Pediatric Cardiology

, Volume 31, Issue 2, pp 181–187

Does Biventricular Pacing Improve Hemodynamics in Children Undergoing Routine Congenital Heart Surgery?


  • Aamir Jeewa
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
  • Alexander F. Pitfield
    • Division of Critical Care, Department of PediatricsBritish Columbia Children’s Hospital, University of British Columbia
  • James E. Potts
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
  • Wendy Soulikias
    • Division of Critical Care, Department of PediatricsBritish Columbia Children’s Hospital, University of British Columbia
  • Eustace S. DeSouza
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
  • A. J. Hollinger
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
  • George G. S. Sandor
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
  • Jacques G. LeBlanc
    • Division of Cardiovascular and Thoracic Surgery, Department of SurgeryBritish Columbia Children’s Hospital, University of British Columbia
  • Andrew M. Campbell
    • Division of Cardiovascular and Thoracic Surgery, Department of SurgeryBritish Columbia Children’s Hospital, University of British Columbia
    • Division of Cardiology, Department of PediatricsBritish Columbia Children’s Hospital, The University of British Columbia
    • Division of CardiologyBritish Columbia’s Children’s Hospital
Original Article

DOI: 10.1007/s00246-009-9581-4

Cite this article as:
Jeewa, A., Pitfield, A.F., Potts, J.E. et al. Pediatr Cardiol (2010) 31: 181. doi:10.1007/s00246-009-9581-4


Biventricular (BiV) pacing or cardiac resynchronization therapy (CRT) is an established therapy for heart failure in adults. In children, cardiac dyssynchrony occurs most commonly following repair of congenital heart disease (CHD) where multisite pacing has been shown to improve both hemodynamics and ventricular function. Determining which patient types would specifically benefit has not yet been established. A prospective, repeated measures design was undertaken to evaluate BiV pacing in a cohort of children undergoing biventricular repair for correction of their CHD. Hemodynamics, arterial blood gas, electrocardiographic (ECG), and echocardiographic data were collected. Pacing protocol was undertaken prior to the patient’s extubation with 20 min of conventional right ventricular (RV) or BiV pacing, preceded and followed by 10 min of recovery time. Multivariate statistics were used to analyze the data with p values <0.05 considered significant. Twenty-five (14 female) patients underwent surgery at a median (range) age of 5.2 (0.1–37.4) months with no early mortality. The Risk-adjusted classification for Congenital Heart Surgery (RACHS) scores were 2 in 14 patients, 3 in eight patients, and 4 in three patients. None had pre-existing arrhythmias, dyssynchrony, or required pacing pre-operatively. No patient required implantation of a permanent pacemaker post-operatively. The median cardio-pulmonary bypass time was 96 (55–236) min. RV and BiV pacing did not improve cardiac index from baseline (3.23 vs. 3.42 vs. 3.39 L/min/m2; p > 0.05). The QRS duration was not changed with pacing (100 vs. 80 vs. 80 ms; p > 0.05). On echocardiography, the time-to-peak velocity difference between the septal and posterior walls (synchrony) during pacing was similar to baseline and was also not statistically significant. BiV pacing did not improve cardiac output when compared to intrinsic sinus rhythm or RV pacing in this cohort of patients. Our study has shown that BiV pacing is not indicated in children who have undergone routine BiV congenital heart surgery. Further prospective studies are needed to assess the role of multisite pacing in children with ventricular dyssynchrony such as those with single ventricles, those undergoing reoperation or those with high RACHS scores.


Biventricular pacingCongenital heart diseaseTissue Doppler echocardiography

Dual-chamber pacing was first used as adjunctive therapy for adults with medically refractory heart failure in the last decade [18]. Acute studies demonstrated that atrioventricular (AV) synchronous pacing with a short AV delay improved cardiac output and exercise duration in patients with heart failure and a prolonged PR interval [22, 25, 27]. The beneficial effects of AV resynchronization were shown to be due to an increased diastolic filling time and a reduction in mitral and tricuspid valve regurgitation. Since then, biventricular (BiV) pacing or cardiac resynchronization therapy (CRT) has been established as a proven therapy for advanced congestive heart failure in adults by improving exercise tolerance, quality of life, and survival [1, 4, 10].

A wide QRS complex is the most commonly used method of evaluating dyssynchrony and is associated with increased long-term mortality in adults with heart failure [17]. The use of BiV pacing in adult patients with narrow QRS complexes and the determination of nonresponders to this therapy are controversial [26]. In addition, the benefits of temporary biventricular pacing in the postoperative adult patient have also been debated. Little evidence exists with regard to the risks or benefits of pacing for heart failure post-cardiac surgery in patients with no other indications, even in the setting of a low ejection fraction and interventricular conduction delay [29].

