Predictors of survival after gastrointestinal complications in bypass grafting

Abstract

Between April 2001 and December 2005, data were collected prospectively from 2,320 consecutive patients who underwent first-time coronary artery bypass. Logistic multiple regression analyses were carried out to determine the independent predictors of gastrointestinal complications and death. There were 65 major gastrointestinal complications identified in 65 (2.8%) patients: paralytic ileus in 15, mesenteric ischemia in 12, upper gastrointestinal hemorrhage in 16, lower gastrointestinal hemorrhage in 8, small bowel obstruction in 5, pseudoobstruction in 5, and others in 4. The 30-day mortality was 21.5% (14 patients). Female sex, preoperative creatinine >200 µmol·L⁻¹, previous gastrointestinal pathology, low cardiac output, readmission to the intensive care unit, postoperative pulmonary complications, arrhythmias, hemofiltration, and reoperation were independent predictors of major gastrointestinal complications. Independent risk factors for death were readmission to the intensive care unit, the need for hemofiltration, reoperation, and ischemic bowel. Careful hemodynamic control and optimization of perioperative organ perfusion are important to minimize the mortality associated with major gastrointestinal complications after coronary artery bypass.

Keywords

coronary artery bypass gastrointestinal diseases gastrointestinal hemorrhage postoperative complications risk factors

Introduction

Gastrointestinal (GI) complications are difficult to diagnose early after coronary artery bypass grafting (CABG) because symptoms may not be evident due to sedation, mechanical ventilation with or without muscle paralysis, and analgesia. The ability to identify patients who are at greater risk of developing GI complications may allow earlier detection and treatment, thereby increasing the probability of a successful outcome. These complications are rare (0.4%–2%), but result in high mortality.¹ Although many retrospective reviews have reported various GI complications after CABG, including GI bleeding, mesenteric ischemia, pancreatitis, cholecystitis, perforated ulcers, and ileus, very few studies have assessed the risk factors for developing GI complications and the associated mortality, using multivariate analysis in the same series of patients.² We aimed to determine the risk factors for developing GI complications and the associated mortality in patients undergoing first-time CABG.

Patients and methods

Between April 2001 and December 2005, data were collected prospectively from 2,320 consecutive patients who underwent first-time CABG at our tertiary referral center. Collating of the data commenced in January 2006, and progressed up to August 2006, mainly to determine the survival data. Information for the study was obtained from our cardiac surgical Patients Analysis and Tracking System database and hospital records. Patients who had a myocardial infarction within 4 weeks, those requiring concomitant valve or other surgery, and patients on inotropic infusions or in shock preoperatively were excluded from the study. Stepwise logistic multiple regression analyses were carried out to determine the independent predictors of GI complications and subsequent death. Two analyses were performed: the patients with GI complications were compared with those without GI complications; and the predictors of survival were determined in 51 patients who survived and 14 who died after GI complications. The definitions of GI complications were derived from the Society of Thoracic Surgeons’ database.³ Upper and lower GI bleeding were determined on the basis of endoscopy findings.

Descriptive statistics were obtained using the SPSS software package version 15 for Windows (SPSS, Inc., Chicago, IL, USA). To generate the unadjusted risk of each factor, continuous variables were compared between groups using the Student t test, and categorical data were compared using Fisher’s exact test and the chi-squared test. To determine the influence of various factors on the incidence of GI complications and subsequent mortality, patient variables that might influence outcome were analyzed using automated forward multiple regression analysis. A value of p < 0.05 was considered significant in all statistical analyses.

