Author + information
- Received March 20, 1998
- Revision received June 18, 1998
- Accepted October 26, 1998
- Published online February 1, 1999.
- Lara S Shekerdemian, MD MRCPa,
- Andrew Bush, MD, FRCPa,
- Darryl F Shore, FRCS∗,
- Christopher Lincoln, FRCSa and
- Andrew N Redington, MD, FRCPa,* ()
- ↵*Reprint requests and correspondence: Professor A. Redington, Professor of Congenital Heart Disease, Royal Brompton Hospital, Sydney Street, London SW3 6NP, United Kingdom
We hypothesized that a period of cuirass negative pressure ventilation (NPV) would augment the cardiac output of patients in the early postoperative period after complete correction of tetralogy of Fallot (TOF).
Diastolic right ventricular dysfunction can lead to a low-output state in an important minority of patients after TOF repair. In these patients, the diastolic pulmonary arterial flow, which characterizes restrictive right ventricular physiology, and on which the cardiac output is so dependent, is highly sensitive to changes in intrathoracic pressure.
The effects of NPV on pulmonary blood flow were investigated in 23 intubated children who were initially ventilated using intermittent positive pressure ventilation after TOF repair. Eight patients had restrictive right ventricular physiology. All children received a 15-min period of NPV, and eight received a prolonged period (45 min) of NPV.
A brief period of NPV increased pulmonary blood flow by 39%, and the improvement further continued if the study period was extended, with a total increase of 67% after 45 min. Patients with restrictive physiology had a somewhat delayed response to NPV, but the ultimate increase during an extended period of NPV was greater in restrictive patients (84%) than nonrestrictive patients (50%).
By manipulating important cardiopulmonary interactions, NPV improves the cardiac output of patients after TOF repair, and has a role as a hemodynamic tool in the management of the low-output state in selected cases.
As advances in surgical techniques, myocardial protection and postoperative care continue to be made in the field of cardiac surgery, complete correction of tetralogy of Fallot (TOF) has become a procedure that is now associated with a low operative morbidity and mortality. However, in a small sub-group of these patients, the early postoperative course can be complicated by a severe low-output state, which is characteristically refractory to conventional therapeutic measures.
We previously reported the presence of restrictive right ventricular physiology in approximately one third of patients in the early postoperative period after TOF repair (1). Restrictive physiology is characterized by antegrade diastolic pulmonary arterial flow that is coincident with atrial systole. This diastolic flow makes an important contribution towards the total cardiac output both directly by increasing forward pulmonary arterial flow, and indirectly by limiting the time available for pulmonary regurgitation. Nonetheless, these patients spend more time in the intensive care unit than their nonrestrictive counterparts, and an important minority can develop a severe low-output state.
In patients who may already have borderline hemodynamics, it would therefore seem essential to preserve this abnormal but valuable source of cardiac output, and one can speculate that any method by which this could be augmented would almost certainly be advantageous for patients with a low-output state. Conversely, a reduction in, or loss of, diastolic pulmonary arterial flow could have potentially devastating consequences. Loss of sinus rhythm, for example, in junctional ectopic tachycardia, may precipitate a rapid clinical decline in these patients, but fortunately this is rare. However, almost all patients require a period of mechanical ventilation, and the diastolic pulmonary arterial flow of patients with restrictive physiology is highly sensitive to changes in intrathoracic pressure: falling during positive pressure inspiration, and increasing during spontaneous inspiration (1). Ideally, postoperative ventilatory settings should include low airway pressures, and management should be aimed at early extubation. Paradoxically, rapid ventilatory weaning is likely to be least feasible in those patients who from a hemodynamic standpoint need it most, and until recently inherent in the need for prolonged mechanical ventilation lay the risk of further potentiating instability in patients with the low-output state.
We previously reported a significant improvement in cardiac output when a brief period of cuirass negative pressure ventilation (NPV) replaced conventional intermittent positive pressure ventilation (IPPV) in a small group of children after right heart surgery (2), of whom five had undergone TOF repair. These preliminary findings prompted further investigation of the effects of NPV in these and other patient groups. Fontan patients (3), and other postbypass patients (4), have been reported elsewhere, and the complete study group of 23 TOF patients, which includes the early cohort, are presented here.
The expansion of the study has allowed a more detailed evaluation of the hemodynamic effects of NPV. First, we have been able to compare the effects of this ventilatory modality on nonrestrictive and restrictive patients. Second, by extending the study period, we have examined the effects of both a longer period of NPV, and also of discontinuing NPV in the early postoperative period after TOF repair.
