Abstract
Background
Pain is often measured by asking people to rate their pain intensity at its worst, on average, and at its least, for the last 7 days using numeric rating scales. The three ratings are summed to produce a composite measure. The validity and reliability of this composite measure has not been examined using modern psychometric methods in any population. We examined the validity and reliability of this pain intensity measure for use with professional musicians, university music and science students, and university staff, all of whom had reported experiencing musculoskeletal symptoms in the last 7 days.
Methods
Data were collected using a questionnaire survey. The validity and reliability of the composite pain measure were examined using Rasch analysis. Differential item functioning was examined for age, gender, student status, musician status, and socioeconomic status.
Results
While the data fit one of the Rasch models, after several response categories were collapsed, differential item functioning was present. There was no solution found that fit one of the Rasch models, without differential item functioning.
Conclusions
Despite the recommendation for the three number ratings scales for pain to be combined, using Rasch analysis showed that this was not a valid approach for our study population. Our findings highlight the importance of using Rasch analysis to examine the utility of measures.
Key messages
Rasch analysis is a useful method for investigating the validity and reliability of scales. Combining pain ratings cannot be assumed to produce a valid and reliable measure.
775Should pain scales be combined
