MRI and SPECT/CT demonstrate, with low certainty of evidence, the highest diagnostic accuracy for aseptic knee arthroplasty loosening: A systematic comparative diagnostic test review and meta‐analysis

Abstract

AbstractPurposeThe purpose of this study was to evaluate and compare the diagnostic accuracy of modalities used to aid the diagnosis of aseptic knee arthroplasty loosening.MethodsA comparative diagnostic test accuracy systematic review and meta‐analysis was conducted following the Cochrane and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) guidelines. PubMed, EMBASE and Cochrane databases were searched for original articles evaluating diagnostic modalities up to March 2024. Included studies compared the modality (index test) to the intraoperative finding as reference test. The QUADAS‐C (Quality Assessment of Diagnostic Accuracy Studies‐Comparative) tool was used to assess the quality of the included studies. The GRADE (Grading of Recommendations, Assessment, Development and Evaluation) approach was used to evaluate the certainty of evidence. Level of evidence was evaluated using the Oxford Levels of Evidence tool. The primary outcome was the summary of diagnostic accuracy metrics for each modality as demonstrated by a summary receiver‐operating characteristic (SROC) curve.ResultsThe search yielded 467 articles. Of these, 14 articles were included. These 14 articles evaluated a total of five different diagnostic modalities: bone scintigraphy (three studies, 146 cases), 18‐fluorodeoxyglucose positron emission tomography (two studies, 50 cases), single‐photon emission computed tomography combined with computed tomography (SPECT/CT) (seven studies, 371 cases), radionuclide arthrogram (three studies, 196 cases) and magnetic resonance imaging (MRI) (one study, 116 cases). Nine studies exhibited a high risk of bias in patient selection, and all studies showed a high risk of bias related to the reference test. The majority of the included studies were classified as Level III evidence, leading to an overall low level of certainty in the evidence. The most accurate tests, as demonstrated by the SROC analysis, were MRI and SPECT/CT, with sensitivities ranging from 0.00 to 1.00 and 0.33 to 1.00 and specificities between 0.31–1.00 and 0.00–1.00, respectively.ConclusionsThis review and meta‐analysis evaluated available diagnostic modalities to aid the diagnosis of knee arthroplasty loosening and based on a low certainty of evidence suggests that MRI and SPECT/CT are currently the most accurate modalities available to aid the diagnosis of aseptic loosening of knee arthroplasty components.Level of EvidenceLevel III.