Results

Sixty-five major GI complications were identified in 65 patients (Table 1); the incidence of GI complications was 2.8%. Tables 2 and 3 show the pre- and postoperative characteristics of patients with and without GI complications, which were compared by univariate analysis. The hospital mortality from any cause in patients who suffered GI complications was significantly higher than in those without GI complications. Mortality was highest in patients with mesenteric ischemia (10/12, 83.3%). The results of multivariate stepwise logistic regression analyses are summarized in Table 4. Multivariate analysis identified female sex, preoperative serum creatinine >200 µmol·L⁻¹, previous history of GI pathology, perioperative low cardiac output, readmission to the intensive care unit (ICU), postoperative pulmonary complications, arrhythmias, postoperative need for hemofiltration, and reoperation for bleeding or tamponade as independent predictors of GI complications. Pulmonary complications included hospital-acquired and aspiration pneumonias, inability to clear secretions, and poor ventilatory reserve leading to hypoxia or ventilatory failure. For the purpose of this study, cardiac dysrhythmias included new-onset atrial fibrillation, supraventricular tachycardia, and ventricular tachycardia or fibrillation. Use of cardiopulmonary bypass (CPB) was not significant. Univariate risk analysis for prediction of death after GI complications was also performed. On multivariate analysis, independent risk factors for death were readmission to the ICU, the need for hemofiltration, reoperation, and ischemic bowel (Table 5). The main cause of readmission to the ICU was cardiorespiratory compromise.

Table 1.

Gastrointestinal complications after first-time coronary artery bypass in 65 patients

Complication	No. of Patients
Paralytic ileus	15 (23.0%)
Mesenteric ischemia	12 (18.5%)
Upper gastrointestinal bleeding	16 (24.6%)
Lower gastrointestinal bleeding	8 (12.3%)
Small bowel obstruction	5 (7.7%)
Pseudoobstruction	5 (7.7%)
Acute pancreatitis	1 (1.5%)
Massive hepatic necrosis	1 (1.5%)
Exacerbation of ulcerative colitis	1 (1.5%)
Acute cholecystitis	1 (1.5%)

Table 2.

Preoperative characteristics of patients with and without GI complications

Variable	GI Complication	No GI Complication	Univariate p
Age (years)	69.4 ± 7.8	64.5 ± 9.5	0.01
Male sex	50 (64%)	1,549 (80.5%)	<0.001
Emergency	4 (5%)	42 (2.2%)	0.1
CCS class 3/4	41 (52.5%)	880 (45.7%)	1
NYHA class III/IV	32 (41%)	503 (26.1%)	0.01
Previous MI	29 (37.1%)	498 (25.9%)	0.1
Diabetes mellitus	26 (33.3%)	374 (19.4%)	0.01
Respiratory problems	9 (11.5%)	139 (7.2%)	0.2
Previous GI problems	26 (33.3%)	293 (15.2%)	0.001
Renal dysfunction	4 (5.1%)	15 (0.8%)	0.001
Low cardiac output	45 (57.7%)	326 (16.9%)	0.001
Impaired LV function	26 (33.3%)	392 (20.4%)	0.05
Peripheral vascular disease	8 (10.3%)	26 (1.4%)	0.001
Previous stroke	14 (17.9%)	126 (6.5%)	0.001
EuroSCORE ≥ 6	11 (33.3%)	123 (14.6%)	0.01
BMI ≥ 30 kg·m⁻²	26 (33.3%)	509 (26.5%)	0.2

BMI = body mass index, CCS = Canadian Cardiovascular Society, GI = gastrointestinal, LV = left ventricular, MI = myocardial infarction, NYHA = New York Heart Association.

Table 3.

Postoperative characteristics of patients with and without GI complications

Variable	GI Complication	No GI Complication	Univariate p
≥3 bypass grafts	54 (69.2%)	1167 (60.6%)	0.2
Cardiopulmonary bypass	29 (37.1%)	456 (23.7%)	0.01
Low cardiac output	45 (57.7%)	317 (16.5%)	0.001
Ventilation time (days)	6.1 ± 8.4	1.16 ± 5.3	<0.001
Mean ICU stay (days)	14.7 ± 22.8	2.1 ± 6.8 days	<0.001
Mean hospital stay (days)	26 ± 26.5	10.0 ± 12 days	<0.001
ICU readmission	12 (15.4%)	68 (3.5%)	0.001
Pulmonary complications	42 (53.8%)	84 (4.3%)	0.001
New onset of arrhythmias	32 (41%)	491 (25.5%)	0.01
Hemofiltration	25 (32%)	56 (2.9%)	0.001
Reoperation for bleeding/tamponade	13 (16.66%)	111 (5.8%)	0.001
30-day mortality	14 (17.9%)	34 (1.7%)	0.001

GI = gastrointestinal, ICU = intensive care unit.

