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): Gideon Cohen Anthony L. Ralph-Edwards Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gordon, R. S. Articles by Ralph-Edwards, A. L. Search for Related Content PubMed PubMed Citation Articles by Gordon, R. S. Articles by Ralph-Edwards, A. L.

Ann Thorac Surg 1998;66:1698-1704
© 1998 The Society of Thoracic Surgeons

Permanent cardiac pacing after a cardiac operation: Predicting the use of permanent pacemakers

^a Division of Cardiovascular Surgery, The Toronto Hospital, Toronto, Ontario, Canada

Accepted for publication May 28, 1998.

Address reprint requests to Dr Ralph-Edwards, Division of Cardiovascular Surgery, Toronto Hospital, General Division, 200 Elizabeth St, EN13-239, Toronto, ON M5G 2C4, Canada
e-mail: (aralph-edwards{at}torhosp.toronto.on.ca)

	Abstract

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
Background. The need for permanent cardiac pacing after cardiac operations is infrequent but associated with increased morbidity and resource utilization. We identified patient risk factors for pacemaker insertion to enable development of a predictive model.

Methods. Data were collected prospectively for 10,421 consecutive patients who had cardiac operations between January 1990 and December 1995. Two hundred fifty-five patients (2.4%) were identified as having received a permanent pacemaker during the same hospitalization. Logistic regression analysis was performed to determine the independent, multivariate predictors of permanent pacing. The predictive accuracy and precision of the logistic regression model was evaluated in the 1996 database of 2,236 consecutive patients by the calculation of Brier scores.

Results. Eight independent predictors of permanent pacemaker requirement were identified. The factor-adjusted odds ratios (OR) with 95% confidence interval (CI) associated with each predictor are as follows: (1) valve replacement surgery (aortic: OR 5.8, CI 3.9–8.7; mitral: OR 4.9, CI 3.1–7.8; tricuspid: OR 8.0, CI 5.5–11.9; double: OR 8.9, CI 5.5–14.6; and triple: OR 7.5, CI 2.9–19.3); (2) repeat operation: OR 2.4, CI 1.8–3.3; (3) age 75 years or older: OR 3.0, CI 2.0–4.4; (4) ablative arrhythmia operation: OR 4.2, CI 1.9–9.5; (5) mitral valve annular reconstruction: OR 2.4, CI 1.4–4.2; (6) use of cold blood cardioplegia: OR 2.0, CI 1.2–3.6; (7) preoperative renal failure: OR 1.6, CI 1.0–2.6; and (8) active endocarditis: OR 1.7, CI 0.9–3.0. A model for postoperative permanent pacemaker requirement using the eight predictors was formulated and tested (Brier score = 0.017 ± 0.003; Z = 0.18).

Conclusions. The proposed predictive model correlated highly with actual pacemaker use, which suggests that the requirement for pacing results from either operative trauma or increased ischemic burden. Preoperative identification of patients at increased risk of conduction disturbances may allow for earlier detection and improved treatment. Patients requiring postoperative pacing had increased morbidity and length of stay.

	Introduction

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
The prevalence of permanent cardiac pacing after operation for heart disease ranges from 0.4% to 1.1% after isolated coronary bypass grafting to 3.0% to 6.0% after a valvular operation (30 or fewer days postoperatively) [1–5].

Damage to the sinoatrial node or conduction system of the heart represents the primary indication for postoperative pacemaker insertion. Mechanical trauma to the conduction system is the most common contributing factor, arising secondary to valvular operation, myectomy for hypertrophic obstructive cardiomyopathy, or repair of ventricular septal defect. Alternatively, ischemic injury to the sinoatrial node or conduction system might occur during any cardiac procedure as a result of inadequate intraoperative myocardial protection. Finally, sinoatrial node dysfunction or heart block may occur in cardiac patients as a result of preexisting, and possibly previously unrecognized, anatomic or physiologic abnormalities.

The indications for permanent pacemaker (PP) insertion after cardiac operations include bradycardia (sinus node dysfunction), atrioventricular conduction block, and fascicular block [6]. Previous studies identified several risk factors associated with a requirement for PP insertion, including advanced age, unstable angina, timing of operation, multivessel coronary artery disease, valvular operation (especially aortic, tricuspid, or multiple valves), preoperative rhythm abnormalities, postoperative myocardial infarction, female sex, and reoperation [1–4].

