Anna Tatakis, MD, Danielle Wilson, MD, Abdul Hafiz Al Tannir, MD, Jason D. Sciarretta, MD, Kelsey Cage, MD, John J. Kessler, II, MD, Rosemary A. Kozar, MD, PhD, Eric O. Yeates, MD, Andrew Medvecz, MD, MPH, Tammy Kopelman, MD, Andrea Slivinski, DNP, APRN, Karla Echeverria-Rosario, MD, Ioana B. Florea, MD, MPHS, Jane Cowan, MD, Abby Wong, MD, Daniel Stephens, MD, Mohamad Abou Chaar, MD, Kendal Weger, MD, Grace Chang, MD, Mark Flage, DO, Lilian Von Husen, BS, Allison E. Berndtson, MD, Jeffrey Anderson, MD, Paige Newman, BS, Brittany Nowak, MD, Brittany Smoot, Michael Farrell, MD, MS, Kristin Oller, DO, MS, Ananth Murthy, BS, Jayden Moore, BS, Sneha Swaminathan, MD, Christine Cocanour, MD, Elizabeth Kim, MD, Donald LaSeur, MD, Kelly Boyle, MD, Jacqueline Blank, MD, and Marc de Moya, MD, Milwaukee, WI
ABSTRACT 
BACKGROUND:
Traumatic hemothorax (HTX) contributes significantly to trauma morbidity and is frequently managed with tube thoracostomy (TT). However, TT placement has complications and some HTXs can be successfully observed. Optimal observation criteria remain unclear, and management practices vary. This multicenter study aimed to characterize current practices, determine observation failure rates, and identify predictors of safe observation.
METHODS:
We conducted a prospective observational study from July 2023 to June 2025 across 16 trauma centers. Adult patients with computed tomography-confirmed HTX were included. Exclusions included age under 18 years, TT before computed tomography, concurrent pneumothorax > 35 mm, death within 48 hours, or observation failure due to operative intervention. HTX volume was calculated using Mergo’s formula (V = d2×L). Each hemithorax was analyzed independently. The primary outcome was observation failure. HTX > 300 mL was evaluated as a predictor of failure in a multivariable logistic regression model.
RESULTS:
Among 962 HTXs in 932 patients, 68% (n = 657) were initially observed. The observation group had shorter hospital (7 vs. 9 d) and intensive care unit stays (1 vs. 3 d) compared with immediate TT (p < 0.001). Observation failure rate was 22% (n = 141), most commonly from HTX progression (54%). Failed observation patients had longer hospital stays (13 vs. 9 d, p < 0.001) but similar complication rates, secondary intervention needs, and 30-day outcomes compared with early TT. HTX volume > 300 mL strongly predicted observation failure (adjusted odds ratio, 16.01; 95% confidence interval, 8.25–31.06). Management practices varied: 52% received antibiotics with TT, 16% underwent irrigation, and thoracostomy tube sizes ranged from 8 to 36 Fr.
CONCLUSIONS:
Initial observation of HTX is frequently successful, with HTX volume > 300 mL being the strongest predictor of failure. Failed observation patients experience longer stays but similar outcomes to early TT placement. Notable practice variation highlights opportunities for research and standardization through unified practice management guidelines. (J Trauma Acute Care Surg. 2026;00: 00–00. Copyright © 2026 Wolters Kluwer Health, Inc. All rights reserved.)
LEVEL OF EVIDENCE:
Therapeutic/Care Management; Level III.
KEY WORDS:
Traumatic hemothorax, hemothorax observation, tube thoracostomy
