Consistency of Repeated Isokinetic Testing: Effect of Different Examiners, Sites, and Protocols

Article type: Research Article

Authors: Byl, Nancy N.^{a; *; **} | Wells, Laurie^{a; ***} | Grady, Deborah^{a; ****} | Friedlander, Anne^{a; *****} | Sadowsky, Steve^{a; ******}

Affiliations: [a] University of California, San Francisco

Note: [*] UCSF Graduate Program in Physical Therapy, School of Medicine, Box 0376, 374 Pamassus, San Francisco, CA 94122.

Note: [**] Address reprint requests to Nancy Byl, PhD, PT, Director, UCSF Graduate Program in Physical Therapy, Assistant Professor, Box 0376, 374 Pamassus Ave., San Francisco, CA 94122.

Note: [***] Department of Veterans Affairs Medical Center, Department of Physical Therapy, Room 117, 4150 Clement St., San Francisco, CA 94121.

Note: [****] Epidemiology and Biostatistics, Division of General Medicine, School of Medicine, Box 0866,74 New Montgomery #600, San Francisco, CA 94104.

Note: [*****] Department of Radiology, Box 0628, U368, School of Medicine, San Francisco, CA 94121.

Note: [******] UCSF Graduate Program in Physical Therapy, School of Medicine, Box 0376, 374 Pamassus, San Francisco, CA 94122.

Abstract: Normative data are needed to permit physical therapists to determine if patient measurements of strength are within normal limits. However, to generate normative data, it is essential to evaluate large numbers of subjects. This is only possible with collaboration to pool measurements across different sites with different examiners. This study evaluated the consistency of repeated isokinetic knee flexion and extension measurements with the same subjects tested on the Cybex 340 Extremity Unit with two different examiners at two different sites following the same protocol and then different protocols. The t test was used to determine if the mean difference scores significantly varied from zero (p = .05). Within the same site, following the same protocol, there were no significant differences in measurement by the two examiners for peak torque (at 90, 120, or 180 deg/sec), work in flexion (120 deg/sec) or endurance (120 deg/sec), but there was a significant difference in work extension. By site, there were no significant differences in repeated measurements for peak torque, work, or endurance (120 deg/sec). There were significant differences in peak torque flexion when different protocols were used, even when the same examiner was testing at the same site. The coefficients of variation for the repeated measurements varied from 4.5% to 13.6%. The results from this study suggest that physical therapists from different sites could collaborate to pool data for developing norms for isokinetic knee flexion and extension torque as long as the same test protocol was followed. However, the large variation in retest measurements and the coefficient of variation suggest that when a therapist wants to determine if a patient has made a significant gain in strength, the remeasurement should be taken by the same examiner at the same site. More research is needed to determine if the consistency of isokinetic torque testing with different examiners and different sites is equally high for other muscle groups and with other types of isokinetic dynamometers.

Keywords: Reliability, isokinetic testing, knee, Cybex 340

DOI: 10.3233/IES-1991-1302

Journal: Isokinetics and Exercise Science, vol. 1, no. 3, pp. 122-130, 1991