Parsimonious collection of pain descriptors: classification and calibration by pain patients

Abstract

Single-word descriptors are commonly used to label and communicate pain in lay as well as clinical settings. Research has shown that the pool of 84 pain descriptors from the McGill Pain Questionnaire (MPQ) can be refined into a parsimonious subset of 36 descriptors that fit into 12 categories. However, the past 3 studies on this issue have been confined to college student samples. The present study investigated the classification structure and calibration of this new system of pain descriptors in 43 chronic pain patients. Employing a 3-point decision rule, a relatively unambiguous classification structure emerged with 3 descriptors for each of the 12 categories. Within and across categories, the intensities implied by these words could be meaningfully rank ordered. The intensities correlated positively and significantly with those previously derived from student samples as well as those of matching MPQ words previously rated by pain patients. This confirms the stability of the intensity ratings of pain words. Information theoretic analysis revealed transmission of 83% of the maximum (3.6 bits) potentially transmissible in a system of such configuration. This lends support to the idea that the 36 pain descriptors are parsimonious and can be used with efficiency to describe chronic pain.

Perspective: This study found that in the English language, 36 words (classified into 12 subcategories) can be efficiently used to describe pain. These words can also be reliably ordered in terms of implied pain intensity. This has implications for the qualitative and quantitative assessment of pain patients.