The Effects of Neuromuscular Paralysis on Systemic and Splanchnic Oxygen Utilization in Mechanically Ventilated Patients

The Effects of Neuromuscular Paralysis on Systemic and Splanchnic Oxygen Utilization in Mechanically Ventilated Patients* Paul E. Marik, MBBCh; and David Kaufman, MD, FCCP Objective: To evaluate the effect of neuromuscular paralysis on systemic and splanchnic oxygen utilization in patients in respiratory failure during controlled mechanical ventilation. Setting: A university-affiliated teaching hospital. Intervention: Mechanically ventilated patients, who were undergoing hemodynamic monitoring and who had a gastric intramucosal pH (pHi) of less than 7.35, were studied. Prior to paralysis, the patients were sedated with lorazepam and morphine to standard end points, and the cardiac output and oxygenation were optimized. The patients were then paralyzed with doxacurium and the ventilator rate adjusted to keep the PaC02 at baseline value. The hemodynamic and oxygenation profUe and pHi were determined prior to paralysis and repeated 2 to 2.5 h later. Results: Eight patients were studied; their mean age was 63 ::±:: 8 years and acute physiology and chronic health evaluation II score was 22::±::4. The mean fraction of inspired oxygen, positive end-expiratory pressure, and venous admixture ratio prior to the study was 0.7::±::0.14, 11.8::±::2.4 em H20, and 26::±::9%, respectively. Prior to paralysis, the mean set assist controlled ventilation rate was 15 : ±: 2 breaths/min and the patient rate was 23::±::5 breaths/min. With neuromuscular paralysis,

the cardiac index fell from 4.6 ::±::2.2 to 4.3 : ±: 2.4 Umin!m2 (p=0.1), the oxygen delivery feU from 537::±::129 to 471::±::95 mUminlm2 (p=0.03), and the oxygen consumption and extraction ratio fell from 200::±::77 to 149::±::35 mUminlm 2 (p=0.03) and 36::±::5 to 31::±::10, respectively (p=0.2). The pHi increased from 7.21::±::0.16 to 7.29::±::0.1 (p=0.02). Conclusion: In critically ill patients in respiratory failure, neuromuscular paralysis decreases whole body oxygen consumption and increases pHi. Presumably, by eliminating the work of breathing, there is a redistribution of blood flow from the respiratory muscles to the splanchnic and other nonvital vascular beds. (CHEST 1996; 109:1038-42)

with the widespread use of advanced organ support, patients with acute life-threatening illness rarely die from their presenting disease, but rather from the development of multisystem organ failure.l·2 It has been postulated that this syndrome is a consequence of inadequate oxygen delivery, often exacerbated by a level of tissue oxy~en extraction that fails to satisfY metabolic demands. ·3 Consequently, in the hope of improving outcome, many investigators have recommended increasing oxygen delivmy. 4-6 However, the benefits of increasing global indexes of oxygen delivery and oxygen consumption remain unproved.7·8 In normal subjects, breathing quietly, the oxygen cost of breathing is less than 5% of total oxygen con-

sumption.9.l 0 In critically ill patients, however, this may increase up to 25%. 9,1 1 Intubation and positive pressure mechanical ventilation reduces the work ofbreathing in these patientsY- 11 It is commonly assumed that the work of breathing is minimal during assisted, volume-cycled mechanical ventilation. However, several studies have demonstrated that significant respiratmy muscle activity persists throughout a mechanically assisted breath, and that the work of breathing during assist-control (AC) ventilation is often a largepercentage of the work performed during spontaneous breathing.12-14 Furthermore, experimental data have demonstrated that in acute respiratory failure, blood flow is diverted from the splanchnic and other nonvital vascular beds to the respiratory muscles.l5-17 We postulated that critically ill patients receiving assisted mechanical ventilation may develop gastric mucosal ischemia due to the redistribution of blood away from the splanchnic blood, particularly if global oxygen de-

*From the Department of Critical Care Medicine, St. Vincent Hospital, and University of Massachusetts, Worcester, Mass. Manuscript received June 1, 1995; revision accepted October 23. Reprint requests: Dr. Marik, Department of Crittcal Care, St. Vincent Hospital, 25 Winthrop Street, Worcester, MA, 01604

