Abstracts
BACKGROUND: the World Health Organization (WHO) and the Brazilian Health Ministry recommend the adoption of the operational classification for therapeutic purposes (multi- or paucibacillary). The validity of that classification is discussed taking as a reference the results of bacilloscopic exams. OBJECTIVE: To verify the sensitivity and specificity of the operational clinical classification that takes into account exclusively the number of cutaneous lesions and that which considers the thickening of neural trunks as well as the number of cutaneous lesions, and relating these with the results from the bacilloscopic exam. METHODS: The information source was consultation of the records referring to patients with a diagnosis of leprosy, from January 2000 to March 2001, at the Fundação Alfredo de Matta, in Manaus, AM. The demographic, clinical and laboratorial data were compiled. RESULTS: The clinical classification based exclusively on the number of cutaneous lesions showed a sensitivity of 73.6% and specificity of 85.6% regarding the bacilloscopy. The classification that combines the number of cutaneous lesions and neural thickening demonstrated 75.8% sensitivity and 71.8% specificity. CONCLUSION: The clinical classification of leprosy based on the number of cutaneous lesions showed, in this work, sensitivity and specificity values similar to those described in studies done in other countries. When the classification was broadened to include the other clinical parameter (the presence of thickened peripheral nerves) there was a significant decrease in the specificity, without a statistically significant increase in sensitivity.
leprosy; leprosy; sensitivity; specificity
FUNDAMENTOS: A Organização Mundial da Saúde - OMS e o Ministério da Saúde do Brasil recomendam a adoção da classificação operacional para fins terapêuticos (multi ou paucibacilar). Discute-se a validade dessa classificação tomando como referência o resultado do exame baciloscópico. OBJETIVO: Verificar a sensibilidade e especificidade da classificação clínica operacional que leva em consideração exclusivamente o número de lesões cutâneas e daquela que considera o espessamento de troncos neurais, além do número de lesões cutâneas, relacionando-as com o resultado do exame baciloscópico. MÉTODO: Como fonte de informação foram consultados os prontuários de pacientes com diagnóstico de hanseníase no período de janeiro de 2000 a março de 2001, na Fundação Alfredo da Matta, em Manaus, AM, onde estavam registrados os dados demográficos, clínicos e laboratoriais. RESULTADOS: A classificação clínica baseada exclusivamente no número de lesões cutâneas mostrou sensibilidade de 73,6% e especificidade de 85,6% com relação à baciloscopia. A classificação que combina o número de lesões cutâneas e de nervos espessados demonstrou sensibilidade de 75,8% e especificidade de 71,8%. CONCLUSÃO: A classificação clínica da hanseníase baseada no número de lesões cutâneas mostrou, neste estudo, valores de sensibilidade e especificidade similares aos descritos em estudos realizados em outros países. Quando se acrescentou à classificação outro parâmetro clínico, a presença de nervos periféricos espessados, ocorreu significativa diminuição da especificidade, sem aumento estatisticamente significativo da sensibilidade.
hanseníase; hanseníase; sensibilidade; especificidade
CLINICAL, EPIDEMIOLOGICAL, LABORATORY AND THERAPEUTIC INVESTIGATION
Correlation between clinical and laboratorial findings of the leprosy cases followed at the "Alfredo da Matta" Dermatology Center, Manaus-AM, Brazil, from January 2000 to March 2001, based on secondary data*
Iara L. F. CrippaI; Maria Conceição SchettiniI; Antonio P. SchettiniII; Silmara N. PenniniIII; Paula F. Bessa RebelloIV
IMD, dermatologist at Fundação Alfredo da Matta
IIMD, dermatologist at Fundação Alfredo da Matta/MSc in tropical pathology
IIIMD, dermatologist at Fundação Alfredo da Matta/ MSc in Public health/epidemiology
IVMD, dermatologist at Fundação Alfredo da Matta/ MSc in Public health/epidemiology
Correspondence Correspondence to Iara L. F. Crippa Av. Codajás nº 24 - Cachoeirinha 69065130 Manaus Amazonas Telefone: (92) 663-4747 - ramal 235 Fax: (92) 663-3155
ABSTRACT
BACKGROUND: the World Health Organization (WHO) and the Brazilian Health Ministry recommend the adoption of the operational classification for therapeutic purposes (multi- or paucibacillary). The validity of that classification is discussed taking as a reference the results of bacilloscopic exams.
