The Effect of Immunosuppression on Emergency Colectomy Outcomes: A Nationwide Analysis
April E Mendoza*, Majed W El Hechi, Jae Moo Lee, George C Velmahos, Noelle N Saillant, Haytham MA Kaafarani
Division of Trauma, Emergency Surgery and Surgical Critical Care, Department of Surgery, Massachusetts General Hospital, Boston, Massachusetts, USA
*Corresponding Author: Dr. April E Mendoza, Department of Surgery, Division of Trauma, Emergency Surgery, and Surgical Critical Care, Massachusetts General Hospital, 165 Cambridge Street, Suite 810, Boston, MA 02114, USA
Received: 02 August 2019; Accepted: 15 August 2019; Published: 02 September 2019
Citation: April E Mendoza, Majed W El Hechi, Jae Moo Lee, George C Velmahos, Noelle N Saillant, Haytham MA Kaafarani. The Effect of Immunosuppression on Emergency Colectomy Outcomes: A Nationwide Analysis. Journal of Surgery and Research 2 (2019): 120-131.View / Download Pdf
Background: The impact of immunosuppression on the outcomes of emergent surgery remains poorly described. We aimed to quantify the impact of chronic immunosuppression on outcomes of patients undergoing emergent colectomy (EC).
Materials and Methods: The Colectomy-Targeted ACS-NSQIP Database 2012-2016 was queried for patients who underwent colectomy for an emergent indication. Chronic immunosuppression was defined as the use of corticosteroid or immunosuppressant medication within the prior 30 days. Univariate analyses were performed to compare 30-day postoperative mortality, morbidity and length of stay between immunosuppressed and immunocompetent patients after emergency colectomy for any indication. Multivariable regression models were then constructed to quantify the independent impact of immunosuppression on outcome, controlling for patient demographics, comorbidities, preoperative laboratory values and operative variables.
Results: Out of a total of 130,963 patients, 17,707 patients underwent an EC. Compared to immunocompetent patients (n=15, 422), immunosuppressed patients (n=2285) had higher ASA classification (47.4% vs 33.5%, p <0.001) On multivariable analyses, controlling for potential confounders, the immunosuppressed patients had higher odds of mortality (OR=1.28, CI 1.05-1.56), total morbidity (OR=1.43, CI 1.20-1.71), wound dehiscence (OR=2.01, 95% CI. 1.49-2.70), unplanned intubation (OR=1.33, 95% CI 1.07-1.66), acute renal failure (OR=1.56, 95% CI. 1.14-2.14), pulmonary embolism (OR=2.05, 95% CI, 1.33-3.12), blood transfusion (OR=1.17, 95% CI=1.01-1.37), pneumonia (OR=1.30, 95% CI=1.07-1.58), and a longer total hospital length of stay (mean difference=1.3 days, p=0.001).
Conclusions: Chronic immunosuppression is independently associated with a significant and quantifiable increase in 30-day mortality and morbidity for patients undergoing emergent colectomy.
Immunosuppression, Emergency Colectomy, Surgery
Immunosuppression articles, Emergency Colectomy articles, Surgery articles
The prevalence of chronic immunosuppression for surgical patients is nearly double that of the general US population at around 5% [1, 2]. This number is expected to rise over the coming years as a result of improved survival in immunosuppressed adults . In the setting of immunosuppression, physical examination can be misleading, and the clinical presentation can be ambiguous, leading to a delay in diagnosis. Additionally, patients requiring immunosuppressive medications tend to be medically frail with multiple co-morbidities, amplifying their risk for multiorgan dysfunction. Emergency surgery in the immunosuppressed adult has been studied in the transplant population  and described in the exclusive setting of complicated diverticulitis . While emergency colon surgery is an independent predictor of less favorable outcomes [4, 5] the degree to which immunosuppression impacts postoperative outcome in the emergency colon surgery patient remains unclear. In this study, we aimed to use the Colectomy-Targeted ACS-NSQIP database to evaluate and quantify the independent impact of chronic immunosuppression on the postoperative outcome of emergency colectomy (EC) patients. We hypothesize that EC in immunosuppressed (IMS) patients would be associated with an increase in 30-day mortality, 30-day morbidity, and would have a greater burden of wound and anastomotic complications when compared to their immunocompetent (IMC) counterparts.
