COST MINIMIZATION OF AFTER HOURS POSTOPERATIVE CARE FOR TRANSPLANT RECIPIENTS
Candidate for the Lee B. Lusted Student Prize Competition
Method: A decision tree modeled care after hours care for recipients of kidney, pancreas and/or liver transplant to analyze a cost-minimization strategy. The model evaluated the cost of care through the emergency department compared with extending transplant clinic hours beyond normal business hours. The analysis was performed from the center (provider) perspective, using direct costs as the outcome of interest. The time course for the model was one day, so no discounting was included. Univariate and multivariate sensitivity analyses were performed to test uncertainty in the model.
A single center secondary database was used to obtain probabilities and cost data. The cost of operating clinic after hours was $2700, corresponding to a $1 million annual operating budget. The corresponding cost for the emergency department was $0, as the emergency department is open 24 hours/day per hospital policy and independent of transplant recipient utilization.
Result: After hours clinic care was the preferred option by $10,661. The difference between high risk and low risk costs in both clinic and the emergency department reflects the relatively high rates of complications seen in high risk transplant patients, as well as the clinical dilemma faced by clinicians in appropriately triaging and managing them. The results were robust to change according to the sensitivity analyses based on expensive additional cost of admission through the emergency department. The results were sensitive to an increased cost of extending clinical hours.
Conclusion: Maintaining an after-hours clinic is a cost minimizing way to extending care to transplant recipients. Major factors that drive costs appear to be relative proportion of high-risk patients presenting for after-hours care, operating costs of clinic and the ability to provide expedited care. Additional research in this topic area could include a cost-utility analysis, and multivariable analyses of the parameters used to populate the model.