• 
• 

Three Part Question

In [patients undergoing cardiac surgery] does [pericardial closure] affect [outcome]?

Clinical Scenario

You have been trained to leave the pericardium open after a routine cardiac surgery procedure because in the early postoperative period the patient's haemodynamic performance is better and there is less incidence of graft failure. In addition there is also said to be a reduced incidence of cardiac tamponade. You begin to question this teaching, especially in view of the benefit of a closed pericardium when it comes to re-do surgery. You decide to scrutinise the published literature with regard to the pitfalls of closing the pericardium.

Search Strategy

Medline 1966–Nov 2004 using the OVID interface
[exp thoracic surgery OR exp cardiac surgical procedures OR heart surgery.mp] AND [exp pericardium OR pericardial.mp] AND [clos$.mp]

Search Outcome

Using the above search strategy 240 publications were found of which 8 were deemed to be relevant. Two of the publications were not directly included as they were letters commenting on the identified studies. One study was in an animal model and was excluded on this basis. No additional papers were identified by widening the search strategy or by looking in the references section of the identified papers. These papers are included in the table.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Rao et al 1999 Canada	42 patients undergoing elective, isolated coronary artery bypass grafting were randomized into two groups: 20 patients underwent closure of the pericardium (closure group) and the pericardium was left open in 22 patients (open group)	Prospective randomised controlled study (Level 1b)	Distance between the epicardial surface and the posterior table of the sternum on CXR	Larger in the closure group compared to the open group at 1 week and 3 months postoperatively (P<0.001).	The patients involved in this study were low-risk, elective patients with preserved preoperative left ventricular function
			Cardiac index and stroke work	Cardiac index and stroke work index in the early postoperative period was lower in the Closure group compared to the Open group (P<0.001) despite similar filling pressures.
Bhatnagar et al 1998 Canada	302 patients having coronary artery bypass graft surgery. Pericardium closed with a gortex membrane [GM] n=138 Pericardium left open [PO] n=164	Prospective randomized controlled study (level 1b)	Mortality	GM n=2, PO n=2	Post operative bleeding was not defined
			Complication	GM 14%, PO 21% P = not significant
			Post operative ischaemic event	GM 2.9%, PO 4.9% P = not significant
			Bleeding	GM 0, PO 1.2% p = not significant
Izzat et al 1994 UK	Patients undergoing an open heart valve procedure. N = 10 Effects of a tension-free pericardial closure technique were studied. Opening the pericardium (1.5 to 2 h after the end of the operation) while the chest remained closed	Prospective Case Series (level 2b)	Cardiac output	8% reduction, p = not significant	Small study Pericardial opening was performed 2 h after the end of surgery No control group
			Systematic vascular resistance	15% reduction, p = not significant
			Mean arterial pressure	13% reduction (p = 0.03)
			Ejection fraction	no change
			Systolic and diastolic left ventricular dimensions	Decreased by 6% and 4% respectively p = not significant
			Opening the pericardium (1.5 to 2 h after the end of the operation)	Not followed by significant change in any of the hemodynamic or echocardiographic variables
Daughters et al 1992	10 patients between 11 and 15 h after cardiac operations, with the pericardium first closed and then open.	Prospective Case Series (Level 2b)	Haemodynamic studies	End-diastolic volume index, peak positive time derivative of pressure, stroke work index, and cardiac index all increased significantly when the pericardium was opened (P<0.001). At physiological pressures, the pericardium had a significant constraining effect on diastolic filling of the left ventricle, and opening of the pericardium resulted in increased cardiac index and stroke work index	Cardiac output studies were performed at varying time intervals No control group
Hunter et al 1992 UK	10 patients who underwent open-heart valve operations. Study of the haemodynamic effect of pericardial closure. Observations were made both while the pericardium was open and after it had been closed, then after closure of the chest after the pericardium had been reopened by removing the pericardial suture through the chest wall.	Prospective Case Series (Level 2b)	Closing the pericardium before closing the chest: cardiac output	Immediate reduction in cardiac output (thermodilution of 1.39 ± 0.24 l/min from 5.09±0.40 l/min (P<0.001).	Small study. No control group. Patients entered into the study acted as their own control group.
			Heart rate	Remained stable
			Stroke volume	Decrease of 29% (P<0.01).
			Systematic vascular resistance	Increase of 34% (P<0.01)
			Mean arterial pressure	Increased of 2% (P = not significant)
			Opening the pericardium (1.5 to 2 h after the end of the operation) while the chest remained closed: cardiac outpupt	Increase in cardiac output of 1.33±0.15 l/min from 4.12± 0.62 l/min (P<0.001).
			Heart rate	P=not significant
			Mean arterial blood pressure	P=not significant
			Stroke volume	Increase of 15±3 ml from 53±5 ml (P<0.01)
			Systematic vascular resistance	Reduction of 473±83 dynes.s.cm-5 from 1,721±181 dyne.s.cm-5 (P<0.01)
Damon and Bolton 1989 Netherlands	30 patients with normal left ventricular function undergoing coronary artery bypass surgery	Prospective Case Series (Level 2b)	MAP, CI, Mean right atrial pressure, Pulmonary Capillary wedge pressure	Closure of the pericardium resulted in decreases in arterial blood pressure (P<0.01), cardiac index (P<0.001), mean right atrial (P<).001), mean pulmonary artery (P<0.001) and pulmonary capillary wedge pressure (P<0.001)	No control group
Jarvinen 1987	Acute haemodynamic effects of a routine pericardial closure after cardiopulmonary bypass was studied in 29 patients undergoing cardiac surgery Coronary artery bypass: N=18 Aortic valve+coronary artery bypass: N=6 Mitral valve+coronary artery bypass: N=1 Mitral valve: N=1 Aortic valve: N=2 Aortic valve+Mitral valve: N=1	Prospective Case Series (level 2b)	Cardiac output after pericardial closure	8% decrease (P<0.01) while cardiac index remained normal (2.9 l/min/m2±0.6 S.D.)	Small study No analysis of subgroups Mixed cohort of patients
			Central venous pressure	Increased from 8±2 mmHg to 9±3 mmHg (P<0.05) after pericardial closure and decreased to 7±3 mmHg (P<0.05) when the pericardium was reopened.
			Left ventricular end-diastolic cavity diameter by echocardiography	Decreased in 19 patients from 46±6 mm to 41± 5 mm (P<0.01) when the pericardium was closed, and increased to 45±6 mm (P<0.01) after re-opening of the pericardiotomy incision.
Nandi et al 1976 UK	821 patients who underwent open heart surgery. 527 cases of congenital heart disease. 278 cases of acquired heart disease. 596 cases the pericardium was left open. 225 cases the pericardium was closed.	Retrospective Cohort Study (Level 4)	Requirement for reoperation due to bleeding/tamponade	Pericardium open 6.87%. Pericardium closed 1.77%. (No level of significance stated).	Study is limited to patients who have not had coronary artery bypass graft surgery. Restrospective study. Age range 8 months to 61 years.
			Mortality in reoperation group	17.77% open vs 0% closed
			Tracheostomy	22.2% open vs 0% closed
			Wound infection	8.8% open vs 0% closed