2. Materials and Methods
The ACS-NSQIP database collects data prospectively and provides information on national trends on surgical outcomes. Collected data included patient demographics, surgical profiles, preoperative risk factors, preoperative laboratory data, and 30-day postoperative outcomes . The Colectomy Targeted ACS-NSQIP database additionally includes colectomy-specific variables such as surgical approach, wound classification, indication for surgery, and colectomy procedure . All patients 18 years or older undergoing an emergency colectomy between 2012 and 2016 were queried. Elective and non-urgent colectomies were excluded. Patients that had received pre-operative bowel preparation or pre-operative oral antibiotics were excluded. The indication for operation included perforation, obstruction, bleeding, toxic colitis, or other. The colonic procedure being performed was determined using Current Procedural Terminology (CPT) codes. Codes corresponding to Primary Anastomosis (44140, 44145, 44160, 44204, 44205, 44207) and Diversion (44141, 44143, 44144, 44146, 44150, 44155, 44188, 44206, 44208, 44210) procedures were identified and included. The post-operative diagnosis of colonic pathology was excluded from our analysis as the variable was missing for more than 30% of patients.
2.1 Definition of immunosuppression
We defined immunosuppression using the ACS-NSQIP database variable definition for ‘steroid or immunosuppressant use’. This includes “patients that have required regular administration of oral or parenteral corticosteroid medications or immunosuppressant medication, within the 30 days prior to the principal operative procedure or at the time the patient is being considered as a candidate for surgery, for a chronic condition (e.g. chronic obstructive pulmonary disease (COPD), asthma, rheumatologic disease, inflammatory bowel disease).” Patients who had a one-time pulse or short course of less than 10 days of steroids in the 30 days prior to surgery were excluded.
All outcomes reported in the analysis occurred within 30 days after the principal operative procedure. Primary outcomes were divided into hospital length of stay, infectious complications, major morbidity, and mortality. Infectious complications included overall surgical site infections, superficial surgical site infections, deep surgical site infections, organ space infections, pneumonia, urinary tract infection, sepsis, and septic shock. Morbidity outcomes included unplanned intubation, myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation (CPR), wound dehiscence, anastomotic leak, acute renal insufficiency (not requiring dialysis), acute renal failure (requiring dialysis), deep venous thrombosis, pulmonary embolism, and stroke/cerebrovascular accident (CVA).
2.3 Statistical analysis
The Shapiro-Wilk test was used to test for normality of the data. Univariate analysis was performed for categorical variables using χ2 tests, results of which were reported as number of observations (percentage). The Wilcoxon rank sum test was used to compare medians of hospital length of stay, results were reported as-median (interquartile range). Backward stepwise multivariable models were constructed to identify the impact of immunosuppression on hospital length of stay, 30-day overall morbidity, and 30-day mortality. Confounders controlled for these models included all preoperative variables (e.g. demographics, comorbidities, pre-operative laboratory values, illness severity, indication for surgery), as well as intraoperative and procedure-specific variables (e.g. wound classification) present in ACS-NSQIP Colectomy Procedure Targeted Database. Statistical significance was set as p<0.05 .All statistical analysis was performed by using the Stata v15.1 (StataCorp 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC).
2.4 Ethical oversight
This study was reviewed and approved by the Institutional Review Board.
Out of 130, 963 patients in the database, 17,707 met inclusion criteria. Of those patients, 15,422 were IMC, and 2,285 were IMS. Compared to IMC patients, IMS patients were older, more likely to be white, female, and fall into higher ASA classes, and have more comorbidities (Table 1). The wound classification in the IMS group was more often dirty/contaminated (63.8% vs. 52.9%; p<0.001). The most common indication for EC was perforation in both the IMC and IMS patients (60.9% & 46.1%) (Figure 1). The IMS patients were more likely to undergo a diversion procedure when compared to IMC patients, in particular a Hartmann’s procedure (CPT code 44143). IMC patients more often underwent primary anastomosis without diverting loop anastomosis (CPT code 44140, 44145, 44160, 44204, 44205, 44207) when compared to the IMS population (Table 2). In terms of outcomes, the univariate analyses demonstrated uniformly poorer outcomes in the immunosuppressed patients (Table 3). Notably, median length of stay (12 vs 10 days, p<0.001), 30-day morbidity (77% vs 65%, p<0.001), and 30-day mortality (19% vs 12%, p<0.001), were higher after EC in IMS patients.
3.1 Multivariable analyses
On multivariable modeling, immunosuppression was found to be independently associated with increased odds of 30-day mortality (OR=1.28, CI 1.05-1.56), 30-day morbidity (OR=1.43, CI 1.20-1.71), wound dehiscence (OR=2.01, 95% CI. 1.49-2.70), unplanned intubation (OR=1.33, 95% CI 1.07-1.66), acute renal failure (OR=1.56, 95%CI. 1.14-2.14), pulmonary embolism (OR=2.05, 95% CI, 1.33-3.12), blood transfusion (OR=1.17, 95% CI=1.01-1.37), and pneumonia (OR=1.30, 95% CI=1.07-1.58) (Figure 2). On multivariable linear regression, IMS patients had a longer total hospital length of stay (mean difference=1.3 days, p=0.001) when adjusting for previously mentioned potential confounders.
