HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Lishan Aklog John G. Byrne David H. Adams Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Anderson, C. A. Articles by Adams, D. H. Search for Related Content PubMed PubMed Citation Articles by Anderson, C. A. Articles by Adams, D. H. Related Collections Cardiac - other

Ann Thorac Surg 2002;73:1484-1488
© 2002 The Society of Thoracic Surgeons

---

Original article: cardiovascular

Liberal use of delayed sternal closure for postcardiotomy hemodynamic instability

^a Division of Cardiac Surgery, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA

Accepted for publication December 5, 2001.

* Address reprint requests to Dr David Adams, Mount Sinai Medical Center, Department of Cardiothoracic Surgery, 1190 Fifth Ave, New York, NY 10029 USA

	Abstract

Top Abstract Introduction Material and methods Results Comment References

 
Background. The purpose of this retrospective study was to evaluate the current incidence, survival, and predictors of mortality for open chest management at our center.

Methods. Our database was analyzed to identify adult postcardiotomy patients who left the operating room without primary sternal closure. Medical records were reviewed to determine mortality, postoperative complications, and pertinent hemodynamic data.

Results. From November 1997 to June 2000, 5,177 adults underwent cardiac procedures at our center. The incidence of open chest management was 1.7% (87 of 5,177), including 0.7% (16 of 2,254) for isolated coronary artery bypass grafting, 1.6% (15 of 912) for isolated valve, and 5.6% (47 of 839) for combined valve/coronary bypass. Hospital survival was 76% (66 of 87). Major complications included deep sternal infection (n = 4), stroke (n = 8), and dialysis (n = 13). Predictors of mortality by univariate analysis included ventricular assist device insertion (p = 0.003), new onset hemodialysis (p < 0.0005), reoperation for bleeding (p = 0.002), sternal infection (p = 0.042), mean length of delay before sternal closure (survivors = 3.2 days, nonsurvivors = 6.2 days; p = 0.031), higher mean dose of epinephrine at the time of chest closure (2.5 µg versus 0.9 µg, p = 0.011), and longer duration of high dose inotropic therapy (110 hours versus 43 hours, p = 0.002). Multivariate analysis showed ventricular assistance and reoperation for bleeding as independent predictors of in-hospital death with odds ratios of 3.8 and 3.4, respectively.

Conclusions. Liberal use of open chest management is useful in patients with postcardiotomy shock, and can be carried out with a relatively low incidence of sternal complications. Patients who require ventricular assistance or exploration for ongoing mediastinal bleeding continue to have a high mortality rate.

	Introduction

Top Abstract Introduction Material and methods Results Comment References

 
Open chest management (OCM) and subsequent delayed sternal closure (DSC) for cardiac surgical patients was first described in the late 1970s [1]. Early hesitance to use this technique stemmed from a concern that prolonged open sternotomy would result in infectious complications. Retrospective analyses have subsequently demonstrated a low incidence of infection, somewhat ameliorating these concerns [2–5]. Since initial description, DSC has become a valuable tool in the management of patients with postcardiotomy instability and mediastinal edema, with a current incidence of 1.2% to 4.2% in the adult cardiac surgical literature [2, 6].

With evidence for low sternal morbidity, and a growing population of patients with complex cardiac disease, DSC will be an increasingly important management option for the cardiac surgeon. Here we report our experience with OCM over a 2.5-year period to examine the outcome and predictors of mortality in a group of patients with postcardiotomy hemodynamic instability.

	Material and methods

Top Abstract Introduction Material and methods Results Comment References

 
Patient characteristics
From November 1997 to June 2000, 5,177 adult cardiac operations were performed at our center. Analysis of the cardiac surgical database revealed that 87 patients (1.7%) left the operating room without primary sternal closure. Preoperative characteristics of these patients include a mean age of 65 years (±13.5 years), male sex (n = 49, 56%), New York Heart Association functional class III/IV (n = 81, 93%), recent myocardial infarction (n = 27, 31%), emergent operation (taken either directly from cardiac catheterization laboratory or requiring intervention within 6 hours of consultation) (n = 25, 29%), and redo cardiac operation (n = 19, 22%). Six patients (7%) required cardiopulmonary resuscitation and 17 patients (19%) required intraaortic balloon pump (IABP) therapy preoperatively. Mild ventricular dysfunction was present with a mean ejection fraction of 41% (±17%, n = 74) and an average mean pulmonary artery (PA) pressure of 31 mm Hg (±10.5 mm Hg, n = 40).

