Are the Benefits of Rapid Source Control Laparotomy Realized after Acute Colonic Perforation?

Patients and Methods

This retrospective cohort analysis was based on three years of data, from 2014 through 2016, compiled from The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The data were used to assess patients who had emergency colectomy for perforation. The different diagnoses are shown in Table 1. The patient populations were separated into two groups: Group 1, patients who had their fascia left open after the initial operation, and Group 2, patients who had their fascia closed primarily (PFC) after initial operation and the skin left open.

Univariable analysis was performed on the baseline patient characteristics including age, sex, and American Society of Anesthesiologists (ASA) class. This was followed by propensity score-matching analysis. Paired matched analysis was performed for death, infection, ventilator days, transfusion rate, and septic shock. The primary outcome of the study was death within 30 days using paired matched analysis with secondary analyses of the rates of complications. Post-operative primary outcomes included length of stay and days until discharge to home or to a rehabilitation facility. Post-operative secondary outcomes included failure to wean, acute renal failure, transfusion requirements, systemic sepsis, and septic shock.

Statistical significance was considered if the p value was <0.05.

Results

The cohort consisted of 1,381 patients. Of these, 396 (28.7%) patients were managed with RSCL and the others had primary fascial closure (PFC). Table 2 shows the 30-day deaths and length of stay in the two groups before matching. Mortality rate, length of stay, and day of death were significantly different in the two groups. Table 3 shows the discharge destination of the two groups before matching. Table 4 shows the results from unadjusted analysis. The patients having RSCL differed by superficial infection, principal procedure, ventilatory support, diabetes mellitus, ascites, dialysis, weight loss, history of congestive heart failure, transfusion, ASA class, and presence of septic shock. Race, smoking history, history of chronic obstructive pulmonary disease (COPD), renal failure, history of cancer, steroid use, and wound classification did not differ between the groups.

After propensity matching, the significant differences are shown in Table 5. Surgical site infections (SSIs) were more than seven times more common in the RSCL group than in the PFC group (2.2% versus 0.3%; p < 0.049). Length of stay was significantly longer in the RSCL than in the PFC group (median 20 [95% CI 18–22] days versus 14 [95% CI 13–16] days; p < 0.001), respectively. A larger number of patients in the RSCL group than in the PFC group went to a rehabilitation facility (18.7% versus 11.2%), whereas the patients in the PFC group were more likely to return home after hospital discharge (44.4% versus 33.7%).

There were more than twice as many patients on ventilator support for more than 48 hours in the RSCL group than in the PFC group (62.5% versus 25.5%; p < 00001). The rate requiring transfusions was higher in patients with open than closed abdomens (41.3% versus 33.4%; p < 0.05). The mortality rate was almost twice as high in the RSCL group than in the PFC group, 129 patients (32.6%) versus 67 patients (16.9%); p < 0.001.

Prior to matching in the presence of septic shock, 46.2% of the RSCL group were affected compared with 18.2% of the PFC group. After matching, the rate of septic shock was 44.2% vs 30.4%. In both conditions, the rate of septic shock was significantly higher in the RSCL than in the PFC group (p < 0.001). Although the matching corrected this difference to some degree, the correction was limited.

Discussion

The RSCL approach for non-trauma abdominal pathology has become an increasingly controversial topic, with some advocating open abdomen [9,10] and others arguing against this technique as routine practice [7,8]. In our study, more than 25% of patients were managed without fascial closure. We believe this is a remarkably high rate, especially given the limited data that support its use. In addition, there as not a standard method or duration of temporary closure. In trauma patients with the lethal triad (hypothermia, coagulopathy, and acidosis), aborting an operative procedure by covering the abdomen to correct the metabolic aberrations may be life-saving. It is difficult, however, to imagine such a large percentage of this patient population would meet the standard criteria for RSCL. Because only 44.2% of the patients in the RSCL group were in septic shock at the time of surgery, 181 patients did not meet the criteria for RSCL. Consequently, those patients were managed with an open abdomen for reasons other than the presence of the lethal triad. One reason would be to avoid an abdominal compartment syndrome. A second motive for justifying RSCL is to re-establish intestinal continuity by serial abdominal washouts in the operating room; in other words, performing a staged procedure to re-establish intestinal continuity. The efficacy of a primary anastomosis in this setting has not been evaluated. One study [10] documented superior outcomes of DCL over PFC in trauma patients. However, the mean age of the patients was 29 years old, which is more than 30 years younger than the mean age of the patients in our study.

