Purpose: To evaluate the diagnostic accuracy of plain film arthrography (PFA) on the presence of intra-articular adhesions of the temporomandibular joint (TMJ).
Patients and methods: The study population included 105 consecutive patients (105 TMJs) and data collected in our department. All patients were examined by PFA and arthroscopy. The findings of interpreting PFA were recorded as positive, suspicious, or negative according to the PFA radiographic criteria. After comparing the findings of PFA with those of arthroscopy, the comparison data were assessed by generating receiver operating characteristic curve (ROC curve) and analyzing the area (A(z)) under the curves with SPSS 11.0 (SPSS, Inc, Chicago, IL), and the area under the ROC curve was calculated. According to the area, the diagnostic value of PFA was evaluated. Meanwhile, the causes of misdiagnosis and underdiagnosis of intra-articular adhesions were analyzed.
Results: Arthroscopic surgery results confirmed that 56 TMJs in 105 patients had adhesions. The area under ROC curve was 0.73 (confidence interval, 0.64, 0.82; P < .05). The adhesions located in the medial groove could hardly be detected through PFA, while those located in the lateral two thirds of upper joint compartment could easily be detected.
Conclusion: Although the diagnostic accuracy of PFA was fair, it was a simple and effective modality to diagnose intra-articular adhesions of the TMJ. The causes of misdiagnosis and underdiagnosis were related to the size and location of the intra-articular adhesions.