objects (34.7%) signifying the present status. However, the positivity was 51% in confirmed cases thus showing a better case detection rate when compared to suspected cases. In lymph node tuberculosis, the positivity is high (53%) as against 21.2% in CNS tuberculosis. Contact tracing revealed 30.2% contact positivity in the probable cases and 30.8% in the confirmed cases. Though contact positivity is lower when compared to Mantoux positivity, the significance of detecting the adult contacts in children with tuberculosis cannot be underestimated. As confirming TB in each case with bacteriological support is difficult, contact tracing being positive in only about 30% of cases and Mantoux positivity in 35% of cases only many a times the diagnosis of tuberculosis is made by the combination of the above factors in the absence of reliable diagnostic tests.
KEY MESSAGES
Mantoux positivity varies from 51% [confirmed cases] to 26.8% [probable cases] of childhood tuberculosis.
Mantoux positivity higher in lymph node tuberculosis [53%]. and lowest in CNS tuberculosis [ 21.2%].
Contact positivity in childhood tuberculosis is 30.4% with no major variation in the different forms of tuberculosis.
Contribution of Authors
Dr. D. Vijayasekaran has conceptualized the study, collected the data and contributed to writing the manuscript. Dr. R. Arvind Kumar was involved in data analysis, outcome assessment, literature review and writing the manuscript.
Dr. N.C. Gowrishankar, Dr. K. Nedunchelian and Dr. S. Sethuraman were involved in critical analysis of the manuscript.
Funding : None
Competing Interests : None
References
1. Starke JR, Correa AG. Management of mycobacterial infection and disease in children. Pediatr Infect Dis J 1995; 14 : 455-470.
2. Kabra S K, Lodha, Seth V. Some current concepts on childhood tuberculosis. Indian J Med Res 2004; 120(4): 387-397.
3. Jonas V, Alden MJ, Curry JI, et al. Detection and identification of Mycobacterium tuberculosis directly from sputum sediments by amplification of rRNA. J Clin Microbiol 1993; 31 : 2410-2416.
4. Mahadevan B, Mahadevan S, Serane VT, Narasimhan R. Tuberculin reactivity in tuberculous meningitis. Indian J Pediatr 2005; 72 : 213-215.
5. Comstock GW, Livesay VT, Woolpert SF. The prognosis of a positive tuberculin reaction in childhood and adolescence. Am J Epidemiol 1974; 99: 131-138.
6. Starke JR, Taylor-Watts KT. Tuberculosis in the pediatric population of Houston, Texas. Pediatrics 1989; 84 : 28-35.
7. Amdekar YK. Consensus Statement of IAP Working Group: Status Report on Diagnosis of Childhood Tuberculosis. Indian Pediatr 2004; 41 : 146-155.
8. Khan EA, Jeffrey R. Starke JR. Diagnosis of Tuberculosis in Children: Increased Need for Better Methods, EID 1995; 1 (4) : 115-123.
9. Huebner RE, Schein MF, Bass JB. The tuberculin skin test. Clin Infect Dis 1993; 17 : 968-975.
10. American Academy of Pediatrics: Committee on Infectious Diseases. Screening for tuberculosis in infants and children. Pediatrics 1994; 93 : 131-134.
11. Ahmad Z, Amin SS. Role of tuberculin test, FNAC and Elisa in the diagnosis of Tuberculous Lymphadenitis. JIACM 2003; 4(4) : 292-295.
12. Oya Uygur-Bayramiηli, Gül Dabak, Resat Dabak. A clinical dilemma: abdominal tuberculosis. World J Gastroenterol 2003; 9(5) : 1098-1101.
13. Ahmet Yarami, Fuat Gurkan, Murat Elevli et al. Central Nervous System Tuberculosis in Children: A Review of 214 Cases, Pediatrics 1998; 102 (5): 49.
14. Rieder BL, Cauthen GM, Comstock GW, Snider DE. Epidemiology of tuberculosis in the United States. Epidemiol Rev 1989; 11: 79-98.
15. Garg RK. Diagnostic criteria for neurocysticercosis: Some modifications are needed for Indian patients. Neurol India 2004; 52 : 171-177.
