Abstract
Introduction
Home sleep apnea tests (HSATs) are convenient alternatives to in-lab polysomnograms (PSGs) but high non-diagnostic rates limit their utility. A clinical decision support tool (CDST) to triage patients to HSAT versus PSG was developed at the Greater Los Angeles VA Healthcare System (GLA-VAHS). It uses a random forest ensemble to reduce non-diagnostic HSAT rates by 46%. While prior studies have found PSGs to be more profitable than HSATs on a per unit basis, these analyses do not factor in relative profitability over time. Additionally, no prior studies have quantified the financial impact of a CDST in diagnostic sleep testing.
Methods
We performed an analysis of the overall profitability of HSATs and PSGs in 2018-2019 within GLA-VAHS which has 6 PSG beds. Revenue was calculated using 2019 Medicare reimbursement rates. Contribution margin (CM) analysis was used to factor out the high fixed costs of healthcare infrastructure, instead focusing on variable direct costs (VDCs). CM analysis is especially useful when calculated on a per diem basis instead of per study, adjusting for number of tests performed in a given day. CM was calculated by subtracting VDCs from revenue under two simulated conditions: with and without the CDST.
Results
PSGs were 2.5 times more profitable than HSATs on a per unit basis (CM $200/study vs. $81/study). However, on a per day basis, PSGs were only 1.4 times more profitable than HSATs at average nightly occupancy rates of 75% (CM $902/day vs. $646/day). Using the CDST to guide testing, 2.2 times more diagnostic HSATs could be performed per day. As a result, HSATs were 1.3 times more profitable than PSGs on a per day basis with CDST use (CM $1,211/day vs. $902/day).
Conclusion
This analysis demonstrates that implementing a CDST and maximizing utilization of HSATs allow hospitals to better allocate limited sleep lab resources, increase diagnostic throughput and generate higher profits. Analyzing costs using contribution margin avoids erroneous assumptions about profitability and leads to better-informed administrative decisions regarding sleep lab expansion.
Support