Confiabilidade e desempenho diagnóstico da elastografia hepática transitória na hepatite c crônica durante fase de treinamento
Data
2023-02-15
Autores
Sister, Flávia Appel 
Orientadores
Carvalho Filho, Roberto José
Tipo
Dissertação de mestrado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Introdução: Estima-se que a infecção crônica pelo vírus da hepatite C (HCV) afete 58 milhões de pessoas em todo o mundo e cerca de 700.000 indivíduos no Brasil Vários marcadores não invasivos de fibrose hepática têm sido propostos, como a elastografia hepática transitória (EHT) por FibroScan®. Entretanto, a confiabilidade da EHT executada por operadores em fase de treinamento é indeterminada, o que dificulta sua aplicabilidade na prática clínica. Objetivos: O presente estudo teve como objetivo principal avaliar o impacto do treinamento de um operador inexperiente para a realização de exames de EHT por FibroScan® em portadores de hepatite C crônica. Especificamente, foram avaliados a frequência de resultados não confiáveis ao longo do treinamento e a concordância da EHT com o índice APRI para estimar a presença de fibrose hepática significativa ao longo do treinamento. Métodos: Trata-se de estudo transversal, com coleta retrospectiva de dados que inclui portadores de hepatite C crônica compensada que tenham sido submetidos a exames de EHT por FibroScan® entre março de 2014 e dezembro de 2016, realizados por dois pesquisadores. Os pacientes incluídos foram divididos em dois grupos, conforme a experiência do operador no exame de EHT: grupo EE (Examinador Experiente) e grupo ET (Examinador em Treinamento), estes últimos divididos em duas fases: fase 1, constituída pelos primeiros 100 exames e fase 2, com os exames subsequentes. Para a análise de confiabilidade da EHT, foram utilizados os parâmetros taxa de sucesso (TS), relação entre intervalo interquartil (IQR) e mediana (Md) das aquisições (IQR/Md) e os critérios de confiabilidade de Castéra et al. (2005) e de Boursier et al. (2013). A análise de desempenho diagnóstico da EHT foi feita de maneira indireta, analisando-se a concordância entre as estimativas do grau de fibrose hepática significativa (FS) pela EHT e pelo APRI. Resultados: Entre março de 2014 e dezembro de 2016, 771 adultos portadores de HCV foram avaliados e divididos nos grupos EE (n = 161), ET fase 1 (n = 100) e ET fase 2 (n = 158), que se mostraram semelhantes em relação às suas características demográficas, clínicas e laboratoriais. A TS e a relação IQR/Md foram semelhantes entre os grupos EE, ET fase 1 e ET fase 2. A proporção de resultados confiáveis foi comparável nos três grupos, utilizando-se os critérios de confiabilidade de Castéra et al. (92,5%, 92,0% e 97,5%, respectivamente) e de Boursier et al. (96,9%, 95,0% e 98,1%, respectivamente). As prevalências de fibrose hepática significativa (FS) conforme estimativa da EHT nos grupos EE, ET fase 1 e ET fase 2 foram semelhantes (60,9%, 62,1% e 63,9%, respectivamente; p = 0,863). Houve proporção expressiva de pacientes com rigidez hepática superior a 7,1 kPa e com APRI inferior a 0,5 (entre 56 e 72%). As prevalências de resultados discordantes foram de 32,3%, 29,0% e 45,7% nos grupos EE, ET fase 1 e ET fase 2, com concordância significativa, mas apenas razoável entre os métodos, nos três grupos (κ = 0,378, 0,429 e 0,207). Conclusões: A confiabilidade dos exames de EHT por FibroScan® executados pelo ET foi comparável àquela observada nos exames realizados pelo EE e semelhante durante as duas fases de treinamento do ET. Houve concordância significativa, embora razoável, entre a EHT e o índice APRI para estimar a presença ou ausência de FS, sem impacto expressivo da experiência do examinador ou do tempo de treinamento na concordância entre os métodos. EHT parece ser um método bastante confiável e acurado, mesmo quando executado por operador com pouco tempo de treinamento. Um curto treinamento (de até 100 exames) parece ser suficiente para que a EHT tenha boa aplicabilidade clínica nas mãos de um operador com experiência inicial no método.