16. Starke JR. Tuberculin skin testing, what, who and when Rep Pediat Infect Dis 1994; 4 : 31-32.
17. Nair NS, Kurian, John TP, Omana Mathew. A review of 1000 cases of childhood tuberculosis at trivandrum. Ind J Tub 1964; 11(2): 82-97.
18. Somu N, Vijayasekaran D, Kannaki M, Balachandran A, Subramanyam L. Adult contacts in children with Tuberculosis. Indian Pediatr 1997; 34 : 819-822.(Vijayasekaran D, Kumar R )
Abstract
Objective: To study the role of Mantoux and contact history in various forms of Childhood tuberculosis. Methods. 605 children registered with TB clinic of Institute of Child Health and Hospital for Children, Chennai over a 5 year period from January 2000 to October 2005 with various forms of tuberculosis were recruited in the study. Clinical examination findings, basic investigations, chest skiagrams, computerized tomography (CT) wherever warranted, sputum or gastric aspirates for AFB smear, histopathology wherever possible were analyzed. Results. The study showed that Mantoux positivity in various forms of tuberculosis studied is 34.7%. The positivity of Mantoux was highest in lymph node tuberculosis (53%) and the lowest with CNS tuberculosis (21.2%). Among the other forms, Mantoux positivity was 36.4% in TB abdomen, 44.4% in Skeletal TB, 30.3% in pulmonary tuberculosis. The contact positivity was 30.4% in the sample studied. Conclusion. The study also reflects that the extra pulmonary forms of tuberculosis seems to be more common in the pediatric population which constituted 79.8% of the cases included in the study.
Keywords: Mantoux; Contact history; Childhood tuberculosis
Despite the availability of effective preventive measures and chemotherapy, the prevalence of tuberculosis (TB) is increasing in the developing world and in much of the industrialized world as well.[1] In developing countries the annual risk of tuberculosis infection in children is 2- 5 per cent. Nearly 8-20 per cent of the deaths caused by tuberculosis occur in children.[2] Early diagnosis plays a vital role in control of TB. Although acid fast bacilli (AFB) microscopy, and conventional Lowenstein Jensen (L-J) culture remain the cornerstone for the diagnosis of TB, these traditional methods are either slow or their sensitivity is quite low.[3]
Among all the available investigations, Mantoux test plays an important role in the diagnosis of tuberculosis as it denotes the particular subject is sensitized to tuberculous antigen. Tuberculous infection is first signaled by a reactive Mantoux test. Its operating characteristics are superior to many other tests commonly used in the diagnosis of tuberculosis, despite the fact that this test is performed with a multiple antigenic mixture and its result is dependent on the variability in immunologic reactivity.[4] The estimated lifetime risk of developing tuberculosis disease for a young child infected with Mycobacterium tuberculosis as indicated by positive tuberculin test is about 10 per cent.[5]
Approximately 10% of otherwise normal children with culture-proven TB do not react to tuberculin initially.[6] Studying Mantoux positivity in various forms of pulmonary and extra pulmonary tuberculosis may give useful information on the present status of Mantoux test. The present study objectives were to estimate the role of Mantoux test (tuberculin skin test) and to determine the magnitude of contact positivity in children suspected and/or confirmed as tuberculosis.
Materials and methods
Six hundred and five children less than 12 years of age diagnosed with various forms of tuberculosis like pulmonary tuberculosis, lymph node tuberculosis, abdominal tuberculosis, skeletal tuberculosis, tuberculous meningitis and tuberculoma, registered at the TB Clinic, Department of Pulmonology, I.C.H & H.C during the period of January 2000 to October 2005 who satisfied our inclusion criteria were retrospectively studied. Emphasis was given for age, Mantoux status and contact history details. The subjects were labeled as confirmed cases when they had either bacteriological or histopathological evidence and highly probable tuberculosis cases that satisfied all the epidemiological criteria but for bacteriological or histopathological support.