OBJECTIVE: To verify the sensitivity and specificity of the operational clinical classification that takes into account exclusively the number of cutaneous lesions and that which considers the thickening of neural trunks as well as the number of cutaneous lesions, and relating these with the results from the bacilloscopic exam.
METHODS: The information source was consultation of the records referring to patients with a diagnosis of leprosy, from January 2000 to March 2001, at the Fundação Alfredo de Matta, in Manaus, AM. The demographic, clinical and laboratorial data were compiled.
RESULTS: The clinical classification based exclusively on the number of cutaneous lesions showed a sensitivity of 73.6% and specificity of 85.6% regarding the bacilloscopy. The classification that combines the number of cutaneous lesions and neural thickening demonstrated 75.8% sensitivity and 71.8% specificity.
CONCLUSION: The clinical classification of leprosy based on the number of cutaneous lesions showed, in this work, sensitivity and specificity values similar to those described in studies done in other countries. When the classification was broadened to include the other clinical parameter (the presence of thickened peripheral nerves) there was a significant decrease in the specificity, without a statistically significant increase in sensitivity.
Key words: leprosy; leprosy/classification; sensitivity; specificity
INTRODUCTION
The World Health Organization (WHO) recommends that in regions where good quality bacilloscopy is not available, the leprosy patients should be classified, for treatment purposes, according to the number of cutaneous lesions, considering paucibacillary to be those with up to five lesions and multibacillary to be those with more than five lesions.1 Although cutaneous bacilloscopy is still the gold standard for this purpose, with the decentralization of the control program, a large part of the patients are seen at peripheral health care centers, which, in general, do not have the physical structure, equipment, materials and technical personnel qualified to perform good quality bacilloscopic exams2
Other systems for clinical classification have been proposed, such as that initially recommended by the Brazilian Health Ministry (HM), that took into account the number of cutaneous lesions and/or nerve trunks involved3,4,5 (Chart 1).
The different systems of clinical classification present disparities in their capacity to detect multibacillary forms (sensitivity) and to correctly identify paucibacillary forms (specificity). Classifications that demonstrate little specificity cause a great number of paucibacillary patients to be included in multibacillary treatment regimens, with serious consequences from the perspective of the individual and public health services: there will be unnecessary exposure of patients to potentially toxic drugs; the diagnosis will lead to psychosocial problems due to the stigma; and the control program will suffer increased costs through the waste of erroneously prescribed drugs. While those that demonstrate little sensitivity cause a significant number of multibacillary patients to be treated in an inadequate manner, which is more serious, since it allows the progression of the disease in the individual and exposes the community to a potential transmission source (Chart 2).
Attempts at a concomitant improvement in sensitivity and specificity of the classification, by introduction of more clinical parameters and addition of complementary exams, have not demonstrated satisfactory results. The increase in sensitivity promotes a decrease in specificity, and vice-versa, while the addition of new technology cannot be absorbed at the most elementary levels of the health care structure.
The present study intends to evaluate, in a sample of leprosy patients resident in an endemic area, the sensitivity and specificity of the clinical classification recommended by WHO, taking the result of the bacilloscopy to be the gold standard, as well as to verify whether the increment of another variable, namely the number of thickened nerves, contributes to improved accuracy of the classification.
MATERIAL AND METHODS
The study was developed in a reference center for leprosy, located in the city of Manaus, AM, Fundação Alfredo de Matta - Fuam, historically responsible for the diagnosis of approximately 50% of the new cases of leprosy in the Amazon State. The source of information was the records of patients with a diagnosis of leprosy from January 2000 to March 2001, from which the demographic, clinical and laboratorial data were registered, according to the routine of the Center, that counts on a team of specialized professionals with experience in the diagnosis and treatment of the disease. In this period, 634 new cases of leprosy were diagnosed, of which only six were not included in the study because the records did not include the results from bacilloscopy.
The bacilloscopic exam was performed with smears of lymph from skin of five locations in each patient: ear lobes, elbow folds and cutaneous lesions. The staining technique and reading followed recommendations by the Brazilian Health Ministry.4
The evaluation of sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) of the clinical classifications, with the respective 95% confidence intervals (95% CI), were performed taking the reference pattern to be the result of the skin bacilloscopy and using Epi Info 6 software.
RESULTS
A total of 628 patients were included with age from three to 82 years, 69 (11%) less than 15 years, and 559 (89%) 15 years or over. The mean age was 33.7 years (SD 18.2 years), and 64% of the patients were male.
Regarding the skin bacilloscopy, 178 (28.34%) presented a positive result, and 450 (71.66%) were negative.