1038

AC=assist controlled ventilation; CO=cardiac output; PCWP=pulmonary artery wedge pressure; pHi=gastric intramucosal pH

Key words: neuromuscular paralysis; oxygen consumption; respiratory failure; splanchnic oxygenation

Clinical Investigations in Critical Care

Table !-Clinical Characteristics of Patients Patient No./ Age, yr/Sex 1/62/F 2/52/M 3/58/M 4176/F

Disease

Pancreatitis/ ARDS Pneumonia Pneumonia Sepsis syndrome/ ARDS 5/55/F Pulmonary vasculitis 6/61/F Abdominal sepsis/ ARDS 7n8/M Pneumonia/septic shock 8/65/F Abdominal sepsis/ ARDS

Flo2f'PEEP* 0.7/10 0.6/15 0.9/15 0.9/12.5

Inotrope, Dose, pglkglmin Dobutamine, 12 Norepinephrine, 0.12

0.6/12.5

o.5n.5 0.7/10

Norepinephrine, 0.08

0.6/12

Dobutamine, 10

*PEEP=positive end-expiratory pressure (em HzO ); Floz=fractional inspired oxygen concentration.

livery is inadequate. Furthermore, reducing the work of breathing with neuromuscular paralysis may increase splanchnic blood flow and reverse gastric mucosal ischemia. MATERIALS AND METHODS

This study was conducted in the Medical and Surgical ICUs at St. Vincent Hospital, a university-affiliated teaching hospital in Worcester, Mass. Approval to conduct this study was obtained from our Institutional Review Board. During the study period, all patients admitted to our ICUs with the sepsis syndrome 18 and acute respiratory failure, who had a pulmonary artery catheter in situ and required neuromuscular blockade to facilitate mechanical ventilation, were screened for inclusion into this study. Prior to inclusion, the patients were volume resuscitated to achieve a pulmonary capillary wedge pressure (PCWP) of between 10 and 14 mm Hg and an inotropic agent was added if the stroke volume waslow. No attempt was made to achieve supranormal values of oxygen delivery.l 9 ·20 Patients in uncompensated septic shock were excluded from the study. Prior to paralysis, an infusion of lorazepam and morphine sulfate was used for sedation and titrated until the patients were asleep with only a sluggish response to pain (Ramsey Score V). 21 Regional tissue oxygenation was monitored using a nasogastric tonometer. Patients who met all of the above inclusion criteria and who had a gastric intramucosal ph (pHi) below 7.35 were entered into the study. A pHi of less than 7.35 has been reported to be a marker of splanchnic ischemia and to be associated with an increased mortality. 22· 24 Once hemodynamic stability was achieved and the end points of sedation had been met, the patients were paralyzed using doxacurium. A 0.03-mglkg loading dose of doxacurium was given followed by an infusion at a rate of 0.03 mglkglh. The degree of neuromuscular blockade was monitored using a p eripheral nerve stimulator. The doxacurium was titrated to achieve two to three twitches following a train-of-four stimulus. All patients were receiving mechanical ventilation with the same type of ventilator (Puritan Bennett 7200). The ventilator rate was adjusted to keep the PaC02 at baseline value. During the study period (of 2 h ), the patients were neither turned nor suctioned, nor did they receive physical therapy or any other intervention that could alter their oxygen d elivery or consumption. Furthermore, during the study period, the infusion rates of the sedative drugs and inotropic agents were held constant. Patients whose temperature changed b ymore than 0.5°C during the study period were excluded from analysis. 25

