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

This Article Abstract 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): Philippe Deleuze Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Diehl, J.-L. Articles by Brochard, L. Search for Related Content PubMed PubMed Citation Articles by Diehl, J.-L. Articles by Brochard, L.

J Thorac Cardiovasc Surg 1994;107:487-0498
© 1994 Mosby, Inc.

Cardiopulmonary Bypass, Myocardial Management, and Support Techniques

Clinically relevant diaphragmatic dysfunction after cardiac operations

Créteil, France

From Réanimation Médicaler, Service de Chirugie Cardiaque, INSERM U296, Université Paris XII, Hôspital Henri Mondor, Creteil, France.

Received for publication April 13, 1992. Accepted for publication July 7, 1993. Address for reprints: Laurent Brochard, MD, Réanimation Médicale, Hôpital Henri Mondor, 94010 Creteil, France.

Abstract

Phrenic nerve injury and diaphragmatic dysfunction can be induced by cardiac operation. The clinical consequences are not well-established. We evaluated 13 consecutive patients over a 2-year period with unexplained and prolonged difficulties in weaning from mechanical ventilation. The mean time of measurement from the operation day was 31 ± 19 days (range 8 to 78). With the same technique we also evaluated 12 control patients: four patients at day 1 after cardiac operation while they were still intubated; four normally convalescing patients at day 7 or 8 after cardiac operation; and four patients who required prolonged mechanical ventilation because of another identified cause after cardiac operation. Diaphragmatic function was evaluated at the bedside from esophageal and gastric pressure measurements. A low or negative ratio of gastric pressure swing to transdiaphragmatic pressure swing, indicative of diaphragm dysfunction, was found in all 13 patients (mean -0.39 ± 0.64). The difference between the 13 patients and all control groups was found to be highly significant. Transdiaphragmatic pressure measured during a maximal voluntary inspiratory effort and transdiaphragmatic pressure measured during a short, sharp sniff were markedly diminished (28 ± 18 cm H₂O and 13 ± 15 cm H₂O, respectively) in the 13 patients, significantly different from values in the four control patients studied at day 7 or 8. Transdiaphragmatic pressure measured after magnetic stimulation in four patients was also markedly reduced (7 ± 5 cm H₂O) as compared with normal theoretic values. Aminophylline infusion had no effect on any of these parameters. In one of two patients evaluated a second time, about 5 weeks later, a marked improvement was observed. Estimating the prevalence of clinically relevant diaphragmatic dysfunction, we found it to be 0.5% when no topical cooling was used and 2.1% when iced slush with no insulation pad was added for myocardial protection (p < 0.005). The most striking finding was that the clinical course of the 13 patients was marked by severe intercurrent events, including cardiorespiratory arrest after early tracheal extubation in 5 patients, nosocomial pneumonia in 11, prolonged mechanical ventilation in all (58 ± 41 days), and a fatal outcome in 3. We conclude that prolonged postoperative diaphragmatic dysfunction may cause severe life-threatening complications after cardiac operation and can be limited to some extent by avoiding the use of iced slush topical cooling of the heart. (J THORAC CARDIOVASC SURG 1994;107:487-98)

Diaphragmatic dysfunction as a result of phrenic nerve injury may follow cardiac operation. The prevalence of this complication is unknown, partly because partial or unilateral impairment may go unrecognized. Several works have suggested that it may induce postoperative atelectasis, reduced exercise capacity, orthopnea and dyspnea, and even the need for prolonged mechanical ventilation. ^1-11 The clinical relevance of this complication, however, is not well-understood. In addition, in previous studies, transdiaphragmatic pressure measurement was never used to confirm and quantify diaphragmatic dysfunction. Such measurement is important because roentgenographic signs of diaphragmatic dysfunction have a poor specificity and alterations in the latency of conduction in the phrenic nerve may have various clinical consequences. ⁶

We report here the cases of 13 patients whose prolonged difficulty in weaning from mechanical ventilation after cardiac operation was a result of diaphragmatic dysfunction. These 13 patients exhibited markedly abnormal results of transdiaphragmatic pressure measurements, as compared with those in 12 control patients. The high rate of morbidity associated with this complication suggests that it may contribute to postoperative mortality.

METHODS

Patients
The patients included in this study had all undergone cardiac operation. The selection criteria were the need for prolonged artificial mechanical ventilation after cardiac operation and unexplained difficulty in weaning the patient from the ventilator, as explained herein.

After cardiac operation, all patients were initially treated in the cardiosurgical intensive care unit of the hospital. The need for prolonged mechanical ventilation, that is, more than 6 days, was an indication for the transfer of the patient to the hospital's medical intensive care unit. For many of these patients it was possible to treat the causes of their need for prolonged ventilation including sepsis, heart failure, fluid overloading caused by renal failure, or stroke. When no explanation could be found for the inability of the patients to breathe by themselves, either initially or after the primary cause was cured, the patients were enrolled in the study and underwent assessment of diaphragmatic function.

Thirteen patients were included and are further mentioned as group 1. When the measurements required for the study were done, no patient showed evidence of cardiac dysfunction (as assessed by echocardiographic and hemodynamic studies), untreated nosocomial pneumonia, or ongoing sepsis. The 13 patients were recruited over a 23-month period, from April 1988 to February 1990. During the same period, a total of 1478 patients had undergone cardiac operation in our hospital, performed by different surgeons. Methods of myocardial protection were similar among surgeons (crystalloid anterograde cardioplegia and whole body hypothermia at 28° C) and, during this time, only one of them added topical hypothermia with iced slush without an insulation pad.