The hemodynamic effects of NPV were prospectively studied in 23 consecutive patients (median age 1.3 years) who were initially receiving IPPV in the pediatric intensive care unit between 4 and 18 h after complete correction of TOF. The study was approved by the Ethics Committee of The Royal Brompton Hospital, and written consent was obtained preoperatively from the parents of all patients.
Anthropometric data with details of prior surgical palliation, placement of a transannular patch to the right ventricular outflow tract and postoperative echocardiographic features are given in Table 1. Three patients had TOF with an atrioventricular septal defect: in the settings of Down’s syndrome (patients 5 and 19) and the CHARGE association (patient 21). All patients underwent a right ventriculotomy, and eight (see Table 1) had enlargement of the right ventricular outflow tract using a transannular patch. Patient 7 had TOF with acquired pulmonary atresia and required insertion of a 21-mm cryopreserved aortic homograft. The sternum of all patients was closed postoperatively.
All patients had a postoperative echocardiogram before the study to confirm the presence or absence of restrictive right ventricular physiology (see below) and any residual ventricular septal defect. No patient had an interatrial communication or a peak instantaneous Doppler gradient from right ventricle to pulmonary artery of greater than 20 mm Hg, which would indicate significant residual right ventricular outflow tract obstruction.
Definition of restrictive right ventricular physiology
Restrictive physiology was present in eight patients at the time of study, and was characterized by antegrade diastolic pulmonary arterial flow that was present throughout the respiratory cycle.
All patients were intubated with cuffed endotracheal tubes (Malinckrodt, Northampton, United Kingdom), and were paralyzed (intravenous infusions of vecuronium 50–80 μg/kg/h) and fully sedated (intravenous infusions of morphine 10–40 μg/kg/h and midazolam 100–300 μg/kg/h) throughout the study period. All patients were initially receiving volume-cycled IPPV (Servo Ventilator 900C; Siemens, Sweden). NPV was delivered using the Hayek oscillator (Breasy Medical Equipment, Hendon, UK). This consists of flexible cuirass and a bedside power unit, and has adjustable rate, inspiratory and expiratory pressures, and inspiratory-to-expiratory ratio.
A cuirass was selected to fit each patient from the level of the clavicles to just below the umbilicus, and using Velcro straps was secured over the chest and abdomen to a bean-filled pillow on which the patient lay. During NPV, the positive pressure ventilator was used to provide the flow of additional inspired oxygen, and set with a trigger sensitivity of −4 cm H2O to deliver a brief period of pressure support; this was necessary to overcome airway resistance from the endotracheal tube. The inspired oxygen fraction was not altered during NPV, and the rate and inspiratory pressures during NPV were adjusted to give a similar expired minute volume and end-tidal carbon dioxide (CO2) to those used during IPPV. Ventilatory settings used during IPPV and NPV are summarized in Table 2.
Pulmonary blood flow measurement
Pulmonary blood flow (Qp) was measured with patients in a cardiorespiratory steady-state, using the direct Fick method. In all cases, mixed venous oxygen content was calculated from pulmonary arterial samples, and arterial oxygen content from systemic arterial or left atrial samples. Oxygen (O2) consumption was measured using the inert gas dilution method, by on-line respiratory mass spectrometry (5). Five patients (Table 1)had echocardiographic evidence of single small residual restrictive ventricular septal defects at the time of study that were hemodynamically insignificant.
All patients had surface electrocardiographic monitoring and pulse oximetry, and invasive monitoring of systemic blood pressure, pulmonary arterial pressure and central venous pressure. In addition, end-tidal CO2was continuously measured using respiratory mass spectrometry.
All patients underwent a standard study, where Qp was measured in a cardiorespiratory steady state during intermittent positive pressure ventilation (IPPV1), and after a 15-min negative pressure ventilation (NPV1).
In 17 children, the study period was extended beyond the standard protocol. In the remainder, therapeutic interventions that could potentially alter our steady-state assumption, such as weaning of inotropes or blood transfusion, precluded extension of the study period. These patients were sub-grouped as follows.
In nine patients after completion of a standard study, a third measurement of Qp was made 15 min after restarting intermittent positive pressure ventilation (IPPV2).
In eight different children, NPV was continued at the end of a standard study, and a third measurement of Qp was made after a further 20- to 30-min negative pressure ventilation (NPV2). Four of these patients had restrictive physiology, and four had nonrestrictive physiology.
Within-patient comparisons at two time-points only were made using the Wilcoxson-signed rank-sum test. Data for sub-groups in whom more than two (repeated) measurements were made in individual patients were compared using the Analysis of Variance with Bonferroni’s correction for multiple comparisons. Between-group comparisons were made using the Mann-Whitney U test. All results are expressed as mean (SD). The null hypothesis was rejected for p values of >0.05.