According to the clinical classification recommended by WHO, 196 (31.21%) patients were multibacillary, considering that they presented more than five skin lesions, and 432 (68.79%) with up to five lesions were classified as paucibacillary. The sensitivity, in relation to the result of the bacilloscopy, was 73.6% (95% CI 66.4-79.8), and the specificity was 85.6% (95% CI 81.9-88.6). The positive predictive value was 66.8% (95% CI 59.7-73.3), and the negative predictive value was 89.1% (95% CI 85.7-91.8) (Table 1).
When the classification was broadened to include the criterion of presence of thickened peripheral nervous trunks, the sensitivity was 75.8% (95% CI 68.8-81.8), and specificity was 71.8% (95% CI 67.3-75.8), with PPV 51.5% (95% CI 45.3-57.7) and NPV 88.3% (95% CI 84.4-91.3) (Table 2).
Table 3 shows the comparison between the values obtained for sensitivity, specificity, PPV and NPV, when used together, for the clinical classification, the number of cutaneous lesions and the presence of nerve trunk involvement, in relation to the result from the cutaneous bacilloscopy. It was verified that there is an increase in sensitivity, although without statistical significance at the 5% level. In relation to specificity, a statistically significant decrease was observed.
DISCUSSION
The clinical classification of leprosy recommended by WHO offers easy applicability and low cost so won rapid acceptance among the various control programs for the disease, and is even used in reference centers that have good quality bacilloscopic exam available.6 The evaluation of the sensitivity and specificity of the different classifications, using a standard test such as bacilloscopy as a reference, is fundamental to validate its use and is of great importance, because even if a single treatment scheme is adopted, it is nevertheless of great value to determine which are pauci- or multibacillary, since the prognostic and clinical management can be different. Various studies have been carried out with this aim, however a comparative analysis between them is rendered difficult due to the different criteria used for the clinical classification and to the use, at times of bacilloscopy and at other times of histopathology, as the reference standard.
Becx-Bleumink7 evaluated a classification that considered multibacillary patients to be those that present more than five cutaneous lesions and compared the results with the bacilloscopy exam, considering multibacillary to be those that presented a bacilloscopic index (BI) greater than one. The data analysis showed 92% sensitivity and 42% specificity.
Groenen et al .9 and van Brakel et al .8,9 obtained similar results (92% sensitivity and 41% specificity in the former, and 93% sensitivity and 39% specificity in the latter study), using different clinical criteria and comparing the clinical classification with the result of the bacilli tests in the histopathological exam of skin biopsies.
These studies demonstrated in common a good sensitivity to detect multibacillary cases, but low specificity, such that a large number of paucibacillary patients would be treated with the multibacillary regimen. Groenen et al.9 attempted to improve the specificity by trying combinations of different clinical criteria, but they were not successful, because the increase in specificity always occurred in detriment to the sensitivity.
Croft et al .,10 Dasananjali et al .11 eand Buhrer-Sekula et al .,12 using the classification proposed by WHO, obtained similar values, differing from the first studies by demonstrating an increase in the specificity, but lower sensitivity. Despite the improvement in the specificity values, the fall in sensitivity meant that a considerable number of patients with positive bacilloscopy were classified as paucibacillary and thus were treated inadequately. Dasananjali et al .11 have suggested that the group of patients inadequately classified as paucibacillary would be at greater risk of recurrence, due to insufficient treatment.
In the present study, when using the classification criterion based solely on the number of cutaneous lesions, it was verified that the values for sensitivity and specificity were close to those found by the last three authors. It was found that 47/178 (26.4%) patients with positive bacilloscopy would be erroneously classified as paucibacillary, with harmful consequences for the patients and for the control program of the disease.
Buhrer-Sekula et al .,12 in a study also carried out in Brazil, tried to improve the sensitivity, by increasing the clinical criterion to include testing for anti-PGL-i antibodies and obtained good results; however a small number of multibacillary patients would receive incorrect treatment, besides which, additional material and training resources would be necessary to perform the antibody tests.
In the present study, we aimed at verifying the possibility of improving sensitivity to detect the multibacillary forms, using the criterion of number of cutaneous lesions together with the presence of thickened nerve trunks, since neural thickening is more common in the multibacillary forms.13 The data showed that this strategy enabled an increased sensitivity, but without statistical significance at the 5% level. There was a statistically significant decrease in specificity, and in this manner the situation of improving one parameter (sensitivity) to the detriment of the other (specificity) was repeated. Besides which, by adding a new criterion (presence of thickened nerve trunks), it was necessary to train the health teams for this evaluation. Nowadays, neither the classification proposed by WHO nor that recommended by the Brazilian Ministry of Health consider the number of thickened nerve trunks in the classification.
