Objective: To describe the diagnostic correlation between clinical and magnetic resonance imaging (MRI) findings in temporomandibular degenerative joint diseases.
Methods: PubMed, Cochrane Library, Scopus and Web of science were searched using specific indexing terms and reference lists were hand-searched. Temporomandibular joint disorders were assessed through clinical diagnosis protocols with the aid of Research Diagnostic Criteria for Temporomandibular Disorders or Diagnostic Criteria for Temporomandibular Disorders. Included publications satisfied pre-established criteria. Primary studies were interpreted using a modification of the second version of Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool.
Results: The literature search yielded 3280 titles and abstracts. 69 studies were eligible for full text reading. After data extraction and interpretation with the QUADAS-2 tool, 11 studies remained. Most studies were methodologically acceptable, although none of them fulfilled all criteria of risk of bias according to QUADAS-2. No clear evidence was found for a relationship between clinical and MRI diagnoses and findings. Pooled overall sensitivity and specificity were 59% and 81%, respectively. Sensitivity and specificity for disk displacement with reduction, disk displacement without reduction and arthralgia were 66% and 72%, 62% and 99%, 43% and 68%, respectively.
Conclusion: Overall diagnostic capabilities of clinical test are of poor-moderate validity, compared to MRI. Specificity was found to be higher than sensitivity, especially for disk displacement without reduction. MRI imaging should only be used when the information provided can influence clinical decisions. Further studies with higher level of evidence are needed.