Table 4.

Multivariate analysis of predictors of GI complications

Variable	Multivariate p
Female sex	0.001
Perioperative low cardiac output	<0.0001
Preoperative history of GI pathology	0.003
Preoperative creatinine > 200 µmol·L⁻¹	<0.0001
Reoperation for bleeding/tamponade	<0.0001
Postoperative arrhythmias	<0.0001
Pulmonary complications	<0.0001
Hemofiltration	<0.001
Readmission to ICU	<0.0001

GI = gastrointestinal, ICU = intensive care unit.

Table 5.

Variables affecting mortality in patients with GI complications

Variable	Died	Survived	Univariate p	Multivariate p
Emergency operation	1 (8.33%)	0 (0%)	<0.05	>0.05
Diabetes mellitus	07 (58.3%)	13 (24.5%)	0.02	>0.05
Mean ventilation time (days)	6.1 ± 8.4	1.16 ± 5.3	<0.001	>0.05
Readmission to ICU	6 (50%)	7 (13.2%)	0.001	<0.0005
Hemofiltration	9 (75%)	12 (22.6%)	0.001	<0.0005
Reoperation for bleeding/tamponade	7 (58.3%)	10 (18.9%)	0.01	<0.001
Ischemic bowel	10 (83.3%)	2 (3.9%)	<0.001	<0.0005

GI = gastrointestinal, ICU = intensive care unit.

Discussion

Previous studies have demonstrated that advanced age (≥65 years), history of ulcer, low cardiac output syndrome, prolonged mechanical ventilation, and reexploration for hemorrhage are risk factors for GI complications after cardiac surgery.^4,5 The literature on this subject emphasizes the interplay of multiple factors leading to splanchnic hypoperfusion and consequent GI complications after CABG.² Published mortality rates vary from 11% to 59%.^1–4,6 The 2.8% incidence of GI complications and 21.5% mortality rate in our series are consistent with those of previous reports. All patients were routinely started on proton-pump inhibitors postoperatively; those who were already taking antacids, H₂ blockers, or proton pump inhibitors had their medication continued up to the day of surgery.

The need for postoperative hemofiltration increases the incidence of GI complications after CABG. Acute renal failure after cardiac surgery is associated with high morbidity and mortality rates, and results from organ hypoperfusion during CPB and in the postoperative phase.⁷ Moderate renal failure in animals alters duodenal-jejunal motility, decreases colonic transit time, and may contribute to GI injury.⁸ A direct relationship between acute renal failure and GI complications after cardiac surgery has previously been determined by multivariate analyses.^2,7,9 History of GI disease increases the likelihood of GI complications after CABG.² Specifically, history of peptic ulcer is an independent risk factor. GI bleeding usually results from stress ulceration and is more common in the elderly and those with preexisting ulcer disease. Previous GI pathology was present in 26 of our patients (peptic ulcer, cholecystitis, pancreatitis, and previous laparotomy). Separately, each of these diagnoses was not significant, but when grouped together, they became significant. Patients with universal arteriosclerosis are at higher risk of mesenteric ischemia postoperatively. None of our patients had preoperative symptoms suggestive of mesenteric ischemia, such as abdominal angina.

Any suspicion of a GI complication was addressed by a GI surgeon or gastroenterologist. Paralytic ileus was diagnosed by a consultant general surgeon on the basis of clinical features (painless abdominal distension) and grossly distended bowel loops on plain abdominal radiographs. This is clearly different from the normal more or less paralytic condition of the bowel postoperatively. Several patients suffered small bowel obstruction or pseudoobstruction. The cause of bowel obstruction or pseudoobstruction was previous laparotomy resulting in subacute obstruction due to adhesions. Once a complication was identified, a specialist referral was made immediately, and appropriate measures were taken. For upper and lower GI bleeding, both esophagogastroduodenoscopy and colonoscopy were performed. Sclerotherapy ± adrenaline injection was carried out in appropriate cases. No patient required a laparotomy for GI bleeding. In suspected ischemic bowel, the on-call general surgeon reviewed the patient, and if felt appropriate, a laparotomy was performed. Paralytic ileus was treated conservatively.

