To simulate what episodes of care in spinal surgery might look like in a bundled payment system and to evaluate the associated costs and characteristics.Summary of Background Data.
Episode-based payment bundling has received considerable attention as a potential method to help curb the rise in health care spending and is being investigated as a new payment model as part of the Affordable Care Act. Although earlier studies investigated bundled payments in a number of surgical settings, very few focused on spine surgery, specifically.Methods.
We analyzed data from MarketScan. Patients were included in the study if they underwent cervical or lumbar spinal surgery during 2000–2009, had at least 2-year preoperative and 90-day postoperative follow-up data. Patients were grouped on the basis of their diagnosis-related group (DRG) and then tracked in simulated episodes-of-care/payment bundles that lasted for the duration of 30, 60, and 90 days after the discharge from the index-surgical hospitalization. The total cost associated with each episode-of-care duration was measured and characterized.Results.
A total of 196,918 patients met our inclusion criteria. Significant variation existed between DRGs, ranging from $11,180 (30-day bundle, DRG 491) to $107,642 (30-day bundle, DRG 456). There were significant cost variations within each individual DRG. Postdischarge care accounted for a relatively small portion of overall bundle costs (range, 4%–8% in 90-day bundles). Total bundle costs remained relatively flat as bundle-length increased (total average cost of 30-day bundle: $33,522 vs. $35,165 for 90-day bundle). Payments to hospitals accounted for the largest portion of bundle costs (76%).Conclusion.
There exists significant variation in total health care costs for patients who undergo spinal surgery, even within a given DRG. Better characterization of impacts of a bundled payment system in spine surgery is important for understanding the costs of index procedure hospital, physician services, and postoperative care on potential future health care policy decision making.Conclusion.
Level of Evidence: N/A