Gastrointestinal Emergencies in Cardiac Surgery

Original Research Cardiology 2008;111:94–101 DOI: 10.1159/000119696

Received: August 5, 2007 Accepted after revision: October 26, 2007 Published online: March 31, 2008

Gastrointestinal Emergencies in Cardiac Surgery A Retrospective Analysis of 3,724 Consecutive Patients from a Single Center

Ioannis Vassiliou a Emmanouel Papadakis c Nikolaos Arkadopoulos a Kassiani Theodoraki b Athanasios Marinis a Theodosios Theodosopoulos a George Palatianos c Vassilios Smyrniotis a Departments of a Surgery and b Anesthesiology, Aretaieio University Hospital, Athens Medical School, University of Athens and c Department of Cardiac Surgery, Onassis Cardiac Surgery Center, Athens, Greece

Key Words Complications, gastrointestinal ⴢ Operation, cardiac ⴢ Ischemia ⴢ Cholecystitis

Abstract Objectives: The aim of this study is to retrospectively analyze risk factors, diagnosis and management of gastrointestinal (GI) complications following cardiac operations. Methods: Patients who developed GI complications after a cardiac operation were studied. Anesthesia protocols, techniques of cardiac surgery, potential risk factors, complications and medical and surgical interventions were reviewed and analyzed. Results: Out of 3,724 consecutive patients undergoing heart operations during an 8-year period, 33 patients developed GI complications. Eleven patients developed ischemic colitis, 8 cholecystitis, 6 GI bleeding, 4 liver failures, 3 pancreatitis and 1 esophageal hernia. Patients with GI complications had a lower mean ejection fraction compared to patients not developing these complications (45.1 vs. 49.7%, p ! 0.01). Also, patients undergoing an urgent cardiac operation were significantly more likely (3.49 times more likely) to develop GI complications postoperatively. Of the 33 affected patients, 18 were treated conservatively and

© 2008 S. Karger AG, Basel 0008–6312/08/1112–0094$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/crd

15 underwent an emergency exploratory laparotomy. Overall mortality was 12% (4 patients). Conclusions: Intestinal ischemia and cholecystitis appear to be the most frequent GI complications associated with cardiac surgery. Risk factors include a low ejection fraction and an urgent cardiac operation. Early recognition and treatment of these complications may reduce mortality. Copyright © 2008 S. Karger AG, Basel

Introduction

Gastrointestinal (GI) complications after cardiac surgery are infrequent but can be associated with significant morbidity and mortality. Some investigators have already identified risk factors for such complications and suggested preventive and therapeutic strategies [1–9]. Based on the identification of risk factors, cardiac surgeons are asked to modify their operative strategy and early postoperative management. The aim of this study is to retrospectively analyze GI complications from a single cardiac surgery unit. Unlike previous reports, all GI complications in this series were evaluated and managed by the same surgical team. Ioannis Vassiliou, MD, FACS 29 El. Venizelou St. GR–15451 Athens (Greece) Tel. +30 210 728 6156, Fax +30 210 728 6128 E-Mail [email protected]

