Diagnostic operating characteristics of PROMIS scales in screening for depression

Abstract

ObjectiveTo determine the diagnostic operating characteristics of The Patient Reported Outcomes Measurement Information Systems (PROMIS) depression scales in screening for major depression. MethodsInterview data from patients enrolled in clinical trials involving patients with chronic pain (2 trials) or post-stroke (1 trial) were analyzed. This included baseline and follow-up interviews in 648 and 586 study patients, respectively. Patients completed PROMIS depression scales of varying lengths (4-item, 6-item, and two 8-item scale versions) as well as the Patient Health Questionnaire 9-item (PHQ-9) depression scale. A Structured Clinical Interview for DSM Disorders (SCID) was administered to establish diagnoses of major depression and any depression. Sensitivity and specificity at various score cutpoints as well as area under the curve (AUC) were calculated. ResultsPROMIS scales of varying lengths had similar diagnostic operating characteristics. The optimal screening cutpoint for PROMIS raw scores on the 4-, 6-, and 8-item scales was 8, 12, and 16, respectively, which corresponds to a PROMIS T-score of 55. The average sensitivity and specificity of the two PROMIS 8-item scales for major depression across the 3 trials using a T-score cutpoint ≥55 was 0.81 and 0.84, respectively. This was almost identical to 0.81 and 0.82 for the PHQ-9 at its standard cutpoint score ≥ 10. The average AUC for major depression was identical (0.91) for the two PROMIS 8-item scales and PHQ-9, and also similar for any depression (0.87 to 0.89). ConclusionPROMIS scales ranging from 4 to 8 items have strong operating characteristics comparable to the PHQ-9 in screening for depression.Trial registration: ClinicalTrials.gov ID: NCT01236521, NCT01583985, NCT01507688