Patient-level payment patterns prior to single level lumbar decompression are associated with resource utilization, postoperative payments, and adverse events.

Background: Understanding patient-specific trends in costs and healthcare resource utilization (HCRU) surrounding lumbar spine surgery is critically needed to better inform surgical decision making and the development of targeted interventions.

Purpose: 1) Identify subgroups of patients following distinct patterns in direct healthcare payments pre- and postoperatively, 2) determine whether these patterns are associated with patient and surgical factors, and 3) examine whether preoperative payment patterns are related to postoperative payments, healthcare resource utilization (HCRU), and adverse events.

Study Design/setting: Retrospective analysis of an administrative claims database (IBM Marketscan Research Databases 2007-2015).

Patient Sample: Adults undergoing primary single-level decompression surgery for lumbar stenosis (n=12,394).

Outcome Measures: Direct healthcare payments, HCRU payments (15 categories), 90-day complications and all-cause readmission, 2-year reoperation METHODS: Group-based trajectory modeling is an application of finite mixture modeling that is able to identify meaningful subgroups within a population that follow distinct developmental trajectories over time. We used this technique to identify subgroups of patients following distinct profiles in preoperative direct healthcare payments. A separate analysis was performed to identify distinct profiles in payments postoperatively. Patient and surgical factors associated with these payment profiles were assessed with multinomial logistic regression, and associations with adverse events were assessed with risk-adjusted multivariable logistic regression.

Results: We identified 4 preoperative patient payment subgroups following distinct profiles in payments: Pre-Low (5.8% of patients), Pre-Early-Rising (4.8%), Pre-Medium (26.1%), and Pre-High (63.3%). Postoperatively, 3 patient subgroups were identified: Post-Low (8.9%), Post-Medium (29.6%), and Post-High (61.4%). Patients following the higher-cost pre- and postoperative payment profiles were older, more likely female, and had a greater physical and mental comorbidity burden. With each successively higher preoperative payment profile, patients were increasingly likely to have high postoperative payments, use more HCRU (particularly high-cost services such as inpatient admissions, ER, and SNF/IRF care), and experience postoperative adverse events. Following risk adjustment for patient and surgical factors, patients following the Pre-High payment profile had 209.5 (95% CI: 144.2, 309.7; p<.001) fold greater odds for following the Post-High payment profile, 1.8 (1.3, 2.5; p=.003) fold greater odds for 90-day complications, and 1.7 (1.2, 2.6; p=.035) fold greater odds for 2-year reoperation relative to patients following the Pre-Low payment profile.

Conclusions: There are identifiable subgroups of patients who follow distinct profiles in direct healthcare payments surrounding lumbar decompression surgery. These payment profiles are related to patient age, sex, and physical and mental comorbidities. Notably, preoperative payment profiles may provide prognostic value, as they are associated with postoperative costs, HCRU, and adverse events.

Level Of Evidence: III.