Assessment of diaphragmatic function was also done in 12 patients recruited among three control groups from November 1992 to February 1993. Group C1 included four patients who were studied 1 day after the cardiac surgical procedure just before extubation. Because these patients were frequently submitted to sedation just before study or had postoperative pain secondary to sternotomy or chest tubes, which could interfere with the measurements, four additional patients were studied at day 7 or 8, just before discharge from the cardiosurgical department. These patients are further mentioned as group C2. Finally, a third group (C3) included four patients who required prolonged mechanical ventilation (i.e., more than 6 days) after the cardiac surgical procedure because of various reasons including sepsis, hemodynamic impairment, or prolonged sedation. Indeed, the consequences of prolonged mechanical ventilation and prolonged stay in the intensive care unit on diaphragmatic function are not well-established. Three patients in this group were still intubated at the time of the study. It is noteworthy that anterograde cardioplegia alone was used for the control patients with no ice slush. All subjects gave informed consent for assessment of diaphragmatic function.

Measurements
Measurements were done with subjects in a semirecumbent position. Data were recorded as follows:

First, esophageal and gastric pressures (Pga) were measured with a double-balloon catheter (Marquat, Boissy St Léger, France) connected to two differential pressure transducers (Sensym SDX001, ±70 cm H₂O, Plaisir, France). The esophageal balloon was positioned in the middle third of the esophagus. It contained 1 ml air and was used as an index of pleural pressure. The gastric balloon was positioned in the stomach. It contained 1 ml air and was used as an index of abdominal pressure. Because gastric pressure and pleural pressure tracings were often found to be similar in patients from group 1, a situation different from that prevailing in normal subjects, three tests were done to ensure correct positioning of the catheters: first, an occlusion test to assess the validity of the esophageal pressure measurement ¹²; second, a search for positive deflections on gastric pressure tracings, made while one of the operators applied gentle pressure to the patient's stomach; and third, the patient was given a drink of water after which the typical tracing showed a sharp rise in esophageal pressure resulting from muscular contraction of the esophagus, without any concomitant modification of gastric pressure (Fig. 1).

View larger version (65K): [in this window] [in a new window]   Fig. 1. Two typical tracings of esophagel pressure (Peso) and Pga indicative of diaphragmatic dysfunction in two patients after cardiac operation. In left panel (patient 12), there is no positive deflection of Pga during inspiration, which can be recognized by negative swing of Peso. In right panel (patient 3), dysfunction is more pronounced , inasmuch as Pga describes negative decay during inspiration. At end of tracing, espophageal spasm occurs without any increase in Pga (see methods section).

Third, diaphragmatic function was assessed by the index of Gilbert, Auchincloss, and Peppi, ¹³ that is, by the Pga/Pdi ratio, which assesses the diaphragmatic contribution to respiratory pressure swings, where Pga is the difference between peak inspiratory gastric pressure minus end-expiratory gastric pressure and Pdi is the transdiaphragmatic pressure change during inspiration. Normally, this index is positive, but when abdominal pressure decreases during inspiration, suggesting a paradoxic motion of the diaphragm, it becomes negative.

Fourth, Pdi was also measured during (1) a maximal inspiratory effort at functional residual capacity, that is, Pdi max, and (2) a maximal short sharp sniff, done during partial occlusion of the endotracheal tube at functional residual capacity in intubated patients, that is, Pdi sniff. Pdi sniff is about 80% to 100% of Pdi max in normal subjects. Several maneuvers were done to measure each of these indexes during which patients were encouraged by being able to watch their performance on the paper recorder. However, some patients from group 1 and group C3 were unable to do these maneuvers. Pdi max and Pdi sniff were used to assess diaphragmatic function.

Fifth, the Pdi/Pdi max index was calculated as the ratio of mean Pdi to Pdi max. Mean Pdi was obtained by assessing the pressure values every 100 msec during the inspiratory phase, defined as the period during which Pdi remained above the baseline level. This ratio determines the time required to produce diaphragmatic fatigue in man. ¹⁴ A Pdi/Pdi max below 0.4 indicates that, theoretically, the Pdi can be generated indefinitely.

Sixth, the Pdi inspiratory time integral, TTdi, was also calculated by the method of Barnard and Levine ¹⁵ as the product of the mean Pdi, expressed as a fraction of Pdi max, and the inspiratory duty cycle Ti/Ttot calculated from the pressure tracing (Ti/Ttot): TTdi = Pdi/Pdi max x Ti/Ttot. TTdi is used to indicate whether diaphragmatic pattern of contraction has reached the fatiguing zone, defined by a TTdi higher than 0.15. Normal subjects breathing at rest have a TTdi of 0.02 to 0.05.

Seventh, a measurement called Pdi stim was also obtained in four patients with the use of cervical magnetic stimulation (stim) at supramaximal levels as described by Similowski and coworkers. ¹⁶

Eighth, the same measurements were repeated in nine patients from group 1 after an infusion of aminophylline (5 mg/kg over a period of 20 minutes).

Ninth, the same measurements were also repeated in two of these patients just more than 5 weeks later.