Surgical procedure
Operations were routinely performed using moderate hypothermic cardiopulmonary bypass with cold blood cardioplegic arrest by intermittent antegrade with or without retrograde delivery. Specific procedures included coronary artery bypass grafting (n = 16), valve replacement or repair (n = 15), combined valve/coronary artery bypass grafting (n = 47), cardiac transplantation (n = 3), aortic dissection repair (n = 2), right ventricular rupture repair (n = 2), pulmonary embolectomy (n = 1) and intracardiac leiyomyoma extraction (n = 1). Further analysis of these subsets revealed an incidence of OCM of 0.7% (16 of 2,254) for coronary artery bypass grafting, 1.6% (15 of 912) for isolated valve, and 5.6% (47 of 839) for combined valve/coronary artery bypass grafting during the study period. Cardiopulmonary bypass time was prolonged, ranging from 51 to 494 minutes (mean, 207 ± 92 minutes) and the mean cross-clamp time was 118 minutes (±56 minutes).

Indication and technique for open chest management
Open chest management was carried out if maneuvers to achieve hemodynamic stability were unsuccessful. Measures taken to stabilize the patient included optimization of preload and afterload, inotropic support and IABP, or ventricular assist device (VAD) insertion. Fifty-two patients (60%) left the operating room with an IABP. Patients with VADs, severe peripheral vascular disease, or atheromatous debris in their aorta did not undergo IABP insertion. Also, if sternal closure was prevented primarily by significant mediastinal edema, IABP insertion was considered to provide little additional benefit. Eleven patients required insertion of a VAD. Ventricular function was routinely assessed by transesophageal echocardiography.

Hemodynamic instability was the primary indication for OCM in all patients. In some patients this instability manifested itself in the setting of a trial chest closure, but not all patients underwent trial closure. Additional indications that influenced the surgeon’s decision to leave the chest open included bleeding/coagulopathy (32 of 87, 36%); cardiac edema (13 of 87, 15%), and arrhythmias (7 of 87, 8%).

Mediastinal isolation was accomplished by suturing or stapling a sheet of Esmark (Medline Industries, Inc, Mundelein, IL) to the skin edge. In cases of more pronounced mediastinal edema the sternal halves were stented apart with either a modified syringe (8 patients) or a sternal retractor (1 patient).

Intensive care unit management
In the intensive care unit patients were ventilated and sedated until the time of chest closure. Patients not closed within 48 hours underwent mediastinal debridement and thorough irrigation in the operating room or in the intensive care unit depending on degree of stability. The irrigation protocol consists of 1 L of normal saline with 500,000 U of Polymyxin B (Bedford Laboratories, Bedford, OH) followed by 3 L of normal saline and a final irrigation with 1 L of Polymyxin B solution. Broad-spectrum antibiotic prophylaxis was maintained while the chest was open. Forty-eight hours after DSC antibiotics were discontinued if the patient was afebrile and cultures obtained before closure remained negative.

Hemodynamics and inotropic agents
Data analysis included inotrope requirements for the immediate postoperative period and just before chest closure. The duration of high dose inotropic therapy, defined as epinephrine more than 2.0 µg or dopamine/dobutamine more than 10 µg/kg/min, was also recorded. Hemodynamic data were documented for the immediate postoperative period, just before chest closure, immediately after chest closure, and at the time of PA catheter removal.

Statistical analysis
Statistical analysis was performed using analysis of variance. Potential predictors of mortality were evaluated using independent samples t-tests and ² analysis for continuous and nominal variables, respectively. Multivariate analysis was performed to determine independent predictors of mortality. Significance was defined as a p value less than 0.05.

