Weave Visualization of CMS Inpatient Hospital Price Data
The CMS MEDPAR inpatient data was explored using Weave, an open source web-based analysis and visualization platform providing high interactivity and performance. Several other public data sources were incorporated:
β’ CMS MEDPAR inpatients
β’ CMS Hospital Value-Based Purchasing measures
β’ U.S. Census American Community Survey
β’ CDC SMART survey
The data were linked in displays to explore their interconnectedness and complexity. Within diagnostic-related group, maps were created showing hospital average charges by hospital referral region (HRR). These spatial representations were connected to scatterplots of provider charge data vs. HRR average charge data and line graphs of the levels of potential indicator variables which included measures of hospital performance and patient satisfaction, demographic and economic variables, and disease prevalence. Additionally, individual provider level data was displayed in table format and linked to the scatterplots, enabling quick visualization of a hospitalβs relative pricing.
Taken together these visualizations help to explain the observed geographic variance in hospital charges and Medicare payments. For example, higher scores on the patient satisfaction survey correlate with lower charges. In contrast, higher per capita income correlates with higher charges.
The public can use this information to better understand the financial implications of provider and treatment selection, in particular by considering the charges in relation to their own insurance coverage provisions. Payers can use the information to assist with contracting decisions and negotiation of discounted pricing, and the formation of tiered networks based on hospital quality. The visualizations can be viewed in any browser for exploration, sharing, and decision support.