A recent publication by Kirkpatrick et al. reviewed the published data regarding RSCL extensively [11]. As those authors point out, many studies mix trauma patients with non-trauma acute care patients, which also mixes age and disease types, making it difficult to draw firm conclusions. For example, patients with acute perforating diverticulitis have a different bacterial milieu in their peritoneal cavity than those with a penetrating bowel injury [12]. A randomized trial to evaluate RSCL in non-trauma patients would provide much information as to where and when this technique should be used.

A putative advantage of RSCL is the control of sepsis. However, in this study, a greater risk of SSI was seen in the RSCL than the PFC group. Additionally, it is presumed that repeated washouts of the abdominal cavity after source control reduce the bacterial load to the site. This did not occur to the degree that the infection rate was reduced. The negative impact of leaving the peritoneal cavity open to the environment needs to be considered as a risk factor for the patients' homeostasis. There has not been any organized assessment of the immune consequences of leaving the abdomen open with a closed drainage system [11]. Not only do the liquid mediators, such as complement and immunoglobulins, leak out through the peritoneal cavity, but the peritoneal surface, a dialyzable membrane, may be exposed to and stimulated by endotoxin [13]. The continued stimulation of the peritoneal surface also may result in neutrophil diapedesis [14]. The delay in restoring homeostasis by closing the peritoneal cavity may differ from surgeon to surgeon from a 24-hour delay to several days. Unfortunately, these individual management concerns cannot be addressed with this database.

The greater ventilator time required for the RSCL patients may impact the length of stay and discharge disposition significantly, but this idea has not been addressed in prior studies. The prolonged ventilatory support results in extended bed rest and the use of paralytics and narcotic sedation, rapidly deconditions patients. The fact that the mean age of the patients in our study was 62 years and that the elderly population is prone to rapid deconditioning is a significant concern [15]. The benefits of early mobilization in this elderly population have been documented with the many enhanced recovery after surgery (ERAS) bundles, which have expedited hospital discharge. Absence of these two factors may have contributed to the prolonged length of stay and higher likelihood of needing rehabilitation facilities after discharge.

The mortality rate was much higher in the RSCL group than in the PFC group, which is consistent with the two prior studies and with our hypothesis [7,8]. This current study, which limited the participants to patients with colon perforations, had a high mortality rate in the RSCL group (31.4%), which is surprisingly similar to the rate in a prior study (32.6%) [8]. The mortality rate in the PFC group was lower in this current study (16.9%) than in the prior study (21.4%). The prior study included a wider cross-section of diagnoses such as gastrointestinal bleeding and severe pancreatitis, which probably impacted the mortality rate for their PFC group.

A weakness of this paper is that before matching, patients treated with RSCL were statistically more likely than PFC patients to be in septic shock (see Table 4) (46.2% versus 18.2%, respectively; p < 00001). After matching (Table 5), the numbers changed slightly, but the difference remained statistically significant (44.2% [RSCL] versus 30.4% [PFC]; p < 00001). A comparison of patients in septic shock treated with RCCL and PFC has been done [16]. A higher mortality rate was seen in the RSCL group (35.7%) than in the PFC group (28.5%), but this difference was not statistically significant. The investigators in the earlier study concluded that there was no significant advantage or disadvantage to RSCL [16]. Such an analysis should be repeated to confirm the earlier findings. These data need to be analyzed carefully to determine whether there is a subset of patients with perforated colon in septic shock who do obtain a significant benefit from RSCL. As mentioned above, these data are limited by the variability among surgeons in the temporary abdominal covering and the interval between the initial source control procedure and definitive abdominal closure.

In conclusion, this study shows that statistically significant morbid and mortal consequences result from RSCL compared with PFC. The limiting factors of the study are: (1) The lack of control for the method and duration of the temporary abdominal visceral covering for the open abdomen group; (2) the application of standard criteria for the use of RSCL; and (3) the performance of enteric anastomoses at the time of surgery in the RSCL group. Because more than half of the patients who had RSCL were not in septic shock, specific criteria need to be established for RSCL. More controlled prospective analysis needs to be done to determine how RSCL best fits for patients with non-traumatic large colon perforations.