Abstract T P302: Analysis of Publicly Available CMS Data on Hospital Charges and Reimbursements for Stroke Care
Background/ Purpose: The Centers for Medicare & Medicaid Services recently made 2011 hospital charge (i.e., list price) and total payment data (CMS reimbursement data and co-payments by patient) publicly available for common Diagnosis-Related Groups, including cerebral infarction/intracranial hemorrhage, at www.cms.gov. We provide descriptive statistics of the data and begin to explore the association of these data with the quality of stroke care hospital’s provide.
Methods: We report the median, mean and extreme hospital charge and reimbursement data for cerebral infarction or intracranial hemorrhage without complications or comorbidity (CC) or major complications and comorbidities (MCC), with CC and with MCC. We report the correlation between charges and reimbursement as well as charge, reimbursement and hospital stroke volumes using a Spearman correlation coefficient. We also report median charge and reimbursement data by state for pooled stroke DRGs.
Results: Data were available for 5735 hospitals. The minimum, median, mean, and maximum charge data were as follows: without CC or MCC: $5392, $19976, $23593, $117831; with CC: $5223, $25151, $29492, $162923; with MCC: $9539, $40953, $48522, $234913. The minimum, median, mean, and maximum reimbursement data were as follows: without CC or MCC: $3916, $5326, $5714, $14744; with CC: $5369, $7280, $7922, $26510; with MCC: $8174, $12084, $13263, $50882. There was modest correlation between hospital charges and reimbursement (without CC or MCC ρ= 0.28; with CC ρ= 0.38, with MCC ρ= 0.46.) There was less correlation between discharge volume and charges or reimbursements (ρ =0.15 and 0.12 respectively). By state, pooled median charges ranged from $10,150 (Maryland) to $58,032 (New Jersey). Pooled median reimbursements ranged from $6,306 (Alabama) to $11,529 (Alaska).
Conclusions: The variability in the amount hospitals charge for stroke admissions is enormous and currently inexplicable. Much more research is needed to understand the reason for this variation and if there is any association to the quality of care provided; otherwise, this information may have unintended consequences for patients and providers.
Author Disclosures: R. Unia: None. A. Kelly: None. R. Holloway: Employment; Modest; Editor, Neurology Today. Consultant/Advisory Board; Modest; Milliman, Inc.
- © 2014 by American Heart Association, Inc.