Materials and Methods In the present study, we retrospectively analyzed the case records of 3,724 consecutive patients who underwent cardiac surgery in the 3rd Department of Surgery of the Onassis Cardiac Surgery Center during an 8-year period (January 1995 to December 2002). Out of these, 2,639 (70.9%) underwent coronary artery bypass grafting (CABG), 961 (25.8%) cardiac valve surgery and 123 (3.3%) combined procedures or various aneurysm operations. Among coronary patients, 2,164 (82%) were treated with ␱n-pump techniques, while 475 (18%) were treated with off-pump techniques. Anesthesia A standard anesthetic protocol was applied in all cases. Patients received preoperative medication that included oral lorazepam (1 mg) 90 min before surgery, morphine sulfate (0.1 mg kg–1 i.m.) and ondansetron (8 mg i.v.) 1 h before surgery. Each patient had a radial artery catheter and a pulmonary artery catheter inserted for continuous pressure monitoring. Anesthesia was induced with etomidate (0.3 mg kg–1), midazolam (0.05–0.075 mg kg–1) and fentanyl (10–15 ␮g kg–1). Muscle relaxation was accomplished by intravenous bolus administration of pancuronium (0.15 mg kg–1) and was maintained by continuous infusion of cisatracurium (1.5–2 ␮g kg–1). Nasogastric tubes were routinely placed in all patients and antacids and H2 blockers were administered for prophylaxis against stress ulcerations. Anesthesia was maintained using sevoflurane (0.5–2% in oxygen/air) and additional bolus doses of fentanyl as needed, or sevoflurane (0.5–2% in oxygen/air) and a continuous infusion of remifentanil (0.2–1 ␮g kg–1 min–1). Cefuroxime (1.5 g) was administered intravenously after the induction of anesthesia. Cardiac output was measured using the thermodilution technique. Ventilation was discontinued when ventricular ejection stopped during cardiopulmonary bypass (CPB). However, lungs were partially inflated with an appropriate oxygen flow to maintain a continuous positive airway pressure of 5 mm Hg. Surgical Techniques On-Pump Technique. CBP was instituted through aortocaval cannulation. The CPB circuit was open and uncoated. It consisted of a multiflow roller pump (Stockett-Shilley, Munich, Germany), a microporous, polypropylene membrane oxygenator (Quadrox; Jostra, Hirrlingen, Germany) and an arterial filter (Safeline; Jostra). Priming was performed with 1,500 ml of Ringer’s lactate solution plus 500 ml of 10% hydroxyethyl starch solution. Anticoagulation was achieved with intravenous heparin (300 U kg–1) 5 min before initiation of CPB and during perfusion, as needed. The activated coagulation time was maintained longer than 480 s throughout CPB. Myocardial preservation during aortic cross-clamping (ACC) was achieved with cold blood cardioplegic solution in anterograde-retrograde fashion. Reconstitution of the crystalloid cardioplegic component was made with 40 or 60 ml sterile concentrate for cardioplegia infusion (Martindale Pharmaceutical, Romford, UK) diluted immediately before use in 1 liter of Ringer’s solution in which potassium chloride (2–4 g) and sodium bicarbonate (20 mEq) were added. The final crystalloid solution was mixed with blood (1 part of crystalloid cardioplegic solution with 4 parts of blood) and was infused in the myocardium in a retrograde manner through the coronary sinus and in an antegrade manner through the aortic root at 8 ° C. Perfusion was

Gastrointestinal Emergencies in Cardiac Surgery

carried out with pump flow at 2.4 liters min–1 m–2 and with moderate hypothermia (28–30 ° C). During the procedure, the mean arterial pressure ranged between 65 and 75 mm Hg. After CPB, protamine chloride was administered at a dose of 3 mg kg–1 to reverse the effect of heparin. Hemodynamic instability throughout the course of the operation was prevented or treated with fluids, inotropic support or both, as needed. Off-Pump Coronary Artery Bypass Technique. Anticoagulation was achieved with 2 mg kg–1 heparin and the aPTT was maintained above 300 s. The heart was stabilized with a suction tissue stabilization system (Octopus; Medtronic Inc., Minneapolis, Minn., USA). A deep pericardial retraction suture was placed at the posterior fibrous pericardium medial to the proximal part of the inferior vena cava to help manipulate and rotate the heart to vertical and lateral positions. A right-sided pericardial incision directed towards the inferior vena cava was selectively performed to facilitate venous return. Intracoronary shunts were used occasionally. During the procedure, mean arterial pressure was maintained at 60–90 mm Hg, using small doses of inotropes (i.e. neosynephrine) if needed. Both surgical techniques were performed through standard midline sternotomy. Selection of Patients GI complications were defined as any postoperative dysfunction of the gastrointestinal system that required therapeutic intervention by means of exploratory laparotomy or for which the patient required prolonged hospitalization. All patients with a suspected GI complication were evaluated and managed by the same surgical team, led by I. Vassiliou, one of the authors. Pancreatitis was diagnosed in patients who presented with acute abdominal pain and had elevated serum amylase and lipase activities. The diagnosis of pancreatitis was confirmed with a CT scan in all cases. GI complications developed postoperatively in a total of 33 patients (tables 1, 2). Patients were divided into 2 groups, according to the selected treatment: medical treatment and laparotomy (tables 1, 2). The vast majority of them were operated with on-pump techniques, except for 2 coronary patients in the laparotomy group and 3 in the medical group who were operated using the off-pump coronary artery bypass (OPCAB) technique. The medical treatment group consisted of 18 patients who received conservative treatment only (table 1). These patients displayed clinical signs of GI complications while on the weaning process from sedative drugs and the ventilator. All patients were kept on a fast, had a nasogastric tube inserted and received intravenous administration of fluids and electrolytes, as well as acid antisecretory and mucosal protective agents. Patients were closely monitored in the intensive care unit (ICU). In the vast majority of cases the laboratory and clinical findings of GI dysfunction improved within the first 24–48 h. The laparotomy group consisted of 15 patients who underwent emergency abdominal procedures (table 2). Laparotomy was performed as soon as a diagnosis of GI complication had been established, or when clinical signs and symptoms failed to improve with medical treatment alone within the first 24–48 h. Statistical Analysis The association between the patients’ characteristics and GI complications was assessed by using the ␹2 test for categorical