	Results

Top Abstract Introduction Material and methods Results Comment References

 
Hemodynamics and inotropic agents
Hemodynamic data are demonstrated in Table 1. Both the central venous pressure and the mean PA pressure decreased from the time of operation to the period before chest closure consistent with fluid mobilization. After approximation of the sternum there was an increase in filling pressures that normalized by the time of PA catheter removal. The cardiac index changed in an inverse fashion to the filling pressures with peak cardiac function at the time of PA catheter removal, although these changes were not dramatic.

Morbidity
Four patients (4.6%) in this study had sternal infection. Three of these infections occurred after normal wire closure of the sternum. The remaining patient developed a pseudomal infection of the mediastinum after VAD insertion before the chest was closed. Three of these 4 patients did not survive. Other complications included stroke (n = 8, 9%), pneumonia (n = 13, 15%), dialysis (n = 13, 15%), leg ischemia from IABP insertion (n = 8, 9%), line sepsis (n = 4, 4.6%), urinary tract infection (n = 2, 2.3%), and pancreatitis (n = 3, 3.4%). Twenty-two patients (25%) required exploration of the mediastinum one or more times to evacuate hematoma before definitive chest closure. Distinct surgical bleeding sites were found in only 4 patients. Blood loss was considerable with patients receiving an average of 36 U of blood by the end of postoperative day 2. Nine of the 11 VAD patients were in this category. Twenty-one patients (24%) failed to wean from the ventilator and required tracheostomy.

Mortality
Our operative mortality, defined as death within 30 days or during the same hospitalization, was 24%. Sixty-six of 87 patients (76%) were discharged from the hospital at a mean of 23 days (±25 days; range, 6 to 194 days). Of the 21 nonsurvivors, 8 patients died before closure and the remaining 13 after closure. Cause of death included multisystem organ failure (n = 7), low cardiac output (n = 5), cardiac arrest (n = 3), stroke (n = 3), and sepsis (n = 3).

Predictors of mortality
Univariate analysis
There were no preoperative predictors of operative mortality (Table 3). We specifically investigated age, gender, functional status, presence of a preoperative myocardial infarction or previous sternotomy, ventricular function, emergent circumstances, and liver or renal disease. Requirement for insertion of a VAD to come off cardiopulmonary bypass was strongly associated with operative mortality (p = 0.003) with death occurring in 7 of 11 patients. Other potential intraoperative determinants such as cardiopulmonary bypass time, cross-clamp time, type of operation performed, and indication for OCM did not influence mortality.

The magnitude of inotropic support in the immediate postoperative period did not predict a poor outcome, but patients who did not survive were on much higher doses of epinephrine just before chest closure (2.5 µg versus 0.9 µg, p = 0.011). Similarly, patients who suffered mortality had a longer duration of high dose inotropic therapy (110 hours versus 43 hours, p = 0.002).

The only hemodynamic measure to reach significance was the mean blood pressure at the time of PA catheter removal, which was 10 mm Hg lower in nonsurvivors (82 mm Hg versus 92 mm Hg, p = 0.017).

Multivariate analysis
Multivariate analysis demonstrated that both VAD insertion and reoperation for bleeding were independent predictors of in hospital death, with odds ratios of 3.8 and 3.4, respectively (Table 4).

	Comment

Top Abstract Introduction Material and methods Results Comment References

In this study we have evaluated our outcome using OCM in patients with postcardiotomy hemodynamic instability. Much of the literature to date on OCM is in the form of case reports or small case series [5–7]. Larger reports are few in number and evaluate outcomes from patients who underwent operation in the late 1970s to early 1990s [2–4, 8–10] (Table 5).

Poor functional class and advanced cardiac disease requiring a lengthy operation are risk factors for postoperative instability and the need for OCM. This is evident in the preoperative characteristics of these 87 patients. Ninety-three percent of patients in this series were in New York Heart Association functional class III/IV, 53% of patients underwent combined valve/coronary artery bypass grafting procedures, and the mean cardiopulmonary bypass time was 207 minutes. These risk factors for OCM, however, did not further stratify patients into groups of varying risk for death. We found no preoperative risk factors for in-hospital death, and the only intraoperative predictor of poor outcome was the requirement for VAD insertion.