Recently, the technique of multisite pacing has been applied to pediatric patients. A large retrospective study has shown that children with chronic heart failure, secondary to cardiomyopathy or congenital heart disease (CHD), have successfully utilized CRT as adjunctive therapy. However, the disease substrates that would specifically benefit from this therapy have not been established [15, 16]. In the immediate postoperative period, children who have undergone cardiac surgery, specifically cardiopulmonary bypass, often experience significant myocardial dysfunction [9]. The literature on the role of multisite ventricular pacing in the postoperative CHD patient is limited. A few prospective studies of multisite ventricular pacing have shown positive effects on hemodynamics and Tissue Doppler imaging (TDI) in postoperative pediatric patients [28, 31]. Traditional two-dimensional (2D) echocardiographic and Doppler indices have been used to assess the efficacy of biventricular pacing and include measuring cardiac output, measuring ventricular ejection times, and assessing wall motion abnormalities [7, 8, 18].

The aim of this study was to compare the effects of biventricular pacing versus conventional pacing on hemodynamics, electrocardiographic (ECG), and echocardiographic data, including TDI, in patients who have undergone routine congenital heart surgery in order to determine whether BiV pacing can improve the postoperative course of these patients.


This study was performed at British Columbia Children’s Hospital, Vancouver, and approved by the Research Ethics Board. Informed consent was obtained from all patients. The patients were recruited between October 1, 2006, and May 31, 2008.

Patient inclusion criteria included a morphologic, functioning left ventricle and the need for a biventricular repair. Patients with (1) normal AV conduction, (2) pre-existing intracardiac conduction delay or bundle branch block, (3) pre-existing conduction disease, or (4) pre-existing ventricular dyssynchrony were all included. Patients with atrial septal defects, ventricular septal defects, AV septal defects, tetralogy of Fallot, aortopulmonary window, and simple transposition of the great arteries were included. Exclusion criteria included patients with (1) single-ventricle morphology; (2) postoperative ECMO, (3) sustained atrial or ventricular arrhythmias that would potentially complicate pacing, (4) pacing wires that malfunctioned; (5) patients who were hemodynamically unstable as assessed by the intensivist, cardiologist, or surgeon; and (6) patients who were unwilling to provide informed consent or assent.

Thirty-four patients were enrolled in the study. One was excluded due to pacing wire malfunction. Four consented but we were not able to adequately capture and pace their left ventricle (LV) intraoperatively. Two consented, but underwent early extubation, and we were not able to perform the study in time. One family withdrew from the study. One patient had poor echocardiographic windows.

Subjects enrolled in the study underwent placement of temporary epicardial pacing wires in the operating room at the completion of surgery. Following repair of CHD, it is our standard practice to place two temporary pacemaker leads on the epicardial surface of the right atrium (RA) and two leads on the epicardial surface of the right ventricle (RV). In addition to these conventional leads, study subjects had one or two additional leads placed on the epicardial surface of the LV, in relatively the same location in all patients. The decision whether one or two wires could be placed was made by the operating surgeon and was dependent on the technical feasibility of placing the wires. If a single wire was placed, LV activation was achieved by unipolar activation. If two LV wires were placed, bipolar activation was achieved.

The pacing study was conducted at the Pediatric Intensive Care Unit at British Columbia Children’s Hospital. Pacing was initiated once the subject was deemed hemodynamically stable by the staff intensivist. The subjects’ level of ventilation and inotropic support were recorded and kept constant during the study period. Subjects were monitored using continuous telemetry that included systemic (via an arterial line or Dynamap7 (GE Healthcare, Wauwatosa, WI, USA) recordings and venous (via a central venous line) blood pressure recordings. A 12-lead electrocardiogram was used to assess the baseline rhythm and measure the PR interval, QRS duration, QT interval, and cycle length before and during pacing protocols. Routine postoperative vital signs and laboratory tests were performed. The subjects underwent a baseline study in their native rhythm with the pacemaker turned off (unless pacemaker dependent). Systemic blood pressure, central venous pressure, and heart rate were recorded. Blood was drawn from the arterial and central venous lines for determination of oxygen saturation.