Statistics
Data are shown as means plus or minus the standard deviation. Comparisons of quantitative values were made with a paired t test or the nonparametric Mann-Whitney test as required by the number of patients. Comparisons of qualitative data were made by a ² test. Correlations between quantitative data were made by univariate analysis with a t test.

RESULTS

The main characteristics of the patients and postoperative events are shown in Table I for group 1 and Table II for the control groups. The average period between the day of operation and the start of the study was 31 ± 19 days (range 8 to 78) in group 1 and 20 ± 6 days (range 8 to 42) in group C3 (not significant).

All patients but one had had one or more episodes of pneumonitis before entering the study. Each episode had been proved by fever, leukocytosis, new roentgenographic findings, and protected bacteriologic samples with quantitative cultures. ¹⁷ Mediastinitis occurred in two of the five patients who required the longest period of mechanical ventilation. Five of the 13 patients had one or more episode of cardiorespiratory arrest that required closed-chest cardiopulmonary resuscitation. In all five patients, this episode occurred early in the postoperative period while they were breathing spontaneously and after tracheal extubation. Emergency reintubation was necessary during these episodes, which were usually followed by the patient's transfer to the medical intensive care unit. Tracheotomy was required in five patients. A simplified acute physiologic score was determined, as an index of the severity of illness on admission to the unit. ¹⁸

Radiologic presentation
Postoperative chest x-ray films of the patients from group 1 frequently showed an elevation of one or both hemidiaphragms. However, the x-rays films varied greatly depending on the patient's position and lung inflation. Fluoroscopic examination of diaphragmatic motion during the sniff test indicated diaphragmatic dysfunction in seven patients, but this motion was often made difficult to interpret by the presence of rapid shallow breathing or the concomitant presence of expiratory muscle activity.

One patient from group C2 exhibited transient left lower lobe atelectasis on postoperative x-ray films. Interestingly, this patient had the lowest value of the Gilbert index in this group (see discussion section).

Diaphragmatic function in group 1
Table III gives the results recorded for Pdi, Pga/Pdi, Pdi max, and/or Pdi sniff, Pdi/Pdi max, Ti/Ttot, and TTdi measurements.

Pdi max and Pdi sniff were respectively obtained in 11 and 6 patients. Their mean values, respectively 28 ± 18 cm H₂O (5 to 72) and 13 ± 15 cm H₂O (4 to 44), were markedly below the theoretic values.

A positive correlation was found between Pga/Pdi and Pdi max (r = 0.62, p < 0.05). The mean value of Pdi/Pdi max was 0.24 ± 0.09 (range 0.11 to 0.40). The mean TTdi was 0.14 ± 0.08. Individual values suggested that seven patients were at high risk of diaphragmatic fatigue (TTdi higher than 0.15).

Pdi stim was determined in four patients (Table IV). According to the theoretic values reported by Similowski and associates, ¹⁶ it was markedly below normal. Nevertheless, when Pdi stim values were expressed as fractions of Pdi max or Pdi sniff, they were similar to the values specified by these authors, indicating that both Pdi stim and Pdi max diminished in the same proportions.

View larger version (12K): [in this window] [in a new window]   Fig. 2. Effects of aminophylline infusion on main parameters used to assess diaphragmatic function, measured in nine patients. Left panel: values for Pdi during quiet breathing before and after aminophylline infusion. Right panel:Pga / Pdi and Pdi /Pdi max, before and after aminophylline infusion and for Pdi max in the two conditions. None of the changes was significant. Values presented as means plus or minus standard error of mean.

View larger version (23K): [in this window] [in a new window]   Fig. 3. . Pga / Pdi ratio in patients with unexplained unsuccessful weaning after cardiac operation (group 1) and in control patients at day 1 after cardiac operation (group C1), at day 7 or 8 after cardiac operation (group C2), and after prolonged mechanical ventilation after operation (group C3). Asterisk indicates significant difference with p value <0.01.

Mean values of Pdi/Pdi max were lower in control groups than in group 1 (respectively 0.19 ± 0.08, 0.09 ± 0.03, and 0.16 ± 0.06 in groups C1, C2, and C3). The difference was significant between group 1 and group C2 (p < 0.01). Similar results were observed for the mean TTdi with near-normal values in all control groups. The difference was significant between group 1 and group C1 (p < 0.05) and group 1 and group C2 (p < 0.01).

Relationships between clinical events and measurements in group 1
Three major links between clinical events and measurements in group 1 were observed. As expected, a paradoxic abdominal motion, at best, was seen in patients who exhibited the most severe diaphragmatic dysfunction, as assessed by pressure measurements. One striking observation was that the five patients who had cardiorespiratory arrest were among the seven who exhibited the highest degree of diaphragmatic dysfunction, as assessed by a negative Pga/Pdi index. For all patients, mechanical ventilation and intensive care unit stay lasted for long periods, the longest being 160 days. Finally, among the five patients with a Pga/Pdi index below -0.35, three had a fatal outcome.

Relationships between surgical procedure and measurements
There was no correlation in any of the 13 patients from group 1 between any measurement or clinical event and surgical characteristics, including bypass time, aortic crossclamp time, or type of myocardial protection. However, as mentioned in the methods section, iced slush myocardial protection was only used in our hospital by one surgeon. He did not use an insulation pad. This surgeon operated on 236 of the 1478 patients who underwent cardiac operation with extracorporeal circulation during the first study period. Because five of the 13 patients from group 1 were operated on with this technique, we estimated that the prevalence of clinically relevant diaphragmatic dysfunction was 2.1% when iced slush was used versus 0.5% when it was not used. Therefore the iced slush method multiplied by four the risk of prolonged artificial mechanical ventilation as a result of diaphragmatic dysfunction, a difference that was highly significant (p < 0.005). Nevertheless, there was still a slight risk of clinically relevant diaphragmatic dysfunction with methods other than iced slush protection.

