Intra-operator and inter-operator reliability of surface electromyography in the clinical evaluation of back muscles.

As a prerequisite to the use of a test battery based on electromyographic (EMG) analysis of the paraspinal muscles for identifying and remedying back muscle dysfunction, the intra- and inter-operator reliability was assessed. Fifteen volunteers underwent EMG tests on three occasions. The test subjects were asked to perform 22 exercises, subdivided into four categories: coordination, stabilization, balance and strength exercises. The time interval between the tests was one week. The myoelectric signals of the multifidus (MF) and iliocostalis lumborum pars thoracis (ICLT) were analysed with regard to amplitude (averaged EMG) and frequency (zero cross rate). The results indicated that the reliability was better for the MF than for the ICLT, and also for exercises at higher loads (strength exercises). In the intra-operator condition, the reproducibility of the averaged EMG was good (ICC>0.75), except for the balance exercises (ICC = 0.40-0.74). In general, the averaged EMG in the inter-operator condition and the zero cross rate in both the intra- and inter-operator conditions are less or poorly reliable. These results demonstrate that when back muscle function is evaluated during coordination, stabilization and strength exercises, only the averaged EMG parameter has acceptable reproducibility over time when assessed by the same operator.