Made available in DSpace on 2015-07-22T20:50:05Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-03-31 / Alguns radiologistas experientes consideram que, com o desenvolvimento dos aparelhos de RM e com a experiência adquirida, cada vez menos se precisa da introdução intra-articular de contraste no ombro, para a avaliação das estruturas internas, à semelhança do que ocorreu com o menisco e as demais estruturas internas do joelho. O autor desconhece a existência de estudo comparativo entre os três métodos diagnósticos (RMC, ATCM e ARM), efetuados no mesmo paciente, bem como estudo avaliador da variabilidade intra e interobservador para os três métodos. O presente trabalho teve como objetivos: 1. Avaliar a reprodutibilidade da RMC, ATCM e da ARM, separadamente, por meio da análise da concordância e da discordância intraobservador e interobservador. 2. Avaliar a concordância entre os três métodos (RMC, ATCM e ARM) aplicando o teste G de Cochran. 3. Verificar a sensibilidade, especificidade, VPP, VPN e acurácia dos três métodos, utilizando-se a artroscopia como padrão-ouro. Para tanto, foram feitos, prospectivamente, exames de RMC, ATCM e ARM em 32 pacientes. Cada paciente foi submetido aos três tipos de exame que foram realizados em cada paciente, todos os três realizados no mesmo dia de exame. Foram analisados: Lesão do lábio anterior, lesão do lábio superior, lesão do lábio posterior, Bankart ósseo, lesão de Hill-Sachs, lesão do manguito rotador, lesão do tendão da cabeça longa do músculo bíceps braquial e lesão condral. No presente estudo, a RMC, ATCM e ARM mostraram-se métodos com reprodutibilidade semelhante e significativa, particularmente para detecção de lesão do lábio anterior, Bankart ósseo e lesão de Hill-Sachs. Nas lesões do lábio superior, a ARM apresenta maior reprodutibilidade em relação à ATCM e RMC. Para os demais achados, os dados são mais variáveis. A comparação entre os três métodos sugere que os mesmos são comparáveis nos quesitos lesão do lábio anterior, lesão do lábio superior, lesão do lábio posterior, Bankart ósseo, lesão de Hill-Sachs, lesão do TCLB e lesão condral, para a população estudada. Com relação à lesão de manguito rotador, a ATCM não se mostrou comparável à RMC e ARM. Alicerçado nos dados de sensibilidade, especificidade, VPP, VPN e acurácia dos oito pacientes submetidos à artroscopia, acredita-se que a ARM tende a ser melhor do que a ATCM e RMC. / Purpose: We prospectively compared MR imaging, multi-detector CT arthrography and MR arthrography, each method performed in all patients, to determine the intra-observer and inter-observer agreement of each method. We also analysed the agreement between the methods, and calculated the acuracy of each method using artrhoscopy as the gold standart. Methods: Thirty two patients 13-43 years old (mean, 24,2 years old) engaged in amateur or professional sports who had signs and symptoms of shoulder instability, SLAP lesion and/ou impingement syndrome were refered to diagnostic imaging. Each patient underwent MR imaging, followed by multi-detector CT arthrography and MR arthrography after intraarticular injection of up to 20 ml of a mixed dilute solution of gadolinium and iodinated contrast material. All patients completed the three examinations on the same day. Eight patients also underwent arthroscopy. Results: For anterior labral lesions, the intra-observer agreement was excelent (Kw = 0,834; 0,925; e 0,803 for MRI, MDCTA and MRA respectively), and the interobserver agreement was significant (Kw = 0,739; 0,762 e 0,672 for MRI, MDCTA and MRA respectively). For superior labral lesions, the intra-observer agreement was slight for MRI (Kw = 0,351) and substantial for MRA (Kw = 0,717), and the interobserver agreement was poor for MRI (Kw = 0,067) and moderate for MRA (Kw = 0,434). For posterior labral lesions, the intra-observer agreement was moderate to excelent (Kw = 0,467; 0,84 e 0,784 for MRI, MDCTA and MRA respectively), and the inter-observer agreement was slight to substantial (Kw = 0,636; 0,368 e 0,273 for MRI, MDCTA and MRA respectively). For Hill-Sachs lesions, the intra-observer agreement was excelent (Kw = 0,84; 1 e 0,84 for MRI, MDCTA and MRA respectively), and the inter-observer agreement was substantial (Kw = 0,632; 0,714 e 0,765 for MRI, MDCTA and MRA respectively). For osseous Bankart lesions, the intra-observer agreement was excelent (Kw = 1; 0,92 e 0,925 for MRI, MDCTA and MRA respectively), and the inter-observer agreement was substantial to excelent (Kw = 0,846; 0,642 e 0,92 for MRI, MDCTA and MRA respectively). For rotator cuff lesions, the intra-observer agreement was excelent for MRI (Kw = 0,89) and substantial for MRA (Kw = 0,784), and the inter-observer agreement was substantial for MRI and MRA (Kw = 0,617 and 0,632 respectively). For biceps tendon lesions, the intra-observer agreement was excelent for MRI (Kw = 1) and alight for MRA (Kw = 0,264), and the inter-observer agreement was substantial for MRI (Kw = 617) and poor for MRA (Kw = 0,081). For condral lesions, the inter-observer agreement was slight for MRI (Kw = 0,368). The Cochran G test showed that the three methods are comparable with regard to anterior labral lesion, superior labral lesion, posterior labral lesion, osseous Bankart, Hill-Sachs lesion, biceps lesion and condral lesion. Only for the evaluation of rotator cuff tears, multi-detector CT arthrography did not compare well with MR imaging and MR arthrography. Of the eight patients who underwent surgery, there were 7 anterior labral lesions, 2 superior labral lesions, 2 posterior labral lesions, 1 osseous Bankart and 5 Hill-Sachs lesions by arthroscopy. The acuracy of MR imaging, multi-detector CT arthrography and MR arthrography were respectively 63, 75 and 88% for anterior labral lesions, 83, 50 and 100% for superior labral lesions, 75,50 and 75% for posterior labral lesions, 75, 75 and 88% for osseous Bankart and 63, 75 and 63% for Hill-Sachs lesions. Conclusions: In the sample population studied, MRI, MDCTA and MRA have significant and comparable reproducibility, particularly for detection of anterior labral lesions, osseous Bankart and Hill-Sachs lesions. For superior labral lesions, MRA shows higher reproducibility when compared to MDCTA and MRI. For the other findings, data are more variable. The methods (MRI, MDCTA and MRA) are comparable with the exception for evaluation of rotator cuff tears where MDCTA did not compare well with MRI and MRA. In the subset of patients who underwent arthroscopy, the data of this study although limited, follow the data in the literature showing tendency of MR arthrography performing better than multi-detector CT arthrography and MR imaging. / TEDE