To the Editor:
It has previously been demonstrated that the visual analog scale (VAS) and the 11-point numerical rating scale (NRS-11) have similar properties.1,2 Craig T. Hartrick claims assessments on the VAS and the NRS-11 to be similar during activity-induced pain after hip arthroplasty, whereas the NRS-11 was found to be unsuitable for research purposes when rating mild to moderate static pain. In our view, the statistical analyses performed to support the conclusion do not seem to be appropriate, and we find the arguments hard to follow. It is stated that NRS-11 scores were not significantly different from VAS scores during activity, but this assertion is followed by a p value, which indicates a systematic difference between the two scales. The numbers (mean pain scores?) in the table supports a difference in activity, although perhaps not at rest, but, in the absence of standard deviations, the table is not very informative. The unexpected discrepancy between VAS and NRS-11 in patients with moderate pain at rest might be a result of a true difference or a few extreme values (possibly erroneous ratings) that have affected the mean score and thus invalidated the analysis. It would have been nice to see a scatter plot of the VAS and the NRS-11, with different markers for each VRS-4 score.
Even if we accept the conclusions made by Hartrick, comparison with the results of our study is not necessarily valid. There are several important differences in the two populations. First, patients undergoing hip replacement are older (range 60-80 years), with higher failure rates for VAS and NRS-11 scores3 than patients undergoing third molar surgery, who are young (range 18-40 years), otherwise healthy individuals without previous experience of chronic disease or intense pain. Second, patients undergoing hip arthroplasty are confined to a bed in the hospital, making assessment more difficult, whereas patients in pain after oral surgery procedures are mostly ambulant, with limited restrictions of movement of the jaw and thus not ideal patients for assessment of static pain. Studies have shown that the VAS is treated differently by different populations,4 and there is no reason to assume that the NRS-11 behaves differently in this regard.
Else Kristine Breivik, Ph.D.
Department of Oral Surgery and Oral Medicine; University of Oslo; Oslo, Norway
Eva Skovlund, Ph.D.
Professor; Section of Medical Statistics; University of Oslo; Oslo, Norway
1. Breivik EK, Björnsson GA, Skovlund E. A comparison of pain rating scales by sampling from clinical trial data. Clin J Pain
2. Janal MN. Concerning the homology of painful experiences and pain descriptors: a multidimensional scaling analysis. Pain
3. Kremer E, Atkinson JH, Ignelzi RJ. Measurement of pain: patients preference does not counfound pain measurement. Pain
4. Williams AC deC, Davies HTO, Chadury Y. Simple pain rating scales hide complex idiosyncratic meanings. Pain