Upper limb neurodynamic tests (ULNT) are used to examine patients with suspected cervical radiculopathy (CR), carpal tunnel syndrome (CTS) or cubital tunnel syndrome. Understanding the validity of these tests is critical if they are to be used to diagnose these conditions in clinical practice.
The purpose was to critically and systematically assess the literature on the validity of ULNTs in identifying nerve dysfunction in the upper limb.
A comprehensive literature search was conducted from inception until September 2018. Databases searched electronically were PubMed, Scopus, CINAHL, the Cochrane library, EMBASE and ScienceDirect.
Studies using ULNT on humans in the upper limb and cervical spine, published in English were included. Methodological quality of biomechanical studies was assessed with the Quality Appraisal for Cadaveric Studies and with the Anatomical Quality Assessment tool.
Concurrent validity studies were assessed by two independent researchers with the Quality Assessment tool of Diagnostic Accuracy Studies.
A receiver operating characteristic (ROC) curve was computed to graphically present the relationship between true disease positives (sensitivity, on vertical axis) and false diseases positives (1- specificity, on horizontal axis) for different positivity thresholds.
Twelve biomechanical studies (nine cadaveric studies, three in-vivo studies) underpinned the plausibility (construct validity) of ULNTs. With ULNTs mechanical loads to the nervous system are applied and the intended changes in nerve bed length (strain) and gliding (excursion) are achieved.
Nine diagnostic accuracy studies and one case study assessed the concurrent validity of ULNTs. Agreement between the two reviewers (J.K. & V.M.) was high for all methodological tools scored (k 0.65-0.71).
High variability in diagnostic accuracy across studies was found which could be due to varying methodological quality, differences in reference standard used and varying criteria for a positive ULNT. Therefore, a positive test should reproduce the patient's symptoms and change with structural differentiation.
Overall diagnostic accuracy was good for CR using ULNT3 (80%) and ULNT1 (79%) but not the ULNT2b (59%). Diagnostic accuracy was found for the ULNT1 (71%) but not for the ULNT2b (47%) in diagnosing CTS.
This review provides evidence that ULNT are biomechanically plausible tests since with the application of test components the individual nerves are targeted. Furthermore, there is limited evidence that ULNTs are accurate in identifying nerve dysfunction in the upper limb. If a positive ULNT1 or 3 reproduces the patient´s symptoms and is altered with structural differentiation, then cervical radiculopathy can be ruled in.
Concerning CTS, the ULNT1 with positive test criteria consisting of symptom reproduction and increased or decreased symptom response with structural differentiation may be a useful test.
The ULNT2b is not helpful in establishing a diagnosis of cervical radiculopathy or CTS. Direct conclusions from a case study concerning the diagnostic accuracy of the ULNT3 for cubital tunnel syndrome cannot be made.
ULNT are biomechanically plausible tests. There is limited evidence that ULNT are accurate in identifying nerve dysfunction in the upper limb.
This work was unfunded
This work was presented at Physiotherapy UK 2019