BACKGROUND/RATIONALE:
Current structural and functional changes within VHA largely parallel the widespread movement toward managed care that has occurred throughout the U.S. health care delivery system. The future of VHA as a viable alternative within that competitive environment requires knowledge concerning the relationship between managed care and cost within the new organizational structure.
OBJECTIVE(S):
This project has the major goals of (1) decomposing changes in costs into various contributing factors including managed care practices; (2) determining how much cost reduction is actually being saved through aggressive managed care practices of moving care from inpatient to outpatient settings; (3) evaluating the cost consequences of changes in treatment settings and (4) comparing the cost containment performance of VHA with that of the private sector under managed care.
METHODS:
This project develops patient level costs to be used in a multilevel cost analysis. Data on both individual patients and on facilities and networks are included so that the source of variation in cost can be attributed to differences occurring at each level as well to differences in specific explanatory variables. Individual patient health status is measured using the Diagnostic Cost Group (DCG) methodology that was developed for use within Medicare. A number of concerns in applying this methodology directly to the VA are addressed through a ‘recalibration’ of the models to better account for the type of services sought within the VA system and for different incentive structures facing VA providers.
FINDINGS/RESULTS:
Our annualized VA patient level cost costs differ considerably from the basic CDR cost in that they show much less variation within bedsection. This improves on the CDR method that allocates within cost distribution accounts exclusively by length of stay. Estimation of patient costs from 24 medical centers indicates that facilities more heavily penetrated by the managed care model may be more effective at controlling costs of their sicker patients in FY97. Examination of case- mix adjusted inpatient and outpatient workload and real adjusted unit costs reveals dramatic shifts in workload but does not show cost shifting at the national level of analysis. We have estimated risk and average costliness of an average VA patient compared to an average Medicare patient. We have in addition reconstructed Medicare risk models for VA experience and have thereby been able to allocate cost across diagnoses and to estimate risk internally for VA patients. Analyses of concurrent year risk show the VA user population to be highly stable. Relative to the Medicare population, VA risk appears to be .73 that of the Medicare population for FY98, FY99, and FY00.
IMPACT:
Knowledge of the structure of costs contained in this study is of major value to VHA budget decision makers. The originality in applying multilevel analytic techniques to hospital cost will be of great interest to health economists and health service researchers. Finally, the scope of DCG work is very important given the interest expressed in it by VA policy officials as well as its potential for adoption in the reform of VERA.
PUBLICATIONS:
Journal Articles
- Carey K, Burgess J. Hospital Costing: Experience from the VHA. Financial Accountability and Management. 2001; (16): 289-308.
- Carey K. A multilevel modelling approach to analysis of patient costs under managed care. Health Economics. 2000; 9(5): 435-46.
- Carey K, Burgess JF. On measuring the hospital cost/quality trade-off. Health Economics. 1999; 8(6): 509-20.
