By Michael H. Crawford, MD
Synopsis: An unblinded, multicentered, randomized trial of patients with acute heart failure hospital admissions and significant pleural effusions showed that early thoracentesis, in addition to recommended medical therapy, did not reduce mortality or length of stay, or increase days alive out of the hospital for 90 days. However, it was relatively safe and could be employed in selected persistently symptomatic patients with very large effusions.
Source: Glargaard S, Thomsen JH, Tuxen C, et al. A randomized controlled trial of thoracentesis in acute heart failure. Circulation. 2025;151(16):1150-1161.
Although pleural effusions are common in patients with acute decompensated heart failure, there is no recommendation in the guidelines concerning when to deploy thoracentesis, probably because there are no randomized trials of this strategy. Thus, these investigators from Denmark conducted a pragmatic, unblinded, multicentered, randomized controlled trial at 10 hospitals with heart failure specialists.
The study population included adults with a left ventricular ejection fraction (LVEF) of ≤ 45% hospitalized for acute heart failure with a non-negligible pleural effusion in whom thoracentesis was deemed feasible. Excluded were patients with pulmonary infections, severe aortic stenosis, estimated glomerular filtration rate (eGFR) < 15 mL/m or on dialysis, and a recent thoracic procedure. Control patients received guideline-directed medical therapy (GDMT). Patients randomized to thoracentesis had ultrasound-guided pig-tail catheter placement and fluid drainage in addition to GDMT.
At 14 and 90 days post-hospital discharge, patients were contacted by phone and administered the Kansas City Cardiomyopathy Questionnaire (KCCQ). There were no protocol-driven follow-up visits. Clinical outcomes were sourced from the Danish National Health Databases. The primary outcome was the number of days out of the hospital up to 90 days. Secondary outcomes included hospital length of stay (LOS), all-cause mortality, and time to second hospitalization or death. Safety outcomes were the common complications of thoracentesis and thromboembolic events in patients who had their anticoagulation therapy interrupted for thoracentesis.
Between 2021 and 2023, 135 patients were recruited (median age 81 years, 33% women, mean EF 25%). Pleural effusions were bilateral in 73% of patients and 49% of patients were taking anticoagulants. The median time to thoracentesis was 22 hours; 30 hours if anticoagulants needed to be withheld. The primary outcome was 84 vs. 82 days (P = NS). LOS was five days for both groups. Hospital mortality was 3% and 90-day mortality was 13% for both groups. There was no difference in GDMT or diuretic use at hospital discharge.
The thoracentesis group pneumothorax rate was 5%. Twenty-five percent of patients experienced mild chest discomfort, but there were no major complications. Also, there was no difference in the KCCQ scores between the two groups at 14 or 90 days post-discharge. The authors concluded that in patients admitted for acute heart failure with significant pleural effusions, a strategy of routine thoracentesis in addition to GDMT did not reduce all-cause mortality or duration of the index admission, or increase days alive out of the hospital for 90 days.
Commentary
The results of this study are similar to those of previous observational studies. There are no prior randomized or controlled trials. In this Danish study, there was a very low incidence of pneumothorax and there were no other major complications. Pneumothorax was readily handled; there was no difference in LOS or mortality. One-quarter of the thoracentesis patients had minor complications, mainly mild chest discomfort during and after the procedure.
Considering this was an older patient population with considerable comorbidities and frailty, thoracentesis was remarkably safe in this population. However, there was no reason to subject these patients to this small risk and minor discomfort for no overall gain in outcomes. Thus, thoracentesis should be reserved for those with very large effusions in whom dyspnea cannot be rapidly improved by diuresis and oxygen administration, and in whom there are no contraindications.
There are limitations to this study. It was unblinded, which may have influenced the physiciansʼ decisions on discharge and readmission. Also, patient-reported outcomes could have been influenced by which group they were in. Also, quality of life results (KCCQ) could have been affected by survivor bias. In addition, there were no data about the completeness of pleural fluid drainage.
About half of the patients were taking anticoagulants. Their thoracentesis was delayed for an average of eight hours to improve safety, but this delay could have diminished the observed benefit of thoracentesis since GDMT was being administered during this time. Finally, the number of patients was too small for subgroup analyses. In summary, this was a well-conducted, pragmatic randomized controlled trial that could help inform future iterations of acute heart failure management guidelines regarding thoracentesis.
Michael H. Crawford, MD, is Professor Emeritus of Medicine and Consulting Cardiologist, UCSF Health, San Francisco.
