Background:
Given the rising costs of medical care in the US, efforts at improving quality of care seek also to control expenditures with high-value healthcare - better quality and lower cost. Prior studies have used variations in Medicare reimbursements at the regional level, rather than individual hospital costs, to account for the unique patient case-mix of individual hospitals. To date, higher expenditures have been unrelated to clinical outcomes. We sought to explore variation using costs, rather than payment, and to study the links between inpatient hospital Medicare expenditures, hospital costs, and Medicare's quality indicators.
Methods:
In New York State, we analyzed hospital discharges in 2008 for acute myocardial infarction (AMI) in the New York State Planning and Resource Cooperative System (SPARCS) database. Using SPARCS reported hospital charges and the Medicare cost-to-charge ratio we calculated, at the individual hospital level, the total costs and departmental-level ancillary costs (costs of specific hospital services) for patients discharged with a primary diagnosis of AMI. We compared total inpatient cost to Dartmouth Atlas of Health Care Medicare inpatient reimbursements, analyzed variation in ancillary costs by hospital cardiac program peer groups (based on highest-level capability), and related total costs to Medicare AMI quality scores.
Results:
The study analyzed data that represented 56,000 AMI cases in 150 hospitals. When comparing total costs to Medicare inpatient reimbursements, there was high correlation with an R-squared of 0.84. The mean ancillary costs per AMI discharge were $12,006 [standard deviation (SD) $4,301, coefficient of variation (COV) 0.358] for cardiac surgery-capable, $6,452 [SD $1,732, COV 0.268] for percutaneous coronary intervention-capable, $5,104 [SD $2,548, COV 0.499] for diagnostic catheterization-capable, and $4,167 [SD $2,756, COV 0.661] for non-invasive hospitals. There was no overall correlation between AMI total costs and Medicare composite AMI quality scores, yet 19.7% (28 out of 142 hospitals with AMI quality score reporting in 2008) of hospitals had above average quality scores, with lower than average total costs.
Conclusions:
In AMI hospitalizations, regional Medicare reimbursements have a strong correlation with total hospital costs. Hospitals vary greatly in their costs of caring for AMI patients. Though there is no overall association between the level of costs and performance on standardized quality scores, high-value healthcare in AMI does exist with numerous high-quality, low-cost hospitals. Examining the expenditures of these hospitals may be useful in identifying patterns of quality and cost-effective patient care. These findings support that high-spending is not necessary for high-quality, and that high-value healthcare can be realized.
Declines in Hospital Admissions for Acute Myocardial Infarction in New York State After Implementation of a Comprehensive Smoking Ban
