Numerical rating scale for dysmenorrhea-related pain: a clinimetric study

Objective: To evaluate the numerical rating scale (NRS) measurement properties in women with dysmenorrhea. Methods: This was an online clinimetric study. Brazilian women aged over 18 years old with internet access to respond to online instruments were included in the study. We evaluated criterion validity (comparing women with and without dysmenorrhea), construct validity between the NRS and the bodily pain domain of the SF-36, test-retest reliability, and measurement errors (in women with dysmenorrhea). Results: Two hundred thirty-eight women with and 192 without dysmenorrhea participated in the study. For criterion validity, the area under the receiver operating characteristic curve was 0.902 (95%CI, 0.873-0.931), and a cutoff point of 3 was considered to have the best sensitivity (83%) and specificity (86%). For construct validity, the NRS showed a moderate negative correlation with the SF-36 bodily pain domain (r=-0.46; p < 0.001). For test-retest reliability and measurement errors, 105 women whose symptoms did not change between 7 and 10 days of retest, with intraclass correlation coefficient = 0.90, standard error of measurement = 0.97, and smallest detectable change = 2.76 points. Conclusions: The NRS can be considered a valid and reliable patient-reported outcome measure for assessing dysmenorrhea-related pain intensity.