Cardiology 2008;111:94–101

95

Table 1. Patient characteristics in the medical treatment group

Number

Age

Gender

NYHA stage

Urgent operation

Redo

Operation

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

51 74 57 65 64 63 73 62 60 64 76 75 66 58 54 69 68 72

M F F M M M M M M M M M M M M M M M

I I II I II I I I I II I I I III I I II II

– + – – + + – – – + – – + – – + + –

– – – – – – – – – – – – – – – – – –

CABG CABG CABG CABG CABG CABG CABG CABG AVR CABG CABG CABG CABG AVR CABG CABG CABG AVR

CPB time min 76 58 80 86 79 74 93 132 128 79 181 114 0 186 109 0 0 94

ACC time min GI complications 54 32 49 56 50 38 37 70 60 45 101 77 0 157 69 0 0 67

GI bleeding cholecystitis pancreatitis pancreatitis ischemic colitis cholecystitis liver failure cholecystitis cholecystitis ischemic colitis GI bleeding pancreatitis cholecystitis liver failure cholecystitis liver failure ischemic colitis liver failure

AVR = Aortic valve replacement.

Table 2. Patient characteristics in the laparotomy treatment group Number Age Gender

NYHA stage

Urgent operation

Redo

Operation

CPB time, min

ACC time, min

GI complications

Operative procedure

64 55

ileus acute cholecystitis ischemic bowel ischemic bowel ileus ischemic bowel upper GI bleeding ischemic bowel ischemic bowel, GI bleeding ischemic bowel acute cholecystitis ischemic bowel ileus upper GI bleeding upper GI bleeding

enterectomy cholecystectomy sigmoidectomy left hemicolectomy enterectomy right hemicolectomy vagotomy + pyloroplasty right hemicolectomy colectomy, antrectomy, vagotomy, cholecystectomy colectomy cholecystectomy colectomy Richter hernia vagotomy + antrectomy vagotomy + pyloroplasty

1 2 3 4 5 6 7 8 9

65 77 55 57 67 59 73 71 74

M F M M M M M M M

III II II I I I I I I

– – + – + – – + –

– – – – + + + – –

AVR CABG!3 CABG!2 CABG!2–IABP CABG!1 CABG!3 CABG!2–IABP CABG!2 CABG!2

81 80 OPCAB 80 OPCAB 91 179 38 37

62 126 55 57

10 11 12 13 14 15

75 50 67 47 64 67

M M M F M F

I I II III I III

+ – – – – +

– – – – + –

CABG!2 AVR CABG!3 Bentall CABG!3 Aortic aneurysm type A

113 128 126 188 184 280

65 91 28 122 106 250

54

AVR = Aortic valve replacement; IABP = intra-aortic ballon pump.

variables, and where appropriate Fisher’s exact p values are reported due to small sample sizes. Mann-Whitney tests were used for continuous variables. Univariate and multivariate relative risks for GI complications with 95% confidence intervals (CI) were estimated by the use of a univariate and multivariate logistic regression model, respectively. Those risk factors that were found

96

Cardiology 2008;111:94–101

to be significantly associated with GI complications in the univariate model were also examined simultaneously with a multivariate model. In this way, we could assess whether they were independently associated with GI complications. Values of p ! 0.05 were considered to be statistically significant.

Vassiliou et al.