Data Collection

A data set was collected at baseline and repeated 2 to 2.5 h after neuromuscular paralysis. This time interval was chosen as it represents the shortest time in which a therapeutic intervention is likely to be reflected by a change in pHi. 24 •26 •27 Each data set included hemoglobin, arterial lactate concentration, core body temperature, a h emodynamic and oxygenation profile, and a pHi measurement. Hemodynamic data included pulmonary arterial pressure, heart rate, central venous pressure, mean arterial pressure, PCWP, and cardiac output (CO). CO was determined by the thermodilution method. Five measurements were made using 10 mL of injectate at room temperature and a monitoring system (Merlin 55; HewlettPackard; Andover, Mass). If the CO measurements differed by more than 10%, the set of CO determinations were repeated. Arterial and mixed v enous blood were sampled simultaneously with the CO measurements for determination of blood gas values (BGM Blood Gas Manager; Instrumentation Laboratory; Lexington, Mass) and hemoglobin saturation (482 Co-Oximeter; Instrumentation Laboratory). Systemic oxygen delivery and consumption, pulmonary venous admixture, and oxygen extraction ratio were calculated using standard formulas. 28 ·29 All flow and volume measurements were indexed to body surface area. Hemoglobin concentration was measured using an analyzer (Coulter Electronics; Hialeah, Fla). Arterial blood specimens for lactate determination were analyzed using an enzymatic method (2300 Stat Analyzer; YSI Inc; YellowSpring, Ohio). The normal range of plasma lactate for our laboratory is 1.2 to 2.2 mmol!L. The acute physiology and chronic health evaluation II score was used as an index of disease severityao GI oxygenation was indirectly assessed by measuring the pHi using a nasogastric tonometer (Tonometries Inc; Hopkinton, Mass ). The balloon of the tonometer was filled with 2.5 mL of normal saline solution and allowed to equilibrate (starting 90 min prior to each data set) with the gastric mucosa. 24 Once equilibrated, the balloon was aspirated and the carbon dioxide tension of the fluid determined (using the BGM Blood Gas Manager). The pHi was then calculated by substituting the partial pressure of carbon dioxide and arterial bicarbonate concentration into the Henderson-Hasselbalch equation. 24 ·31 To improve the accuracy of the measurement of the pHi, all patients received IV cimetidine and enteral tube feedings were withheld. 31 A pHi of less than 7.35 was used as indirect evidence for the presence of splanchnic ischemia. 22· 24 At the end of the data collection, summary statistics were compiled to allow a description of the study population. Data are presented as means::+:SD, ranges, or proportions, as appropriate. The x2 test was used for the analysis of categorical data. Continuous variables were analyzed by Student's t test. RESULTS

Eight patients were studied: three men and five women. The mean age of the patients was 63±8 years. The patients' clinical characteristics are listed in Table 1. The patients were all seriously ill, with a mean acute physiology and chronic health evaluation II score of 22±4. All patients had significant lung injury requiring a mean fraction of inspired oxygen of 0.7±0.14 and positive end-expiratory pressure of 11.8±2.4 em HzO to maintain adequate oxygenation (arterial oxygen saturation >90%). The pulmonary venous admixture ratio at the time of the study was 26±9%. Prior to paralysis, the mean set AC rate was set at 15±2 breaths/min, with the mean patient rate being 23±5 breaths/min. During the study period, there was no significant change in the patients' PCWP and core CHEST I 109 I 4 I APRIL, 1996

1039

temperature. Following neuromuscular paralysis, oxygen delivery and consumption fell in all patients, with a highly significant increase in the pHi. The hemodynamic, oxygenation, and pHi changes with paralysis are presented in Figure 1 and Table 2.

10

DISCUSSION

4

In this study, we have demonstrated that neuromuscular paralysis will decrease total body oxygen consumption and increase pHi in clitically ill patients with evidence of splanchnic ischemia, duling assisted, volume-cycled mechanical ventilation. Presumably, by eliminating the work of breathing duling AC ventilation, there is a redistlibution of blood from the respiratmy muscles to the splanchnic and other nonvital vascular beds. The results of this study are supported by both clinical and expelimental data. Madger and coworkers17 evaluated respiratory muscle blood flow after oleic acid-induced acute lung injury in dogs. In this model, respiratory muscle blood flow increased almost threefold, whereas splanchnic and renal blood flow decreased substantially despite the maintenance of a normal CO and BP. Hussain and Roussos 16 studied respiratory muscle blood flow after the induction of endotoxemia in spontaneously breathing dogs. They observed that significant blood flow diversion to the respiratory muscles occurred, whereas blood flow to the splanchnic bed decreased significantly. Vires and colleagues 15 demonstrated that respiratory muscle blood flow increased by up to 350% in spontaneously breathing dogs duling respiratory failure, and this was accompanied by a significant fall in hepatic, renal, and splanchnic blood flow. Furthermore, these changes were essentially reversed with the institution of mechanical ventilation. In a recent study, Manthous and coworkers32 investigated the effects of valious modes of ventilation on oxygen consumption in eight patients in respiratory failure. These authors demonstrated that the addition