Multiple postoperative factors were associated with patient death including high epinephrine requirement just before chest closure, duration of high dose inotropic support, sternal infection, dialysis, reoperation for mediastinal bleeding, and delay before sternal closure. Preoperative requirement for hemodialysis did not predict a poor outcome.

Twenty-two patients required reexploration in either the operating room or in the intensive care unit for ongoing mediastinal bleeding. Distinct surgical sites of bleeding were found in only 4 patients. A large proportion of the patients requiring reexploration for bleeding (9 of 22) were dependent on VADs. Despite this close association between bleeding and VAD placement, these factors were independently predictive of death.

On the basis of our findings, a longer delay before DSC is associated with an increased incidence of infection and higher mortality. Therefore, we believe that it is important to close patients as soon as possible. For this reason we perform flap closure when persistent mediastinal edema prevents sternal closure in patients who are otherwise stable. Persistent mediastinal edema was the indication for flap closure in 6 patients.

Sternal infection rates for patients undergoing routine cardiac operations are in the range of 1% to 2%. Critically ill patients with open chests, tracheostomies, VADs, and prolonged stays in the intensive care unit have a heightened risk for this complication. Four patients in this series developed deep sternal infections. Of the 4 patients, 3 were redo operations, 2 had VADs and 2 required tracheostomies. Also, these patients were closed at a mean of 6.3 days compared to a mean of 3.2 days for the overall group. The high mortality associated with this complication in our series reemphasizes the need to take all precautions to avoid sternal infection and treat it aggressively and early when it occurs. Aggressive debridement of devitalized tissue with flap closure if possible was performed. It is our policy to keep patients on broad-spectrum antibiotics while the chest is open. Since 1997 that regimen has consisted of vancomycin, levoquine, and metronidazole.

Inotrope requirements decreased from the immediate postoperative period to the time just before chest closure, indicating patient stabilization. Surviving patients went to the operating room for closure on lower doses of epinephrine (0.09 µg versus 2.5 µg, p = 0.01) and required a shorter duration of high dose inotropic therapy (43 hours versus 110 hours, p = 0.002). Interestingly, the level of inotropic support required in the immediate postoperative period had no bearing on outcome.

The mean cardiac index remained greater than or equal to 2.5 L/min/m² at all times, improving only slightly from the immediate postoperative period to just before closure and then dipping only slightly after closure. Viewing these indices in isolation is misleading, however, as they are a reflection of aggressive measures to optimize perfusion and cardiac performance. Many of the patients with satisfactory CI and blood pressure with open chests had prompt hemodynamic decompensation when undergoing trial closure, although trial closure was not always performed. The subsequent mild alterations in CI, blood pressure, and filling pressures have to be considered in the context of simultaneous weaning of inotropic agents, balloon counterpulsation, as well as fluid mobilization.

Several factors have to be examined when determining the timing of DSC. Transesophageal echocardiography is used in the operating room for any unstable patient to give periodic information about ventricular function and response to various therapeutic maneuvers. Other considerations regarding timing of closure include inotrope requirements, degree of fluid mobilization, and level of dependency on IABP therapy. Although the decision of when to close is made on a case-by-case basis, patients who are successfully closed are generally on less than 2.0 µg of epinephrine and in negative fluid balance.

In summary, OCM with delayed closure of the sternum remains an important management tool in patients with postcardiotomy instability. Delayed closure can be achieved in the majority patients with a relatively low incidence of sternal morbidity and low mortality. We believe that liberal application of OCM provides the best opportunity for survival in patients with postcardiotomy instability after cardiac operations.