With respect to the echocardiographic protocol, a transthoracic echocardiogram was done with tissue Doppler to determine synchronization of the RV and LV. Specifically, the LV outflow tract diameter was measured. From the parasternal long-axis view, M-mode measurements of the time to peak contraction of the posterior wall (PW) and intraventricular septum (IVS) were also taken during baseline, RV pacing, and BiV pacing stages. 2D apical four- and two-chamber views (or a triplane view) were used to measure biplane/geometric ejection fractions. Continuous-wave Doppler of the pulmonary and aortic outflow tracts (from the best views) was used to measure the pre-ejection intervals. Pulse-wave Doppler of the mitral valve inflow tract was used to measure diastolic filling time (onset of E to end of A wave). Pulse-wave Doppler of the LV outflow tract (from the best views) was used to obtain the velocity-time integral, and stroke volume and cardiac output were calculated. TDI was done on all 25 patients from the apical four-chamber, two-chamber, and long-axis views. TDI was sampled at the base of the RV free wall, IVS, and base of the LV free wall. TDI was also used to measure tissue velocities, tracking, and strain rates.

Pacing Protocol

The atrial and ventricular pacing leads were attached to an external dual-chamber pacemaker (Medtronic Model 5388; Medtronic, Inc., Minneapolis, MN, USA) using standard techniques. The pacing protocols were performed in random order to eliminate an order effect. The principal investigator or designate was at the bedside during the study to assess the subjects’ hemodynamic status and comfort level during pacing. The pacing protocol timeline was approximately 90 min in duration (Fig. 1).
Fig. 1

Pacing protocol schematic

Baseline Measurements

Prior to pacing, the vital signs, blood gases, ECG, and echocardiographic measurements were taken. Minimal adjustment to ventilation, hemodynamics, inotropic support, or patient handling was done at that time. If the patient required postoperative pacing prior to study initiation, the prestudy settings were considered the baseline for that patient. Once data were collected, we proceeded to RV pacing.

Conventional Dual-Chamber Pacing (or Right Ventricular Pacing)

Subjects were paced for 20 min at a rate 10 bpm faster than their baseline rate using the atrial wires and the RV leads in DOO mode with the AV delay set at 80% of the baseline PR interval. Data were collected after 20 min of pacing had elapsed. The temporary pacemaker was then switched off or set back to baseline settings, and after 15 min baseline measurements were repeated as above.

Biventricular Pacing

Subjects were paced for 20 min at a rate 10 bpm faster than their baseline rate in the DOO mode utilizing the atrial wires and the RV and LV leads. The pacing mode was DOO, with the AV delay set the same as for conventional dual-chamber pacing. Data were collected after 20 min of pacing. Again, the temporary pacemaker was switched off or set back to baseline conventional pacing and data were collected after 15 min had elapsed.


Values are reported as the median (range). A repeated-measures analysis of variance was used to determine if the dependent variables differed between baseline measurements and pacing; p-values < 0.05 were considered statistically significant. Statistical analysis was completed using SAS Statistical Software Version 9.1.3 (SAS Institute, Cary, NC, USA).


The median age of the 25 patients studied was 5.2 months (range, 0.1 to 37.4 months). Fourteen patients underwent the pacing protocol on the day of surgery after transfer to the pediatric intensive care unit and all patients underwent the pacing protocol within 2 days postsurgery. Fourteen of the patients were The Risk Adjusted classification for Congenital Heart Surgery (RACHS) score 2, eight patients were RACHS score 3, and three patients were RACHS score 4, with the majority of the lesions comprising ventricular sepal defects, AV septal defects, and tetralogy of Fallot [21]. The median cardiopulmonary bypass time was 96 min (range, 55–236 min). None of the patients had any ECG evidence of dyssynchrony preoperatively, with a median QRS duration of 80 ms (range, 60 to 100 ms). Postoperatively, there was no significant change in QRS duration. Although the median QRS duration of the BiV pacing group was narrower, with a median value of 80 ms (range, 80 to 100 ms), compared to the median value of the conventional pacing group, 100 ms (range, 60 to 160 ms), there was no statistical difference between conventional pacing and BiV pacing (see Table 1). With respect to the hemodynamic parameters there was no statistical difference in the systolic blood pressure, mean arterial pressure, or central venous pressure compared to that with conventional pacing and BiV pacing (see Table 2). The mixed venous saturation, an indicator of cardiac output, was also unaffected by the BiV pacing versus conventional pacing.
Table 1

Electrocardiographic data


Baseline 1

RV pacing*

Baseline 2

BiV pacing*

Baseline 3

Atrial rate (bpm)

140 (107–165)

150 (115–167)

140 (107–165)

150 (115–167)

140 (107–165)

Ventricular rate (bpm)

140 (107–165)

150 (115–167)

140 (107–165)

150 (115–167)

140 (107–165)

PR interval (ms)

100 (80–180)

80 (40–120)

100 (80–180)

80 (40–120)

100 (60–160)

QRS duration (ms)

80 (60–140)

100 (60–160)

