L. E. Coleman1, C. M. Theodorou1, D. Anaya1, S. Mahdavi1, S. Zakaluzny1, C. S. Cocanour1 1University Of California – Davis,General Surgery,Sacramento, CA, USA
Introduction: Trauma patients undergoing damage control laparotomy are frequently left with an open abdomen (OA). While often physiologically necessary, there are many potential complications with OA, including formation of enterocutaneous fistula and intra-abdominal infection. This study focuses on factors predictive of intra-abdominal abscess (IAA) development, specifically with regards to both fixed patient factors and modifiable operative factors.
Methods: We performed a retrospective review of adult trauma patients admitted to a level 1 trauma center June 2016 – June 2019 who were initially managed with OA following damage control laparotomy. Patients who survived at least 7 days after surgery, allowing time for the development of IAA, were included. Univariate analysis of comorbidities, injuries, and surgical decision-making was performed. Subgroup analysis was performed on all patients who achieved fascial closure within three operations. Univariate factors approaching significance (p <0.20) were included in multivariate logistic regression. Six-month follow-up data was collected on development of enterocutaneous fistula (ECF) and ventral hernia.
Results: Ninety-one patients were included (79% male), with a median age of 40 years old (IQR 27-57). Twenty-four patients (26%) developed an IAA. All examined comorbidities had no significant association with IAA. The time to fascial closure was not significantly different between those who developed IAA and those who did not (50 vs. 45 hours, p= 0.42). Furthermore, receiving antibiotics throughout the entirety of having an open abdomen was not associated with IAA (58% vs. 61%, p= 0.8). Factors significant for IAA were abdominal AIS score (4 vs. 3, p= 0.03), interrupted facial closure (75% vs. 46%, p = 0.02), and any gastrointestinal (GI) injury (92% vs. 60%, p < 0.01). On multivariable regression, interrupted fascial closure (OR 3.8) and any GI injury (OR 7.4) remained significant. Within the subgroup who achieved closure within three operations, interrupted fascial closure was significantly associated with IAA (67% vs. 38%, p= 0.05). Follow-up data was available for 65 patients; development of an enterocutaneous fistula was higher in those with prior IAA (18.8% vs 0%, p=0.01). Ventral hernia was not significantly higher (25% vs 8.2%, p=0.09).
Conclusion: Comorbid conditions historically associated with higher rates of infection showed no association with IAA formation after OA. Furthermore, administration of antibiotics did not impact the formation of IAA. However, both GI injury and fascial closure using an interrupted technique were significantly associated with IAA. The association of an interrupted fascial closure with IAA is unexpected and requires further study.
