University of Utah | USA
Title: Comparing automated mental health screening to manual processes in a health care system
Biography: Josh Biber
Aim: The implementation of Patient Reported Outcomes (PROs) to screen for mental health conditions has traditionally used a manual process of paper and pencil with little standardization throughout a Healthcare System. Patients are typically screened when a provider identifies them as at risk; many patients are not screened. As we move forward into an era of health technology, we can leverage this capability to provide a standardized health outcomes assessment using PROs for mental health screening to all patients.
Methods: We currently administer the PROMIS Depression v1.0 instrument as part of a standard PRO assessment to all patients. PRO assessments are either completed at home or in clinic with a tablet computer. We customize the time between assessments to minimize patient burden, while maximizing relevant data for clinicians. The PHQ-9 is still used as a manual screening process within primary care and mental health. ICD-10 codes identified patients diagnosed with depression; anti-depressant prescriptions identified treated patients. Patients with a PROMIS score of ≥65 or PHQ-9 score of ≥15 were identified as at risk for depression.
Results: Since September 2016, 205,813 unique patients have been seen at the University of Utah Health System. Of the 33,484 (16.3%) patients screened using PROMIS Depression in an automated electronic process, 2,312 were identified as at risk for depression. Of the 6,039 (2.9%) patients screened using the PHQ-9 in a manual process (2.9%), 2,212 being identified as at risk for depression Compared to the manual process, the automated process identified more patients at risk for depression and not diagnosed (982 vs 402) and at risk for depression and not treated with medication (852 vs 226).
Conclusions: The use of an automated PRO platform for mental health screening enables a clinician to identify patients at risk that may have otherwise gone unidentified. Our automated platform screened more unique patients than a manual process in the same time frame. The implementation of the automated process is implemented in just under 70% of our total health system’s appointments. By implementing automated PROs, we hope to identify our total population of patients at risk for mental health conditions.