Introduction: Several non-invasive markers of hepatic fibrosis have been proposed, such as a transient hepatic elastography (THE) by FibroScan®. However, the reliability of THE performed by operators in the training phase is undetermined, which makes its applicability in clinical practice difficult. Objectives: The main objective of the present study was to evaluate the impact of training an inexperienced operator to perform THE exams by FibroScan® in patients with chronic hepatitis C. Specifically, the frequency of unreliable results during training and the agreement of THE with the APRI index were evaluated to estimate the presence of significant hepatic fibrosis during training. Methods: This is a cross-sectional study, with retrospective data collection that includes patients with compensated chronic hepatitis C who underwent THE exams by FibroScan® between March 2014 and December 2016, performed by two researchers. The patients included were divided into two groups, according to the operator's experience in the THE exam: EE group (Experienced Examiner) and ET group (Training Examiner), the latter divided into two phases: phase 1, consisting of the first 100 exams and phase 2, with subsequent exams. For the THE reliability analysis, the parameters success rate (TS), relationship between interquartile range (IQR) and median (Md) of acquisitions (IQR/Md) and the reliability criteria of Castéra et al. (2005) and de Boursier et al. (2013). The analysis of the diagnostic performance of THE was performed indirectly, analyzing the agreement between the estimates of the degree of significant liver fibrosis (SF) by the THE and by the APRI. Results: Between March 2014 and December 2016, 771 adults with HCV were evaluated and divided into groups EE (n = 161), ET phase 1 (n = 100) and ET phase 2 (n = 158), which showed to be similar in relation to their demographic, clinical and laboratory characteristics. The TS and the IQR/Md ratio were similar between the EE, ET phase 1 and ET phase 2 groups. The proportion of reliable results was comparable in the three groups, using the reliability criteria of Castéra et al. (92.5%, 92.0% and 97.5%, respectively) and by Boursier et al. (96.9%, 95.0% and 98.1%, respectively). The prevalence of significant hepatic fibrosis (SF) as estimated by THE in the EE, ET phase 1 and ET phase 2 groups were similar (60.9%, 62.1% and 63.9%, respectively; p = 0.863). There was a significant proportion of patients with liver stiffness greater than 7.1 kPa and an APRI less than 0.5 (between 56 and 72%). Prevalence of discordant results were 32.3%, 29.0% and 45.7% in the EE, ET phase 1 and ET phase 2 groups, with significant but only reasonable agreement between the methods in the three groups (κ = 0.378, 0.429 and 0.207). Conclusions: The reliability of the THE exams by FibroScan® performed by the ET was comparable to that observed in the exams performed by the EE and similar during the two phases of ET training. There was significant, although reasonable, agreement between the THE and the APRI index to estimate the presence or absence of SF, with no significant impact of the examiner's experience or training time on the agreement between the methods. THE seems to be a very reliable and accurate method, even when performed by an operator with little training time. A short training (up to 100 exams) seems to be enough for the THE to have good clinical applicability in the hands of an operator with initial experience in the method.
Introduction: Several non-invasive markers of hepatic fibrosis have been proposed, such as a transient hepatic elastography (THE) by FibroScan®. However, the reliability of THE performed by operators in the training phase is undetermined, which makes its applicability in clinical practice difficult. Objectives: The main objective of the present study was to evaluate the impact of training an inexperienced operator to perform THE exams by FibroScan® in patients with chronic hepatitis C. Specifically, the frequency of unreliable results during training and the agreement of THE with the APRI index were evaluated to estimate the presence of significant hepatic fibrosis during training. Methods: This is a cross-sectional study, with retrospective data collection that includes patients with compensated chronic hepatitis C who underwent THE exams by FibroScan® between March 2014 and December 2016, performed by two researchers. The patients included were divided into two groups, according to the operator's experience in the THE exam: EE group (Experienced Examiner) and ET group (Training Examiner), the latter divided into two phases: phase 1, consisting of the first 100 exams and phase 2, with subsequent exams. For the THE reliability analysis, the parameters success rate (TS), relationship between interquartile range (IQR) and median (Md) of acquisitions (IQR/Md) and the reliability criteria of Castéra et al. (2005) and de Boursier et al. (2013). The analysis of the diagnostic performance of THE was performed indirectly, analyzing the agreement between the estimates of the degree of significant liver fibrosis (SF) by the THE and by the APRI. Results: Between March 2014 and December 2016, 771 adults with HCV were evaluated and divided into groups EE (n = 161), ET phase 1 (n = 100) and ET phase 2 (n = 158), which showed to be similar in relation to their demographic, clinical and laboratory characteristics. The TS and the IQR/Md ratio were similar between the EE, ET phase 1 and ET phase 2 groups. The proportion of reliable results was comparable in the three groups, using the reliability criteria of Castéra et al. (92.5%, 92.0% and 97.5%, respectively) and by Boursier et al. (96.9%, 95.0% and 98.1%, respectively). The prevalence of significant hepatic fibrosis (SF) as estimated by THE in the EE, ET phase 1 and ET phase 2 groups were similar (60.9%, 62.1% and 63.9%, respectively; p = 0.863). There was a significant proportion of patients with liver stiffness greater than 7.1 kPa and an APRI less than 0.5 (between 56 and 72%). Prevalence of discordant results were 32.3%, 29.0% and 45.7% in the EE, ET phase 1 and ET phase 2 groups, with significant but only reasonable agreement between the methods in the three groups (κ = 0.378, 0.429 and 0.207). Conclusions: The reliability of the THE exams by FibroScan® performed by the ET was comparable to that observed in the exams performed by the EE and similar during the two phases of ET training. There was significant, although reasonable, agreement between the THE and the APRI index to estimate the presence or absence of SF, with no significant impact of the examiner's experience or training time on the agreement between the methods. THE seems to be a very reliable and accurate method, even when performed by an operator with little training time. A short training (up to 100 exams) seems to be enough for the THE to have good clinical applicability in the hands of an operator with initial experience in the method.