Table 1: Patient characteristics undergoing colectomy in the emergency setting.
*“Primary Anastomosis” was performed more frequently than “Diversion” in the Immunocompetent population, whereas “Diversion” was performed more frequently than “Primary Anastomosis” in the Immunosuppressed population (P<0.001).
Table 2: Operations (by CPT code) performed in the emergency setting.
*Total length of stay is the median with interquartile range in brackets
Table 3: Morbidity after emergent colectomy (univariate analysis).
To our knowledge, this is the first study that quantifies the effects of chronic immunosuppression on the outcomes of EC using the Colectomy-Targeted ACS-NSQIP database. Immunosuppression independently increases the risk of 30-day mortality and morbidity. Specifically, the odds of mortality were 28% higher and the odds of morbidity were 49% higher for the immunosuppressed patient undergoing EC. The elevated morbidity observed in the emergency setting for the IMS patient is largely due to the independent increase in the risks of unplanned intubation and wound dehiscence. While avoidance of intubation is generally advocated for IMS patients , studies have shown that COPD, pre-operative ventilator dependence, smoking, emergency surgery, and older age confer higher risk for unplanned intubation after surgery . Additionally, this increased ventilator dependence contributes to the higher mortality seen in the IMS patients within the emergency colectomy group as intubation has a well-recognized association with mortality in the IMS population [9, 10].
Emergency bowel resection has been shown to have higher mortality and an association with deep organ space infections . While the literature suggests that steroid use is associated with anastomotic leaks , we did not observe an increase in anastomotic complications nor deep organ space infections between the two groups. This can be attributed to selection bias, whereby surgeons were opting for fecal diversion in emergency settings in a high-risk patient population. Additionally, however, no difference was seen in anastomotic leaks when comparing only patients who underwent a primary anastomosis procedure. Moreover, unlike colectomy procedures performed in the elective setting [13, 14] our analysis showed that IBD did not independently increase the risk of anastomotic leaks or wound dehiscence in the emergency setting.
Our data further supports what has been observed in diverticulitis population. Al- Khamis et al. used the ACS-NSQIP database and included solely patients undergoing colectomy for diverticulitis. They observed an increased mortality risk in the emergency setting (OR 1.79; 95% CI, 1.17-2.75), but did not detect an increase in morbidity or wound dehiscence. This difference may be due to sample size and patient selection. Nevertheless, in the elective setting, they detected an increase in morbidity (OR 1.46; 95% CI, 1.17-1.83) and wound dehiscence (OR 2.69; 95% CI, 1.63-4.42) in IMS patients. Brandl et al. found that IMS was associated with higher mortality after diverticulitis (20% vs 4.7%, p=0.045), but determined that age and emergency status were the main contributors to mortality after multivariate analysis . Our study, however, demonstrates that immunosuppression in EC is associated with the development of important complications; namely pneumonia, pulmonary embolism, unplanned intubations, transfusions, and wound dehiscence. Interestingly, there was no evidence of an increased risk of wound infections, or anastomotic leaks for the immunosuppressed patients.
While emergency abdominal surgery has been studied in the transplant population , the vast majority of studied cases include cholecystectomy. However, in the transplant demographic, morbidity and mortality has been shown to be unacceptably high after emergency colectomy in the setting of acute diverticulitis . In our study, we do not focus on a single colon pathology, but rather evaluate a heterogeneous population of colon disease that could reflect the diversity of cases that the emergency surgeon manages on a day-to-day basis. As this is based on a multi-institutional, national database, there are inherent limitations with this study. The indication for immunosuppression is unreported, and the database is not explicit about the inclusion of patients immunosuppressed due to a medical condition (e.g. HIV/AIDS). Moreover, the specific immunosuppressant medication for each patient is not specified. Finally, as is encountered frequently in ‘big data’, missing data is not uncommon, whereby in our study sample, one third of colectomy data have no associated post-operative diagnosis, preventing us from stratifying patient groups by diagnosis.
Nevertheless, this work quantifiably validates that IMS poses serious challenges to the well-being of the colectomy patient in the emergency setting. Quantifying the independent impact of immunosuppression on outcome is essential to guide emergency surgeons, patients, and families through difficult discussions in the emergency room and to set realistic expectations of recovery postoperatively if the decision was made to proceed with surgery.
