Premium
Measuring the Intensity of Chronic Pain: Are the Visual Analogue Scale and the Verbal Rating Scale Interchangeable?
Author(s) -
Kliger Mark,
Stahl Shy,
Haddad May,
Suzan Erica,
Adler Rivka,
Eisenberg Elon
Publication year - 2015
Publication title -
pain practice
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.899
H-Index - 58
eISSN - 1533-2500
pISSN - 1530-7085
DOI - 10.1111/papr.12216
Subject(s) - visual analogue scale , rank correlation , ordinal scale , intensity (physics) , statistics , medicine , spearman's rank correlation coefficient , mathematics , correlation , ordinal data , rating scale , audiology , physical therapy , physics , geometry , quantum mechanics
Objectives The 0 to 100 mm visual analogue scale ( VAS ) and the five‐category verbal rating scale ( VRS ) are commonly used for measuring pain intensity. An open question remains as to whether these scales can be used interchangeably to allow comparisons between intensities of pain in the clinical setting or increased statistical power in pain research. Methods Seven hundred and ninety‐six patients were requested to rate the present intensity of their chronic pain on the two scales. Spearman's rank correlation coefficients between VAS and VRS were calculated. For testing interchangeability, VAS was transformed into a discrete ordinal scale by dividing the entire VAS into five categories, either equidistantly (biased) or using frequency distributions of VAS (unbiased). We used G oodman– K ruskal's gamma and W ilson's e measures of ordinal association quantified the relationships between the transformed VAS and VRS scores and Svensson method to evaluate agreement between biased and unbiased discrete VAS and VRS scales. Results Average VAS and VRS scores were 76 ± 18 mm and “severe,” respectively. Spearman's rank correlation coefficient values between continuous VAS and VRS were 0.77 to 0.85. G oodman– K ruskal's gamma ordinal associations between discrete VAS and VRS were 0.82 to 0.92 and 0.90 to 0.98 for the biased and unbiased VAS , respectively. Wilson's e measures were 0.51 to 0.61 and 0.54 to 0.65, accordingly. Svensson analysis showed low probability of agreement between both biased (0.66 to 0.76) and unbiased (0.75 to 0.82) VAS and VRS . Discussion Regardless of the relatively high Spearman correlations between original VAS and VRS , the low ordinal association and low probability of agreement between discrete VAS and VRS suggest that they are not interchangeable. Therefore, VAS and VRS should not be used interchangeably in the clinical setting or for increased statistical power in pain research.