It is important to identify patients at increased risk of GI complications, but it is equally important to examine the outcomes of those who suffer such complications. There was a significant correlation between GI complications and postoperative cardiac rhythm, pulmonary, and renal complications. The factors associated with increased mortality after GI complications in our series were readmission to the ICU, need for hemofiltration, reoperation, and ischemic bowel. GI complications are among the causes of ICU readmission after cardiac surgery.¹⁰ We found that readmission to the ICU was associated with increased mortality in patients with GI complications after CABG. However, it is difficult to establish the cause-effect association between different complications after CABG. It is worth noting that these patients frequently have multiple complications. Significant associations between GI complications and the length of ICU stay, hospitalization, and the period of mechanical ventilation were found in our study and that of Recht and colleagues.⁹ This has important economic implications in these times of over-stretched health resources and the risk of acquiring nosocomial infections.

Prolonged CPB time has been suggested as a risk factor for the development of GI complications.^5,11,12 Off-pump CABG has been reported to reduce the inflammatory response after conventional CABG.¹³ Based on this argument, there is a belief that off-pump CABG can reduce end-organ damage, but the BHACAS study, the Octopus group study, and another recent study from Canada did not reveal any difference between the 2 approaches in terms of GI complications.^14–16 In our study, CPB was not found to be a significant risk factor for GI complications; 37% of patients with GI complications and 23.7% of those without GI complications underwent CABG with CPB. Multivariate analysis did not identify CPB as a significant factor. Hospital mortality among those who developed GI complications was also similar for the 2 surgical approaches. This study was unable to address the difference in pathogenesis of GI complications between conventional and off-pump CABG. However, a study of markers of gut injury (gastric intramucosal pH, tumor necrosis factor-α, interleukin-6, endotoxin) may identify factors involved in gut damage during and after CABG with and without CPB. This may improve our overall understanding of gut physiology during CABG.

Although the pathogenesis of GI complications after CABG is complex, the key factor implicated is reduced systemic blood flow with end-organ hypoperfusion, leading to anaerobic metabolism at the cellular level.¹⁷ The reasons for the increased susceptibility to hypoxia of the GI tract include its inability to autoregulate during periods of hypotension or persistent vasoconstriction during normotension. This is why low ejection fraction, peripheral vascular disease, and postoperative arrhythmias (noted in our study) predispose to mesenteric ischemia.⁶ Furthermore, low cardiac output requiring inotropic or balloon pump support is a predictor of mesenteric ischemia. The pathogenesis of ischemic and bleeding complications is different, and also of upper and lower GI bleeding, but the low incidence of GI hemorrhage in our series means that a much larger study is needed for a meaningful analysis.

This report included only patients who underwent CABG since 2001, with a relatively large sample size from a center with a complex case mix, but it was not randomized. On the basis of this study and previous reports, patients with identified risk factors should be alerted to the possibility of GI complications and the associated mortality, to ensure that informed consent has been obtained. Specific clinical measures tailored to individual patients at risk of developing GI complications should be implemented to optimize their hemodynamic status. The implementation of improved risk stratification with risk models, and the application of goal-directed hemodynamic therapy should be considered.^18,19 The early opinion of a GI surgeon regarding early intervention should be sought. A diagnostic laparoscopy for quick inspection carries less morbidity than an exploratory laparotomy. GI complications after CABG may occur at a higher rate than previously described.²⁰ With CABG being performed in older and sicker patients, a higher level of vigilance is mandatory to ensure swift referral of suspected cases to specialists for early diagnosis and rapid treatment of GI complications.

Footnotes

Acknowledgments

We would like to thank Michael Roughton, statistician at the Royal Brompton Hospital, London, and Joy Spanner who maintains the Patients Analysis and Tracking System database.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statement

None declared.

References

Ott

Buchman

Baumgartner

(1995) Postoperative abdominal complications in cardiopulmonary bypass patients: a case-controlled study. Ann Thorac Surg 59: 1210–1213.

