N. W. Kugler1, P. Knechtges2, D. Milia1, T. W. Carver1, L. Goodman2, J. S. Paul1 1Medical College Of Wisconsin,Trauma And Critical Care,Milwaukee, WI, USA 2Medical College Of Wisconsin,Radiology,Milwaukee, WI, USA
Introduction: Hemothorax (HTx), pneumothorax (PTx), or both (HPTx) can be managed with tube thoracostomy (TT) in the majority of cases. Improperly placed tubes are common with rates near 30%. Management includes observation, repositioning, replacement, additional TT placement, or early surgical intervention. This study was performed to determine whether TT position affects the rate of secondary intervention.
Methods: Using the trauma registry a retrospective review of all adult trauma patients undergoing bedside TT placement over a 4-year period was performed. Staff radiologist classified the position of original TT as ideal, non-ideal, or kinked based on AP chest x-ray. Ideal TTs were apically directed, terminating in the lateral or mid thoracic cavity. Non-ideal TT was defined within the fissure or supra-diaphragmatic position. TTs with sentinel hole outside the thoracic cavity were excluded. The primary outcome was any secondary intervention (TT replacement, additional TT tube placement, or surgical intervention).
Results:486 adult trauma patients (547 hemithoraces) underwent TT placement and met inclusion criteria. Indications for placement were HPTx (37.2%), HTx (28.8 %), and PTx (34.0%). The majority of patients were male (76%), median age of 41 years (IQR 26-55 years), and blunt (67.9%) trauma. Ideal TT positioning in 429 (78.4%) and non-ideal in 118 (21.6%) hemithoraces. Secondary intervention rate was 27.8% including 109 (19.9%) additional / replaced TT, 31 (5.7%) VATS, and 12 (2.2%) thoracotomies. Rate of secondary intervention for ideal and non-ideal TT position was 25.1% and 37.3% (p=0.009) respectively. Kinked TTs were noted in 33 (6%) hemithoraces with a 45.5% secondary intervention rate. Due to likelihood for treatment bias, kinked TTs were removed from final analysis. Subsequently, the rate of secondary intervention was no longer significant (25.1% vs 34.1%, p=0.09).
Conclusion: Position of a non-kinked TT with the sentinel hole within the thoracic cavity does not affect secondary intervention rates, including the rate of surgical intervention. Inherent practice bias demonstrates with ideal position surgeons are significantly more likely to proceed with early operative intervention. Given over 20% of individuals with additional TT placement required operative intervention for definitive management, early operative intervention in the setting of non-kinked TT provides ideal patient care.