Introduction

Damage control strategies (DCS) are accepted and supported by significant scientific evidence when applied to severely injured patients in the context of major trauma^1-3.

This strategy has occupied a prominent place since 2008 due to the interest of surgeons experienced in non-traumatic abdominal emergencies (NTAEs) and has produced a paradigm shift in the management of these patients, as the benefit is greater with the restoration of physiologic normality than by solving the lesion in a single procedure^4-9. Several reviews have supported this indication, but some authors consider that the evidence is still insufficient and limited^4-11.

To understand the benefits of DCS in NTAEs in our environment, the aims of the present study were: 1) to describe the main etiopathogenic causes and the pathophysiologic conditions to implement this strategy, 2) to associate mortality with etiology, criteria used to implement the method and associated risk factors, and 3) to compare the observed mortality with the expected mortality according to the APACHE II score.

Material and methods

We conducted an observational and retrospective study of consecutive patients >18 years with NTAEs operated on by specialists in a highcomplexity institution between January 2009 and December 2019.

A total of 6632 patients were evaluated due to non-traumatic abdominal emergencies. The criteria used to implement a DCS were the presence of signs, symptoms or results of tests consistent with abdominal compartment syndrome (ACS), abdominal or systemic sepsis, metabolic acidosis and hypovolemic shock.

The risk factors for severity considered were age ≥70 years, sex, white blood cell (WBC) count ≥ 10 000 x mm³, hemoglobin (Hb) ≤ 9 g/dL, creatinine level ≥ 1.3 mg/dL, pH ≤ 7.25, lactic acid ≥ 2.5 mmol/L, diabetes, American Society of Anesthesiologists (ASA) score (≤ 3 versus ≥ 4), number of reoperations (≤ 3 versus ≥ 4), and wall closure versus open abdomen.

Patients were grouped according to the APACHE II score before DCS and expected mortality was compared with observed mortality, both overall mortality and mortality grouped by scores.

Data were incorporated into a Microsoft Excel 97^® database and analyzed using R Core Team (2018) 4.0 statistical software package.

All the variables included were dichotomized and expressed as frequency and percentage.

The differences between groups were analyzed using the asymptotic test for comparing proportions and, if the assumptions were not met, the chi-square test and Fisher’s exact test were used.

A p value < 0.05 with a 95% confidence interval was considered statistically significant.

Results

Following the inclusion criteria, 118 patients were operated on for NTAE, 112 due generalized secondary peritonitis (95%) and 6 due to severe intraabdominal bleeding (5%). Peritonitis was secondary to infected necrotizing pancreatitis (INP) in 42 cases (37.5%), non-ischemic bowel perforations (NIBP), diverticular disease, tumor or anastomotic dehiscence in 38 (34%), and perforations due to severe mesenteric ischemia (SMI) in 32 (28.5%).

Severe intra-abdominal bleeding episodes occurred after hepato-pancreato-biliary or pelvic surgeries (5%).

Overall mortality was 43.1% (51/118) in the series.

Mortality (by etiology) was 65.6% (21/32) in SMI, 42.8% (18/42) in INP and 31.5% (12/38) in NIBP. None of the patients with severe intra-abdominal bleeding died.

Mortality associated with SMI was significantly higher (p = 0.0027). When the influence on mortality was analyzed according to the criteria used for implementation, mortality was 0% using 1 criterion, 20% with 2 criteria, 50% with 3 criteria, 78% with 4 criteria and 100% with 5 criteria (Figure 1).

Univariate analysis of the risk factors showed that white blood cell count ≥ 10 000/mm³, creatinine ≥ 1.3 mg/dL, hemoglobin ≤ 9 g/dL, pH ≤ 7.25, lactic acid ≥ 2.5 mmol/L, presence of diabetes, ASA score ≥ 4, ≥ 4 operative procedures, and open abdomen were significantly associated with mortality (Table 1).

Mean APACHE II score in the series was 27.7 points (range 17- 38) with a median of 28. The expected and observed mortality for a score between 25 and 29 points was 53% and 43.1%, respectively (p = 0.118) (Table 2).

When the scores were compared by groups, there were significant differences in the groups between 20 and 24 points (p = 0.004) and between 25 and 29 points (p = 0.028).

There were no differences with the other scores (Table 3).

Discussion

Damage control strategy for the treatment of NTAEs is still controversial in several aspects. According to some international authors, the main indications for this strategy are generalized peritonitis secondary to perforated hollow viscus, abdominal vascular catastrophes (mesenteric ischemia, complicated aortic aneurysms and intraoperative bleeding), abdominal compartment syndrome (ACS), severe wall infections and acute grade III incisional hernias^1-16.

In the present series many of these conditions were the causes involved. We did not have any cases of complicated aortic aneurysms.

