ASSESSING TRENDS IN PHARMACY COST IN SUPER-UTILIZERS AND HIGH-COST PATIENTS IN THE U.S. FROM 1996 THROUGH 2013: EVIDENCE FROM MEDICAL EXPENDITURE PANEL SURVEY DATA

OBJECTIVES: To analyze trends in pharmacy costs among super-utilizers and very high-cost non-institutionalized U.S. patients. METHODS: This was a retrospective analysis using Medical Expenditure Panel Surveys (MEPS) administered by Agency for Healthcare Research and Quality (AHRQ). MEPS data from 1996 through 2013 were used to construct the top 1% of the costliest patient cohorts. Super-utilizers were denned according to AHRQ definition: Medicare or Medicaid patients with four or more hospitalizations per year; or, privately insured patients with three or more hospital admissions per year. Healthcare costs were in ated to 2013 prices. Multivariate quantile regression was used to estimate pharmacy cost for the top 1%, and generalized linear modeling (GLM) was used to estimate mean pharmacy cost for super-utilizers. RESULTS: The share of top 1% of the costliest patients steadily grew from 073% of the U.S. population to 1.35% ( 4.26 million) during the study period. The average age for these patients varied between 51 and 56 during the study period. Mean pharmacy cost rose from $2,275 in 1996 to a high of $16,831 in 2011, then declined to $11,962 in 2013. Mean out-of-pocket pharmacy cost during this period rose from $676 to a high of $2,122 in 2004, and then declined to $743 in 2013. For super-utilizers, mean pharmacy cost rose from $2,618 to a high of $6,901 in 2007 and then declined to $5279 in 2013. Mean out-of-pocket pharmacy cost for super-utilizers started with $930 in 1996, peaking to $1,720 in 2003 then declined to $677 in 2013. Both quantile regression of pharmacy costs for the top 1% and the GLM regression for super-utilizers confirm these descriptive findings. CONCLUSIONS: Mean pharmacy cost increased by 425% for the top 1% of the high-cost patients in the U.S. while it rose by 102% for the super-utilizers during the study period.