A New Approach for Safe Tube Thoracostomy Insertion: An Objective and Subjective Comparison versus Established Techniques

Abstract

Introduction: Tube thoracostomy is an invasive procedure, which may result in life-threatening injury to major organs and blood vessels. We propose a new approach for inserting tube thoracostomies to improve the safety of this procedure, termed guided blunt dissection. In this article, we compared the safety of this new approach with traditional blunt dissection and two commercially available Seldinger tube thoracostomy kits in an ex vivo model. Methods: We recruited 32 clinicians from a variety of medical specialties with a range of experience in performing tube thoracostomy. Each clinician was required to perform tube thoracostomy using all four approaches in a randomised order. Objectively, each insertion was categorised as “safe” if the lung remained intact and “unsafe” if the lung deflated. Subjectively, participants were asked to rank each approach in order of perceived safety on a four-point scale. Statistical analysis was performed using a Fisher’s exact test. Results: Objectively, guided blunt dissection was significantly safer than both Seldinger approaches ( p < 0.0001), but not traditional blunt dissection ( p = 0.71). Subjectively, none of the approaches were felt to be superior. Conclusions: These data support the conclusions that, in this ex vivo model, the new guided blunt dissection approach provided a safe method for tube thoracostomy. Guided blunt dissection produced less lung deflations relative to competing methods, certainly when compared objectively to Seldinger techniques. Of note, the Seldinger approaches were perceived by the participants to be as safe despite there being an increased incidence of lung injury associated with their use in this model. This indicates that it was not always possible for the clinician to determine when lung injury had occurred. This potential for lung injury when using Seldinger approaches for tube thoracostomy should be emphasised.