Although some evidence exists to suggest that blood cardioplegia may increase the risk of PP insertion, reports are conflicting [7]. Compromised septal blood flow has been shown to promote the development of conduction abnormalities and might, thus, increase the risk of PP insertion [8]. Ventricular dilatation secondary to aortic insufficiency has been associated with progression of conduction defects long after valve replacement [9]. Valve operations for active endocarditis and surgical re-exploration are also associated with the subsequent need for PP insertion. Although the occurrence of new atrioventricular conduction defects after coronary artery bypass grafting is related to the number of vessels bypassed as well as cardiopulmonary bypass time and aortic cross-clamp time, most new atrioventricular conduction defects in such circumstances resolve and do not require PP insertion [10].

Patients requiring permanent pacemaker implantation after cardiac operation have prolonged hospital stays compared with nonpaced counterparts [1, 4]. Ventilation times and intensive care unit stays are also longer in patients requiring postoperative permanent pacing [1]. Preoperative identification of patients at increased risk for PP insertion may permit modification of operative techniques or cardioplegic maneuvers to prevent postoperative conduction abnormalities.

	Material and methods

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
Patient population
Prospective clinical, operative, and outcome data were collected on 10,421 consecutive patients undergoing cardiac operations necessitating cardiopulmonary bypass at The Toronto Hospital between January 1, 1990, and December 31, 1995. Among these, 255 (2.4%) consecutive patients were identified as having required postoperative implantation of a permanent pacemaker (paced group) during the same hospitalization. This paced group was compared with 10,166 patients who had cardiac operations during the same time period without the need for postoperative PP implantation. Cardiac operations were subdivided into three groups, including isolated coronary bypass grafting, valve replacement with or without coronary bypass grafting, and other procedures consisting of valvular repair, congenital malformation repair, left ventricular aneurysmectomy, and myectomy. Ablative arrhythmia operations includes cryoablation, endocardial resection, and aneurysm resection for treatment of recurrent inducible nonischemic ventricular arrhythmias. Preoperative renal failure is defined as a preoperative creatinine level higher than 150 mmol/day. A predictive model was generated using the 1996 patient database for validation. Details of this database have been published previously [11].

Postoperative indications for permanent pacing included persistent second- and third-degree heart block (more than 5 days), intermittent heart block, symptomatic bradycardia, or heart rate less than 40 beats per minute (persisting more than 5 days postoperatively).

The outcome of interest was the postoperative requirement for a new permanent pacemaker. One hundred thirty-four variables from the cardiac operation database were analyzed, and the variables considered to be potential risk factors were identified (Appendix 1).

Statistical analysis
Data were collected and managed in dBASEIV datasets. The SAS for PC (SAS Institute, Cary, NC) and BMDP/DYN LR (BMDP Statistical Software Inc, Los Angeles, CA) programs were used for statistical analyses [12–15].

The contemporary univariate association between the insertion of a PP and explanatory variables in our clinical database was examined by t tests for continuous variables and ² or Fisher’s exact test for categoric variables. To determine the independent predictors of postoperative pacing, all explanatory variables with a univariate p value less than 0.25 or those of known biologic importance but failing to meet the critical alpha level were submitted to logistic regression analyses using forward selection combined with backward elimination (BMDP LR program).

Model accuracy was evaluated by calculating the area under the receiver-operator characteristic curve [12, 13]. Model precision was evaluated by calculating the Hosmer-Lemeshow goodness-of-fit statistic [16] as well as Brier scores, as previously described [14]. Briefly, the Brier score quantifies accuracy and precision, at the individual patient level, for predicted probabilities versus the actual outcomes. This penalty score ranges in value from 0 to 1. The smaller the Brier score, the more accurate the judgment for each individual patient. We calculated Z scores for overall (or mean) Brier scores for the entire patient sample. A Brier score with a Z score of -1.96 or less, or greater than 1.96 indicated a significant difference between the predicted probability and observed outcome, and thus, poor precision. Regression coefficients from the predictive model were applied to the 1996 dataset, and a predictive probability of pacing was calculated for each patient.

	Results

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
Table 1 summarizes the preoperative demographics of the patient population considered in our analysis. A comparison of perioperative parameters and outcomes by procedure between paced and nonpaced groups is demonstrated in Table 2. Among 10,421 patients who had cardiac operations at our institution between January 1990 and December 1995, 255 (2.4%) required PP insertion. When analyzed by procedure, 0.73% of patients undergoing isolated coronary bypass operation (50 of 6,859 patients) and 6.2% of patients undergoing valve replacement with or without coronary bypass (131 of 2,119 patients) required PP implantation. The remainder of the patients included those who had such procedures as valvular repair, congenital malformation repair, left ventricular aneurysmectomy, and myectomy, of which 5.1% (74 of 1,443 patients) required postoperative permanent pacing.