	References

Top Abstract Introduction Material and methods Results Comment References

 

1. Riahi M., Tomatis L.A., Schlosser R.J., Bertolozzi E., Johnston D.W. Cardiac compression due to closure of the median sternotomy in open-heart surgery. Chest 1975;67:113-114.[Abstract/Free Full Text]
2. Christenson J.T., Maurice J., Simonet F., Velebit V., Schmuziger M. Open chest and delayed sternal closure after cardiac surgery. Eur J Cardiothorac Surg 1996;10:305-311.[Abstract]
3. Mestres C.A., Pomar J.L., Acosta M., et al. Delayed sternal closure for life-threatening complications in cardiac operations: an update. Ann Thorac Surg 1991;51:773-776.[Abstract]
4. Freeman R.K., Daily P.O., Dembitsky W.P., Adamson R.M., Moreno-Cabral R.J. The treatment of low cardiac output syndrome following cardiopulmonary bypass using delayed sternal closure. Am Surg 1997;63:882-884.[Medline]
5. Josa M., Khuri S., Braunwald N., et al. Delayed sternal closure. An improved method of dealing with complications after cardiopulmonary bypass. J Thorac Cardivasc Surg 1986;91:598-603.[Abstract]
6. Milgater E., Uretzky G., Shimon D.V., Silberman S., Appelbaum A., Borman J.B. Delayed sternal closure following cardiac operations. J Cardiovasc Surg 1986;27:328-331.[Medline]
7. Donatelli F., Triggiani M., Benussi S., Grossi A. Advantages of delayed sternal closure in cardiac-compromised adult patients. J Cardiol Surg 1995;10:632-636.
8. Gielchinsky I., Parsonnet V., Krishnan B., Silidker M., Abel R.M. Delayed sternal closure following open-heart operation. Ann Thorac Surg 1981;32:273-277.[Abstract]
9. Furnary A.P., Magovern J.A., Simpson K.A., Magovern G.J. Prolonged open sternotomy and delayed sternal closure after cardiac operations. Ann Thorac Surg 1992;54:233-239.[Abstract]
10. Fanning W.J., Vasko J.S., Kilman J.W. Delayed sternal closure after cardiac surgery. Ann Thorac Surg 1987;44:169-172.[Abstract]

This article has been cited by other articles:

				T. Fleck, B. Kickinger, R. Moidl, F. Waldenberger, E. Wolner, M. Grabenwoger, and W. Wisser Management of open chest and delayed sternal closure with the vacuum assisted closure system: preliminary experience Interactive CardioVascular and Thoracic Surgery, October 1, 2008; 7(5): 801 - 804. [Abstract] [Full Text] [PDF]

				A. L. Estrera, E. E. Porat, C. C. Miller III, R. Meada, P. E. Achouh, A. D. Irani, and H. J. Safi Outcomes of delayed sternal closure after complex aortic surgery Eur. J. Cardiothorac. Surg., June 1, 2008; 33(6): 1039 - 1042. [Abstract] [Full Text] [PDF]

				Z. I. Khalpey, R. B. Ganim, and J. D. Rawn Postoperative Care of Cardiac Surgery Patients Card. Surg. Adult, January 1, 2008; 3(2008): 465 - 486. [Full Text]

				T. Shibata, K. Hattori, H. Hirai, H. Fujii, T. Aoyama, and S. Seuhiro Rectus abdominis myocutaneous flap after unsuccessful delayed sternal closure Ann. Thorac. Surg., September 1, 2003; 76(3): 956 - 958. [Abstract] [Full Text] [PDF]

				Y. Misawa Liberal use of delayed sternal closure: sternal infection does not increase Ann. Thorac. Surg., February 1, 2003; 75(2): 638 - 638. [Full Text] [PDF]

				F. Filsoufi, L. Aklog, and D. H. Adams Liberal use of delayed sternal closure: sternal infection does not increase: Reply Ann. Thorac. Surg., February 1, 2003; 75(2): 638 - 639. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Lishan Aklog John G. Byrne David H. Adams Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Anderson, C. A. Articles by Adams, D. H. Search for Related Content PubMed PubMed Citation Articles by Anderson, C. A. Articles by Adams, D. H. Related Collections Cardiac - other