Table 3. Demographic and operative characteristics of patients

Table 4. Univariate associations between potential risk factors

with or without GI complications

and GI complications

Demographic and clinical characteristics Age, years Gender, n Male Female NYHA stage, n I II III and IV Ejection fraction, % Operation procedure, n CABG Valve/other CPB time, min Median (range) ACC time, min Median (range) Operation, n Elective Urgent Reoperation, n No Yes

GI complications

p value

Risk factor

Univariate relative risk (95% CI)

p value

Agea Gender Males Females NYHA I II III and IV Ejection fractionb Operation procedure CABG Valve/other CPB timec ACC timed Operation Elective Urgent Reoperation No Yes

1.22 (0.98–1.06)

0.30

2.65 (0.82–9.1) 1.0

0.10

no (n = 3,691)

yes (n = 33)

63829.7

64.988.1

0.30

30 (91) 3 (9)

0.13

21 (64) 9 (27) 3 (9) 45.189.8

0.58

26 (79) 7 (21)

0.44

2,916 (79) 775 (21) 2,436 (66) 775 (21) 480 (13) 49.7811.7 2,613 (71) 1,078 (29)

0.01

95 (0–1,138) 93 (0–298)

0.96

62 (0–280)

59 (0–157)

0.42

3,172 (86) 519 (14)

21 (64) 12 (36)

0.001

3,438 (93) 253 (7)

28 (85) 5 (15)

0.07

Values are mean 8 SD and percentages are in parentheses unless otherwise stated.

Results

1.0 1.34 (0.62–2.99) 0.72 (0.21–2.35) 1.37 (1.11–1.85) 1.53 (0.66–3.54) 1.0 1.02 (0.55–1.45) 0.87 (0.67–1.14)

0.44 0.56 0.03 0.32 0.91 0.34

1.0 3.49 (1.71–7.14)

0.001

1.0 2.42 (0.93–6.33)

0.07

a Increased

age by 10 years. Decreased ejection fraction by 10%. c Increased CPB time by 120 min. d Increased ACC time by 30 min. b

Table 5. Multivariate association between ejection fractions, ur-

gent operations and GI complications

Out of 3,724 patients, 33 (0.88%) developed GI complications. Table 3 demonstrates demographic and clinical characteristics of patients with and without GI complications. The 2 groups of patients, those with and without GI complications, were different in terms of ejection fraction (p = 0.01) and the type of cardiac operation (urgent vs. elective, p = 0.001), while there was little evidence for a difference between those who had a reoperation compared to those who had not (p = 0.07). A decrease of 10% in the ejection fraction was significantly associated with an increased risk of GI complications (1.37 times higher or a 37% increase). Similarly, patients that underwent an urgent heart operation were 3.49 times more likely to develop GI complications than those undergoing an elective procedure. Age, gender, NYHA stage, type of cardiac operation, CPB time and ACC time were not associated with a statistically significant increase of relative risk of GI complications (table 4). Gastrointestinal Emergencies in Cardiac Surgery

Risk factor

Multivariate relative risk (95% CI)

p value

Ejection fractiona Operation Elective Urgent

1.35 (0.95–1.05)

0.10

1.0 3.38 (1.58–7.26)

0.002

a

Decreased ejection fraction by 10%.

Subsequently, the above risk factors were examined in a multivariate model to assess whether they were independently associated with GI complications (table 5). These results demonstrated that only an urgent operation is independently associated with an increased risk of GI complications. More specifically, patients operated urgently were 3.38 times more likely to develop GI complications than those who were not. Cardiology 2008;111:94–101

97

Table 6. Median CPB times for each type

of operative procedure

CPB time

GI complications, median (range)

CABG, min Valve/other, min a

Table 7. Odds of being in the laparotomy

group for each GI risk factor

yes

84 (0–1,138) 132 (0–460)

85.5 (0–181) 179 (81–298)

p value

1.74a (0.43–2.57) 2.12a (0.55–7.56)

0.90 0.24

Relative risk (95% CI)

p value

Increased CPB time by 120 min.

GI risk factors

Operation Elective Urgent Ejection fractiona a

no

Relative risk (95% CI)

Treatment medical (n = 18)

laparotomy (n = 15)

11 (61%) 7 (39%) 45.5811.5

10 (67%) 5 (33%) 44.587.5

1.0 0.78 (0.18–3.29) 1.12 (0.56–2.35)

0.74 0.78

Decreased ejection fraction by 10%.

