Background/Aim: Cervical radiculopathy is a clinical condition that occurs due to damage to the spinal nerve roots and manifests as pain, tingling and motor weakness of the arm muscles. Besides magnetic resonance imaging, electromyoneurography is one of the most significant
diagnostic methods in the evaluation of cervical radiculopathies. The dilemma for electromyographers is determining the number of muscles necessary to test in order to obtain a reliable electromyographic finding that
will detect nerve root damage. The aims of this study were to determine
the correlation between electromyographic findings and magnetic resonance imaging of the cervical spine and to examine the specificity and
sensitivity of electromyographic protocols that include different numbers of
muscles.
Methods: Forty subjects undergoing electromyographic examination and
magnetic resonance imaging of the cervical spine were analysed. Electromyography involved testing ten muscles, including the paraspinal muscles.
Muscle screens were made from the tested muscles, comprising 5, 6 and
7 muscles with paraspinal muscles. Correlation of positive radiological
findings with all muscle screens was performed and the specificity and
sensitivity of magnetic resonance imaging with all muscle screens were
determined.
Results: Optimal testing involved six muscles, including paraspinal muscles in the myotomal distribution defined by clinical presentation. Screen
6A-PS provided positive findings in 83 % of subjects. The sensitivity of
magnetic resonance imaging was 79.31 %, the highest in screen 6D-PS
and the same screen showed the highest specificity of 72.73 %.
Conclusion: For confirming the diagnosis of cervical radiculopathy, it is
optimal to perform an electromyographic examination of six muscles, including the paraspinal muscles. Electromyographic examination is a more
sensitive method in the evaluation of patients with cervical radiculopathy
compared to magnetic resonance imaging, as it detects a greater number
of relevant electrophysiological abnormalities even in patients without reliable morphological correlation of the lesion.