All studies proceeded without complication or any clinical evidence of hemodynamic instability. Comparable gas exchange, as indicated by end-tidal CO2and arterial CO2tension, was easily achieved during NPV in all cases.
(IPPV1vs NPV1). All tetralogy patients (N = 23)
The results for standard studies are given in Table 3. Mixed venous O2saturation and content increased significantly during NPV1(p = 0.0002). Systemic arterial saturation and O2content were unchanged during NPV1, and so the arteriovenous O2difference fell from 7.2 (2.5) ml/dl to 6.1 (2.4) ml/dl (p = 0.0001).
Qp during IPPV1ranged from 1.1 to 5.9 liters/min/m2, and ranged from 1.4 to 8.4 liters/min/m2at NPV1. In patient 10, Qp decreased by 4.5% during NPV1, and in the remainder, it increased by between 7% and 99%. The mean increase in Qp for all TOF patients was 38.8% (28%) (p < 0.0001; Fig. 1).
Restrictive vs nonrestrictive patients
Restrictive right ventricular physiology was present in eight (35%) of the children at the time of study. Baseline data for restrictive and nonrestrictive patients are compared in Table 4. Systemic arterial CO2tension was similar in both sub-groups; however, patients with restrictive physiology had a more significant metabolic acidosis. Interestingly, the increase in Qp for restrictive patients during standard studies tended to be lower than the increase in nonrestrictives, although this did not reach statistical significance (p = 0.1).
Sub-group 1: IPPV1→ NPV1→ IPPV2(Table 5)
In these nine patients, an increase in Qp of 40% during NPV1was followed by a fall back towards baseline after NPV had been discontinued; the changes are illustrated in Figure 2. There was no group difference between the initial and final Qp during IPPV1and IPPV2, respectively.
Subgroup 2: IPPV1→ NPV1→ NPV2
Table 6shows the results for this sub-group of eight children who received an extended period of NPV. The total increase in Qp from IPPV1to NPV2was 67% (30%), and the relative increase during NPV2was significantly higher than that during NPV1. These changes are illustrated in Figure 3a.
Restrictive versus nonrestrictive patients in sub-group 2
The changes in Qp in restrictive (n = 4) and nonrestrictive (n = 4) patients receiving an extended period of NPV are compared in Table 7, and illustrated in Figure 3b. The results of statistical analysis for this sub-group should be interpreted with relative caution, as they reflect comparisons between small numbers of patients.
The initial increase in Qp during NPV1tended to be higher for nonrestrictives than restrictives (26% vs 19%; p = 0.09), but the mean increase during NPV2was significantly higher in the restrictive patients (54%) than nonrestrictives (20%). Thus, the total increase in Qp during an extended period of NPV was higher for restrictive patients (84%) than nonrestrictive patients (50%).
Although baseline Qp at IPPV1tended to be lower for restrictives than nonrestrictives, there was no difference in final values at NPV2(3.8 liters/min/m2for restrictives vs 4.0 liters/min/m2in nonrestrictives; p > 0.5).
A brief period of NPV increased the pulmonary blood flow of children after repair of TOF by 39%, and this improvement further continued if the period of NPV was extended. Patients with restrictive physiology tended to respond more slowly at first, but after a prolonged period of NPV, their pulmonary blood flow increased by over 80%.
Patients with restrictive right ventricular physiology have been previously shown to spend an average of 5 d longer on the intensive care unit than their nonrestrictive counterparts (1). Their early postoperative progress can be complicated initially by a transient low-output state, and subsequently hindered by fluid overload: a direct consequence of the aggressive volume resuscitation that they typically require in order to maintain an adequate cardiac output. In addition, these children will inevitably be exposed to the adverse aspects of prolonged intensive care stay, such as the increased risk of infection, and pulmonary consequences of mechanical ventilation.
The hypothesis central to this study was that a period of NPV might improve the early postoperative hemodynamics of children with a low-output state after repair of TOF, and more importantly, that this ventilatory modality may have a potential application as a hemodynamic tool in selected cases. In these patients, an unusual relationship exists between pulmonary blood flow and intrathoracic pressure: that is, with higher airway pressures, less antegrade pulmonary blood flow and more pulmonary regurgitation occur, thus reducing the cardiac output. Ventilation with a negative mean airway pressure was hypothesized to be particularly indicated in this sub-group.
The study group was of sufficient size to allow a more detailed investigation of the manipulation of the cardiopulmonary interactions in children receiving NPV after TOF repair than was previously possible (2). First, restrictive and nonrestrictive patients could be compared; second, the study gave an opportunity to examine any temporal influences on the hemodynamic changes induced by a period of NPV in the acute postoperative period. Third, the extended studies provided valuable data indicating the potential scope for improvement in pulmonary blood flow in patients with a low cardiac output.
