Postoperative Residual Curarisation (PORC): Incidence, Aetiology and Associated Morbidity
G. D. Shorten FFARCS(I) FRCA.
How Should Postoperative Residual Curarisation Be Detected?
If the neuromuscular blocking effects of a nondepolarising muscle relaxant, which has been administered intraoperatively, persist postoperatively to a degree that neuromuscular function is inadequate, then postoperative residual curarisation is said to be present. Of course, this definition depends on the criteria of "adequate" neuromuscular function. Suggested criteria include those based on simple clinical tests and those based on commonly used techniques for monitoring of neuromuscular function. Clinical indicators of adequate neuromuscular function include the abilities to sustain a headlift for five seconds, to sustain a firm hand-grip or tongue protrusion, to keep the eyes open for five seconds, to cough effectively and the absence of double vision.
These tests must be interpreted in the light of the well-recognised differential sensitivities of different skeletal muscles to the effects of muscle relaxants. Since the most important neuromuscular functions are pulmonary ventilation, airway protection and maintenance of airway patency, ideally recovery in these muscles should be measured directly. A reasonable and practicable alternative is to assess muscle groups in which recovery is consistently preceded by that in the diaphragm and muscles of the upper airway.
Voluntary hand-grip force can be quantified and used as a measure of postoperative residual muscle weakness.1,2 This measure is very sensitive and may be substantially decreased, even when a five-second headlift can be sustained.2 Corresponding to a degree of skeletal muscle weakness at which maximum inspiratory pressure (MIP) and vital capacity (VC) are decreased to 40% and 50% of control values, handgrip strength is reduced to 6% of its control value3 and is undetectable when MIP is -20 cm H20.4 The major disadvantage of the use of hand-grip force for this purpose is that it is dependent on all the factors that govern patient co-operation (including residual anaesthetic effects) and thus is non-specific. It is diminished following the use of inhalational anaesthesia alone to 77% of a pre-anaesthetic value.1
The MIP consistent with adequate minute ventilation is - 25 cm H20.5,6 In healthy adult volunteers, when MIP is -39 cm H20 or greater (i.e. less negative), airway obstruction may occur, and when MIP > -43 cm H20, difficulty in swallowing is encountered. 4 Thus, in the presence of a partial neuromuscular blockade, adequate minute ventilation and inadequate airway protection may coexist.
If a patient can sustain a five-second headlift, it is likely that adequate airway protection is present .4 Although sustained headlift is not as reliable an indicator of recovery from tubocurarine-induced neuromuscular blockade as sustained contraction in response to a 30 Hz tetanus,7 the latter technique is too painful to be used in awake patients. Of the clinical tests available, inability to sustain a five-second headlift is the most useful indicator of residual muscle weakness.8
As early as 1962, it was recognised that clinical judgement alone might not be adequate in assessing ventilatory function following anaesthesia and the use of muscle relaxants. 5 In 1989, Miller explained the logic of using the response to peripheral nerve stimulation to estimate residual neuromuscular blockade as follows:9 "First, the diaphragm recovers from the effects of non-depolarizing neuromuscular blocking drugs more rapidly than does the adductor pollicis and, second, at a train-of-four ratio of 0.7, vital capacity returns to normal (15-20 ml/kg) and maximum inspiratory and expiratory force are only slightly depressed (-20 to -25 cm H20)."
While a train-of-four (TOF) ratio of 0.7 or greater is commonly accepted to indicate adequate recovery from neuromuscular blockade, it is worth examining the evidence on which this assumption is based as well as the criticism that has been levelled at its use. Good correlation exists between the response to TOF stimulation and clinical recovery of neuromuscular function in healthy anaesthetised patients,10 non-medicated conscious volunteers,11 and patients of ASA classes III and IV after major surgery.12 In the healthy patients, clinical tests indicated adequate recovery was present when the TOF ratio was at least 0.75.10 In eight adult volunteers, when the TOF ratio was 60% or greater, vital capacity and peak expiratory flow rate were greater than 90% of control values and maximum inspiratory force was less (better) than -70 cm H20.10 At a TOF ratio of approximately 0.7, ten ASA III or IV patients recovering from major surgery were able to sustain eyeopening, tongue protrusion and hand-grip; nine of the ten could sustain a five-second headlift.12 During recovery from neuromuscular blockade induced by pancuronium, tubocurarine or metocurine, when TOF ratio is 0.7 or greater, a sustained response to 50 Hz tetanus for five seconds is present.13 Because the degree of tension developed by a 50 Hz tetanus is equal to that which occurs with a maximum voluntary effort,14 it is concluded that a TOF ratio of 0.7 or greater reliably indicates adequate clinical recovery.