Cardiac Index llmin/M2

8

6

2 0 Baseline 800

Post paralysis

Oxygen delivery (mllmin/M2)

700 600 500 400 300 200~----------------.---------------~

Baseline

Post paralysis

Oxygen consumption (mllmin/M2)

400~----------------------------------~

300

200

100 7.5

Baseline

Post paralysis

Baseline

Post paralysis

pHi

7.4 7.3 7.2

Table 2-Hemodynamic, Oxygenation, and pHi Changes With Neuromuscular Paralysis Cardiac index, Umin!m 2 Heart rate PCWP, mm Hg MAP,* mm Hg 02 extraction ratio, % 02 delivery, mUmin!m 2 0 2 consumption, mUmin!m 2 Te mpe rature, °C PaC0 2, mm Hg Lactate, mmoVL pHi *MAP=mean arte1ial pressure.

1040

Baseline

Postparalysis

p V alue

4.6::+::2.2 106::+::24 13::+::4 75::+::12 36::+::5 537::+::129 200::+::77 37.4::+::0.3 39.1::+::10.1 2.4::+::1.7 7.21::+::0.16

4.3::'::2.4 105::+::30 13::+::3 74::+::9 31::'::10 471::+::95 149::+::35 37.6::'::0.4 40.7::+::10.1 2.0::+::0.8 7.29::+::0.1

0.1 0.7 1.0 0.9 0.2 0.03 0.03 1.0 0.5 0.6 0.02

7.1 7 6.9 FIGURE 1. Changes in cardiac index, oxygen delivery, oxygen consumption, and pHi with neuromuscular paralysis.

of muscle relaxation to AC ventilation decreased oxygen consumption in six of the eight patients studied, with an 8% fall in the mean oxygen consumption. These authors, however, did not concomitantly measure changes in hemodynamics or indexes of tissue Clinical Investigations in Critical Care

oxygenation. Oxygen consumption fell, on average, by 25% in our study. The larger fall in oxygen consumption in our study presumably reflects the selection of a sicker group of patients with severe respiratory compromise in whom the work of breathinf had profoundly increased. Pohlman and colleagues3 measured the effect of sedation with propofol and morphine sulfate on oxygen consumption in critically ill patients undergoing mechanical ventilation. In this study, the mean oxygen consumption fell from 306 mUmin to 271 mUmin (p<0.05) with sedation. However, the addition of a paralytic agent in five of their patients did not result in a further r eduction in oxygen consumption. One can assume that all spontaneous respiratory effects were inhibited by the sedative agents used by these authors, and it is, therefore, not surprising that oxygen consumption did not fall further with neuromuscular paralysis. The end point of sedation in our study was not the inhibition of all respiratory efforts but a level of narcosis adequate for patient comfort. It is probable that using the same end point of sedation, the results of our study would have duplicated those of Pohlman and colleagues. 33 However, we have observed that in patients with severe respiratory compromise, it is often not possible to eliminate all respiratory efforts with the use of hypnotic agents alone, without producing hemodynamic embarrassment. Mohsenifar and colleagues34 measured pHi in a group of patients being weaned from mechanical ventilation. These authors demonstrated that the pHi fell significantly in patients who failed a weaning trial. These authors postulate that the increased work of breathing in this group of patients diverted blood from the splanchnic bed to the respiratory muscles, resulting in gastric mucosal ischemia. This study provides further supportive evidence that increasing the workof breathing in critically ill patients with decreased compensatory reserves may result in regional tissue dysoxia. A limitation of our study is that oxygen consumption was calculated rather than measured directly.35 However, we believe that it is unlikely that this factor significantly influenced the results of our study. Furthermore, we surmise that the fall in cardiac index and oxygen d elivery were a consequence of the fall in systemic oxygen requirements following the elimination of the excessive work of breathing. It could also be argued that because of the delay in repeating the data set (2 to 2.5 h) , the observed changes may not necessarily represent those due to the institution of neuromuscular paralysis. When measuring pHi by balloon tonometry, an equilibration period of at least 60 min is required. 24 Furthermore, once a change in therapy is made, the effect of this change on tissue oxygen flux and tissue pH is not instantaneous, but rather there is an equilibration phase before a new steady state is

