Method
Electrodiagnostic data of 152 patients who had been diagnosed as lumbosacral radiculopathy with the findings of operative record were obtained retrospectively. The findings of needle electromyography were reviewed and the frequency of abnormal spontaneous activities in L5 and S1 myotomes was investigated. We selected 8 individual muscles which had high sampling rate. These muscles were combined into different muscle screens and the detection rates were calculated that the frequency with which one or more muscles in the screen displayed abnormal spontaneous activity was divided by the total number of radiculopathies.

Results
The detection rates of lumbosacral radiculopathy were compared according to the number of muscle screens. Including paraspinal muscle, the detection rate of 6 muscle screens was higher than 5 muscle screens (p＜0.05), but there was no significant difference of detection rate between 6 muscle screens and 7 muscle screens. The detection rates of each muscle screens without paraspinal muscle were lower than those including paraspinal muscle for all screens (p＜0.05).

Conclusion
Although there is controversy about selection of muscles, six muscle screen including paraspinal muscles may be optimal number for detecting lumbosacral radiculopathy. (J Korean Acad Rehab Med 2003; 27: 63-69)