|
BACKGROUND/RATIONALE:
VA annually provides $351 million of specialized treatment for substance use disorders to some 139,000 patients. This treatment may be at least as cost-effective as other types of health care, but research is limited by the difficulty in finding treatment cost. Improved methods of determining costs, with information on outcomes, can help guide decisions about the types and location of care that should be offered.
OBJECTIVE(S):
This project will improve methods of determining the cost of specialized treatment of substance use disorders, and help three QUERI projects evaluate the economic effect of practices that are thought to enhance treatment effectiveness.
METHODS:
This study will combine prospectively gathered primary data from three QUERI projects with retrospective analysis of VA databases. It will evaluate care received by 3,250 individuals enrolled in 92 programs. QUERI data will include surveys of program managers and staff, as well as outcomes and other data on a sample of newly enrolled patients. VA cost and utilization data will be obtained from the VA Austin Automation Center, including two new sources: the department-level and encounter-level extracts of Decision Support System (DSS). In addition, a special extract of intermediate product detail will be obtained from DSS. Methods will be improved by analyzing survey data, by linking surveys to databases with new data fields, and by evaluating DSS data. Alternative sources of VA cost and utilization data will be compared to each other and to private-sector reimbursement rates. Cost-effectiveness of care will be improved by evaluating the cost of factors that are believed to affect the quality of care. These include compliance with methadone treatment guidelines, continuity-of-care practices, staff training, program size, and the duration, quantity, and type of treatment sessions. QUERI data provide a unique opportunity to understand how factors that determine cost also affect patient outcomes, controlling for case-mix.
FINDINGS/RESULTS:
We compared VA costs reported in DSS to reimbursements rates for psychiatric care from three private insurance databases and Medicare. VA costs were much higher than private reimbursement rates, in part because VA care is provided in the hospital setting. Medicare payments, with facility fees included, were comparable to the costs reported in DSS. The mix of VA services, as recorded by CPT codes, is quite different from that recorded in private insurance datasets.
Our evaluation of the outpatient visits of a random sample of patients enrolled in the MOST study revealed that the National Patient Care Database captures the vast majority of services recorded in VISTA. We found that VA methadone treatment programs do not use the new HCPCS code for methadone dispensing, and instead use procedure codes for services that are more expensive to provide than methadone. Cost estimates based on these codes, such as the HERC average cost data set, overstate the cost of this care. We learned that some VA methadone programs record psychosocial services using clinic stop codes other than the code for methadone care. Complete characterization of services from these programs must include visits to other psychiatric and substance abuse clinic stops.
We used micro-cost methods to estimate the cost of treatment provided by programs studied by QUERI projects. We combined information on the number of different types of staff from program manager survey with salary estimates from FMS and workload data from the VA administrative data. We learned that it is not feasible to use the provider identification field or the DSS intermediate product department to identify a specific outpatient substance use treatment program in VA datasets. The most feasible method to characterize program workload is to identify services received by a sample of patients who have identified as program participants from a separate data source. We then compared DSS and micro-cost estimates of the direct staff cost per substance use treatment visit. Most DSS cost data was consistent with the micro-cost estimates.
IMPACT:
This project will help identify the cost of VA substance abuse treatment programs, and determine the economic consequences of three specific initiatives to improve the quality of care.
PUBLICATIONS:
None at this time.
DRA:
Health Services and Systems
DRE:
none
Keywords:
Cost, Cost effectiveness, Decision support
MeSH Terms:
none
|