The total number of patients/voluteers in these studies is small and the muscle relaxants employed (pancuronium, tubocurarine, metocurine) are long-acting and now much less commonly used. A "true" TOF ratio of greater than 0.7 may be difficult to identify without mechanomyographic (MMG) or electromyographic (EMG) monitoring equipment. Tactile evaluation of a TOF ratio between 0.41 and 0.7 has been shown to be associated with a high percentage of failure to detect fade.15 Another potential source of error in the interpretation of the response to TOF stimulation lies in the variability of the peripheral nerve stimulation units used: at higher load impedance, reduction in current output occurs to a variable extent and may result in less than supramaximal stimulation.16
It has been suggested that following antagonism of neuromuscular blockade induced by the intermediateacting atracurium, a TOF ratio of 0.5 indicates clinically adequate recovery of muscle function.17 Bevan et al.18 found a TOF ratio of 0.7 or greater to be an insensitive test of residual weakness as defined by clinical tests such as sustained headlift and recommended that a TOF ratio >0.9 should be taken to indicate adequate recovery. When electromyography is used, a TOF ratio of 0.9 is necessary before tests of respiratory function return to normal.19 The disparity in these alternative "threshold" values for the TOF ratio illustrates some of the problems in studying PORC. In each of the three studies quoted, the association between precise measurement of the response to peripheral nerve stimulation and the performance of clinical tests by patients recovering from anaesthesia are examined. No attempt is made to quantify the residual anaesthetic effects on patient comprehension or co-operation. How does one distinguish between residual anaesthetic effects and residual neuromuscular weakness when a patient will open his eyes to command but will not sustain a headlift for five seconds? For a patient who has undergone abdominal surgery, headlifting is painful and may not be sustained for this reason. Of the 150 patients studied by Bevan et al.,18 37 were insufficiently awake to co-operate with clinical testing. It is not clear how this assessment was made and, in particular, how it was distinguished from postoperative neuromuscular weakness.
One solution to this problem is to study unsedated volunteers. To date, volunteer studies have been carried out by Ali et al.,11 which supports a safe threshold TOF ratio of 0.7, and by PavIin,4 in which TOF ratios were not measured (clinical tests were performed at various levels of maximum inspiratory pressure). Electromyographic and mechanomyographic measurements of the response to TOF stimulation are not interchangeable.20-22 During vecuronium infusion and during alcuronium-induced blockade (but not during tubocurarine-induced blockade) the EMG single twitch response was consistently greater and thus less sensitive than the simultaneous MMG response.20,21 Kopman22 demonstrated that when the EMG TOF ratio measured at the hypothenar muscles is 0.7-0.85, simultaneously recorded MMG TOF ratio at the adductor pollicis may be less than 0.6. While measuring post-tetanic potentiation and tetanic fade during tubocurarine-induced blockade, Epstein and Epstein23 found EMG to be consistently more sensitive than MMG. The evidence presented by Jones et al.,13 in support of the suggestion that a TOF ratio of 0.5 or greater is compatible with safe reversal of atracurium-induced block, is incomplete. All TOF ratio measurements were made while the patients were anaesthetised with nitrous oxide/oxygen and 1% enflurane. No attempt was made to simultaneously measure ventilatory function. The time between the final measurement of the TOF ratio and application of standard tests of adequate neuromuscular function in the recovery area, such as sustained headlift, is not documented. Further increase in the TOF ratio is likely to have occurred during this interval.
