Analysis of cost differences between comparing Percutaneous Transforaminal Endoscopic Discectomy (PTED) and Unilateral Biportal Endoscopy (UBE) in the treatment of lumbar disc herniation under Diagnosis-Related Group (DRG) payment: a retrospective cohort study
Background: Under the Diagnosis-Related Group (DRG) payment system, controlling medical costs while ensuring quality is essential. This study compares the economic differences between Percutaneous Transforaminal Endoscopic Discectomy (PTED) and Unilateral Biportal Endoscopy (UBE) for lumbar disc herniation (LDH), two minimally invasive techniques with similar efficacy but unclear cost profiles under DRG. Method: A retrospective cohort study was conducted using data from 364 LDH patients treated between 2019 and 2023. Propensity score matching was applied to balance baseline characteristics. Costs were adjusted using provincial Consumer Price Index (CPI). Clinical outcomes and cost components were compared, and predictive models (LASSO and Elastic Net regression) were used to identify factors influencing hospitalization costs. Result: No significant differences were found in clinical outcomes, including recurrence rates (PTED: 13.8% vs UBE: 12.9%, p = 0.806) and reoperation rates (PTED: 1.7% vs UBE: 0.8%, p = 0.446). However, total costs were significantly lower in the PTED group (median: 25,794 Yuan) compared to UBE (median: 33,247 Yuan, p < 0.001), with significant savings in treatment, anesthesia, surgery, medication, and consumables fees. Subgroup and regression analyses confirmed the cost advantage of PTED across different patient profiles. Discussion: PTED provides a more cost-effective option than UBE for LDH treatment under DRG payment, without compromising clinical outcomes. The economic benefit of PTED is consistent across various patient subgroups, supporting its use especially in resource-sensitive settings. These findings offer valuable insights for clinicians and policymakers aiming to optimize resource allocation and reduce financial burden within DRG frameworks.