Mechanisms for the increase in pulmonary blood flow
There are a number of ways in which NPV may influence the circulation, but a direct augmentation of myocardial contractility by NPV should first be discounted. Biventricular systolic performance is usually well preserved in children who have undergone repair of TOF, and in the absence of a change in serum catecholamines, acid-base balance or heart rate, it would seem unlikely that a short period of NPV could directly affect the myocardium. Indeed, a theoretical concern regarding the use of NPV would be its potential to increase the left ventricular afterload (6,7), and while not an issue in the patients studied here, this ventilatory technique may perhaps be less applicable to those with significant left ventricular dysfunction.
A much more likely explanation for the mechanism underlying the augmentation of cardiac output during NPV is its influence on the passive properties of the heart and pulmonary circulation. The influences of spontaneous and mechanical ventilation on systemic venous return are well understood (8–10): in brief, a reduction in intrathoracic pressure during normal (or negative pressure) inspiration increases the venous pressure gradient between the peripheries and the thorax. This increases the systemic venous return, and thus augments the cardiac output. The converse happens during positive pressure ventilation, with a reduction in systemic venous return accompanying the increase in intrathoracic pressure. Moreover, the effects of mechanical ventilation on the right heart are not confined to its influence on preload. If the lungs are inflated beyond functional residual capacity, positive pressure ventilation can also increase right ventricular afterload, and hence reduce stroke volume (11). This can be particularly important in patients with restrictive right ventricular physiology whose cardiac output is exquisitely dependent on a high central venous pressure and the maintenance of a low pulmonary vascular resistance.
Comparison between restrictive and nonrestrictive patients
It was initially surprising to find that the improvement in pulmonary blood flow during standard studies tended to be lower in restrictives than nonrestrictive patients. However, the data from patients receiving an extended period of NPV suggested that the rate of increase in pulmonary blood flow subsequently caught up with and then overtook the nonrestrictives as the period of NPV was extended. Thus, the overall increase was greater in restrictive patients, presumably a manifestation of an ultimately profound manipulation of diastolic pulmonary arterial flow.
Why was the initial increase lower in restrictive patients? Our explanations for this finding must remain speculative. Metabolic acid-base balance is a nonspecific marker of hemodynamic status, clinical stability and global oxygen delivery. A mild metabolic acidosis is not uncommon in children after cardiopulmonary bypass, but in this study, restrictive patients were more acidotic than the nonrestrictives. This may have reflected relatively more constricted systemic and pulmonary vascular beds, which increased resistance or impedance to flow, and could perhaps explain the rather “delayed” hemodynamic response to NPV.
Many other mechanisms may exist, but our data clearly show that the hemodynamic effects of NPV are sustained and may improve further with longer term application.
NPV led to a significant improvement in the cardiac output of children after repair of TOF. In restrictive patients, the ultimate potential for improvement of cardiac output with the additional manipulation of diastolic pulmonary blood flow was greater than in nonrestrictives, but the onset of this clinically important effect was slightly delayed, presumably reflecting the more marked hemodynamic and metabolic abnormalities seen in this sub-group.
There is little doubt that the sequelae of restrictive right ventricular physiology last for much longer than its direct adverse hemodynamic effects. If intervention with NPV is early, then many of the adverse effects of medical therapy, and the multi-organ consequences of a low-output state, can potentially be avoided in selected patients.
☆ This work was supported by The British Heart Foundation, and LIFFEbenefits.
- carbon dioxide
- intermittent positive pressure ventilation
- negative pressure ventilation
- pulmonary blood flow index
- tetralogy of Fallot
- Received March 20, 1998.
- Revision received June 18, 1998.
- Accepted October 26, 1998.
- American College of Cardiology
- Cullen S,
- Shore D,
- Redington A
- Shekerdemian L.S,
- Shore D.F,
- Lincoln C,
- Bush A,
- Redington A.N
- Shekerdemian L.S,
- Bush A,
- Shore D,
- Lincoln C,
- Redington A.N
- Shekerdemian L.S,
- Bush A,
- Lincoln C,
- Shore D.F,
- Petros A.J,
- Redington A.N
- Scharf S.M,
- Brown R,
- Tow D.E,
- Parisi A.F
- Peters J,
- Fraser C,
- Stuart R.S,
- Baumgartner W,
- Robotham J.L
- Guyton A.C,
- Lindsey A.W,
- Abernathy B
- Pinsky M.R
- Whittenberger J.L,
- McGregor M,
- Berglund E,
- Borst H.G