DISCUSSION

Our data show that after cardiac operation the need for prolonged mechanical ventilation and unexplained difficulties in weaning from the ventilator are strongly correlated with the presence of diaphragmatic dysfunction. The clinical consequences of the dysfunction include atelectasis and recurrent pneumonia, orthopnea, and a need for prolonged mechanical ventilation. ^1-11 Although they may be asymptomatic or unrecognized, our data suggest that severe, life-threatening complications may result from diaphragmatic dysfunction after cardiac operation.

It is noteworthy that five of our patients from group 1 underwent emergency reintubation for respiratory failure and cardiac arrest. Arterial blood gas samples taken shortly before intubation showed hypoxemia and respiratory acidosis, suggesting that the respiratory failure occurred first, as a result of diaphragmatic dysfunction, and was followed by hypoxemic cardiac arrest. It is therefore likely that these patients were extubated too early and that secondary respiratory failure was a result of gradual exhaustion, leading to the cardiac arrest. Clinical signs of diaphragmatic dysfunction are, however, difficult to diagnose in the early postoperative period, because of the presence of chest tubes, of abdominal muscle recruitment, and of frequent postoperative pain. This partly explains why diaphragmatic dysfunction was only suspected several days or weeks after operation, when patients had recovered from the initial episode and its consequences.

Nosocomial pneumonia was also a frequent event that was a direct result of the need for prolonged mechanical ventilation. ¹⁹ Emergency reintubation, cardiac arrest, and nosocomial pneumonia all carry a high mortality rate, ^19-21 and therefore the true prevalence of severe diaphragmatic dysfunction may have been underestimated in our study, inasmuch as part of the postoperative mortality may have been a result of unrecognized diaphragmatic dysfunction. Three of the patients in this series had a fatal outcome after the diagnosis had been made, and one died shortly after discharge from the hospital, of uncertain cause.

As far as we know, there have never been reports of the assessment of diaphragmatic function by transdiaphragmatic pressure measurement in patients with postoperative complications after cardiac operation. Moreover, the course of transdiaphragmatic pressure measurements in patients with uncomplicated postoperative periods had never been investigated.

In fact, diaphragmatic function can be assessed in the intensive care unit in different ways. Transdiaphragmatic pressure measurement represents one of these ways. It is not possible to directly measure the forces produced by the respiratory muscles. Thus diaphragmatic strength is usually assessed indirectly by measuring the pressure generated across the diaphragm. This is achieved by the balloon catheter system used in this study, which allows the determination of the Pdi under different conditions:

1. Peak and mean Pdi are usually recorded at inspiration during quiet tidal breathing: extremely low values of Pdi could suggest a diaphragmatic paralysis, but more sensitive indexes are available.
   
2. The determination of the ratio Pga/Pdi over inspiration during quiet tidal breathing represents a good index of the contribution of the diaphragm to the respiratory pressure swings. ¹³ This index, termed Gilbert's index, is not dependent on the patient's cooperation. In addition, a negative value of this index clearly indicates isolated diaphragmatic dysfunction or paralysis in a specific manner.
   
3. Pdi max and Pdi sniff are also reduced in cases of diaphragmatic dysfunction. However, Pdi max and Pdi sniff are dependent on the subject's ability and motivation to cooperate, which represents certainly a major limitation in cases of acutely ill patients. In addition, these measurements do not differentiate isolated diaphragmatic dysfunction from global respiratory muscle weakness, whereas Pga/Pdi does.
   
4. Finally, it is also possible to record the Pdi "twitch" after a phrenic cervical stimulation at a supramaximal voltage level. The electromagnetic method used in the present study is a painless method for stimulation. ¹⁶ Patient cooperation is not critical for this measurement.
   

Some other limitations besides the patient's cooperation must be considered in obtaining measurements. The double-balloon catheter should be properly positioned. This can be relatively easily controlled by the means mentioned previously. Another important point is the dependance of the pressure developed on muscle length and geometry. Therefore major determinants of Pdi are diaphragmatic contractility or activity, but also lung volume and chest wall configuration. Interpretation of the Pdi and Pga tracings should also consider the possibility in some patients of an expiratory activity of the abdominal muscles that is manifest as an increase in the Pga curve during expiration followed by a rapid relaxation at the end of expiration. Thus correct interpretation of the alterations in Pdi swings necessitates screening for these parameters before a change in diaphragmatic activity or contractility is considered.

