• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Avaliacao de diferentes abordagens para o diagnostico da leishmaniose visceral humana

Assis, Talia Santana Machado de January 2012 (has links)
Submitted by Nuzia Santos (nuzia@cpqrr.fiocruz.br) on 2012-08-21T18:19:33Z No. of bitstreams: 1 Tese_versao_final_Tália.pdf: 1946064 bytes, checksum: 3f43e249b71311bf82e36d76fdcdf61d (MD5) / Made available in DSpace on 2012-08-21T18:19:33Z (GMT). No. of bitstreams: 1 Tese_versao_final_Tália.pdf: 1946064 bytes, checksum: 3f43e249b71311bf82e36d76fdcdf61d (MD5) / A leishmaniose visceral (LV) e uma doenca parasitaria grave que afeta anualmente 500.000 pessoas em 65 paises. No Brasil, mais de 15.000 casos de LV foram registrados entre os anos de 2007 e 2010, com 880 mortes. Esta elevada taxa de letalidade esta, em parte, relacionada a ineficiencia dos servicos de saude em diagnosticar e tratar precocemente os pacientes. O diagnostico da LV exige suspeicao clinica e confirmacao laboratorial eficientes. Assim, uma reestruturacao do algoritmo e do arsenal de metodos diagnosticos disponiveis nos servicos de saude pode ajudar a melhorar este cenario, aumentando o acesso e a disponibilizacao de testes nos municipios. O objetivo deste estudo foi avaliar o desempenho de diferentes abordagens para o diagnostico da LV humana no Brasil, atraves da avaliacao do desempenho do teste de aglutinacao direta (DAT) e do teste rapido IT-LEISH; do desenvolvimento e validacao de modelos preditivos clinico-laboratoriais e da avaliacao de metodos diagnosticos pela analise de classes latentes (ACL). O grupo de estudo compreendeu 404 pacientes, residentes no Piaui, Maranhao, Bahia e Minas Gerais, com suspeita de LV, submetidos ao exame do aspirado de medula ossea e a coleta de sangue para a realizacao dos testes sorologicos. Para a avaliacao do desempenho do DAT e do IT-LEISHR foi utilizada a analise de validacao classica (AVC) utilizando como padrao de referencia o exame do aspirado de medula ossea. Para o desenvolvimento e a validacao dos modelos preditivos foram utilizados dados epidemiologicos, clinicos e laboratoriais dos pacientes avaliados. Os modelos preditivos utilizando regressao logistica multipla (RLM) foram desenvolvidos no software STATAR e aqueles utilizando arvore de classificacao e regressao (CART) foram desenvolvidos no software SPLUSR. Para a avaliacao da aplicabilidade da ACL foram utilizados dados do estudo de validacao do IT-LEISHR e do DAT e a analise foi realizada no software R. O DAT e o IT-LEISH apresentaram, pela AVC, sensibilidades e especificidades de 90 e 93%; 93 e 97%, respectivamente. O modelo preditivo composto pelas variaveis clinicas e laboratoriais associadas ao teste rapido foi o que apresentou melhor desempenho, tanto pela RLM, quanto por CART (sensibilidade de 90,1% e especificidade de 97,2-97,4%). A ACL mostrou-se util na validacao de testes diagnosticos para LV no Brasil, estimando para o DAT e o IT-LEISH sensibilidades de 88,5 e 94% e especificidades de 95,4 e 100%, respectivamente. Diante do elevado desempenho apresentado, o DAT e o teste rapido IT-LEISH estao indicados para o diagnostico da LV humana no Brasil. Os modelos preditivos desenvolvidos sao uteis e poderiam ser utilizados em centros de referencia para a doenca no pais. A ACL esta indicada para validacao de testes diagnosticos no pais e quando possivel deve ser utilizada. / Visceral leishmaniasis (VL) is a severe parasitic disease that affects 500 000 people yearly in approximately 65 countries. In Brazil, more than 15.000 VL cases were reported between 2007 and 2010, with 880 deaths. This high fatality rate is partly related to the inefficiency of health services to diagnose and treat early patients. The diagnosis of VL requires effective clinical suspicion and laboratory confirmation. Thus, a restructuring of the algorithm and the arsenal of available diagnostic methods in health services can help improve this scenario, by increasing the access and availability of tests in the municipalities. The objective of this study was to evaluate the performance of several approaches for the diagnosis of human VL in Brazil, through of evaluation of performance of DAT and rapid test IT-LEISH; of developing and validating of predictive models and of evaluation of diagnostic methods by the latent class analysis (LCA). The study group was composed by 404 patients, from the states of Piaui, Maranhao, Bahia and Minas Gerais, with suggestive clinical of VL, submitted to the examination of bone marrow aspirate and blood collection for the performance of serological tests. To evaluate the performance of the DAT and the rapid test IT-LEISHR was used the classic validation analysis, using as reference standard the examination of bone marrow aspirate. To develop and validate the predictive models were used epidemiological, clinical and laboratory data of the patients. The predictive models using logistic regression were developed in STATAR software and those using classification and regression tree (CART) were developed in the software SPLUSR. To evaluate the applicability of the LCA were used data from the validation study of the rapid test IT-LEISHR and the DAT and the analysis was performed in software R. The DAT and the rapid test IT-LEISHR showed, by the classic validation analysis, sensitivities and specificities of 90 and 93%, 93 and 97% respectively. The predictive model composed of clinical-laboratory variables and rapid test IT-LEISHR developed using both logistic regression and the model CART showed the best performance (sensitivity 90.1% and a specificity of 97.2 to 97.4%). The LCA proved to be a useful tool for the validation of diagnostic methods for human VL in Brazil, estimating for the DAT and the rapid test IT-LEISHR sensitivities of 88.5 and 94% and specificities of 95.4 and 100% respectively. The DAT and the rapid test IT-LEISHR showed high performance and are indicated for the diagnosis of VL in Brazil. The predictive models developed are useful and could be used in reference centers for VL in the country. The LCA is indicated for validation of diagnostic tests in Brazil and when possible should be used.

Page generated in 0.0575 seconds