Comment(s)

Various institutions have attempted to answer the question we posed. Only two groups implemented a prospective randomized study. Rao et al randomized 42 patients who were having coronary artery bypass grafting to pericardial closure or leaving the pericardium open. They found that cardiac index and stroke work index were lower in the closure group compared to the open group (P<0.001), however, these difference were only present for one hour post operatively and at 4 h and 8 h post operatively no difference could be determined. Bhatnagar et al conducted a prospective randomised study to assess the impact of a tension-free pericardial closure with the use of a gortex membrane and found no significant difference with or without its use on early mortality, complications, bleeding or post operative ischaemic events in a cohort of patients who had all had coronary artery bypass graft surgery. Bhatnagar et al did not report any data regarding differences between groups in terms of their haemodynamic performance in the early post operative period. The adverse impact of pericardial closure haemodynamically was confirmed by several of the other studies [Rao, Daughters, Hunter, Izzat, Jarvinen, Damen]. No study reported an adverse clinical outcome due to the closure of the pericardium. Daughters et al measured cardiac output and stroke work index in patients immediately after operation and found that removal of the pericardial suture immediately improved left ventricular haemodynamics. This finding raises concerns about pericardial closure in patients with marginal preoperative left ventricular function or in those patients with postoperative ventricular dysfunction who require high preloads to maintain cardiac output. Only three studies [Rao, Bhatnagar, Damen] concentrated on patients who had coronary artery bypass graft surgery, two studies included a mixture of cases [Jarvinen,Nandi], and the remainder included only patients who had had valve surgery [Daughters, Hunter, Izzat, Damen]. None of the studies followed up patients to find out if the mortality was lower in patients having re-sternotomy with a closed or open pericardium.

Clinical Bottom Line

The adverse haemodynamic impact of pericardial closure is confirmed in several studies; however, no study has yet reported an adverse clinical outcome due to the closure of the pericardium.

References

1. Rao V, Komeda M, Weisel RD, Cohen G, Borger MA, David TE. Should the pericardium be closed routinely after heart operations? Ann Thorac Surg 1999;67:484–488.
2. Bhatnagar G, Fremes SE, Christakis GT, Goldman BS. Early results using an ePTFE membrane for pericardial closure following coronary bypass grafting. J Card Surg 1998;13:190–193.
3. Izzat MB, Anderson M, Wilde P, Wisheart JD, Bryan AJ, Angelini GD. Hemodynamic effects and echocardiographic consequences of tension-free pericardial closure after heart valve surgery. J Heart Valve Dis 1994;3:295–299.
4. Daughters GT, Frist WH, Alderman EL, Derby GC, Ingels NB Jr., Miller DC. Effects of the pericardium on left ventricular diastolic filling and systolic performance early after cardiac operations. J Thorac Cardiovasc Surg 1992;104:1084–1091.
5. Hunter S, Smith GH, Angelini GD. Adverse hemodynamic effects of pericardial closure soon after open heart operation. Ann Thorac Surg 1992;53:425–429.
6. Damen J, Bolton DT. Acute hemodynamic effects of pericardial closure in man. Acta Anaesthesiol Scand 1989;33:207–209.
7. Jarvinen A, Peltola K, Rasanen J, Heikkila J. Immediate hemodynamic effects of pericardial closure after open-heart surgery. Scand J Thorac Cardiovasc Surg 1987;21:131–134.
8. Nandi P, Leung JS, Cheung KL. Closure of pericardium after open heart surgery. A way to prevent postoperative cardiac tamponade. Br Heart J 1976;38:1319–1323.