Others tools are available to assess diaphragmatic function that were not used in the present study including bilateral cervical phrenic stimulations associated to the recording of a diaphragmatic electromyogram by cutaneous, esophageal, or needle electrodes. This method, however, is more appropriate for laboratory settings than for bedside application. Unilateral phrenic nerve stimulation allows the determination of the right and left phrenic latencies, the functional consequences of which, however, are not clear. It is also possible to record the motion course of both hemidiaphragms by roentgenologic, fluoroscopic, or echographic methods. The latter could represent a promising noninvasive approach, but has not been validated against the reference methods for these indications. ^{11, 22}

Despite these limitations, it was nevertheless possible to diagnose severe diaphragmatic dysfunction in most of our patients from group one. The Pga/Pdi ratio (Gilbert's index) was significantly reduced in group 1 as compared with that in the other groups. The negative values we obtained in seven patients from group 1 strongly suggested a very low diaphragmatic contribution to respiratory pressure swings. Four patients had a Pga/Pdi index of 0, which was consistant with the absence of inspiratory diaphragmatic motion and might therefore be indicative of less severe diaphragmatic dysfunction. Positive values for the Pga/Pdi index were observed in two patients from group 1, but were lower than the normal values. The possible explanations for this finding are conflicting, because an abnormally low positive Pga/Pdi index may be connected with minor diaphragmatic dysfunction, but it may also signify increased rib cage activity. The same explanations can be proposed for patients from group C1 who all exhibited positive but reduced values. One patient in the C2 group exhibited a marked reduction of the Gilbert's index at day 8 after cardiac operation: interestingly, the radiologic course was suggestive of transient left phrenic dysfunction in this patient.

The falls in Pdi max and Pdi sniff values also indicated severe diaphragmatic dysfunction in most patients in group 1. The difference was significant between this group and the patients with uncomplicated recovery studied at day 8 (group C2). However, no clear difference was observed between group 1 and patients from groups C1 and C3. In the early postoperative period (group C1), patient cooperation and ability to perform these maneuvers was not always optimal and postoperative pain could also have contributed to the observed reduced values. In addition, a transient mild diaphragmatic dysfunction has been described in the first postoperative hours after cardiac operations, similar to what is observed after upper abdominal operations. ²³ In patients from group C3, a global alteration of contractility of respiratory muscles is a possible explanation for these reduced values associated with near-normal values of the Gilbert's index. This group, however, was the most difficult to select. Indeed, mechanical ventilation was required for a number of medical reasons that resulted in inability of the patient to tolerate spontaneous breathing. First, these factors generally made difficult the performance of tests that required patient cooperation and, second, we cannot exclude the possibility that mild diaphragmatic dysfunction participated in their difficult weaning.

The values observed for the TTdi index indicated that seven patients from group 1 were exposed to diaphragmatic fatigue while spontaneously breathing. This was not the case in any of the control patients and the difference was significant between group 1 and groups C1 and C2. However, it is difficult to interprete the results of this index in terms of diaphragmatic fatigue in the presence of diaphragmatic dysfunction. At least, it indicates that patients performed work in considerable excess of their force reserve. Pdi stim also displayed a marked drop in four patients from group 1. It is of interest to note that the ratios of Pdi stim/Pdi max and Pdi stim/Pdi sniff were comparable to the values obtained in normal subjects by Similowski and associates. ¹⁶ The fall in Pdi stim might be a result of unilateral phrenic nerve paralysis or unilateral or bilateral paresis or a decrease in diaphragmatic contractility. Because no electromyogram or latency was recorded, it was impossible to distinguish between these mechanisms.

Aminophylline infusion had no effect on any parameter of diaphragmatic function in nine patients from group 1. By contrast, such infusion was reported to be beneficial after upper abdominal operation, ²⁴ which is also reported to induce diaphragmaticdysfunction. ^25-28 Nevertheless, dysfunction that occurs after abdominal operation is generally transient and its causes are probably different from those observed in prolonged dysfunction after cardiac operation. The absence of effect of aminophylline indirectly suggests that a lesion of the nerve, and not of the muscle itself, was the cause of the dysfunction.

This dysfunction indeed could have been caused by many factors including the following:

1. Altered diaphragmatic contractility: patients from group 1 were subjected to many conditions that are known to potentially impair diaphragmatic contractility, such as sepsis or denutrition. Nevertheless, the study was done in patients who were in stable condition, showed no signs of sepsis, and were given conventional enteral nutrition. Moreover, these conditions are more likely to result in global respiratory muscle weakness than in isolated or more pronounced diaphragmatic dysfunction, as shown by the negative Pga/Pdi index. On the other hand, such an alteration in respiratory muscle contractility may explain the results observed in group C3.
   
2. Mechanical impairment of diaphragmatic motion. This may be induced by sternotomy, postoperative gaseous gastric distension, dynamic hyperinflation, or mediastinitis. However, mediastinitis was observed in only two patients from group 1 and in one from group C3, the mechanical consequences of sternotomy are likely to be transient, and gaseous gastric distention was not observed at the time of our measurements in any patient.
   
3. Neural reflex regulation of breathing, which could be affected by sternotomy itself or by postoperative pain. This could be the cause of the slight alteration observed in group C1, but it is unlikely to explain the more severe alteration in group 1: indeed, no pain was mentioned in any patient in this group and, in most cases, sternotomy itself does not affect weaning from mechanical ventilation after cardiac operation. The near-normal values of diaphragmatic explorations observed in group C2 are consistent with this view.
   