achieved (probably >1 h). As a consequence, most researchers have recommended an interval of not less than 2 h , after an intervention, before repeating the pHi measurement. 22-26 However, it is important in the study design to ensure that no additional changes are made during the study period that could alter the patient's oxygen delivery or consumption. In conclusion, our study suggests that reducing the work of breathing during AC ventilation by neuromuscular paralysis may reduce whole body oxygen consumption and redistribute blood flow to the splanchnic and other nonvital vascular beds. Patients with evidence of an "oxygen debt" should first be adequately resuscitated and relaxed on the ventilator using sedatives to the limits of hemodynamic stability. If lactic acidemia and/or a low pHi persist despite these primary therapies, it may be justifiable to attempt a therapeutic trial of muscle relaxation. If the lactic acidemia and/or pHi do not improve, treatment with muscle relaxants should not be continued to avoid complications associated with their use.36•37 When muscle relaxants are used, neuromuscular monitoring should be employed to minimize the doses and treatment with them should be discontinued as soon as signs of oxygen debt abate. REFERENCES

1 Knaus WA, Draper EA, Wagner DP, et al. Prognosis in acute organ-system failure. Ann Surg 1985; 202:685-93 2 Marshall JC, Meakins JL. Multiorgan failure. In: Rubenstein E, Federman DD, eds. New York: Scientific America Inc, 1989; 1-17 3 Fry DE. Multiple system organ failure. SurgCiin North Am 1988; 68:107-22 4 Edwards JD, Brown GCS, Nightingale P, et al. Use of survivors cardiorespiratory values as therapeutic goals in septic s hock. Crit Care Med 1989; 17:1098-1103 5 Bland RD, Shoemaker WC, Abral1am E , etal. Hemodynamic and oxygen transport patterns in surviving and non-surviving postoperative patients. Crit Care Med 1985; 13:85-90 6 Shoemaker W, Appel P, Kram H, etal. Prospective trial of supranormal values of survivors as therapeutic goals in high risk surgical patients. Chest 1988; 94: 1176-86 7 W egJG. Oxygen transport in adult respiratory distress syndrome and other acute circulatory problems: relationship of oxygen delively and oxygen consumption. Crit Care Med 1991; 19:650-57 8 Parillo JE , Parker MM , Natanson C, etal. Septic shock in humans: advances in understanding of pathogenesis, cardiovascular dysfunction and therapy. Ann Intern Med 1990; 113:227-42 9 Viale J, A nnat G, Bouffard Y, et al. Oxygen cost of breathing in postoperative patients: pressure support venti.lation vs continuous positive airway pressure. Chest 1988; 93:506-09 10 FieldS, Kelly SM, Mackiem PT. The oxygen cost of breathing in patients with cardiopulmonary disease. Am Rev Respir Dis 1982; 126:9-13 11 Annat GJ, Viale JP, D ereymez CP, eta!. Oxygen cost of breathing and diaphragmatic pressure-time index: measurement in patients with COPD during weaning with pressure support ventilation. Chest 1990; 98:411-14 12 Marini JJ, Rodriguez M, Lamb V. The inspiratory workioad of patients-initiated mechanicai venti.lation. Am Rev Respir Dis 1986; 134:902-09 CHEST I 109 I 4 I APRIL, 1996