Inclusion criteria for various forms of tuberculosis
Tuberculous lymphadenitis: Cervical and axillary lymph nodes (>1 cm), inguinal lymph nodes (> 1.5 cm) with / without matting, not responding to antibiotic therapy for 2 weeks, with histopathological evidence by FNAC or by excision biopsy.
Abdominal tuberculosis: Children presenting with ascites, sub acute obstruction or pyrexia of unknown origin with or without hepato-splenomegaly supported by barium meal follow through showing pulled up caecum with or without peritoneal or lymph node biopsy.
Pulmonary tuberculosis: Pulmonary tuberculosis may be classified as primary, progressive primary and post primary tuberculosis.
Children presenting with persistent fever and/or cough >3 weeks, unexplained recent loss of weight or appetite with persistent radiological lesions (unilateral exudative pleural effusion; pneumonia with an enlarged mediastinal lymph node and miliary lesion highly suggestive of tuberculosis) beyond 4 weeks inspite of antibiotic therapy supported by AFB positivity in gastric lavage, bronchoalveolar lavage or sputum were diagnosed as progressive primary complex.
Children presenting with the cavitory lesions, supported by AFB in gastric lavage, bronchoalveolar lavage or sputum were diagnosed as post primary tuberculosis (adult type tuberculosis).
Tuberculous meningitis: Tuberculous meningitis was diagnosed in children with the following clinical criteria and investigations as suggested by fever for more than 2 weeks and/or altered behavior / change in personality of recent origin, headache, vomiting, movement disorder, focal deficit, seizures supported by CSF studies and/or CT scan.
Tuberculoma: Children presenting with focal neurological deficit (seizures) with CT evidence of granulomas larger in size (>20mm), with significant edema persisting more than 3 months as evidenced by repeat CT. Granulomas at the gray-white junction, smaller in size were excluded. Since histopathology is not routinely implicated in the diagnosis of tuberculomas, we have included only highly probable cases.
Skeletal tuberculosis: Spinal tuberculosis and joint tuberculosis diagnosed by skiagram supported by CT and MRI with/without histopathology evidence.
Exclusion criteria : Tuberculosis in children with immunosuppressive states including HIV.
Mantoux test
1 TU PPD with RT 23 Tween 80 was injected intra-dermally and the reaction read between 48-72 hours. Induration of 10 mm or more in largest diameter was suggestive of natural infection, irrespective of prior BCG vaccination. In case a patient presented late but within 7 days of the test, then any reaction above 10 mm is was considered as a positive test.
Contact: Contact was defined as any child who lived in a household with an adult taking anti-TB therapy or had taken such a therapy in the past 2 years.[7]
Statistical analysis
Proportions of Mantoux and contact positives in childhood tuberculosis as a whole and in various forms were arrived. To compare these among confirmed against probable cases, chi-square test wsa used. P< 0.05 was considered statistical significance. Sensitivity, specificity, positive and negative predictive value for Mantoux test and contact history were arrived. For this, histopathology or bacteriological proof was considered as gold standard.
Results
Of 605 children diagnosed for tuberculosis, 309 (51.1%) children were less than 6 years and 296 (48.9%) were of 7-12 year age group. Of the total study group, 198 (32.7%) were confirmed by bacteriology or by histopathology. The remaining 407 (67.3%) cases were included as highly probable cases of various forms of tuberculosis. The number of confirmed and highly probable cases in various forms of tuberculosis is shown in table1 Mantoux was positive in a total of 210 (34.7%) cases which included 101 (51%) confirmed cases and 109 (26.8%) probable cases (p < 0.05). Mantoux positivity in less than 6 year age group was in 103 (33.3%) in comparison to 107 (36.15%) in 7-12 year age group (p = 0.47).
Of 122 (20.2%) children with pulmonary tuberculosis, only 46 (37.7%) had bacteriological evidence (38 Progressive primary and 8 Post primary), the rest were diagnosed on epidemiological factors. Mantoux was positive in 44 (36.1%) children with pulmonary tuberculosis with progressive primary complex having 37 (36.6%) and post primary tuberculosis having 7 (33.3%) cases of Mantoux positive children. In progressive primary complex, the Mantoux positivity was 20 (52.6%) in confirmed cases in contrast to 17 (27%) in probable cases (p = 0.01). Similarly in post primary tuberculosis Mantoux was positive in 4 (50%) of confirmed cases and in 3 (23.1%) of probable cases (p = 0.20). Totally among pulmonary tuberculosis, Mantoux was positive in 24 (52.2%) confirmed cases in comparison to 20 (26.3%) probable cases (p = 0.004).
