Methodological problems in the measurement of pain: A comparison between the verbal rating scale and the visual analogue scale

TLDR

The verbal rating scale (VRS) is widely used for pain assessment but has several disadvantages compared to the visual analogue scale (VAS). The study used a double‑blind, cross‑over design to compare analgesic effects on pain measured by VRS and VAS. The VRS produced higher statistical significance and appeared to overestimate analgesic effects, whereas the VAS more accurately reflected patient pain changes; the two scales were strongly correlated (r = 0.81) but the regression line and variance heterogeneity indicated systematic bias, suggesting the need for variance homogeneity checks and use of non‑parametric tests.

Abstract

The effect of analgesics on pathological pain in a double-blind, complete cross-over design was assessed by means of two rating scales, a verbal rating scale (VRS) and visual analogue scale (VAS). The VRS is widely used, but has several disadvantages as compared to the VAS. The results obtained by means of the VRS showed higher F-ratios (analysis of variance and Kruskall-Wallis H-test) than those obtained by means of the VAS. The VRS, which transfers a continous feeling into a digital system, seems to augment artificially the measurement of effects produced by analgesics, and the VAS seems to assess more closely what a patient actually experiences with respect to change in pain intensities. The correlation between the two scales was highly significant (r = 0.81, P < 0.001). The calculated regression line (y = −29.6 + 0.55 · x) was not similar to the line of identity and showed much lower values for the VAS, supporting our intepretation. The distribution of the variances of the values obtained by means of both scales was not homogenous. This indicates that the homogeneity of the distribution of variances should always be checked and a Kruskall-Wallis H-test used, if they are inhomogenously distributed.