1041

13 Marini JJ, Smith TC, Lamb V. External work output and force generation during svnchronized intennittent mechanical ventilation: effect of machine assistance on breathing effort. Am Rev Respir Dis 1988; 138:1169-79 14 Marini JJ, Capps JS, Culver BH. The inspiratory work of breathing during assisted mechanical ventilation. Chest 1985; 87:612-18 15 Vires N, Sillye G, Au bier M, et al. Regional blood flow distribution in dog during induced hypotension and low cardiac output: spontaneous breathing versus artificial ventilation. J Clin Invest 1983; 72:935-47 16 Hussain SN, Roussos C. Distribution of respiratory muscle and organ blood flow during endotoxemic shock in dogs. J Appl Physiol 1985; 59:1802-08 17 Madger S, Erian R, Roussos C. Respiratory muscle blood flow in oleic acid-induced pulmonary edema. J Appl Physiol 1986; 6:1849-56 18 Society of Critical Care Medicine Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992; 20:864-74 19 Edwards JD, Brown GCS, Nightingale P, et al. Use of survivors cardiorespiratory values as therapeutic goals in septic shock. Crit Care Med 1989; 17:1098-1103 20 Hayes MA, Timmins AC, Yau EHS, et al. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Eng! J M ed 1994; 330:1717-22 21 Ramsey MAE, Savege TM, Simpson BJR, et al. Controlled sedation with alphaxolone-alphadalone. BMJ 1974; 2:656-59 22 Doglio GR, Pusajo JF, Egurrola MA, et al. Gastric mucosal pH as a prognostic index of mortality in critically ill patients. Crit Care Med 1991; 19:1037-40 23 Gutierrez G, Palizas F, Doglio G, et al. Gastric intramucosal pH as a therapeutic index of tissue oxygenation in critically ill patients. Lancet 1992; 339:195-99 24 Gutierrez G, Brown SD. Gastric tonometry: a new monitoring modality in the intensive care unit. J Intensive Care Med 1995; 10:34-44

1042

25 Mathous CA, Hall JB, Olson D, et al. Effect of cooling on oxygen consumption in febrile critically ill patients. Am J R espir Crit Care Med 1995; 151:10-4 26 Marik PE, Mohedin M. The contrasting effects of dopamine and norepinephrine on systemic and splanchnic oxygen utilization in hyperdynamic sepsis. JAMA 1994; 272:1354-57 27 Marik PE, Sibbald WJ. Effect of stored-blood transfusion on oxygen delivery in patients with sepsis. JAMA 1993; 269:3024-29 28 Groeneveld ABJ, Bronsveld W, Thijs LG. He modynamic determinants of mortality in human septic shock. Surgery 1986; 99:140-43 for t 29 Ganz W, Donoso R, Marcus HS, et al. A new echnique measurement of cardiac output by thermodilution in man. Am J Cardiol 1971; 27:392-96 30 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13:818-29 31 Heard SO, H elsmoortel CM, Kent JC, et a!. Gastric tonometry in healthy volunteers: effects of ranitidine on calculated intramural pH. Crit Care Med 1991; 19:271-74 32 Manthous CA, Hall JB, Kushner R, et al. The effect of mechanical ventilation on oxygen consumption in critically ill patients. Am Rev R espir Crit Care Med 1995; 151:210-14 33 Pohlman A, O'Connor M, Olson D, et a!. Sedation with propofollowers V02 in critically ill patients. Am Rev Respir Crit Care Med 1995; 151:A325 34 Mohsenifar Z, Hay A, Hay J, eta!. Gastric intramucosal pH as a predictor of success or failure in weaning patients from mechanical ventilation. Ann Intern Med 1993; 119:794-98 35 Bartlett RH, Dechert RE. Oxygen kinetics: pitfalls in clinical research. J Crit Care 1990; 5:77-80 36 Giosta E, Magistris MR, Pizzolato G, eta!. Neuromuscular disorder in intensive care unit patients treated with pancuronium bromide: occurrence in a cluster group of seven patients and two sporadic cases, with electrophysiologic and histologic examination. Chest 1994; 106:210-20 37 Hansen-Flashen J ,Cowan J, Raps EC. Neuromuscular blockade in the intensive care unit: more than we bargained for. Am Rev Respir Dis 1993; 147:234-36

Clinical Investigations in Critical Care