Extra pulmonary tuberculosis constituted 483 (79.8%) cases. This included lymph node tuberculosis 117 (19.3%), skeletal tuberculosis 99 (16.4%), abdominal tuberculosis 22 (3.6%). CNS tuberculosis which included tuberculous meningitis 160 (26.4%) and tuberculoma 85 (14%) constituted 245 (40.5%) cases.
Mantoux positivity was observed in 62 (53%) of lymph node TB of which 55 (56.1%) were confirmed cases and 7 (36.8%) probable cases (p = 0.123). Eight (36.4%) of TB abdomen cases were Mantoux positive which included 7 (43.8%) confirmed cases and 1 (16.7%) probable case (p= 0.24). In skeletal tuberculosis, 44 (44.4%) cases were Mantoux positive of which 9 (52.9%) were confirmed and 35 (42.7%) probable cases (p = 0.439). Six (28.6%) confirmed cases and 28 (20.1%) probable cases of tuberculous meningitis were Mantoux positive (p=0.379). The overall Mantoux positive in tuberculous meningitis cases was 34 (21.3%). Mantoux was positive in 18 (21.2%) cases of tuberculoma. Contact tracing revealed 184 (30.4%) of children had contact with adult tuberculosis that included 61 (30.8%) confirmed and 123 (30.2%) probable cases (p = 0.883).
Outcome
All the children who were included in the study received anti tuberculous drugs from the hospital. Of the 407 children with probable TB 28[6.8%] were lost for follow up. The remaining 379 completed their treatment and had recovered from their illness. Of the 198 confirmed cases 14 [7.1%] children were lost for follow up. And the rest 184 had recovered.
Discussion
Most recently developed sensitive and specific diagnostic tests for tuberculosis like Bactec, PCR, have not found a place in the routine evaluation of children with suspected tuberculosis as they are not widely available and too costly to be included in the routine screening. In spite of extensive research it is yet to evolve a simple diagnostic test in place of Mantoux. Even after hundred years, Mantoux stands to be a simple, easily available test, making it very useful for the diagnosis of TB though it has its own limitations.
Even in developed countries, the diagnosis of tuberculosis in children is largely based on indirect epidemiological evidences where Mantoux test still plays a significant role. Clinical criteria tuberculin skin test results, radiographic changes, and documented exposure to an infectious adult, remains the standard diagnostic methods.[8] No single ancillary test is 100 percent accurate in making or excluding the diagnosis of tuberculosis. Although a Mantoux test should always be performed, it may be negative in 10 to 25 percent of patients with active disease.[9] The relatively low sensitivity and specificity of this test makes the test very useful for persons at high risk for TB infection or disease but undesirable for use in persons at low risk.[10]
Mantoux positivity in various forms of childhood tuberculosis studied was found to be 34.7%. Mantoux positivity in confirmed cases was 51% in contrast to 26.8% in probable cases which is statistically significant.
In pulmonary tuberculosis, the overall Mantoux positivity was 36.1%. Among pulmonary tuberculosis, Mantoux test was significantly positive in higher proportion in confirmed cases (52.2%) in contrast to 26.3% in probable cases which is statistically significant (p = 0.004).
