Interpretation of PROMIS Depression and Anxiety Measures Compared with DSM-5 Diagnostic Criteria in Musculoskeletal Patients

Abstract

Background: There is growing awareness among orthopaedic clinicians that mental health directly impacts clinical musculoskeletal outcomes. The Patient-Reported Outcomes Measurement Information System (PROMIS) is increasingly used for mental health screening in this context, but proper interpretation of patient scores remains unclear. The purpose of the present study was to compare musculoskeletal patients’ PROMIS Depression and Anxiety scores with a board-certified clinical psychologist’s assessment of their depression and/or anxiety diagnoses, as defined by Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria. Methods: In this cross-sectional analysis, existing medical records were reviewed for 50 patients who presented to an interdisciplinary program within a tertiary care orthopaedic department for the treatment of ≥1 musculoskeletal condition. All patients completed PROMIS Depression and Anxiety measures and were evaluated by a board-certified clinical psychologist. Receiver operating characteristic (ROC) curve analyses were performed to assess the diagnostic accuracy of PROMIS Depression and Anxiety scores as compared with the psychologist’s diagnosis of a DSM-5 depressive or anxiety disorder. Results: Twenty-eight patients (56%) were diagnosed by the psychologist with a DSM-5 depressive disorder, and 15 (30%) were diagnosed with a DSM-5 anxiety disorder. The ROC analysis for PROMIS Depression had an area under the curve (AUC) of 0.82. The optimal score cutoff to predict a diagnosis of a DSM-5 depressive disorder was ≥53 (sensitivity, 79% [95% CI, 63% to 94%]; specificity, 86% [72% to 100%]; positive predictive value [PPV], 88% [75% to 100%]; negative predictive value [NPV], 76% [59% to 93%]). The ROC analysis for PROMIS Anxiety had an AUC of 0.67. The optimal score cutoff to predict a diagnosis of a DSM-5 anxiety disorder was ≥59 (sensitivity, 60% [95% CI, 35% to 85%]; specificity, 74% [60% to 89%]; PPV, 50% [27% to 73%]; and NPV, 81% [68% to 95%]). Conclusions: Modestly elevated PROMIS Depression scores were suggestive of the presence of a DSM-5 depressive disorder, whereas elevations in PROMIS Anxiety scores seemed to have less association with DSM-5 anxiety disorders. Nevertheless, neither PROMIS measure demonstrated adequate discriminant ability to definitively identify patients who met DSM-5 criteria. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.