The refinement of relative resource weights for non-admitted patients
journal contributionposted on 05.06.2017, 06:48 by Jackson, Terri, Sevil, Petia
This study reports the development of cost weights for non-admitted patient services in Victorian hospitals based on patient-level cost data from six Victorian hospitals. The Victorian Ambulatory Classification System (VACS) developed by the authors in 1995 attributes related ancillary services to an index clinic visit. The base classification reported here has been adopted by the Victorian Department of Human Services for introduction of an activity-related variable payment system for public hospital outpatient services in 1997/98. Demonstration of the capacity of Victorian hospitals to generate patient and cost information to support a new payment system based on VACS was also an important research objective of the current study. A purposive sample of seven hospitals was recruited, with hospitals selected on the basis of well developed clinical costing systems; one of the seven was later excluded because data could not be provided. Two months of 1995 activity and related cost data were extracted from each hospital’s information system for analysis.Financial data and allocations of costs to non-admitted patient activities were reasonably consistent across hospitals, and represented accurate estimates of the costs of the services. Particular attention was paid to the problem of costing privatised clinics and ensuring that all costs were attributed to any encounter costed as part of the study. In contrast to the earlier study, assignment of hospital clinics was undertaken by an expert Clinical Panel, with reference to detailed descriptions of all outpatient activities provided by the hospitals. Recommendations are made to continue this process on an annual basis in order to give flexibility to the classification, and ensure the integrity of clinic assignment. Data were not available for four of the original 40 VACS categories, but three additional categories were recommended by the Clinical Panel, and a fourth by the Study Team. The final count of categories in this year's version of VACS would be 44 if data were available for all clinical specialties. Cost data from Allied Health and Emergency Departments were again judged to be too unreliable to support extensive analysis, and the Research Team supports the Department of Human Services' decision to continue block funding of these areas until more reliable data for weight setting is available. The Refinement of Relative Resource Weights for Non Admitted Patients The use of the 41 VACS categories for which data were available and recommended in the Interim Report of the project yielded a reduction in variance of 19.8%. This is low, but comparable to the results of other casemix models developed on the basis of individual level data. An alternative model using four additional subgroups, and collapsing two of the original VACS categories, yielded only 8% additional variance reduction, but may make the model more clinically meaningful. Detailed analysis of the components of the bundled encounter costs was undertaken to provide profiles of the cost of ancillary services for VACS categories. The average cost of pathology, imaging and pharmacy across all patients in each category is reported as well as the ‘conditional’ average. The latter takes into account only those encounters where the relevant ancillary cost is reported, eg the average cost of pathology for those diabetes encounters where a pathology test was undertaken. The study also investigated the implication for payment policy of multidisciplinary clinics. The issue touches on larger controversies about the appropriate role of public outpatient clinics which remain to be resolved. Structured discussions in four Melbourne teaching hospitals were organised to gain a better understanding of the way in which multidisciplinary clinics are organised, and the likely hospital responses to alternative payment policy approaches for these complex services. In order to guarantee fairness in payment between hospitals which structure multiple encounters and those which organise single multidisciplinary encounters, counting rules and audit procedures are proposed. Finally, the Commonwealth’s proposed ‘Developmental Ambulatory Classification System’ is compared with the VACS with particular emphasis on the extent to which new/review patient status improves the explanatory power of the VACS in accounting for total encounter cost. Systems for measurement of differences in staff time for these two categories of encounter type are still not well developed, and thus results must be interpreted with caution. Eighteen of the 31 VACS categories tested showed significant differences in total cost between new and review encounters. When the explanatory power of the classification model using this distinction was tested using PC-Group, however, the split yielded only very small increments in total variance explained. Benefits and problems of collecting patient-specific information and of expanding the classification system are discussed.