Thus, in order to interpret a TOF ratio, it is necessary to know the site of nerve stimulation, the method of quantifying the muscular response (i.e. electromyography or mechanomyography) and possibly the muscle relaxant used. The available evidence suggests that, with any of the currently used muscle relaxants, a mechanomyographically measured TOF ratio of 0.7 or greater following stimulation of the ulnar nerve is a reliable indicator of adequate clinical recovery.
A "true" TOF ratio of greater than 0.7 may be difficult to identify without special monitoring equipment. Tactile evaluation of a TOF ratio between 0.41 and 0.7 has been shown to be associated with a high percentage of failure to detect fade.14 Alternative patterns of peripheral nerve stimulation may be useful for the detection of residual muscle relaxant effects. The skeletal muscle response to indirect tetanic stimulation is a more sensitive monitor of neuromuscular blockade than the single twitch contraction.24 When TOF response recovers to its control level, 25-30% of postsynaptic receptors are unoccupied compared with 50% unoccupied receptors when a five-second 100 Hz tetanus returns to contro1.25 This suggests that the response to 100 Hz tetanic stimulation is a more sensitive indicator of receptor block than that to TOF stimulation, but is too painful for use in awake or awakening patients. Clinical assessment of the response to 5-second 50 Hz tetanic stimulation of the ulnar nerve is inaccurate except when the tetanic fade is less than 0.3 and thus is not useful in identifying moderate residual muscle relaxant effect.19 Finally, the use of tetanic stimulation alters the response to subsequent peripheral nerve stimulation, increasing single twitch response, TOF and Double Burst ratios26 (see below) as well as increasing the speed of recovery from bolus doses of atracurium and vecuronium.27
Double burst stimulation (DBS) is another pattern of peripheral nerve stimulation which has been described for identification of residual neuromuscular block.28 It comprises two "bursts" of stimulation, each one made up of three impulses at 50 Hz (i.e. at 20 msec intervals) separated by 750 msec. This timing results in a response consisting of two discrete contractions of short duration. Of the "double burst" options studied, this was found to be the most sensitive and least painful. The authors suggest that when TOF ratio is > 0.4, more instances of fade are detectable with DBS than with TOF stimulation. The same investigators subsequently found manual evaluation of the response to DBS to be superior to that of TOF stimulation in assessment of residual neuromuscular blockade.29 Using a range of stimulating currents, Brull et al. demonstrated that visual detection of fade was more sensitive following DBS than following TOF stimulation.30 To date, DBS has not become widely used in clinical practice. The reasons for this may be:
In 1976, Kopman32 and Pue33 demonstrated that pregelled surface electrodes can be safely used in place of transcutaneous needle electrodes for stimulation of the uInar nerve. In order to achieve the supramaximal currents considered necessary for evaluation of neuromuscular function,34 higher currents are necessary with the cutaneous electrodes. A wide range of currents is necessary to produce supramaximal stimulation of the ulnar nerve,35 ranging from 15-60 mA in one study of 38 patients. The current necessary to produce supramaximal stimulation of the ulnar nerve is associated with more patient discomfort than lower currents. 31 Brull et al.36 have demonstrated that TOF stimulation with 20 and 30 millampere (mA) currents consistently provided TOF ratios comparable with those obtained using supramaximal (50 mA) currents. If all four twitches are present at lower current stimulation, the difference between TOF ratios obtained in response to 20, 30 and 50 mA is less than 2%.36 Submaximal current TOF stimulation appears to solve the problem of patient discomfort during peripheral nerve stimulation while testing for residual curarisation. Its clinical applicability is limited by a number of factors, namely:
In summary, it is likely that a patient who is capable of sustaining a five-second headlift is also capable of airway protection and adequate ventilation. Currently, the response to TOF stimulation has several perioperative uses, including assessment of residual curarisation both prior to, and after, extubation. When the mechanical response is assessed, a TOF ratio of at least 0.7 reliably indicates adequate neuromuscular recovery. Residual curarisation can be present without clinically detectable TOF fade. The newer DBS pattern of stimulation is a sensitive clinical monitor of intra- and postoperative neuromuscular function. Lower, submaximal currents for both TOF stimulation and DBS may have a role in the postoperative assessment of patients recovering from neuromuscular blockade.