80 (60–120)

80 (80–100)

80 (60–100)

Range given in parentheses

RV right ventricular, BiV biventricular

p-value not significant when comparing RV pacing to BiV pacing or comparing pacing to baseline measurements

Table 2

Hemodynamic data


Baseline 1

RV pacing*

Baseline 2

BiV pacing*

Baseline 3

Blood pressure (mm Hg) Systolic

88 (66–120)

85 (51–120)

85 (54–111)

87 (52–109)

87 (59–111)


50 (37–68)

50 (29–64)

48 (33–56)

46 (30–62)

47 (33–71)


63 (43–89)

62 (38–86)

61 (40–76)

62 (34–76)

62 (43–89)

CVP (mm Hg)

13 (7–18)

11 (6–17)

12 (7–18)

11 (5–19)

11 (6–19)

SVI (ml/m2)

23.7 (6.3–58.0)

24.3 (7.4–49.5)

25.0 (8.2–56.0)

21.7 (6.5–52.3)

25.0 (8.4–52.9)

CI (L/min/m2)

3.39 (1.28–9.56)

3.23 (1.14–7.36)

3.26 (1.15–6.83)

3.42 (1.03–8.40)

3.28 (1.14–7.45)

Mixed venous sat. (%)

63.0 (36–81)

57.5 (31–75)

54.5 (32–76)

55.0 (33–75)

54.0 (33–74)

Range given in parentheses

BiV biventricular, CI cardiac index, CVP central venous pressure, RV right ventricular, sat saturation, SVI stroke volume index

p-value not significant when comparing RV pacing to BiV pacing or comparing pacing to baseline measurements

Transthoracic echocardiograms done on all patients provided cardiac output and stroke volumes. The cardiac index during BiV pacing was 3.42 L/min/m2 (range, 1.03 to 8.40 L/min/m2), which was statistically no different from the 3.23 L/min/m2 (range, 1.14–7.36 L/min/m2) observed during RV pacing; nor was it statistically different compared to baseline measures. With respect to M-mode assessment of synchrony, during RV pacing the difference between PW-to-IVS time to peak contraction (∆PW to IVS) was −10 ms (range, −230 to 170 ms), which was, interestingly, narrower than the initial baseline measure, −30 ms (range, −199 to 80 ms). During BiV pacing the ∆PW to IVS was 0 ms (range, −157 to 90 ms), implying synchronous movement of the two measured walls of the ventricle. However, none of these values were statistically different compared at baseline or after pacing (see Table 3). The TDI sampled at the RV, IVS, and LV free wall during baseline, RV pacing, and BiV pacing was not statistically significant during any of the stages. No complications were observed during pacing of either the RV or both ventricles or during placement or removal of the pacing wires. All patients survived to discharge from hospital.
Table 3

Echo M-mode data


Baseline 1

RV pacing*

Baseline 2

BiV pacing*

Baseline 3

Time to peak PW (ms)

220 (150–285)

257.5 (150–331)

225 (160–294)

237.5 (150–330)

224 (153–300)

Time to peak IVS (ms)

220 (143–458)

255 (208–460)

225 (190–485)

240 (160–434)

220 (200–433)


−30 (−199 to 80)

−10 (−230 to 170)

0 (−200 to 60)

0 (−157 to 90)

0 (−190 to 70)

Range given in parentheses

BiV biventricular, IVS interventricular septum, PW posterior wall, RV right ventricular

p-value not significant when comparing RV pacing to BiV pacing or comparing pacing to baseline measurements


This is the largest prospective study that we are aware of focusing on multisite pacing in children with BiV CHD repairs. In our study of 25 postoperative pediatric patients, in contrast to the other multisite pacing studies currently published, we did not show any benefits of BiV pacing versus conventional pacing with respect to cardiac index, systolic blood pressure, or TDI. Janousek et al. reported a decrease in QRS duration and improved blood pressure in patients with left bundle branch block [20]. Similar findings with regard to systolic blood pressure were also found by Zimmerman et al., who showed that multisite pacing increased the cardiac index and decreased the QRS duration. However, this study included patients with both single-ventricle and BiV anatomy and no comparison was made to conventional pacing protocols [31].

Pham et al., in their study of 19 patients who underwent BiV pacing in the postoperative period, showed an increase in synchrony during BiV pacing, as assessed by TDI [28]. However, no increase in systolic blood pressure was noted as described in the previous pediatric multisite pacing studies. In addition, the cardiac index did improve with BiV pacing compared to conventional pacing, however, no difference in cardiac index was observed compared to patients who were paced only in the atrium (AOO) [28]. Similarly to our population, the patients in that study did not have significant interventricular conduction delay, suggestive of a patient group that was already synchronous postoperatively. Another study of multisite ventricular pacing by Vanagt et al. [30], compared LV apex pacing to LV free wall pacing and conventional RV apex pacing in children post congenital heart surgery. Interestingly, that study showed an improvement in contractility as measured by LV dp/dt with LV apex pacing compared to LV free wall and RV pacing, with no significant changes in QRS duration [30].