Table 8. Number of deaths categorized by GI complications and treatment group

GI complication

GI bleeding (n = 6) Cholecystitis (n = 8) Ischemic colitis (n = 11) Liver failure (n = 4) Pancreatitis (n = 3) Postoperative esophageal hiatal hernia (n = 1) Total, patients treated/deaths (%)

There was no difference demonstrated in the length of CPB time between those patients who manifested GI complications and those who did not (table 6). Additionally, relative risk analysis showed that for those undergoing a CABG procedure, an increase in CPB time by 120 min was associated with an increased relative risk of 1.74 (74% increase) for GI complications, a finding that was not statistically significant. For those that underwent a valve replacement or other procedure, an increase in CPB time by 120 min was associated with an increased risk (2.12, an increase of more than double but not statistically significant) of GI complications.

98

Cardiology 2008;111:94–101

Number of patients treated/number of deaths (mortality rate, %) medical group

laparotomy group

2/0 (0) 6/1 (17) 3/1 (33) 4/0 (0) 3/0 (0) –

4/1 (25) 2/0 (0) 8/1 (12.5) – – 1/0 (0)

18/2 (7)

15/2 (13)

Overall disease mortality % 17 13 18 0 0 0 33/4 (12)

We also examined the association between risk factors (operation urgency and ejection fraction) and treatment modalities (table 7). There was no significant association between the examined risk factors and the subsequent treatment. Patients undergoing an urgent cardiac operation were 22% less likely to be in the laparotomy group than the medical group (p = 0.74). Also, a decreased ejection fraction by 10% was associated, though not significantly, with an increased odds ratio of 1.12 (95% CI: 0.56– 2.35, p = 0.78) of being in the laparotomy rather than in the medical group. The treatment protocol in the medical therapy group appeared effective; 16 of 18 patients responded to therapy. Vassiliou et al.

There were 2 deaths (11.1%) in this group of patients due to complicated ischemic colitis and cholecystitis (table 8). Diagnosis was delayed due to atypical clinical manifestations, and the patients were managed conservatively and finally died after 15 days in the ICU from sepsis and multiorgan failure. The laparotomy treatment group consisted of 15 patients who underwent emergency abdominal procedures for GI complications. There were also 2 deaths (13.3%) in this group: Case I: A 71-year-old male patient (No. 8, table 2) underwent an urgent double CABG operation. He had an uncomplicated postoperative course until the 6th postoperative day when he developed acute abdominal pain. The diagnosis of ischemic bowel disease was delayed due to nonspecific clinical manifestations. However, due to worsening sepsis an urgent right hemicolectomy was performed. The patient died 2 days postoperatively. Case II: A 67-year-old female patient (No. 15, table 2) was operated urgently for a type A aortic aneurysm (Stanford classification). Unfortunately, the patient’s postoperative course was complicated with multiorgan failure and emboli affecting the toes. On the 14th postoperative day, while still in the ICU, she developed massive upper GI bleeding and underwent an urgent vagotomy and pyloroplasty. Postoperatively, the patient remained unresponsive, was hospitalized in the ICU and finally died 3 months later from overwhelming sepsis.

Discussion

Intra-abdominal complications after heart surgery are rare, with an incidence ranging from 0.53 to 5.5% [1–9] and mortality rates varying from 13.9 to 44% [2, 4, 5, 7, 9]. Our series had a 0.88% incidence of GI complications and a 12% mortality rate, both of which are consistent with published reports. Although complex pathogenetic mechanisms are implicated in GI complications of cardiac surgery, reduced systemic blood flow seems to be the major event underlying almost every complication. Atheromatosis of the splanchnic vascular network, a common finding in CPB patients, can also play an important role. It has been reported that blood flow to the jejunal mucosa decreases by 40% and blood flow to the serosa decreases by 50% upon the institution of a hypothermic CPB. However, during rewarming, oxygen delivery further decreases by approximately 50% while oxygen consumption increases, with a proportionate increase in inGastrointestinal Emergencies in Cardiac Surgery