4. Unilateral or bilateral phrenic nerve paresis or paralysis. This was probably the main factor accounting for diaphragmatic dysfunction in patients in group 1. Phrenic nerve paralysis after cardiac operation is a well-known complication and is generally related to methods of myocardial protection that use topical cooling. ^2-5,29,30 In studies assessing phrenic nerve conduction, the incidence of this complication seems relatively low, from 1% when it was diagnosed retrospectively from clinical signs ⁷ to 10% when studied systematically afteroperation. ⁶ Here, too, we found a low incidence, about 1%, but this was probably underestimated, as previously discussed, because it may have been a factor contributing to other postoperative deaths. Note that difficulties in weaning occurred more frequently in the patients who were given myocardial iced slush protection during operation without an insulation pad than in those for whom intermittent saline protection was used. Cardioplegia that uses ice chips or iced slush without an insulation pad is believed to be followed more often by postoperative atelectasis or diaphragmatic dysfunction than other methods. ^2-5,29 However, phrenic nerve paralysis after cardiac operation may result from factors other than hypothermic injury, such as direct damage or stretch injury during the operation. ^1,30 Accordingly, neither ice chips nor iced slush was used in a study of 13 patients with diaphragmatic dysfunction after cardiac operation done by Abd and coworkers. ⁷ In that report, an internal mammary artery graft was suggested to have caused the dysfunction. It is noteworthy that mammary artery harvesting was not correlated in our study with diaphragmatic dysfunction, inasmuch as one patient only was involved during a 2-year period, despite frequent use of this technique.
   

Unilateral or bilateral phrenic lesions may have occurred in our patients. Some authors have studied inspiratory muscle function in patients with unilateral diaphragmatic paralysis, with or without cardiopulmonary disease. ³¹ A decrease of Pdi max wasobserved in half of these patients and a negative Pga/Pdi index in more than half. This is consistent with the present data from group 1, suggesting the possibility of unilateral or bilateral phrenic nerve paralysis. On the other hand, a positive Pga/Pdi index does not entirely exclude unilateral diaphragmatic paralysis, which might have accounted for the diaphragmatic dysfunction in two of these patients.

Experimental and pathologic findings are consistent with the possibility of phrenic nerve demyelination, leading to a transient diaphragmatic dysfunction. ³² Nevertheless, more severe damage, such as axonal damage, might be caused by prolonged exposure to cold and therefore explain a more prolonged phrenic nerve dysfunction. ³³ In the first situation, phrenic dysfunction might account for postoperative atelectasis without serious clinical consequences and could be similar to the postoperative diaphragmatic dysfunction observed after upper abdominal operation. ^{10, 34} On the other hand, more severe postoperative damage to the phrenic nerve could generate prolonged phrenic paralysis or paresis with serious clinical consequences such as persistent unexplained dyspnea, pneumonitis, cardiorespiratory arrest, or, more commonly, difficulties in weaning and the resulting need for prolonged ventilatory assistance. ^{7, 29, 35}

The frequent occurrence of cardiorespiratory arrest, with its strong correlation with the degree of diaphragmatic dysfunction, is a feature of predominant importance in our study. It suggests that respiratory arrest may be a direct consequence of ventilatory failure that results from diaphragmatic dysfunction. Moreover, the three in-hospital deaths in this study occurred among the five patients with the more severe diaphragmatic dysfunction. All three had previously had a cardiorespiratory arrest. Their age, severity index on admission to the medical intensive care unit, and surgical characteristics were not different from those of the other patients. It is therefore likely that their severe diaphragmatic dysfunction and the previous episode of cardiorespiratory arrest contributed to their fatal outcome.

Recovery from such paralysis can take many months after operation. It is of interest to note that 2 months after operation, one of the two patients for whom repeated measurements were made exhibited improved diaphragmatic function, consistent with recovery of the phrenic nerve function.

In conclusion, most of the unexplained difficulties in weaning from mechanical ventilation after cardiac operation were caused in our series by severe diaphragmatic dysfunction, leading to high morbidity and mortality. Such dysfunction may be attributed, at least in part, to cold-induced phrenic paralysis. The use of iced slush cardioplegia without an insulation pad should be avoided in an effort to reduce the prevalence of this complication.

References

1. Markand ON, Moorthy SS, Mahomed Y, King RD, Brown JW. Postoperative phrenic nerve palsy in patients with open heart surgery. Ann Thorac Surg 1985;39:68-73.[Abstract]
   
2. Benjamin JJ, Cascade PN, Rubenfire M, Wajszczuk W, Kerin NZ. Left lower lobe atelectasis and consolidation following cardiac surgery: the effect of topical cooling on the phrenic nerve. Radiology 1982;142:11-4.[Abstract/Free Full Text]
   
3. Rousou JA, Parker T, Engelman RM, Breyer RH. Phrenic nerve paresis associated with the use of iced slush and the cooling jacket for topical hypothermia. J THORAC CARDIOVASC SURG 1985;89:921-5.[Abstract]
   
4. Wheeler WE, Rubis LJ, Jones CW, Harrah JD. Etiology and prevention of topical cardiac hypothermia-induced phrenic nerve injury and left lower lobe atelectasis during cardiac surgery. Chest 1985;88:680-3.[Abstract/Free Full Text]
   
5. Esposito RA, Spenser FC. The effect of pericardial insulation on hypothermic nerve injury during open-heart surgery. Ann Thorac Surg 1987;43:303-8.[Abstract]
   
6. Wilcox P, Baile EM, Hards J, et al. Phrenic nerve function and its relationship to atelectasis after coronary artery bypass surgery. Chest 1988;93:693-8.[Abstract]
   
7. Abd AG, Braun NMT, Baskin MI, O'Sullivan MM, Alkaitis DA. Diaphragmatic dysfunction after open-heart surgery: treatment with a rocking bed. Ann Intern Med 1989;111:881-6.
   