Incidence And Aetiology
Viby-Mogensen et al.37 studied the incidence of postoperative residual curarisation (PORC) in the following way. On three separate, randomly chosen days, all patients in each of three hospitals who received a nondepolarising muscle relaxant were assessed on arrival in the recovery room. No attempt was made to standardise the anaesthetic technique, nor were the anaesthetists involved aware that their patients would be monitored postoperatively. In no case was a peripheral nerve stimulator used intra-operatively. Thirty (42%) of the 72 patients studied were found to be inadequately recovered as evidenced by a TOF ratio of less than 0.7. Of those patients who were awake enough to co-operate, 24% were unable to sustain a headlift for five seconds. The authors concluded that PORC was a significant problem, that anaesthetists tend to use high doses of muscle relaxants and that neostigmine 2.5 mg is an inadequate dose in some patients. Subsequent studies by Lennmarken and Lofstrom38 and Beemer and Rozental39 supported these findings and confirmed that, in widely disparate settings, PORC occurred with unacceptable frequency.
As newer, intermediate-acting, nondepolarising muscle relaxants (atracurium and vecuronium) became widely used, it appeared likely that the incidence of PORC would diminish. Andersen et al.40 compared 30 pairs of patients, each pair similar in terms of sex and type of operation. One of each pair received pancuronium and the other atracurium. Postoperatively, none of the atracurium group had a TOF ratio less than 0.7 and six of the thirty patients who received pancuronium had TOF ratios less than 0.7. Ten of the pancuronium group (compared with only two of the atracurium group) were clinically judged to be inadequately reversed. Bevan et al.41 studied the association between PORC and a number of variables (choice of relaxant, whether neuromuscular monitoring was employed, and the reversal agent chosen). The mean TOF ratio in patients who received pancuronium (0.74) was significantly less than that in either the atracurium (0.93) or the vecuronium (0.99) groups. However, since the patients who received pancuronium underwent longer procedures, this study does not rule out the possibility that the duration of the procedure and thus of the neuromuscular blockade is the factor responsible for the higher incidence of PORC. A
peripheral nerve stimulator had been used in a greater proportion of patients with a TOF ratio <0.7 than in patients with a TOF ratio >0.7 (92% and 71%, respectively). The choice of anticholinesterase did not affect the frequency of PORC. O'Connor and Russell2 studied postoperative hand-grip strength using a dynanometer in patients who had received pancuronium and atracurium. When the patient was able to sustain headlift for five seconds, the mean grip force was 60% of the control value in the atracurium group and 44% in the pancuronium group. As has been pointed out above, residual anaesthetic effects may have influenced these findings. The past decade has provided substantial evidence 42- 45 that, in terms of residual neuromuscular blockade, the newer intermediate-acting muscle relaxants are safer than the older long-acting agents. However, the use of these newer muscle relaxants has not eliminated the problem of PORC.
Many of the preceding quoted articles have included a recommendation that perioperative neuromuscular monitoring be used to reduce the frequency of PORC. A number of investigators have provided data which does not support this recommendation. 40,45,46 The limitations of the study by Pedersen et al .46 have been detailed elsewhere.47 Recently, the incidence of PORC (as defined by a TOF ratio of 70% or less) following the intra-operative use of pancuronium was found to be significantly less (47% vs 15%) when tactile TOF monitoring was employed. 48 The routine intraoperative use of TOF monitoring is to be recommended as a reliable and non-invasive guide to muscle-relaxant administration and a means of reducing the incidence of PORC.