testinal mucosal pH [10]. Eventually, macrovascular embolism or thrombosis (such as proximal superior mesenteric artery embolism) or microvascular disease can be implicated in mesenteric ischemia, pre-, intra- and postoperatively. In on-pump procedures, gut ischemia also seems to be due to the nonpulsatile flow of the pump during CPB, which acts as a potent stimulus that releases endogenous splanchnic vasoconstrictors and activates platelets and leukocytes to form cellular aggregates capable of occluding vessels in the microcirculation. In addition, on-pump CPB techniques contribute to the development of systemic inflammatory response syndrome, due mostly to the interaction of blood with the nonendothelial surfaces of the CPB circuits. The use of pulsatile flow during CPB has been shown to be associated with a lower incidence of GI complications [10]. Off-pump techniques appear to significantly reduce the inflammatory reactions associated with the on-pump procedures and may be expected to reduce complication rates [11]. Preoperative patient comorbidities, such as a low left ventricular ejection fraction and peripheral vascular disease [12], as well as risk factors for a prothrombotic state (history of smoking or COPD), embolic event (type II heparin-induced thrombocytopenia, atrial fibrillation) or indicators of microvascular disease (prior cerebrovascular accident, peripheral vascular disease), are all implicated in causing splanchnic hypoperfusion and strongly predict the occurrence of GI complications. In our study, a reduced ejection fraction was significantly associated with an increased risk of GI complications. It appears that the resulting reduction in blood flow leads to inappropriate oxygen delivery, a state that the GI tract is not able to compensate for, due to its inability to autoregulate; furthermore, splanchnic hypoperfusion due to persistent vasoconstriction may be sustained even after restoration of hemodynamic stability. Moreover, according to our findings, the urgent nature of the cardiac operation was an additional risk factor for the occurrence of GI complications. It seems that the preoperative workup of the patient’s clinical status is suboptimal in the case of an urgent heart operation. The patient may be brought to surgery severely hypovolemic and not pharmacologically pretreated to optimize cardiac output. Such factors can lead to intraoperative hypoperfusion. Moreover, in urgent cases intraoperative vasoconstrictor needs may be higher compared to elective operations where more adequate patient preparation is the norm. In fact, parameters like the ones described, such as hypovolemia, prolonged CPB and administration of vasoconstrictors have been shown to correlate strongly with reduced GI blood perfuCardiology 2008;111:94–101

99

sion, mucosal acidification and increased GI tract permeability [13–15]. Injury of the bowel, gallbladder, liver or pancreas during CPB can result from regional splanchnic malperfusion secondary to hypoperfusion, vasoactive substances, cytotoxins, microemboli and angiotensin-IImediated splanchnic vasoconstriction [16, 17]. In our study, the type of procedure, CPB time, ACC time and reoperation appeared to increase the incidence of GI complications although the associations were not statistically significant, probably due to the small complication rate. Although rare, GI complications following coronary revascularization are invariably associated with high morbidity and mortality [5, 18, 19]. The most common include GI bleeding (especially if anticoagulants are administered), bowel ischemia and pancreatitis. Perforation of a duodenal ulcer, pseudomembranous colitis, hepatic failure and cholecystitis occur less frequently. Hepatic injury may also occur from manual compression of the liver, a technique that is sometimes used in order to augment venous return. The spectrum of GI complications reported in the literature is similar to the spectrum in our series, in which ischemic colitis was the most common. The most devastating complication is intestinal ischemia with a high mortality varying from 64 to 80% [18, 19]. In fact, ischemic colitis was a cause of death in 3 out of 4 patients who died of GI complications in our series. Specifically, 2 patients died from overwhelming sepsis due to delayed diagnosis of an ischemic bowel, 1 patient died after conservative treatment of ischemic colitis and acute cholecystitis and, finally, another patient died after developing multiorgan failure and a massive upper GI hemorrhage, despite early surgical intervention. Early recognition of a postoperative intra-abdominal complication and prompt consultation from the GI sur-

geon is crucial for the improvement of morbidity and mortality rates of patients undergoing heart operations. Early complaint of abdominal pain is a very sensitive clinical indicator of GI disorders. Other clinical indices that should prompt efforts for early recognition of intestinal ischemia include abdominal bloating, persistent ileus, sepsis or lower GI bleeding. The surgeon has to identify the presence of every possible risk factor and keep a heightened awareness. Concerning intubated and sedated patients, a delay in diagnosis is common, despite markers such as lactate, which is usually increased in very advanced disease. In fact, diagnosis was delayed in 3 out of the 4 patients in our series who eventually died, due to atypical clinical manifestations. Especially in the case of intestinal ischemia, the diagnostic workup should aim to provide the surgeon with a safe diagnosis, in order to schedule prompt surgical intervention when necrotic bowel is suspected. Some patients will require no surgical intervention and treatment may be confined to GI endoscopy and medical management. In conclusion, suspicion of the development of postoperative GI complications in patients undergoing heart operations should be based upon risk stratification of patients and early postoperative abdominal complaints. Early diagnosis of an intra-abdominal complication should prompt the GI surgeon to decide between further surgical intervention or appropriate medical therapy, in order to improve patient outcome and reduce mortality rates. Optimization of anesthetic and surgical procedures, extensive utilization of off-pump techniques and improvement of postoperative critical care can all contribute to the prevention of GI complications and their associated mortality and morbidity.