Musleh

Patel

Grayson

Pullan

Keenan

Fabri

(2003) Off-pump coronary artery bypass surgery does not reduce gastrointestinal complications. Eur J Cardiothorac Surg 23: 170–174.

The Society of Thoracic Surgeons (1996) Data Analyses of The Society of Thoracic Surgeons National Cardiac Surgery Database, January 1996. Summit Medical Systems.

Christenson

Schmuziger

Maurice

Simonet

Velebit

(1994) Gastrointestinal complications after coronary artery bypass grafting. J Thorac Cardiovasc Surg 108: 899–906.

Zacharias

Schwann

Parenteau

Riordan

Durham

Engoren

(2000) Predictors of gastrointestinal complications in cardiac surgery. Tex Heart Inst J 27: 93–99.

Rady

Kodavatiganti

Ryan

(1998) Perioperative predictors of acute cholecystitis after cardiovascular surgery. Chest 114: 76–84.

D’Ancona

Baillot

Poirier

Dagenais

de Ibarra

Bauset

(2003) Determinants of gastrointestinal complications in cardiac surgery. Tex Heart Inst J 30: 280–285.

Lefebvre

Ferre

Watson

Brown

Serthelon

Laroute

(2001) Small bowel motility and colonic transit are altered in dogs with moderate renal failure. Am J Physiol Regul Integr Comp Physiol 281: R230–R238.

Recht

Smith

Woods

Engel

Hiratzka

(2004) Predictors and outcomes of gastrointestinal complications in patients undergoing coronary artery bypass graft surgery: a prospective, nested case-control study. J Am Coll Surg 198: 742–747.

10.

Vohra

Goldsmith

IRA

Rosin

Briffa

Patel

(2005) The predictors and outcome of recidivism in cardiac ICUs. Eur J Cardiothorac Surg 27: 508–511.

11.

Fitzgerald

Kim

Karakozis

Alam

Provido

Kirkpatrick

(2000) Visceral ischemia after cardiopulmonary bypass. Am Surg 66: 623–626.

12.

Halm

(1996) Acute gastrointestinal complications after cardiac surgery. Am J Crit Care 5: 109–118.

13.

Matata

Sosnowski

Galinanes

(2000) Off-pump bypass graft operation significantly reduces oxidative stress and inflammation. Ann Thorac Surg 69: 785–791.

14.

Angelini

Taylor

Reeves

Ascione

(2002) Early and midterm outcome after off-pump and on-pump surgery in Beating Heart Against Cardioplegic Arrest Studies (BHACAS 1 and 2): a pooled analysis of two randomized controlled trials. Lancet 359: 1194–1199.

15.

van Dijk

Nierich

Jansen

Nathoe

Suyker

Diephuis

(2001) Octopus Study Group. Early outcome after off-pump versus on-pump coronary bypass surgery. Results from a randomized study. Circulation 104: 1761–1766.

16.

Croome

Kiaii

Fox

Quantz

McKenzie

Novick

(2009) Comparison of gastrointestinal complications in on-pump versus off-pump coronary artery bypass grafting. Can J Surg 52: 125–128.

17.

Gaer

Shaw

Wild

Swift

Munsch

Smith

(1994) Effects of cardiopulmonary bypass on gastrointestinal perfusion and function. Ann Thorac Surg 57: 371–375.

18.

Andersson

Brandt

Höglund

Algotsson

Nilsson

(2010) Gastrointestinal complications after cardiac surgery—improved risk stratification using a new scoring model. Interact Cardiovasc Thorac Surg 10: 366–370.

19.

Giglio

Marucci

Testini

Brienza

(2009) Goal-directed haemodynamic therapy and gastrointestinal complications in major surgery: a meta-analysis of randomized controlled trials. Br J Anaesth 103: 637–646.

20.

Rodriguez

Nguyen

Galanko

Morton

(2007) Gastrointestinal complications after coronary artery bypass grafting: a national study of morbidity and mortality predictors. J Am Coll Surg 205: 741–747.