8. Shapira N, Zabatino SM, Ahmed S, Murphy DMH, Sullivan D, Lemole GM. Determinants of pulmonary function in patients undergoing coronary bypass operations. Ann Thorac Surg 1990;50:268-73.[Abstract]
   
9. Laub GW, Muralidharan S, Chen C, et al. Phrenic nerve injury: a prospective study. Chest 1991;100:376-9.[Abstract/Free Full Text]
   
10. Tulla H, Takala J, Alhava E, Huttunen H, Kari A, Manninen H. Respiratory changes after open-heart surgery. Intensive Care Med 1991;17:365-9.[Medline]
    
11. De Vita MA, Robinson LR, Rehder J, Hattler B, Cohen C. Incidence and natural history of phrenic neuropathy occurring during open heart surgery. Chest 1993;103:850-6.[Abstract/Free Full Text]
    
12. Baydur A, Behrakis PK, Zin WA, Jaeger M, Milic-Emili J. A simple method for assessing the validity of the esophageal balloon technique. Am Rev Respir Dis 1982;126:788-91.[Medline]
    
13. Gilbert R, Auchincloss JH, Peppi D. Relationship of rib cage and abdomen motion to diaphragm function during quiet breathing. Chest 1981;80:607-12.[Abstract/Free Full Text]
    
14. Roussos C, Macklem PT. Diaphragmatic fatigue in man. J Appl Physiol 1977;43:189-97.[Abstract/Free Full Text]
    
15. Barnard PA, Levine S. Critique on application of diaphragmatic time-tension index to spontaneously breathing humans. J Appl Physiol 1986;60:1067-72.[Abstract/Free Full Text]
    
16. Similowski T, Fleury B, Launois S, Cathala MP, Bouche P, Derenne JP. Cervical magnetic stimulation: a new painless method for bilateral phrenic nerve stimulation in conscious humans. J Appl Physiol 1989;67:1311-8.[Abstract/Free Full Text]
    
17. Pham LA, Brun-Buisson C, Legrand P, et al. Diagnosis of pneumonia in mechanically ventilated patients: comparison of a plugged telescoping catheter with the protected specimen brush. Am Rev Respir Dis 1991;143:1055-61.[Medline]
    
18. Le Gall JR, Loirat P, Alperovitch A, et al. A simplified acute physiology score for ICU patients. Crit Care Med 1984;12:975-7.[Medline]
    
19. Fagon JY, Chastre J, Hance AJ, Montravers P, Novara A, Gibert C. Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am J Med 1993;94:281-8.[Medline]
    
20. Taryle DA, Chandler JE, Good JT, Potts DE, Sahn SA. Emergency room intubations: complications and survival. Chest 1979;75:541-3.[Abstract/Free Full Text]
    
21. Pingleton SK. Complication of acute respiratory failure. Am Rev Respir Dis 1988;137:1463-93.[Medline]
    
22. Fedullo AJ, Lerner RM, Gibson J, Shayne DS. Sonographic measurement of diaphragmatic motion after coronary artery bypass surgery. Chest 1992;102:1683-6.[Abstract/Free Full Text]
    
23. Clergue F, Pansard JL. Respiratory changes induced by upper abdominal and cardiac surgery. In: Vincent JL, ed. Update in intensive care medicine. Berlin, Heidelberg, New York: Springer, 1990:166-73.
    
24. Dureuil B, Desmonts JM, Mankikian B, Prokocimer P. Effects of aminophylline infusion on diaphragmatic dysfunction after upper abdominal surgery. Anesthesiology 1985;62:242-6.[Medline]
    
25. Ford GT, Whitelaw WA, Resenal TW, Cruse PJ, Guenter CA. Diaphragmatic function after upper abdominal surgery in humans. Am Rev Respir Dis 1983;127:431-6.[Medline]
    
26. Simonneau G, Vivien A, Sartene R, et al. Diaphragm dysfunction induced by upper abdominal surgery: role of postoperative pain. Am Rev Respir Dis 1983;128:899-903.[Medline]
    
27. Dureuil B, Viires N, Cantineau JP, Aubier M, Desmonts JM. Diaphragmatic contractility after upper abdominal surgery. J Appl Physiol 1986;61:1775-80.[Abstract/Free Full Text]
    
28. Pansard JL, Mankikian B, Bertrand M, Kieffer E, Clergue F, Viars P. Effects of thoracic extradural block on diaphragmatic electrical activity and contractility after upper abdominal surgery. Anesthesiology 1993;78:63-71.[Medline]
    
29. Kohorst WR, Schonfeld SA, Altman M. Bilateral diaphragmatic paralysis following topical cardiac hypothermia. Chest 1984;85:65-8.[Abstract/Free Full Text]
    
30. Serraf A, Planche C, Lacour Gayet F, Bruniaux J, Nottin R, Binet JP. Post cardiac surgery phrenic nerve palsy in pediatric patients. Eur J Cardiothorac Surg 1990;4:421-4.[Abstract]
    
31. Lisboa C, Pare PD, Pertuze J, et al. Inspiratory muscle function in unilateral diaphragmatic paralysis. Am Rev Respir Dis 1986;134:488-92.[Medline]
    