It has been suggested that, to date, three-dimensional synchrony has not been adequately assessed by any of the modalities currently used [15]. The recent results of the PROSPECT trial have shown that 2D echocardiographic modalities, such as pulse-wave Doppler and TDI, may not be adequate modalities to assess mechanical dyssynchrony [12]. Temporary BiV pacing did not affect cardiac index, systolic blood pressure, or TDI in our patient population. This may have been affected by the fact that the majority of our patients did not have widened QRS complexes prior to surgery. In addition, the current methods of evaluating dyssynchrony may not be adequate for the postoperative pediatric patient.

In adults, CRT responders are most often those with wide QRS complexes, ejection fractions of <35%, and New York Heart Association (NYHA) class III–IV heart failure [13, 24]. Typically, nonresponders to BiV pacing have a narrow complex QRS. Approximately 40%–50% of patients have been found to be nonresponders [6]; however, even those with a narrow complex QRS may obtain some hemodynamic benefit, highlighting the fact that QRS duration itself is not a reliable marker for dyssynchrony. Other reasons that patients do not respond to BiV pacing include suboptimal LV lead placement and myocardial scarring [2, 5]. While the majority of patients undergoing congenital heart surgery will not have pre-existing myocardial scarring due to infarction, many will have localized areas of myocardial ischemia following cardiac arrest and reperfusion due to abnormal preoperative loading conditions, or from post bypass low cardiac output states and, therefore, the potential for reduced LV activation in the area of the pacing leads remains [3, 5]. Isolated RV dysfunction is not uncommon in the postoperative congenital heart patient due to the effects of cardiopulmonary bypass, postoperative arrhythmias, and the need for a higher preload, which leads us to suspect that a global process is responsible for depressed function rather than attributing it to dyssynchrony alone [11, 23].

The findings presented here are new in that cardiac index, blood pressure, and TDI were unchanged in the postoperative patient undergoing BiV pacing. While the previously published literature highlights some of the observed effects of multisite pacing in the postoperative child, results have been quite varied with respect to improvements in the quantitative measures of dyssynchrony [19]. The question remains whether improvements in systolic blood pressure, QRS duration, or echocardiographic measures of synchrony truly translate into clinical benefits to the pediatric patient having undergone congenital heart surgery. The lack of any significant benefit from multisite ventricular pacing in our study should obviate the need for the placement of extra epicardial pacing wires and their concomitant risks such as bleeding, tamponade, and/or retention of wires [14]. It is also important to recognize the importance of a negative study such as ours in emphasizing the clinical limitations of measuring cardiac performance.

Study Limitations

Our study was limited by the small sample size of patients. It may be postulated that since BiV pacing with the Medtronic 5388 external dual-chamber pacemaker did not allow us to make subtle adjustments to AV or interventricular conduction delay, this represented nonphysiologic ventricular conduction and myocardial function. Our study patients underwent BiV pacing just prior to extubation, by which time myocardial dysfunction may have recovered, so they represented a potentially healthier subset in comparison to adult patients with chronic heart failure. It can also be postulated that our study patients were inherently nonresponders to BiV pacing since there was no evidence of preoperative or postoperative dyssynchrony, recognizing that there is a lack of agreement with regard to the adequacies of the current measures of dyssynchrony in pediatric patients. ECG and echocardiographic optimization was not done at the time of epicardial lead placement. Finally, transthoracic echocardiograms in the immediate postoperative patient can be technically challenging and patients often have poor acoustic windows, limiting our ability to assess the different measures of dyssynchrony.


Our study showed that BiV pacing did not improve the hemodynamics or TDI components in children having undergone routine BiV congenital heart surgery. This study was an important one in addressing the question of BiV pacing versus conventional pacing in a low-risk group. It is unlikely that further studies are needed in this group, although there are parameters not studied in detail here (e.g., LV-only pacing, variable interventricular delays). Future prospective studies attempting to define responders versus nonresponders in pediatric populations with dyssynchrony or in those undergoing repeat congenital heart surgery are needed.


The authors would like to acknowledge the assistance of Mrs. Mary T. Potts and Mrs. Karen Gibbs for their roles in the completion of this study.

Copyright information

© Springer Science+Business Media, LLC 2009