Pietz K (Houston Center for Quality of Care and Utilization Studies)
Petersen LA (Houston Center for Quality of Care and Utilization Studies)
Woodard L (Houston Center for Quality of Care and Utilization Studies)
Byrne MM (University of Miami)
Nelson HA (Houston Center for Quality of Care and Utilization Studies)
Kuebeler M (Houston Center for Quality of Care and Utilization Studies)
Sookanan S (Houston Center for Quality of Care and Utilization Studies)
To determine whether VAMCs that accept significant numbers of patients transferred from other VAMCs experience insufficient funding.
The study population was all inpatients treated at the 123 VAMCs during FY 2003. Patients were classified as transfers if they were discharged from one VAMC and admitted to another within the same network within 24 hours. For each VAMC, we calculated the number of patients transferred in, transferred out, and not transferred, as well as the total cost to treat these patients. Linear regression with no intercept at the VAMC level was used to calculate the average cost to treat patients of each type. We did the analysis on all 361,833 inpatients as well as a special group of 147,244 inpatients with a selection of serious medical conditions that each VAMC should be expected to treat regardless of the procedure technology. A Diagnostic Cost Group funding allocation method was used to generate allocations for all patients, taking into account the proportion of their care that occurred at each VAMC. We estimated the gain or loss to the VAMC for patients of each type.
The mean cost for all patients for those not transferred and for those transferred in were $26,138 and $140,106 respectively. For the special condition group the mean costs were $32,499 for those not transferred and $145,185 for those transferred in. The parameters for patients transferred out were not significant. The mean loss to the receiving VAMC per patient transferred in was $89,309 for all patients and $69,661 for patients in the special group.
Our results indicate that transfer patients require more costly care at the receiving VAMC. This occurs even when the analysis is restricted to serious conditions that every VAMC should be able to treat. A funding allocation system based on a diagnosis-based risk adjustment scheme may not provide adequate funding to medical centers that accept large numbers of transfer patients, even after the allocations are prorated for the proportion of care received at each VAMC.
When allocating funds to medical centers, the percentage of patients transferred in should be taken into account.