Predictors of permanent pacemaker insertion
Table 3 demonstrates the independent predictors of PP requirement. Nineteen prognostic variables were identified by univariate analysis. Stepwise logistic regression analysis of these variables identified a model of eight independent predictors for PP requirement. The factor-adjusted odds ratio (OR) and 95% confidence intervals (CI) associated with each predictor were as follows: (1) valve replacement operations: (a) aortic: OR 5.8, CI 3.9–8.7 (PP-dependent, 46.7%; non–PP-dependent, 15.4%); (b) mitral: OR 4.9, CI 3.1–7.8 (33.7%, 10.9%); (c) tricuspid: OR 8.1, CI 5.5–11.9 (8.6%, 1.7%); (d) double valve replacement: OR 8.9, CI 5.5–14.6 (14.9%, 22.5%); (e) triple valve replacement: OR 7.5, CI 2.9–19.3 (2.6%, 4.6%); (2) repeat operation: OR 2.4, CI 1.8–3.3 (38.5%, 18.4%); (3) age 75 years or older: OR 3.0, CI 2.0–4.4 (36.5%, 23.2%); (4) ablative arrhythmia operation: OR 4.2, CI 1.9–9.5 (3.1%, 0.4%); (5) mitral valve annular reconstruction: OR 2.4, CI 1.4–4.2 (6.3%, 0.66%); (6) use of cold blood cardioplegia: OR 2.0, CI 1.2–3.6 (93.7%, 85.4%); (7) preoperative renal failure: OR 1.6, CI 1.0–2.6 (10.6%, 4.9%); and (8) active endocarditis: OR 1.7, CI 0.92–3.0 (6.6%, 1.1%).

	Comment

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
Patients requiring PP implantation after cardiac operations represent a population with higher morbidity and increased resource use [4]. Such patients might have longer hospital and intensive care unit stay because of the late diagnosis of rhythm disturbances and the requirement for prolonged monitoring.

The need for postoperative pacing likely arises by one of two routes: (1) operative procedures in close physical proximity to the sinoatrial or atrioventricular nodes or the His bundle might result in trauma to the conduction system; or (2) extensive coronary artery disease might compromise myocardial protection during intraoperative cardioplegic arrest, thus facilitating ischemic injury to the conduction system.

Increased risk of physical damage to the conduction system is present in patients who have redo valve operations, multiple valve replacement, and debridement or reconstructive operation for active endocarditis. Similarly, debridement of a heavily calcified aortic annulus after excision of the aortic valve might be the source of significant trauma to the conduction system. Finally, during complex congenital lesions, the precise location of conduction pathways might not be fully ascertained, and inadvertent or unavoidable damage may result.

Preexisting compromise to the conduction system is more common in patients with extensive coronary artery disease. Moreover, adequate and uniform intraoperative myocardial protection might be difficult to achieve in such patients, leading to perioperative myocardial infarction and low output syndrome, further exacerbating postoperative conduction disturbances. Although cold blood cardioplegia represents an independent risk factor for PP requirement, this finding might be related to differences in the technique of cardioplegic delivery (in comparison to warm or tepid techniques) rather than temperature alone. Whereas hypothermic blood cardioplegia is most often administered as intermittent antegrade doses at our institution, normothermic cardioplegia requires continuous delivery to ensure its effectiveness [17–19]. Such continuous methods provide improved myocardial protection and might reduce the incidence of postoperative conduction disturbances. The finding of transient pacer dependency in some patients suggests a short-term ischemic origin, which might improve with adequate reperfusion [1].

In the current series, the requirement for postoperative permanent pacing tended to occur in older patients who had nonelective procedures. The typical complex nature of the operative procedures performed in such high-risk patients was reflected in significantly prolonged cardiopulmonary bypass and aortic cross-clamp times. Additional preoperative risk factors for permanent pacing included advanced New York Heart Association status, renal insufficiency, and congestive heart failure. Postoperatively, such patients experienced significantly longer intensive care unit and hospital stays in comparison to their nonpaced counterparts. Although mortality was not significantly different between groups, coronary artery bypass grafting patients requiring postoperative pacing demonstrated an increased incidence of low output syndrome, postoperative stroke, and perioperative myocardial infarction. In patients who had valvular or other operations, there was an increased incidence of permanent pacing in cases of reoperation or active endocarditis.

