Neurological examination for cervical radiculopathy: a scoping review
The short answer
How accurate is the bedside neurological examination for diagnosing cervical radiculopathy?
The bedside neurological examination has high specificity but low sensitivity for cervical radiculopathy, meaning a positive finding moderately increases the probability of the diagnosis but a negative finding does not rule it out. The evidence base is small, heterogeneous, and lacking a gold-standard reference test, so firm recommendations cannot yet be made.
DescriptiveRead paper
Systematic review6 TrialsLimited evidence
Key points
- Only 6 cross-sectional studies met inclusion criteria, reflecting a major gap in the literature
- All BNE components (reflexes, sensation, muscle strength) showed low sensitivity but generally high specificity compared to EMG/NCS or MRI
- Reduced tendon reflexes were the most specific component; sensory testing was least specific
- Combining BNE components increased specificity (up to 99%) but further reduced sensitivity
- BNE procedures were poorly and inconsistently reported across studies, limiting comparability
How it was conducted
- Design
- Scoping review following Arksey and O'Malley framework and JBI methodology, reported using PRISMA-ScR
- Search
- PubMed, Embase, Scopus, CINAHL, DiTA from inception to January 23, 2024; plus grey literature and citation tracking
- Included studies
- 6 cross-sectional studies, all in English
- Population
- Patients with confirmed or clinically suspected cervical radiculopathy
- Index test
- Bedside neurological examination components: tendon reflexes, somatosensation (light touch, pin prick), and muscle strength testing
- Reference standards
- Needle EMG and nerve conduction studies (4 studies, 67%) or MRI (2 studies, 33%)
What they found
- Tendon reflex testing vs. EMG/NCS: sensitivity 21-22% (95% CI 11-27%; -LR 0.80-0.84), specificity 94-97% (95% CI 92-99; +LR 3.5-7.33)
- Tendon reflex testing vs. MRI: sensitivity 28-67% (95% CI 18-40% to 43-85%), specificity 81% (95% CI 69-89%; +LR 1.38)
- Pin prick sensation vs. EMG/NCS: sensitivity 25-38% (-LR 0.84-1.35), specificity 46-89% (95% CI 0.83-0.95%; +LR 0.7-2.27)
- Light touch vs. MRI: sensitivity 42-52% (95% CI 30-54% to 30-74%); combined sensory (light touch + pin prick) sensitivity 67% (95% CI 43-85%); specificity 72% (95% CI 59-82%; +LR 1.42-1.56)
- Muscle strength testing vs. EMG/NCS: sensitivity 54-73% (95% CI 38-65% to NR; -LR 0.44-0.49), specificity 61-93% (95% CI 85-97%; +LR 1.87-7.71)
- Muscle strength testing vs. MRI: sensitivity 30-81% (95% CI 20-43% to 58-95%), specificity 72% (95% CI 60-82%; +LR 1.05)
- Combination of tendon reflex deficit + sensory loss vs. EMG/NCS: specificity 99% (95% CI 94-100%; +LR 14-22), sensitivity 9-21%
- Full BNE (all 3 components) vs. EMG/NCS: sensitivity 7-83% (95% CI NR to 52-98%; -LR 0.61-0.95), specificity 28-99% (95% CI 13-47% to 95-100%; +LR 1.15-14)
- Full BNE vs. MRI: sensitivity 91% (95% CI 70-99%)
- No studies reported on muscle atrophy inspection despite it being a BNE component
Limitations
- No gold standard reference test for cervical radiculopathy exists; EMG/NCS and MRI each have known limitations including inability to assess small nerve fibers (EMG) and detection of structural rather than functional nerve compromise (MRI)
- BNE procedures were poorly described in 50% of studies and key muscles, reflexes, and sensory tests were non-consistent across studies
- Only 6 studies were identified, with small sample sizes and significant heterogeneity preventing meta-analysis or firm conclusions
- Study participants were recruited with heterogeneous diagnostic criteria, some using arm pain as the CR criterion rather than the IASP definition of loss of nerve function
Why it matters
- For patients
- If your clinician finds abnormal reflexes, sensation, or muscle strength in a dermatomal pattern, a positive result meaningfully raises the likelihood of a pinched nerve in your neck, but a normal examination does not rule it out.
- For clinicians
- BNE components have high specificity for cervical radiculopathy (useful for confirming the diagnosis) but low sensitivity (a normal BNE cannot exclude CR); combining components further raises specificity but should be interpreted alongside the full clinical picture and valid reference tests.
- For readers
- This scoping review is the first to systematically map BNE accuracy for cervical radiculopathy and identifies critical gaps: standardized procedures, accepted diagnostic criteria, and a valid reference standard are all needed before firm clinical recommendations can be issued.
Source
doi:10.1186/s12891-025-08560-9
Read the original paperClinically assessing this area? See the neck & cervical spine special tests.
More Neck & Cervical Spine studies
- Consensus-based dosage recommendations for sensorimotor training in the management of neck pain: a Delphi studyConsensus
- Risk stratification scoring system for femoral neck bony stress injuries in military recruits: a pilot studyPrimary study
- Physical and psychological predictors for persistent and recurrent non-specific neck pain: a systematic reviewSystematic review
- Effect of median nerve neural mobilisation and cervical lateral glide on pain, disability and function in nerve-related neck and arm pain: a systematic review and meta-analysisMeta-analysis
- Cervical flexion posture during smartphone use was not a risk factor for neck pain, but low sleep quality and insufficient physical activity were: a longitudinal investigationCohort study
- Combining evidence and practice to optimise neck training aimed at reducing head acceleration eventsPrimary study