Of 117 children diagnosed as lymph node tuberculosis, Mantoux positivity was observed in 62(53%) of lymph node tuberculosis with 56.1% in histopathology proven cases. The present study results are comparable with the study conducted by Ahmed et al which showed 65% positivity in lymph node tuberculosis.[11]
Skeletal TB included spinal TB (76) and various other bone and joint TB (23).44.4% of Skeletal TB was Mantoux positive. Though bacteriological support confirms TB, it is not possible in bone TB where imaging techniques play an important role. The high prevalence of tuberculosis in India precludes the need of histopathological diagnosis especially in skeletal TB as other alternative lesions including malignancy are rare in pediatric population. Radiographs are the first line of investigations to substantiate or refute a clinical diagnosis of tuberculosis of the spine. In spinal tuberculosis, the commonest site of involvement is the dorsolumbar vertebra in 62 cases (81.6%). The diagnosis of spinal tuberculosis was substantiated by CT evidence. In 5 children, the initial spine X-rays were normal inspite of highly suggestive clinical presentation and in these, CT confirmed the early tuberculous changes. Though the wide availability of CT scan and MR imaging has increased the use of these modalities in supporting the diagnosis of tuberculosis of the spine, the significance of Mantoux positivity is important in the absence of confirmatory evidence like histopathology.
Only 8 of the 22 cases of TB abdomen were Mantoux positive which constituted 36.4%. The contact positivity in confirmed cases of abdominal tuberculosis was found to be 22.7%. This is very similar to the study done earlier which shows 25.6%.[12]
21.3% of tuberculous meningitis was Mantoux positive which are comparable with studies done by Ahmet Yarami et al and Reider et al.[13], [14] Cases of tuberculomas included in the study were based on strong criteria which exclude the possibility of neurocysticercosis as suggested by Garg et al .[15] The inclusion criteria adopted was based on the consensus adopted by Amdekar et al.[7]
From the present study, it is evident that Mantoux positivity in childhood tuberculosis varies from 21.2 % to 53%. Also positivity of Mantoux is higher in lymph node tuberculosis 62 (53%) than other forms of tuberculosis. The significance of Mantoux is least in CNS tuberculosis 52 (21.2%) which includes tuberculous meningitis 34 (21.3%) and tuberculoma 18 (21.2%). The results obtained from the study are similar to the study done by Mahadevan et al.[4]
Contact investigation remains an important mechanism to identify potentially infected and infective patients. Although the most efficient way of identifying infected children is through contact investigations of adults with pulmonary tuberculosis, the reverse i.e., associate investigation, which includes tracking contacts of infected children, is also important.[16] So the authors studied the contact details of the children included in the study. It was found that only 184 (30.4%) children had contact history positive. This included 61(30.8%) children with confirmed tuberculosis and 123 (30.2%) children with highly probable tuberculosis (p =.883). The present study results are comparable to the studies done by Nair et al[17] and Somu et al[18] which shows the contact positivity of 35.5% in 1000 children and 61.2% in 227 children respectively.
It was found that the sensitivity of Mantoux was 51% and the specificity was 71.7%. The positive predictive value of the test was 52.6% and negative predictive value was 70.4%. The percentage of false negatives was 49% and percentage of false positives was 28.3%. Thus with such low sensitivity and specificity, the reliability of diagnosing tuberculosis with Mantoux alone is limited. With a higher false negativity rate, it cannot be reliable as a very good screening test for tuberculosis. The sensitivity of contact positives was found to be 30.8% and the specificity was 69.8%. The positive predictive value was 33.2% and the negative predictive value was 67.5%. With contact history many a times underreported, the reliability of contact history alone in diagnosing a case as tuberculosis is very low. Considering the above facts and figures, there arises definitely a need of an alternative test for Mantoux both for screening and for diagnosing tuberculosis.
Conclusion
False positive and false negative test make significant variation in Mantoux reactivity during the diagnostic work up in children with suspected tuberculosis. In the present study, the Mantoux positivity was in 210 objects (34.7%) signifying the present status. However, the positivity was 51% in confirmed cases thus showing a better case detection rate when compared to suspected cases. In lymph node tuberculosis, the positivity is high (53%) as against 21.2% in CNS tuberculosis. Contact tracing revealed 30.2% contact positivity in the probable cases and 30.8% in the confirmed cases. Though contact positivity is lower when compared to Mantoux positivity, the significance of detecting the adult contacts in children with tuberculosis cannot be underestimated. As confirming TB in each case with bacteriological support is difficult, contact tracing being positive in only about 30% of cases and Mantoux positivity in 35% of cases only many a times the diagnosis of tuberculosis is made by the combination of the above factors in the absence of reliable diagnostic tests.
