OBJECTIVES : To evaluate the association between the PROMIS-Preference (PROPr) score and utilization in outpatient Neurology clinic patients in a large medical system.

METHODS : The seven PROMIS domains necessary to calculate the PROPr score have been routinely collected in Neurology practices across a large medical system since May 2018. We use the first scores collected from any patient between May and December 2018 to examine the association to inpatient stays and emergency department (ED) visits within 30 days of the PROPr data collection. Inpatient stays exclude inpatient rehab, inpatient hospice, and maternity admission. Number of inpatient stays were categorized as 0, 1, or 2+. ED visits exclude transfers, direct observation visits, and labor and delivery. Number of ED visits were categorized as 0, 1, or 2+.

RESULTS : We analyze 13084 unique patients from 16 outpatient clinics. The average age was 56 and 59% were female. There were 485 patients with 1 inpatient stay (4%) and 46 patients with 2+ inpatient stays (<1%). There were 995 patients with 1 ED visit (8%) and 287 patients with 2+ ED visits (2%). PROPr scores were statistically significantly different across number of visits for both inpatient and ED visits (p<0.01 by Kruskal Wallis tests). As number of visits increased, PROPr scores decreased. The average PROPr score for individuals without an inpatient stay was 0.35, with a single inpatient stay was 0.31, and with 2+ inpatient stays was 0.27. The average PROPr score for individuals without an ED visit was 0.36, with a single ED visit was 0.29, and with 2+ ED visits was 0.24.

CONCLUSIONS : We found large differences in PROPr scores between groups of patients with different healthcare utilization. This study provides evidence of construct validity for the PROPr score and evidence that PROPr scores may provide useful information in clinical analytics and identifying patients for targeted interventions.