We have identified eight independent risk factors for PP insertion, five of which relate to physical trauma to the conduction system (valve operation, repeat operation, ablative arrhythmia operation, mitral annulus reconstruction, and active endocarditis), the rest of which relate to increased ischemic burden (age 75 years or older, renal failure) or suboptimal intraoperative myocardial protection (use of cold blood cardioplegia). By applying the multivariate regression coefficients, a model to determine the probability of subsequent pacemaker requirement was successfully formulated and tested. Consecutive patients who had cardiac surgical procedures between January and December 1996 (none of whom were included in the original analysis) were used to test the predictive model. This analysis found the model to be highly correlated with actual pacemaker requirement.

Using the probability formula with the derived coefficients for each factor, we calculated the risk of postoperative permanent pacing for several common patient scenarios (Fig 1). Not unexpectedly, the highest probability was calculated for aortic and mitral valve cases involving reoperation for active endocarditis (29.4% and 26%, respectively). In this situation, the requirement for additional mitral annular reconstruction increased the associated risk to 45.4%. The calculated risk of postoperative pacing after uncomplicated coronary bypass procedures was 0.9%.

View larger version (29K): [in this window] [in a new window]   Fig 1. Predicted probabilities of permanent pacemaker requirement based on common operative scenarios. Each column has a set of variables representing a scenario of a cardiac patient. Above each column is a bar graph representing the calculated probability, P, of permanent pacemaker requirement based on the logistic regression model (P = ex/1 + ex), expressed as a percentage. Ablation = ventricular mapping with resection, cryoablation, or both; ACB = aortocoronary bypass; Act Endo = active endocarditis; AVR = aortic valve replacement or repair; Cold pleg = cold blood cardioplegia; Double Valve = double valve repair or replacement; MV ann = mitral valve annulus reconstruction; MVR = mitral valve repair or replacement; Reop = reoperation; RF = preoperative renal failure; Triple Valve = triple valve repair or replacement.

	Acknowledgments

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
Doctor Cohen is a Fellow of the Heart and Stroke Foundation of Canada.

	Appendix 1

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
The following variables were identified as potential risk factors for PP insertion after the procedures in question: Preoperative: age, sex, timing of operation (elective, urgent, emergent), New York Heart Association classification, unstable angina, congestive heart failure, myocardial infarction, left ventricular grade (I–IV), diabetes, peripheral vascular disease, renal failure, active endocarditis, previous endocarditis, and intraaortic balloon pump requirement; Intraoperative: number of bypass grafts, number and type of valve(s) repaired or replaced, mitral valve annular reconstruction, aortic valve annular enlargement, ablative arrhythmia operation, myectomy, cross-clamp time, cardiopulmonary bypass time, intraaortic balloon pump requirement, method of myocardial protection, and systemic hypothermia; Postoperative: requirement for reoperation, duration of ventilation, duration of intensive care unit stay, hospital stay, inotropic agent requirement, intraaortic balloon pump requirement, low output syndrome, stroke, transient ischemic attack, and myocardial infarction.

	Appendix 2

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 
To calculate the predicted probability of permanent pacemaker requirement for a given patient, we added the regression coefficients describing the patient’s characteristics to the constant for a coefficient total (x). This value was applied to the formula . For example, the predicted probability of permanent pacemaker requirement for a 76-year-old male patient with renal failure undergoing repeat aortic valve replacement with cold blood cardioplegia would be calculated as follows: x = -5.357 + 1.096 + 0.4802 + 0.8822 + 2.187 + 0.716 + 0 + 0 = -0.4228; P = e^-0.4228/(1 + e^-0.4228) = 0.396 (or 39.6%).

	References

Top Abstract Introduction Material and methods Results Comment Acknowledgments Appendix 1 Appendix 2 References

 

