Background
Damage control (DC) laparotomy was developed to quickly control exsanguinating hemorrhage and gross contamination in injured patients with severe physiologic derangements [
1]. It was first adopted by American trauma centers in the 1970s–1990s and then increasingly used worldwide as it was felt to be associated with an increase in unexpected survival among the most critically injured patients [
1‐
4]. However, systematic reviews of randomized and non-randomized studies have found insufficient evidence supporting that use of DC instead of definitive laparotomy improves mortality or other patient-important outcomes [
5,
6]. Further, use of the procedure is resource-intensive and associated with increased morbidity when compared to definitive laparotomy [
7‐
13].
The above suggests that there is likely insufficient evidence to support the high DC laparotomy utilization rates reported by some trauma centers [
5]. Limited data exist suggests that there is substantial variation in the frequency of use of DC laparotomy across level-1 trauma centers [
14‐
17]. In a post-hoc analysis of the Pragmatic, Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial, DC was used among 33–83% of patients requiring urgent laparotomy across 12 participating American level-1 trauma centers between 2012 and 2013 [
17]. While there was no significant mortality difference between the participating trauma centers, the risk of complications was higher among those treated with DC laparotomy [
5,
17]. This finding is supported by two other studies which reported that use of DC laparotomy among lower risk cohorts of injured patients is associated with increased risks of complications and longer hospital lengths of stay [
5,
16,
18].
Reasons for variation in use of DC laparotomy between level-1 trauma centers in the United States are unknown. It is also unknown whether variation in use of DC laparotomy exists across trauma centers outside of the United States and how often the procedure is used across level-2 and -3 trauma centers (e.g., to stabilize a critically injured patient before transport to a level-1 centers). To address these knowledge gaps, a cross-sectional survey of trauma centers located in the United States, Canada, and Australasia (Australia and New Zealand) was conducted [
19]. This survey had two objectives. First, it sought to determine if variation in use of DC laparotomy across trauma centers may be partially driven by surgeon uncertainty as to when the procedure was appropriately indicated. Results of this part of the study have been reported, which suggested that practicing surgeons have relatively consistent impressions of the appropriateness of using DC surgery in certain clinical scenarios [
19]. Second, it sought to study variation in the frequency of use of DC laparotomy across level-1, -2, and -3 trauma centers in these regions and predictors of more frequent use of the procedure (the subject of the current study). The study hypothesis was that the reported frequency of use of DC laparotomy would be highly variable across trauma centers, including level-1 trauma centers, and that this variation would be predicted by trauma center and program characteristics, including differences in setting, institutional characteristics, and patient mix.
Discussion
In this large, international, cross-sectional survey of trauma program directors or leaders, the reported frequency of use of DC laparotomy was highly variable across trauma centers. DC laparotomy was used more often in level-1 than level-2 or -3 trauma centers in the United States, Canada, and Australasia. The frequency of use of DC laparotomy also varied significantly across level-1 trauma centers, especially between those that were high- versus lower volume (based on the number of severely injured patients assessed in the last year). Nearly half of high-volume, level-1 trauma centers reported using DC laparotomy at least once weekly. Trauma center and program characteristics that independently predicted higher reported use of DC laparotomy included country of origin (with centers in the United States reporting using DC laparotomy significantly more often than those in Canada), level-1 trauma verification status, and the assessment of a higher number of severely injured (ISS score > 15) patients and patients with penetrating injuries in the last year.
Studying variation in use of DC laparotomy is important because surgeons are at risk of confirmation bias when only those with whom they work reflect their practice [
29]. However, to date, only one other study has examined variation in use of DC surgery between trauma centers [
17]. In a post-hoc analysis of the PROPPR trial, DC laparotomy was reportedly used among a highly variable 33–83% of patients requiring urgent laparotomy across 12 American level-1 trauma centers between 2012 and 2013 [
17]. In the current study, the trauma program directors who were surveyed also reported that the frequency of use of DC laparotomy varied widely across trauma centers, including level-1 trauma centers. Most level-1 trauma centers reported using DC laparotomy at least once a month, and nearly half of high-volume, level-1 trauma centers reported using it at least once weekly. However, 6% of high-volume, level-1 trauma centers reporting never using DC laparotomy during the past year. Reasons for variation in use of DC laparotomy between level-1 trauma centers may include differences in patient injury mechanisms, injury severity, and/or physiology; trauma center experience; trauma surgeon capabilities; and trauma provider education.
