Intra-rater and inter-rater reliability of the Penn Spasm Frequency Scale in People with chronic traumatic spinal cord injury

Abstract

Study design: Psychometric study. Objectives: Evaluate the intra-rater and inter-rater reliability of the Penn Spasm Frequency Scale (PSFS) in people with chronic traumatic spinal cord injury (SCI). Setting: Community setting. Methods: Participants with chronic traumatic SCI were administered the PSFS, a self-reported spasticity measure in which the individual documents the frequency (Part 1) and severity (Part 2) of their muscle spasms. One rater administered the PSFS at baseline, 5 to 10 days, and 4 to 6 weeks after baseline. The second rater administered the PSFS within 3 days of the first rater. Results: Reliability was assessed using weighted-kappa (95% confidence interval) statistics. In 61 participants, the intra-rater reliability between 5 to 10 days and 4 to 6 weeks after baseline was 0.822 (0.709, 0.935) and 0.734 (0.586, 0.883), respectively, for PSFS Part 1. With the addition of Part 2, the intra-rater reliabilities were 0.812 (0.705, 0.919) and 0.729 (0.586, 0.872) for 5 to 10 days and 4 to 6 weeks, respectively. The PSFS inter-rater reliability within a 3-day time interval was 0.862 (0.759, 0.965) for Part 1 and 0.857 (0.762, 0.952) with the addition of Part 2. Conclusions: PSFS Part 1 intra-rater and inter-rater reliability weighted-kappa statistics are in the "almost perfect" category, with lower confidence bounds in the "substantial" range. By adding Part 2 the reliability decreases, but the kappa statistics remain "substantial," with a lower confidence bound in the "moderate" range. The PSFS is probably a reliable outcome measure to assess self-reported spasticity after chronic traumatic SCI.