KEY MESSAGES
Mantoux positivity varies from 51% [confirmed cases] to 26.8% [probable cases] of childhood tuberculosis.
Mantoux positivity higher in lymph node tuberculosis [53%]. and lowest in CNS tuberculosis [ 21.2%].
Contact positivity in childhood tuberculosis is 30.4% with no major variation in the different forms of tuberculosis.
Contribution of Authors
Dr. D. Vijayasekaran has conceptualized the study, collected the data and contributed to writing the manuscript. Dr. R. Arvind Kumar was involved in data analysis, outcome assessment, literature review and writing the manuscript.
Dr. N.C. Gowrishankar, Dr. K. Nedunchelian and Dr. S. Sethuraman were involved in critical analysis of the manuscript.
Funding : None
Competing Interests : None
References
1. Starke JR, Correa AG. Management of mycobacterial infection and disease in children. Pediatr Infect Dis J 1995; 14 : 455-470.
2. Kabra S K, Lodha, Seth V. Some current concepts on childhood tuberculosis. Indian J Med Res 2004; 120(4): 387-397.
3. Jonas V, Alden MJ, Curry JI, et al. Detection and identification of Mycobacterium tuberculosis directly from sputum sediments by amplification of rRNA. J Clin Microbiol 1993; 31 : 2410-2416.
4. Mahadevan B, Mahadevan S, Serane VT, Narasimhan R. Tuberculin reactivity in tuberculous meningitis. Indian J Pediatr 2005; 72 : 213-215.
5. Comstock GW, Livesay VT, Woolpert SF. The prognosis of a positive tuberculin reaction in childhood and adolescence. Am J Epidemiol 1974; 99: 131-138.
6. Starke JR, Taylor-Watts KT. Tuberculosis in the pediatric population of Houston, Texas. Pediatrics 1989; 84 : 28-35.
7. Amdekar YK. Consensus Statement of IAP Working Group: Status Report on Diagnosis of Childhood Tuberculosis. Indian Pediatr 2004; 41 : 146-155.
8. Khan EA, Jeffrey R. Starke JR. Diagnosis of Tuberculosis in Children: Increased Need for Better Methods, EID 1995; 1 (4) : 115-123.
9. Huebner RE, Schein MF, Bass JB. The tuberculin skin test. Clin Infect Dis 1993; 17 : 968-975.
10. American Academy of Pediatrics: Committee on Infectious Diseases. Screening for tuberculosis in infants and children. Pediatrics 1994; 93 : 131-134.
11. Ahmad Z, Amin SS. Role of tuberculin test, FNAC and Elisa in the diagnosis of Tuberculous Lymphadenitis. JIACM 2003; 4(4) : 292-295.
12. Oya Uygur-Bayramiηli, Gül Dabak, Resat Dabak. A clinical dilemma: abdominal tuberculosis. World J Gastroenterol 2003; 9(5) : 1098-1101.
13. Ahmet Yarami, Fuat Gurkan, Murat Elevli et al. Central Nervous System Tuberculosis in Children: A Review of 214 Cases, Pediatrics 1998; 102 (5): 49.
14. Rieder BL, Cauthen GM, Comstock GW, Snider DE. Epidemiology of tuberculosis in the United States. Epidemiol Rev 1989; 11: 79-98.
15. Garg RK. Diagnostic criteria for neurocysticercosis: Some modifications are needed for Indian patients. Neurol India 2004; 52 : 171-177.
16. Starke JR. Tuberculin skin testing, what, who and when Rep Pediat Infect Dis 1994; 4 : 31-32.
17. Nair NS, Kurian, John TP, Omana Mathew. A review of 1000 cases of childhood tuberculosis at trivandrum. Ind J Tub 1964; 11(2): 82-97.
18. Somu N, Vijayasekaran D, Kannaki M, Balachandran A, Subramanyam L. Adult contacts in children with Tuberculosis. Indian Pediatr 1997; 34 : 819-822.(Vijayasekaran D, Kumar R )