(SLC-015) Clinical and Economic Burden Associated with Prolonged Air Leaks Among Patients Undergoing Lobectomy: A Retrospective Database Analysis

Background: Prolonged air leak (PAL), typically defined as an air leak persisting longer than five days post-operatively, is one of the most serious complications following thoracic resection.

Objectives: To estimate the incremental economic and clinical burden of PAL among patients undergoing lobectomy using a large US healthcare database.

Methods: Retrospective analysis of hospital data from the US Premier Healthcare Database (PHD). Study subjects were age ≥18 years and had an inpatient hospital encounter for lobectomy between January 1, 2016 and December 1, 2022 (first of such admissions=index). The study sample was limited to patients for whom their institution continues to contribute data to the PHD for at least 90 days post discharge. PAL was defined by a diagnosis of post procedural air leak or postprocedural pneumothorax with length of stay exceeding 5 days post lobectomy procedure. Outcomes included, length of stay (LOS), total hospital costs during index, discharge status to home (vs skilled nursing/other), all-cause 30-, 60-, and 90-day readmission, and inpatient death. Generalized linear models were used to examine the association of PAL with the outcomes, accounting for hospital-level clustering and adjusting for patient demographics and clinical characteristics (e.g., age, sex, race, malignancy), procedural (e.g., admission type, surgical approach, lobe location of resection), and hospital/provider characteristics (e.g., hospital bed size, teaching status, procedural volume).

Results: A total of 47,383 subjects were included for study: 54.1% females, 83.0% white, mean (SD) age of 66.5 (10.5) years. Majority (88.4%) of patients had a malignancy surgical indication. Nearly two-thirds (64.6%) of lobectomies were performed in teaching hospitals and a third (33.5%) were performed with a robotic approach. Overall, 9.1% had evidence of PAL, which was associated with significant incremental increases in LOS (5.2 days, p< 0.001, 10.6 vs. 5.4 days) and incremental total hospital cost ($12,434, p< 0.001, $40,637 vs. $28,203). Having PAL significantly increased the risk of non-home discharge by 3.6% (from 7.6% to 11.3%, p< 0.001) and significantly increased the risk of being readmitted at 30 days by 3.4% (a 39.5% increase from 8.5% to 11.9%), at 60 days by 3.6% (a 33.9% increase from 10.7% to 14.3%), and at 90 days by 3.6% (a 29.1% increase from 12.3% to 15.9%), all p< 0.001. Although the absolute risk of death was low, it was 46% higher in patients with PAL versus those without PAL (1.8% vs. 1.2%, p=0.005).

Conclusions: This analysis demonstrates that PAL was associated with substantial patient clinical and economic burden among patients undergoing lobectomy, emphasizing a greater need for preventative measures to reduce the risk of PAL.