Trauma centers in the United States reported using DC laparotomy for trauma more often than those in Canada despite adjustment for level-1 verification status and the volume of severely injured patients and patients with penetrating injuries. This could be because of differences in institutional cultures regarding use of DC laparotomy between countries. It could also be because of unmeasured differences in patient mix between countries aside from injury mechanism or ISS scores (e.g., a higher percentage of patients with high-risk injury patterns, deranged physiology, or who receive significant volumes of resuscitation fluids) or beliefs regarding appropriate indications for use of the procedure [
30]. Of the American, Canadian, and Australasian trauma centers included in this study, a nearly equal percentage reported using the procedure less than once every 3 months or never using it, using it once monthly or once every 2–3 months, or using it more than once monthly. Further, more than one-third of level-2 centers reported using DC laparotomy once monthly or more than once monthly and even some level-3 trauma centers reported using the procedure. We assume that many of these level-3 trauma centers may be using DC laparotomy to stabilize critically injured patients before transport to a higher level of trauma care [
7].
In addition to country of origin, other independent predictors of an increased reported use of trauma DC laparotomy included level-1 trauma verification status and the assessment of a higher number of severely injured patients and patients with penetrating injuries in the last year. In the post-hoc analysis of the PROPPR study, the ISS score (OR per-1 point increase = 1.05; 95% CI 1.02–1.07) of the patients assessed at level-1 trauma centers also predicted an increased odds of use of DC laparotomy [
17]. This is likely because high-energy blunt torso trauma often results in high ISS scores and also may produce some of the high-risk injury patterns considered by many surgeons to be appropriate indications for DC laparotomy (e.g., massive pelvic fracture-related hemorrhage or multiple injuries spanning across more than one body cavity that each require surgery) [
19,
30,
31]. Further, while patients with penetrating injuries may have a lower ISS score, those with gunshot wounds (and especially shotgun wounds) more often present with certain injury patterns that have been suggested to be appropriate indications for DC laparotomy [
30]. These may include a major abdominal vascular injury and multiple associated hollow organ injuries or an injured pancreaticoduodenal complex [
19,
32].
The study findings should be considered in the context of its strengths and limitations. First, the opinions of trauma program directors could be argued to be only estimates of the frequency of use of DC laparotomy. Surveys are a practical strategy to capture practice variation across hundreds of level-1, -2, and -3 trauma centers in the United States, Canada, and Australasia. However, because of the limitations of using cross-sectional survey data, they should be confirmed by future observational studies. Second, although we used several techniques shown to increase response rates to surveys (and the response rate is above what has been reported by many surveys reported in the trauma or surgery literature), it is possible that respondents’ answers on the reported frequency of use of DC laparotomy differ systematically from those who did not respond to the survey [
27]. Third, as the data used in this study are now over 5-years old, it is unclear whether our findings reflect current practice. However, based on evidence that adoption of new evidence-informed practices is slow in medicine (typically quoted to be 18 years), and evolving literature suggesting that deadoption of practices is likely even slower [
33,
34], there may have been little change in use of DC laparotomy. This highlights the importance of conducting follow-up studies like ours to track practice patterns and ensure that practice is evidence-informed.
This study has important implications for future research, trauma surgery practice, and quality improvement efforts. First, a systematic review of 36 cohort studies found very little evidence to support that use of DC instead of definitive laparotomy in trauma patients was associated with an improvement in mortality or other patient-important outcomes [
5]. However, use of the procedure is associated with an increased risk of morbidity, a longer length of intensive care unit and hospital stay, and possibly a reduced quality of life among survivors and an increase in healthcare utilization [
7,
11‐
13,
35‐
37]. As equipoise now exists among some surgeons about the effectiveness of DC for improving mortality in many patients undergoing urgent laparotomy, there is a need for rigorously-designed randomized trials comparing it to definitive trauma laparotomy [
38]. Second, although DC laparotomy is supported by insufficient evidence, nearly half of high-volume, level-1 trauma centers reported using it at least once weekly. One other study reported utilization rates exceeding 80% in some level-1 trauma centers (with most level-1 centers using it among 30% of those undergoing urgent laparotomy) [
17]. Some authors have suggested that more comprehensive indications guiding patient selection for use of DC laparotomy may decrease its associated morbidity and costs [
17]. However, our group previously compiled a comprehensive list of indications for the procedure that both experts and practicing trauma surgeons consistently agree appropriately indicate its use [
7,
31,
32]. Further, a cohort study suggested that these comprehensive indications that highly predicted use of DC laparotomy in practice had an incidence of 2% or less [
30]. Collectively, the above may indicate that DC laparotomy is presently overused in trauma centers. Reasons for this are largely unknown, but may include surgeons’ training, differences in patient mix between institutions, and/or institutional characteristics or culture [
19,
29]. However, efforts to decrease use of DC laparotomy across trauma centers may be necessary until further evidence becomes available. The centers that may need to be targeted first include high-volume, level-1 trauma centers, particularly those that often manage penetrating injuries. Indeed, some data suggests that utilization rates of DC laparotomy can be safely reduced through quality improvement efforts such as audit-and-feedback without adversely influencing patient outcomes [
14,
39,
40].
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.