References 1 Bolcal C, Iyem H, Sargin M, Mataraci I, Sahin MA, Temizkan V, et al: Gastrointestinal complications after cardiopulmonary bypass: sixteen years of experience. Can J Gastroenterol 2005;19:613–617. 2 Mangi AA, Christison-Lagay ER, Torchiana DF, Warshaw AL, Berger DL: Gastrointestinal complications in patients undergoing heart operation: an analysis of 8,709 consecutive cardiac surgical patients. Ann Surg 2005;241:895–901.

100

3 McSweeney ME, Garwood S, Levin J, Marino MR, Wang SX, Kardatzke D, et al: Adverse gastrointestinal complications after cardiopulmonary bypass: can outcome be predicted from preoperative risk factors? Anesth Analg 2004;98:1610–1617. 4 D’Ancona G, Baillot R, Poirier B, Dagenais F, de Ibarra JI, Bauset R, et al: Determinants of gastrointestinal complications in cardiac surgery. Tex Heart Inst J 2003;30:280–285.

Cardiology 2008;111:94–101

5 Perugini RA, Orr RK, Porter D, Dumas EM, Maini BS: Gastrointestinal complications following cardiac surgery. An analysis of 1,477 cardiac surgery patients. Arch Surg 1997;132:352–357. 6 Grap MJ, Savage L, Ball GB: The incidence of gastrointestinal symptoms in cardiac surgery patients through six weeks after discharge. Heart Lung 1996;25:444–450. 7 Yilmaz AT, Arslan M, Demirkilic U, Ozal E, Kuralay E, Bingol H, et al: Gastrointestinal complications after cardiac surgery. Eur J Cardiothorac Surg 1996; 10:763–767.

Vassiliou et al.

8 Johnston G, Vitikainen K, Knight R, Annest L, Garcia C: Changing perspective on gastrointestinal complications in patients undergoing cardiac surgery. Am J Surg 1992; 163: 525–529. 9 Krasna MJ, Flancbaum L, Trooskin SZ, Fitzpatrick JC, Scholz PM, Scott GE, et al: Gastrointestinal complications after cardiac surgery. Surgery 1988;104:773–780. 10 Ohri SK, Becket J, Brannan J: Effects of cardiopulmonary bypass on gut blood flow, oxygen utilization and intramucosal pH. Ann Thorac Surg 1994;57:1193–1199. 11 Matata MB, Sosnowski WA, Galinanes M: Off-pump bypass graft operation significantly reduces oxidative stress and inflammation. Ann Thorac Surg 2000;69:785–791.

Gastrointestinal Emergencies in Cardiac Surgery

12 Zacharias A, Schwann TA, Parenteau GL, Riordan CJ, Durham SJ, Engoren M, et al: Predictors of gastrointestinal complications. Tex Heart Inst J 2000;27:93–99. 13 Winter DC, O’Sullivan GC, Harvey BJ: Direct effects of dopamine on colonic mucosal pH: implications for tonometry. J Surg Res 1999;83:62–68. 14 Heikkinen LO, Ala-Kulju KV: Abdominal complications following cardiopulmonary bypass in open-heart surgery. Scand J Thorac Cardiovasc Surg 1987;21:1–7. 15 Halm MA: Acute gastrointestinal complications after cardiac surgery. Am J Crit Care 1996;5:109–118.

16 Reilly PM, Bulkley GB: Vasoactive mediators and splanchnic perfusion. Crit Care Med 1993;21(suppl 2):S55–S68. 17 Plestis KA, Gold JP: Importance of blood pressure regulation in maintaining adequate tissue perfusion during cardiopulmonary bypass. Semin Thorac Cardiovasc Surg 2001; 13:170–175. 18 Huddy SP, Joyce WP, Pepper JR: Gastrointestinal complications in 4,473 patients who underwent cardiopulmonary bypass surgery. Br J Surg 1991;78:293–296. 19 Byhahn C, Strouhal U, Martens S: Incidence of gastrointestinal complications in cardiopulmonary bypass patients. World J Surg 2001;25:1140–1144.

Cardiology 2008;111:94–101

101