Thus the results of the present study are in agreement with previous studies published and also demonstrate the necessity to look for new parameters, with simple implementation, that would allow improved sensitivity in the detection of multibacillary cases, without significant detriment to specificity.13,14 Good quality bacilloscopic exam, that potentially reaches 100% sensitivity and specificity13 still seems irreplaceable for the correct allocation of the various treatment regimes, while the classification based only on clinical criteria should be restricted to places where this exam is not available.
CONCLUSION
The clinical classification of leprosy recommended by WHO, when compared to the results of the bacilloscopic exam, presented sensitivity and specificity of 73.6% (95% CI 66.4-79.8) and 85.6% (95% CI 81.9 88.6), respectively. The concomitant use of another clinical parameter, the presence of peripheral nerve involvement, lead to a significant decrease in the specificity, at the 5% level, without a significant increase in sensitivity. Therefore, it is only justifiable to use this clinical classification in regions where bacilloscopy is not available.
REFERENCES
Received on May 28, 2004.
Approved by the Consultive Council and accepted for publication on August 31, 2004.
References
- 1 World Health Organization. Chemoterapy of leprosy for control programmes. WHO Technical report Series nr. 874, Geneva, 1998.
- 2. Andrade V LG, Moreira TA, Avelleira JCR, Marques A B. Bayona, M. Paucibacilar ou multibacilar? Uma contribuiçăo para os serviços de saúde. Hans Int 1996;21:6-13.
- 3 Ministério da Saúde do Brasil. Secretaria de Políticas de Saúde. Coordenação Nacional de Dermatologia Sanitária. Guia para o controle da Hanseníase-1993, Brasília.
- 4 Ministério da Saúde do Brasil. Secretaria de Políticas de Saúde. Departamento de Atenção Básica. Área Técnica de Dermatologia Sanitária. Legislação sobre o controle da Hanseníase no Brasil-2000, Brasília.
- 5 Ministério da Saúde do Brasil. Secretaria de Políticas de Saúde. Departamento de Atenção Básica. Guia para o controle da Hanseníase-2002, Brasília.
- 6. Opromolla DVA. Clínica da Hanseníase. Hans Int. 2001; 26:1-4.
- 7. Becx-Bleumink M. Allocation of pacients to paucibacillary or multibacillary drug regimenens for the treatment of leprosy a comparison of methods based mainly on skin smears as opposed to clinical methods for classification of patients. Int J Lepr. 1991; 5 9 : 2 9 2 - 3 0 3 .
- 8. Van Brakel WH, Soldenhoff R, McDougall AC. The allocation of leprosy cases into paucibacillary and multibacillary groups for multidrug therapy taking into account the number of body areas a ffected by skin and nerve lesions. Lepr Rev. 1992;63:231-46.
- 9. Groenen G, Saha NG, Rashid MA, Hamid MA, Pattyn SR. Classification of leprosy cases under field conditions in Bangladesh. II. Realiability of clinical criteria. Lepr Rev.1 9 9 5 ; 6 6 : 1 3 4 - 4 3 .
- 10. Croft RP, Smith WC, Nicholls P, Richardus JH. Sensitivity and specificity of methods of classification of leprosy without use of skin-smear examination. Int J Lepr. 1998;66:445-50.
- 11. Dasananjali K, Schreuder PA, Pirayvaraporn C. A study on e ffectiveness and safety of the W H O / M D T regimen in the north-east of Thailand: a prospective study.1984-1996. Int J Lepr. 1 9 9 7 ; 5 : 2 8 - 3 6
- 12. Buhrer-Sékula S, Sarno EN, Oskam L et al. Use of MLd i p s t i c k as a tool to classify leprosy patients. Int J Lepr. 2001; 68:456-63.
- 13. The Diagnosis and Classification of Leprosy. Report of the international Leprosy Association Technical Forum. Int J Lepr. 2002; S23-S31 .
- 14. Pereira, MG. Epidemiologia Teoria e Prática. 1Ş ed. Brasília: Editora Guanabara Koogan S. A. Rio de Janeiro, 1995. 583p.
Publication Dates
-
Publication in this collection
29 May 2006 -
Date of issue
Oct 2004
History
-
Accepted
31 Aug 2004 -
Received
28 May 2003