Journal of General Internal Medicine | 2019
Use of PROMIS-29® in US Veterans: Diagnostic Concordance and Domain Comparisons with the General Population
Abstract
BackgroundPROMIS® items have not been widely or systematically used within the Veterans Health Administration (VA).ObjectiveTo examine the concordance of PROMIS-29® scores and medical record diagnosis in US Veterans and to compare Veteran scores relative to US population norms.Design/ParticipantsCross-sectional multi-site survey of Veterans (n\u2009=\u20093221) provided sociodemographic and PROMIS-29® domain data. Electronic medical records provided health condition (depression, anxiety, sleep disorders, pain disorders) diagnosis data.Main MeasuresFor each domain, we calculated PROMIS® standardized T scores and used t tests to compare PROMIS® scores for Veterans diagnosed with each targeted health condition vs. those without that documented clinical diagnosis and compare mean Veterans’ PROMIS-29® with US adult population norms.Key ResultsVeterans with (vs. without) a depression diagnosis reported significantly higher PROMIS® depression scores (60.3 vs. 49.6, p\u2009<\u2009.0001); those with an anxiety diagnosis (vs. without) reported higher average PROMIS® anxiety scores (62.7 vs. 50.9, p\u2009<\u2009.0001). Veterans with (vs. without) a pain disorder reported higher pain interference (65.3 vs. 57.7, p\u2009<\u2009.0001) and pain intensity (6.4 vs. 4.4, p\u2009<\u2009.0001). Veterans with (vs. without) a sleep disorder reported higher sleep disturbance (55.8 vs. 51.2, p\u2009<\u2009.0001) and fatigue (57.5 vs. 51.8, p\u2009<\u2009.0001) PROMIS® scores. Compared with the general population norms, Veterans scored worse across all PROMIS-29® domains.ConclusionsWe found that PROMIS-29® domains are selectively sensitive to expected differences between clinically-defined groups, suggesting their appropriateness as indicators of condition symptomology among Veterans. Notably, Veterans scored worse across all PROMIS-29(R) domains compared with population norms. Taken collectively, our findings suggest that PROMIS-29® may be a useful tool for VA providers to assess patient’s physical and mental health, and because PROMIS® items are normed to the general population, this offers a way to compare the health of Veterans with the adult population at large and identify disparate areas for intervention.