Abstract
INTRODUCTION AND OBJECTIVE: Legislators and regulatory bodies have indicated a desire to accelerate the transition to risk-based payment paradigms. To succeed in these models, providers must first measure and standardize utilization; such attempts have historically met with limited success. We sought to determine if prospective behavior-shaping tools implemented in a single large urology practice (Integrated Medical Professionals, PLLC; IMP) was effective in reducing variability and improving accuracy in evaluation and management (E/M) codes. METHODS: After educating IMP providers in billing accuracy from 2013-15, CPT codes for new (99201-5) and established (99211-5) E/M services for Medicare beneficiaries seen at IMP were compared to those urologists practicing nationally (US) as well as those practicing in groups of 10 or more (G10), using data from CMS Medicare Public Use Files for 2016. In addition to mean code level, coding variability for each CPT code was determined for each group. Coding inaccuracy was assessed determining frequency of a cluster coding (provider use of single E/M code >70% of the time) overall and for level 4/5 codes in aggregate. Statistical analysis was performed using Student’s pooled t-test. Variability in coding patterns was measured using f-test, and cluster coding was compared using two proportion z-test. RESULTS: We reviewed 1,032,623 new and 7,045,239 established E/M visits billed to CMS in 2016 by a total of 8,651 urologists nationally. Results are presented in Table 1. When compared to both G10 and US for new and established E/M visits, IMP mean code levels were both significantly shifted left and more uniform (p=0.00 and f=0.00, respectively for both visit types). Overall cluster coding rates for both new and established E/M codes were significantly lower for IMP when compared to both G10 and US (z=0.00 all categories); these differences were even more marked for level 4/5 new and established E/M codes (z=0.00 all categories). CONCLUSIONS: Institution of provider education and standardized review both improved accuracy and reduced variability in E/M coding. Institution of such programs may be an important stepping stone to developing value-based care models.Source of Funding: none
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have