The agents and techniques used to reverse neuromuscular blockade have been comprehensively reviewed .49-51 Evidence of spontaneous recovery should be present before reversal is attempted.52 The degree of recovery necessary is dependent on the mode of administration and choice of muscle relaxant, on the anaesthetic technique used, on the choice and dose of reversal agents used and the degree of neuromuscular function considered adequate. A lower dose of neostigmine is required to antagonise the neuromuscular blockade induced by a bolus dose of vecuronium than a similar degree of block induced by a bolus of the longer-acting pancuronium. 53 When these muscle relaxants are administered by continuous infusion, there is no difference in neostigmine requirement.54 Pyridostigrnine and edrophonium are more effective antagonists of pancuronium- than tubocurarine-induced neuromuscular blockade at several different levels of spontaneous recovery.55 When four responses are present to TOF stimulation, edrophonium (0.3 mg/kg) will reliably reverse (i.e. produce a TOF ratio > 0.7) an atracurium-induced block in ten minutes; when vecuronium is used an edrophonium dose of 0.75 mg/kg is required to produce the same effect.56 When administered at 10% single-twitch recovery from pancuronium-induced neuromuscular blockade, the reversing potency of neostigmine is significantly lower during enflurane than during halothane or nitrous oxide/fentanyl anaesthesia.57 The edrophonium dose-TOF recovery relationship is flatter than that of neostigmine or pyridostigmine.55 Thus, if adequate reversal is determined by achievement of a TOF ratio of 0.7 or greater, edrophonium is less likely to be effective than if single twitch recovery alone were measured. The effectiveness achieved with a reversal agent is inversely related to the degree of neuromuscular block at which it is administered.58-62 Furthermore, the potency ratio of neostigmine and edrophonium changes markedly at different levels of blockade, being 16.6 and 35.3 at 90% and 99% atracurium-induced neuromuscular blockade.58 Deeper levels of block may be reversed by increasing the dose of anticholinesterase used. The upper limit of anticholinesterase dosage is limited by a "ceiling effect" at high doses and deep levels of block 63 and potential adverse effects on neuromuscular function when high or repeated doses are used . 64
The factors enumerated above influence the degree of neuromuscular blockade at which reversal should be attempted. With respect to safe clinical practice, certain conclusions can be drawn:
"There are, in fact, no outcome data assessing the role of residual paralysis, residual anaesthetic or surgery or some combination of the three when the occasional case of airway obstruction or inability to dispose of vomitus or orpharyngeal secretions occurs in the recovery room" (Miller RD, 1989).9 Any attempt to rectify this must contend with three major difficulties. Firstly, since poor patient outcome is rare, any study of patient morbidity or mortality will require that a large number of patients be studied. Secondly, it is unlikely that most postoperative complications can be attributed to a single cause. Thus, if a patient in whom PORC had been demonstrated were to aspirate gastric contents into the lungs in the immediate postoperative period, it would be misleading to conclude that residual weakness alone caused the aspiration. Thirdly, an ethical dilemma exists if one is to study PORC. If postoperative neuromuscular function is judged to be inadequate, the investigator is obliged to intervene and correct this, and in so doing has altered the postoperative course which would otherwise have taken place.
In "A study of deaths associated with anaesthesia and surgery", which included 600,000 anaesthetics administered over five years (1948-1952), Beecher and Todd67 claimed that the use of curare was associated with a sixfold increase in the perioperative death rate. This study has been heavily criticised because of indisputable and major design faults.68 In particular, the sixfold difference in mortality occurred between two groups of patients in whom comparison is not valid, because relevant variables such as the site and duration of the surgical procedures have not been taken into account. In 1949, a committee was appointed by the Association of Anaesthetists of Great Britain and Ireland to collect and analyse clinical data on anaesthesia-related deaths. One early report69 from the committee described nineteen cases in which, "following the use of a relaxant, respiratory insufficiency (other than obstruction) was present postoperatively". In 1956,70 one thousand such deaths were reviewed and, of 589 in which anaesthesia was implicated, 115 occurred within 30 minutes of the end of the operation. Thirty-three were mainly due to postoperative respiratory obstruction, of which 23 were attributed to "pharyngeal relaxation". Dinnick 71 reviewed a further 600 such deaths in 1956. Of 400 in which anaesthesia was implicated, 13 were mainly due to postoperative respiratory obstruction. In 43 patients, adequate respiration could not be restored at the end of the operation. Several of these patients received massive doses of neostigmine (10-12.5 mg), certainly enough to produce a neostigmine-induced blockade. One patient died when carbon dioxide was added to the inspired gas mixture in an attempt to stimulate respiration. Five patients aspirated gastric contents postoperatively - at least one due to unreversed gallamine NMB. In 1961, in reviewing the causes of inadequate post-anaesthetic ventilation, Dam and Guldman72 stated: "Today, the most common cause of postoperative respiratory inadequacy is the use and misuse of muscle relaxant drugs".
