A45-03 Evaluation of Institutional Practices on Thoracentesis in Heart Failure Patients - Are We Tapping Too Much?

Abstract Rationale Pleural effusions affect about 1.5 million patients annually in the U.S., with heart failure (HF) accounting for over one-third of cases. While most transudate effusions improve with diuresis, thoracentesis is often performed for refractory or symptomatic cases, yet clear guidance on optimal timing remains limited. Furthermore, data on mortality impact and overall benefit of thoracentesis in HF are sparse. We sought to evaluate institutional practices to assess the appropriateness of thoracentesis in heart failure-related effusions. Methods We conducted a retrospective chart review of patients with HF who underwent thoracentesis at the University of Florida Jacksonville between January 2023 and January 2025. Only transudative effusions (TE) were included; cases with incomplete data were excluded. Effusions initially classified as exudative were further evaluated for pseudoexudative features using a serum-pleural albumin gradient 1.2 g/dL or a serum-pleural protein gradient 3.1 g/dL, and those meeting these criteria were reclassified as transudates. We assessed whether patients received a trial of diuresis—defined as 24 hours of diuretic therapy—prior to thoracentesis and stratified results by the clinical service performing the procedure. Results A total of 80 patients with HF and TE were identified. Of these, 26 patients (33%) received a trial of diuresis before thoracentesis, with the internal medicine procedure team performing most of these procedures (57%). The remaining 54 patients (67%) underwent thoracentesis prior to trial of diuresis, most commonly performed by interventional radiology (IR) (35%). In the latter group, the mean days to thoracentesis is 4.7 ± 3.3. Across both groups, only 12% of patients required repeat thoracentesis during the same admission. As IR functions primarily as a procedural service, we examined the source of their consults and found that 98% originated from the private hospitalist group. Conclusions Our analysis revealed that over half of thoracentesis performed at our institution occurred without an adequate trial of diuresis and within 24 hours of admission. This likely reflects considerable variability in physicians’ subjective assessment of severe dyspnea, timing of thoracentesis, and overall clinical decision-making. Although no formal guidelines discourage this practice, recent studies associate thoracentesis with increased mortality, complications, and length of stay, while a randomized trial showed no benefit from early intervention. Through evaluating our institutional practices, we aim to compare outcomes between patients undergoing early versus delayed thoracentesis. Ultimately, our goal is to identify opportunities to improve patient selection and timing of thoracentesis to enhance patient outcomes. This abstract is funded by: None