1. Del Rizzo D.F., Nishimura S., Lau C., Sever J., Goldman B.S. Cardiac pacing following surgery for acquired heart disease. J Card Surg 1996;11:332-340.[Medline]
2. Goldman B.S., Hill T.J., Weisel R.D., et al. Permanent cardiac pacing after open-heart surgery: acquired heart disease. PACE 1984;7:367-371.
3. Goldman B.S., Ivanov J., Irwin M., Cameron D., Harris L. Permanent pacing following cardiac surgery. PACE 1990;13:554.
4. Lipton I.H., Cameron D.A., David T.E., Mackall J.A. Morbidity of permanent pacing following valvular surgery. PACE 1995;18:1768.
5. Gaillard D., Lespinasse P., Vanetti A. Cardiac pacing and valvular surgery. PACE 1988;11:2142-2148.
6. Kusumoto F.M., Goldshlager N. Cardiac pacing. N Engl J Med 1996;334:89-98.[Free Full Text]
7. Gundry S.R., Sequeira A., Coughlin T.R., McLaughlin J.S. Postoperative conduction disturbances: a comparison of blood and crystalloid cardioplegia. Ann Thorac Surg 1989;47:384-390.[Abstract]
8. Mosseri M., Meir G., Lotan C., et al. Coronary pathology predicts conduction disturbances after coronary artery bypass grafting. Ann Thorac Surg 1991;51:248-252.[Abstract]
9. Poveda J., Vazquez J., Larman M., et al. Conduction defects in aortic valve diseases after isolated aortic valve replacement. PACE 1983;6:A-15.
10. Baerman J.M., Kirsh M.M., de Buitleir M., et al. Natural history and determinants of conduction defects following coronary artery bypass surgery. Ann Thorac Surg 1987;44:150-153.[Abstract]
11. Rao V., Ivanov J., Weisel R.D., Ikonomidis J.S., Christakis G.T., David T.E. Predictors of low cardiac output syndrome after coronary artery bypass. J Thorac Cardiovasc Surg 1996;112:38-51.[Abstract/Free Full Text]
12. Fletcher R.H., Fletcher S.W., Wagner E.H. Clinical epidemiology: the essentials. Baltimore: Williams & Wilkins, 1988.
13. Sackett D.L., Haynes R.B., Tugwell P. Clinical epidemiology: a basic science for clinical medicine. Boston: Little Brown, 1985.
14. Hosmer D.W., Lemeshow S. Applied logistic regression. New York: John Wiley, 1989.
15. Dixon W.J. BMDP statistical software manual, part 2. Berkeley: University of California Press, 1992.
16. Redelmeier D.A., Bloch D.A., Hickam D.H. Assessing predictive accuracy: how to compare Brier scores. J Clin Epidemiol 1991;44:1141-1146.[Medline]
17. Naylor C.D., Lichtenstein S.V., Fremes S.E., Warm Heart Investigators. Randomized trial of normothermic versus hypothermic coronary bypass surgery. Lancet 1994;343:559-563.[Medline]
18. Yau T.M., Ikonomidis J.S., Weisel R.D., et al. Ventricular function after normothermic versus hypothermic cardioplegia. J Thorac Cardiovasc Surg 1993;105:833-844.[Abstract]
19. Yau TM, Weisel RD, Mickle DAG, et al. Optimal delivery of blood cardioplegia 1991;84(Suppl 3):380–8.

This article has been cited by other articles:

				D. Berdajs, U. P. Schurr, A. Wagner, B. Seifert, M. I. Turina, and M. Genoni Incidence and pathophysiology of atrioventricular block following mitral valve replacement and ring annuloplasty. Eur. J. Cardiothorac. Surg., July 1, 2008; 34(1): 55 - 61. [Abstract] [Full Text] [PDF]

				N. Piazza, Y. Onuma, E. Jesserun, P. P. Kint, A.-M. Maugenest, R. H. Anderson, P. P. T. de Jaegere, and P. W. Serruys Early and Persistent Intraventricular Conduction Abnormalities and Requirements for Pacemaking After Percutaneous Replacement of the Aortic Valve J. Am. Coll. Cardiol. Intv., June 1, 2008; 1(3): 310 - 316. [Abstract] [Full Text] [PDF]

				P. Lukac, V. E. Hjortdal, A. K. Pedersen, P. T. Mortensen, H. K. Jensen, and P. S. Hansen Superior Transseptal Approach to Mitral Valve Is Associated With a Higher Need for Pacemaker Implantation Than the Left Atrial Approach Ann. Thorac. Surg., January 1, 2007; 83(1): 77 - 82. [Abstract] [Full Text] [PDF]

				R. Hall Cased based learning on the web Heart, September 1, 2003; 89(9): 979 - 979. [Full Text] [PDF]

				R. Hall and M. Earley Learning on the Web. Case 1: syncope and palpitation in a patient following aortic valve replacement. Heart, September 1, 2003; 89(9): 979 - 979. [Full Text] [PDF]

				G. Abuin, A. Nieponice, S. Martinez, and C. Fernando The role of atrial vessels in aortic root and mitral valve operations Ann. Thorac. Surg., October 1, 2000; 70(4): 1234 - 1237. [Abstract] [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): Gideon Cohen Anthony L. Ralph-Edwards Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gordon, R. S. Articles by Ralph-Edwards, A. L. Search for Related Content PubMed PubMed Citation Articles by Gordon, R. S. Articles by Ralph-Edwards, A. L.