The circumstantial evidence implicating PORC in patient morbidity or mortality is not confined to the era when anaesthetists were unfamiliar with the use of muscle relaxants. Although residual neuromuscular weakness is not stated to be the underlying cause, upper airway obstruction in the immediate postoperative period accounted for deaths in the surveys conducted by Marx et al.73 and Utting et al.,74 and was the main cause of death in the latter. In 1983, in a study conducted by the Association of Anaesthetists of Great Britain and Ireland,75 only 16% of 197 (anonymously reported) deaths that occurred within six days of anaesthesia were found by assessors to be totally due to anaesthesia. Five deaths were due to postoperative respiratory failure that was attributed to "neuromuscular causes". In 1987, Derrington and Smith 76 reviewed the studies of anaesthetic risk, morbidity and mortality which were then available. These authors cite postoperative ventilatory failure due to myoneural blockade as a "factor commonly reported as contributing to anaesthetic mortality". Pedersen et al.77 prospectively studied the incidence of postoperative pulmonary complications (POPC) following the use of either pancuronium or atracurium. A greater proportion of patients who had received pancuronium (11%) developed POPC than the 3% of patients who had received atracurium. The difference between the two muscle relaxant groups was more marked (and thus the choice of pancuronium would appear ill-advised) in patients with other risk factors for POPC such as the elderly or those with chronic obstructive lung disease.
It has been suggested that the degree of PORC that might be overlooked by an anaesthetist is unlikely to adversely affect the patient, particularly if vigilant recovery-room staff are present. ". . . even if episodes of respiratory depression or inability to protect the airway did occur, competent recovery room personnel should be able to prevent a permanent adverse effect from occurring" (Miller RD, 1989).9 Invocation of a safety margin offered by the presence of recovery rooms and trained recovery-room staff is unconvincing in the light of the findings of the Confidential Enquiry into Perioperative Deaths (CEPOD) report.78 This report clearly indicates that, at. least in the three regions of the United Kingdom studied, dedicated recovery rooms are not available in all hospitals where operations under general anaesthesia are being carried out. Some recovery rooms close at night and at weekends because of lack of staff. Monitoring facilities were judged to be inadequate in 21% of cases.
Only 26 of 13,000 retrospectively reviewed postanaesthetic care unit (PACU) admissions required insertion of an endotracheal tube while in the PACU;79 three of these were due to persistent muscle relaxant effects. Of two hundred patients studied by Moller et al.80 in a recent, single-blind, observer study, 55% were noted to have at least one episode of hypoxaemia (Sp02 < 90%) in the PACU. In only 24% of these episodes did clinical observers attribute the hypoxaemia to identifiable causes. This led the authors to conclude that, in many cases, hypoxaemia is the result of multiple and co-existing causes, including the residual effects of neuromuscular blocking drugs.
The accumulated evidence presented here suggests that PORC is a real and current problem and contributes to postoperative patient morbidity. The practical and ethical considerations that have made a definitive study of this subject difficult to perform have been outlined. The widespread use of muscle relaxants of intermediate duration has decreased the incidence of PORC and the introduction of mivacurium chloride and rocuronium is likely to continue this trend. The use of perioperative TOF monitoring is to be recommended both as a guide to muscle relaxant administration intra-operatively and as a means of reducing the incidence of PORC.