Acute compartment syndrome, severe wall infections and acute grade III incisional hernias were common in the cases here described and were always the result of an uncontrolled intra-abdominal sepsis. For this reason, they were considered a consequence of primary disease progression rather than a determining factor.

The mortality reported is highly variable between the series and is strongly associated with the cause and systemic inflammatory response^{1,7,11,13,15-18}.

In previous publications on SMI, the authors stated that non-specific clinical presentation and inconclusive diagnostic tests lead to misdiagnosis or delayed diagnosis, resulting in more advanced stages with viscus perforation, ACS, abdominal or systemic sepsis or both, and multiple organ failure^17,18.

As stated in other publications, the authors agree that the successive activation of this cascade of events in elderly patients with severe comorbidities and a lower physiologic reserve is the cause of this high mortality rate^19-25. This physiologic abnormality occurs to a lesser extent in cases of peritonitis secondary to non-ischemic hollow viscus perforations (diverticular disease, tumors or anastomotic failure). It is possible to consider that lower mortality is due to earlier diagnosis and treatment, less physiologic injury, and may be related with younger patients with lower chronic disease burden. Regarding severe infections in acute pancreatitis, a better understanding of the natural history of the disease, an aggressive initial management, potent antibiotics and the development of minimally invasive strategies improve patient’s physiologic conditions and infections, reducing the need for DCS and mortality^14,20,21.

In the present series, severe intra-abdominal bleeding solved after compression maneuvers and initial resuscitation. In these patients, hemodynamic impairment, metabolic acidosis and, in some cases, hypothermia, were rapidly compensated, and this explains the absence of mortality, which is consistent with the findings of other specialists consulted^19-26.

The selection criteria to indicate this DCS in patients with NTAE is a vital key point in the final outcomes^4-7,9,11.

The pathophysiologic insult and systemic shock differ in trauma patients and in non-traumatic emergencies. Hemorrhage and tissue damage prevail in trauma while in non-traumatic emergencies, inflammation, infection and sepsis are more prevalent^1-7,9,11.

For this reason, the specialists agree that considering only hypotension, hypothermia and metabolic acidosis as an indication in NTAE is insufficient^4,5,21,29-31.

So far, there are no data or evidence to define a set of criteria to support the indication of this strategy in NTAE; therefore, the decision remains complex and challenging, depending mainly on the surgeon’s experience and opinion.

Other variables that may indicate the physiologic decline of these patients should be added to the lethal triad^11,21,27-31. For this reason, in this series we included the presence of signs, symptoms or test results suggestive of ACS, abdominal sepsis and septic shock as variables for indicating DCS, which was activated by the surgeon’s decision in the presence of at least one of the aforementioned criteria.

The results shown in Figure 1 allow us to infer that mortality is related to the number of indication criteria included for each case. When the relative risk of mortality was analyzed according to the variables applied in the entire series, we were able to conclude that, with four or more criteria, the risk of mortality is higher and statistically more significant than in the general population. The same cause-specific analysis found that, in SMI, the relative risk of mortality increased according to three criteria. In addition, in the analysis of the deceased according to cause, number of cases and number of factors involved, SMI also showed a non-significant higher mortality rate than the other conditions.

As in other papers published by other authors, it is possible to consider that these results are closely related with the greater systemic aggression^4,6,28-31.

Many clinical and laboratory variables and associated comorbidities have been evaluated by different authors, with dissimilar results^{6,7,9,11,21,27,28,31}.

In the present study, besides age, sex, biochemical parameters of inflammation, sepsis and multiorgan failure, the ASA score and two technical variables were included, such as the number of surgical procedures and presence or absence of early wall closure. Both variables performed as protective predictors since mortality decreased significantly when patients overcame the early stages of CDS or when the abdomen could be closed at an early stage. Both conditions were not evaluated on other series consulted.

Yet, the results coincide with the international literature reviewed^{7,8,11,14,16-26}.

To determine the benefits of DCS in the series, the observed mortality was compared with the expected mortality, according to the APACHE II score. Although this comparison did not reach statistical significance, the lower percentage of mortality observed, especially in the groups between 20 and 24 points and 25 and 29 points, encourages the authors to continue with this approach, since a greater number of cases could confirm the hypothesis.

Based on the results presented and in response to the objectives established, we may state that, in the population analyzed, SMI, the number of criteria applied, and the inclusion of some risk factors studied, was statistically associated with higher mortality. Although the retrospective nature of this single-center study is a limitation of this study, it shows the results of a surgical team made up of board-certified surgeons with experience in emergency medicine, who shared the same criteria to select patients in a consecutive series in a high-complexity center in Argentina. Future prospective studies with a larger number of patients might reach more strong conclusions in the matter.