32. Marco JD, Hahn JW, Barner HB. Topical cardiac hypothermia and phrenic nerve injury. Ann Thorac Surg 1977;23:235-7.[Abstract]
    
33. Dureuil B, Viires N, Pariente R, Desmonts JM, Aubier M. Effects of phrenic nerve cooling on diaphragmatic function. J Appl Physiol 1987;63:1763-9.[Abstract/Free Full Text]
    
34. Estenne M, Yernault JC, De Smet JM, De Troyer A. Phrenic nerve and diaphragm function after coronary artery bypass grafting. Thorax 1985;40:293-9.[Abstract/Free Full Text]
    
35. Driver AG, Blackburn BB, Marcuard SP, Austin EH. Bilateral diaphragm paralysis treated with cuirass ventilation. Chest 1987;92:683-5.[Abstract/Free Full Text]
    

This article has been cited by other articles:

				S. J. Durham and J. P. Gold Late Complications of Cardiac Surgery Card. Surg. Adult, January 1, 2008; 3(2008): 535 - 548. [Full Text]

				A. O. Karacalioglu, B. Jata, S. Kilic, N. Arslan, S. Ilgan, and M. A. Ozguven A Physiologic Approach to Decreasing Upward Creep of the Heart During Myocardial Perfusion Imaging J. Nucl. Med. Technol., December 1, 2006; 34(4): 215 - 219. [Abstract] [Full Text] [PDF]

				N. S. G. Toledo, S. K. Kodaira, P. C. B. Massarollo, O. I. Pereira, J. C. Dalmas, G. G. Cerri, and C. A. Buchpiguel Left Hemidiaphragmatic Mobility: Assessment With Ultrasonographic Measurement of the Craniocaudal Displacement of the Splenic Hilum and the Inferior Pole of the Spleen J. Ultrasound Med., January 1, 2006; 25(1): 41 - 49. [Abstract] [Full Text] [PDF]

				E. L'Her, N. Deye, F. Lellouche, S. Taille, A. Demoule, A. Fraticelli, J. Mancebo, and L. Brochard Physiologic Effects of Noninvasive Ventilation during Acute Lung Injury Am. J. Respir. Crit. Care Med., November 1, 2005; 172(9): 1112 - 1118. [Abstract] [Full Text] [PDF]

				A J Moore, J Moxham, and M I Polkey Diaphragm weakness as a cause of breathlessness after anatomically distant surgery Thorax, September 1, 2005; 60(9): 786 - 787. [Abstract] [Full Text] [PDF]

				R. Wynne and M. Botti Postoperative Pulmonary Dysfunction in Adults After Cardiac Surgery With Cardiopulmonary Bypass: Clinical Significance and Implications for Practice Am. J. Crit. Care., September 1, 2004; 13(5): 384 - 393. [Abstract] [Full Text] [PDF]

				R. Salenger, J. S. Gammie, and T. J. Vander Salm Postoperative Care of Cardiac Surgical Patients Card. Surg. Adult, January 1, 2003; 2(2003): 439 - 469. [Full Text]

				ATS/ERS Statement on Respiratory Muscle Testing Am. J. Respir. Crit. Care Med., August 15, 2002; 166(4): 518 - 624. [Full Text] [PDF]

				D A Simansky, M Paley, Y Refaely, and A Yellin Diaphragm plication following phrenic nerve injury: a comparison of paediatric and adult patients Thorax, July 1, 2002; 57(7): 613 - 616. [Abstract] [Full Text] [PDF]

				S. JABER, A. CARLUCCI, M. BOUSSARSAR, R. FODIL, J. PIGEOT, S. MAGGIORE, A. HARF, D. ISABEY, and L. BROCHARD Helium-Oxygen in the Postextubation Period Decreases Inspiratory Effort Am. J. Respir. Crit. Care Med., August 15, 2001; 164(4): 633 - 637. [Abstract] [Full Text] [PDF]

				A. PURRO, L. APPENDINI, A. DE GAETANO, M. GUDJONSDOTTIR, C. F. DONNER, and A. ROSSI Physiologic Determinants of Ventilator Dependence in Long-term Mechanically Ventilated Patients Am. J. Respir. Crit. Care Med., April 1, 2000; 161(4): 1115 - 1123. [Abstract] [Full Text]

				I. E.W. van Onna, R. Metz, L. Jekel, S. R. Woolley, and H. J.C.M. van de Wal Post cardiac surgery phrenic nerve palsy: value of plication and potential for recovery Eur. J. Cardiothorac. Surg., August 1, 1999; 14(2): 179 - 184. [Abstract] [Full Text] [PDF]

				N M Siafakas, I Mitrouska, D Bouros, and D Georgopoulos Surgery and the respiratory muscles Thorax, May 1, 1999; 54(5): 458 - 465. [Full Text]

				S.-i. Takeda, S. Miyoshi, H. Maeda, M. Minami, H.-E. Yoon, H. Tanaka, K. Nakahara, and H. Matsuda Ventilatory muscle recruitment and work of breathing in patients with respiratory failure after thoracic surgery Eur. J. Cardiothorac. Surg., April 1, 1999; 15(4): 449 - 455. [Abstract] [Full Text] [PDF]

				D. J. Nikas and J. A. Elefteriades Reply Ann. Thorac. Surg., July 1, 1998; 66(1): 302 - 302. [Full Text] [PDF]

This Article Abstract 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): Philippe Deleuze Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar