Abstract Background Cervical radiculopathy (CR) is a clinical condition caused by compression of the nerve root. In clinical practice, the diagnosis of CR is based on information from the patient’s history, physical examination, and diagnostic imaging. This systematic review aimed to update and summarise the evidence reported in a systematic review published in 2018 on the diagnostic performance of physical examination tests. Methods A literature search was performed in six electronic databases. Selection, assessment of risk of bias (using the QUADAS-2) and data extraction were performed independently by two reviewers. Sensitivity and specificity were calculated, and the certainty of the evidence was assessed using the GRADE framework. For the meta-analysis, a hierarchical bivariate random-effects model was used and, in line with recommendations for sparse data, models were interpreted as bivariate fixed-effect Generalized Linear Mixed Models. Results In total, eight studies were included. Diagnostic value was assessed for six physical examination tests. Slightly different versions of Spurling’s test were assessed in five studies, with a reported high specificity ranging from 0.84 to 1.00 (95% CI range: 0.56–1.00) and sensitivity values ranging from 0.38 to 0.98 (95%CI range: 0.22–0.99). There is low certainty evidence of pooled sensitivity of 0.70 (95%CI 0.60–0.79) and specificity of 0.71 (95%CI 0.63–0.79) for Upper Limb Neurodynamic Test (ULNT) 1. Similary there was low certainty evidence of pooled sensitivity of 0.97 (95%CI 0.88–0.99) and pooled specificity of 0.51 (95%CI 0.40–0.62) for combined ULNTs, and of pooled sensitivity of 0.49 (95%CI 0.39–0.60) and pooled specificity of 0.76 (95%CI 0.66–0.84) for the shoulder abduction relief test. All other tests were assessed in a single study only. Conclusions Clinicians may use the outcome of Spurling’s test and the outcome of the combination of four ULNTs to assist their clinical reasoning in diagnosing CR. However, evidence on the diagnostic accuracy of all physical tests for the diagnosis of CR is sparse, and the certainty of the evidence was very low for all outcomes of all tests, implying that new studies of high methodological value are still required to strengthen these results. Because of the small number of studies, the pooled estimates are valid only for the populations and tests studied in the specific studies included in this review.
Thoomes et al. (Thu,) studied this question.