Emergency colectomy in IMS patients is associated with increased mortality and morbidity. Our results provide the emergency surgeon with quantifiable risk estimates that can help guide better patient counseling while setting reasonable expectations.
A.E. Mendoza, M.W. El Hechi, J. Moo Lee N. Saillant, G.C. Velmahos, and H.M.A. Kaafarani contributed to the conception and design of the research.
A.E. Mendoza, M.W. El Hechi and J. Moo Lee contributed to the acquisition of data.
A.E. Mendoza, M.W. El Hechi, J. Moo Lee and H.M.A. Kaafarani contributed to the data analysis.
A.E. Mendoza, M.W. El Hechi, J. Moo Lee, N. Saillant, G.C. Velmahos, and H.M.A. Kaafarani contributed to the data interpretation.
A.E. Mendoza drafted the manuscript. All authors critically revised the manuscript, read and approved the final manuscript, and agree to be fully accountable for ensuring the integrity and accuracy of the work. Each of the individual authors has sufficiently participated in this study to be listed as author.
Conflict of Interest
The authors report no proprietary or commercial interest in any concept discussed in this article.
- Harpaz R, Dahl RM, Dooling KL. Prevalence of Immunosuppression Among US Adults, 2013. Jama 316 (2016): 2547-2548.
- Al-Khamis A, Abou Khalil J, Demian M, et al. Sigmoid Colectomy for Acute Diverticulitis in Immunosuppressed vs Immunocompetent Patients: Outcomes From the ACS-NSQIP Database. Dis Colon Rectum 59 (2016): 101-109.
- De'Angelis N, Esposito F, Memeo R, et al. Emergency abdominal surgery after solid organ transplantation: a systematic review. World J Emerg Surg 11 (2016): 43.
- Biondo S, Borao JL, Kreisler E, et al. Recurrence and virulence of colonic diverticulitis in immunocompromised patients. American journal of surgery 204 (2012): 172-179.
- Golda T, Kreisler E, Mercader C, et al. Emergency surgery for perforated diverticulitis in the immunosuppressed patient. Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland 16 (2014): 723-731.
- Tam SF, Au JT, Chung PJ, et al. Is it time to rethink our management of dialysis patients undergoing elective ventral hernia repair? Analysis of the ACS NSQIP database. Hernia: the journal of hernias and abdominal wall surgery 19 (2015): 827-833.
- De Jong A, Calvet L, Lemiale V, et al. The challenge of avoiding intubation in immunocompromised patients with acute respiratory failure. Expert review of respiratory medicine 12 (2018): 867-880.
- Alvarez MP, Samayoa-Mendez AX, Naglak MC, et al. Risk Factors for Postoperative Unplanned Intubation: Analysis of a National Database. The American surgeon 81 (2015): 820-825.
- Huang HB, Xu B, Liu GY, et al. Use of noninvasive ventilation in immunocompromised patients with acute respiratory failure: a systematic review and meta-analysis. Critical care (London, England) 21 (2017): 4.
- Hilbert G, Gruson D, Vargas F, et al. Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure. The New England journal of medicine 344 (2001): 481-487.
- Benjamin E, Siboni S, Haltmeier T, et al. Deep organ space infection after emergency bowel resection and anastomosis: The anatomic site does not matter. The journal of trauma and acute care surgery 79 (2015): 805-811.
- Midura EF, Hanseman D, Davis BR, et al. Risk factors and consequences of anastomotic leak after colectomy: a national analysis. Diseases of the colon and rectum 58 (2015): 333-338.
- Lightner AL, Raffals LE, Mathis KL, et al. Postoperative Outcomes in Vedolizumab-Treated Patients Undergoing Abdominal Operations for Inflammatory Bowel Disease. J Crohns Colitis 11 (2017): 185-190.
- Lightner AL, Mathis KL, Tse CS, et al. Postoperative Outcomes in Vedolizumab-Treated Patients Undergoing Major Abdominal Operations for Inflammatory Bowel Disease: Retrospective Multicenter Cohort Study. Inflamm Bowel Dis 24 (2018): 871-876.
- Brandl A, Kratzer T, Kafka-Ritsch R, et al. Diverticulitis in immunosuppressed patients: A fatal outcome requiring a new approach? Canadian journal of surgery Journal canadien de chirurgie 59 (2016): 254-261.
- Reshef A, Stocchi L, Kiran RP, et al. Case-matched comparison of perioperative outcomes after surgical treatment of sigmoid diverticulitis in solid organ transplant recipients versus immunocompetent patients. Colorectal disease: the official journal of the Association of Coloproctology of Great Britain and Ireland 14 (2012): 1546-1552.
© 2016-2019, Copyrights Fortune. All Rights Reserved