Discriminative and Predictive Analysis of the Brazilian Version of the Örebro Musculoskeletal Pain Screening Questionnaire (ÖMPSQ) Short-Form in Patients With Low Back Pain

Objective

The purpose of this study was to analyze the discriminative and predictive validity of the Brazilian Portuguese version of the Örebro Musculoskeletal Pain Screening Questionnaire Short-Form (ÖMPSQ-Short).

Methods

In this prospective longitudinal cohort study, the ÖMPSQ-Short was administered to 150 patients with acute, subacute, and chronic nonspecific low back pain. A 6-month follow-up was also completed to assess the predictive ability of the questionnaire for risk of pain, disability, and absence from work. The Roland-Morris disability questionnaire, pain numerical rating scale, and additional questions related to work and absence from work were also used.

Results

The ÖMPSQ-Short cutoff point was determined by analyzing the receiver operating characteristic curve, sensitivity, and specificity, which yielded a value of 44 points for the Brazilian population. Analysis of the 2 × 2 contingency tables showed that the ÖMPSQ-Short correctly classified high-risk patients with pain. Predictive capacity was assessed by univariate linear regression analysis. The ÖMPSQ-Short was able to predict pain, disability, and absence from work in patients with acute pain and disability in patients with subacute pain, and disability and absence from work in patients with chronic pain, but with a small predictive capacity.

Conclusion

The ÖMPSQ-Short cutoff demonstrated the ability to classify patients with low and high risk of involvement of psychosocial factors. However, due to its small predictive capacity, we suggest that the ÖMPSQ-Short should not be used to predict pain, disability, and absence from work over 6 months. However, we recommend using it as a guide to discuss the patient's clinical condition.