Aprotinin therapy

British Journal of Anaesthesia 1995; 75: 245–253

CORRESPONDENCE

Subarachnoid spread of isobaric tetracaine in adolescents Sir,—In a recent report, Hirabayashi and colleagues [1] showed that subarachnoid administration of hyperbaric amethocaine (tetracaine) produced an unexpectedly higher level of analgesia in adolescents than in adults. The authors postulated that when adolescents lie in the supine horizontal position, the thoracic kyphotic curvature decreases because of the flexibility of their vertebral columns which become straighter at the upper part of the thoracic vertebrae. In our recent investigation [2], we studied 50 patients (17 adolescents, aged 13–16 yr and 33 adult patients aged 17–68 yr) who received isobaric tetracaine 10 mg in 2 ml. The anaesthetic solution was prepared from 1.0 % tetracaine solution with CSF. As in the study of Hirabayashi and colleagues, the drug was administered with the patient in the lateral position, at the L3–4 interspace, at a speed of 0.2 ml s⫺1, with the patient remaining in the supine horizontal position during the study. The extent of analgesia was assessed by loss of sensation to pinprick, and motor block according to the Bromage scale at 2-min intervals for 20 min and then every 15–20 min until full recovery. There were no differences in physical characteristics between the adolescent and adult groups; the median peak/highest sensory levels were T7 (range T11–3) and T7 (T10–3), respectively. This difference was not statistically significant (P 9 0.5). We also used 2 ml of anaesthetic solution, as did Hirabayashi and colleagues but the tetracaine dose that we used was 2 mg greater. When a patient is in the supine horizontal position, a hyperbaric solution spreads cephalad because of gravity and is therefore affected by the anatomical configuration and flexibility of the vertebral column. Adolescent patients have greater flexibility and thus a flatter thoracic curvature, which leads to a higher anaesthetic level. As gravity has little, if any, effect on the spread of isobaric solutions, the peak/highest levels of anaesthesia in our study did not differ significantly. The difference in anaesthetic levels between adolescent and adult patients observed by Hirabayashi and co-workers did not occur in our patients with isobaric tetracaine. A review of the literature has shown some correlation between patient age and subarachnoid spread of isobaric anaesthetic solution [3, 4]. However, these studies did not compare adolescents with adults, but young (:30 yr) with elderly adults (970 yr). In addition, they were performed with 0.5 % plain bupivacaine solution which is slightly hypobaric at body temperature [5] and the results were poorly predictable [4, 5]. Although not a formal study, our data support the results of Hirabayashi and colleagues who concluded that the higher anaesthesia level in adolescent patients with hyperbaric spinal anaesthesia was probably caused by a flatter thoracic curvature. This problem does not arise with an isobaric anaesthetic solution which may therefore be preferable in adolescent patients. H.-K. KING D. J. WOOTEN Department of Anesthesiology King/Drew Medical Center Charles R. Drew University of Medicine and Science Los Angles, CA, USA 1. Hirabayashi Y, Shimizu R, Saitoh K, Fukuda H. Spread of subarachnoid hyperbaric amethocaine in adolescents. British Journal of Anaesthesia 1995; 74: 41–45. 2. King HK, Wooten DJ. Effects of drug dose, volume, and concentration on spinal anesthesia with isobaric tetracaine. Regional Anesthesia 1995; 20: 45–49. 3. Pitkanen M, Haapaniemi L, Touminen M, Rosenberg PH. Influence of age on spinal anaesthesia with isobaric 0.5 % bupivacaine British Journal of Anaesthesia 1984; 56: 279–284. 4. Cameron AE, Arnold RW, Ghoris MW, Jamieson V. Spinal analgesia using bupivacaine 0.5 % plain. Variations in the extent of block with patient age. Anaesthesia 1981; 36: 318–322.

5. Logan MR, McClure JH, Wildsmith JAW. Plain bupivacaine—An unpredictable spinal anaesthetic agent. British Journal of Anaesthesia 1986; 58: 292–296.

Sir,—We agree with the conclusion of Drs King and Wooten that the difference in the extent of spinal anaesthesia with hyperbaric tetracaine between adolescent and adult patients [1] does not occur when isobaric tetracaine is used [2]. In addition to hyperbaric tetracaine, we have previously studied the effect of age on spread of spinal anaesthesia with 0.5 % plain bupivacaine 3 ml in 185 patients aged 12–94 yr [3]. The maximum spread of analgesia did not differ between adolescents (median T8 (range T3.5–T12)) and adults (T8 (T3–L1)). The overall range of analgesia was very wide at all ages (fig. 1). The unpredictability of isobaric (or slightly hypobaric at body temperature) anaesthetic solution is of clinical importance. We therefore believe that this anaesthetic solution produces reliable anaesthesia for perineal and lower limb surgery.

Figure 1 Analgesic spread of subarachnoid injection of 0.5 % plain bupivacaine 3 ml with age (from [3] with permission).

Gravity has no effect on the spread of isobaric solutions, while hyperbaric solutions do spread under the influence of gravity [4]. As isobaric anaesthetic solutions injected into the midlumbar subarachnoid space may be expected to stay in the vicinity of the puncture site, the difference in curvature of the thoracic hollow between adolescents and adults does not affect the cephalad spread of isobaric solutions. On the other hand, with respect to hyperbaric anaesthetic solutions, when a patient is turned to the supine position after midlumbar subarachnoid injection, the injectate splits into a part migrating cephalad and a part spreading caudad under the influence of gravity. The solution migrating cephalad can be expected to pool in the thoracic hollow. Accentuation or elimination of thoracic kyphosis affects the distribution of this pool [5]. A flatter thoracic hollow in adolescent patients, the lowest point of which is located in the higher thoracic region, may allow the drug to spread more easily to higher thoracic segments. In addition, the reduced incline in the upper thoracic spinal canal may also enhance cephalad spread. Y. HIRABAYASHI R. SHIMIZU K. SAITOH H. FUKUDA Department of Anaesthesiology Jichi Medical School Tochigi, Japan

246 1. Hirabayashi Y, Shimizu R, Saitoh K, Fukuda H. Spread of subarachnoid hyperbaric amethocaine in adolescents. British Journal of Anaesthesia 1995; 74: 41–45. 2. King HK, Wooten DJ. Effects of drug dose, volume, and concentration on spinal anesthesia with isobaric tetracaine. Regional Anesthesia 1995; 20: 45–49. 3. Hirabayashi Y, Shimizu R, Saitoh K, Fukuda H, Nakaigawa Y, Saitoh J, Ikeno S, Horiguchi Y, Togashi H, Mitsuhata H, Akazawa S, Kasuda H. Spread of spinal anesthesia with 0.5 % isobaric bupivacaine (in Japanese with English abstract). Masui (Japanese Journal of Anesthesiology) 1993; 42: 1628–1634. 4. Wildsmith JAW, McClure JH, Brown DT, Scott DB. Effects of posture on the spread of isobaric and hyperbaric amethocaine. British Journal of Anaesthesia 1981; 53: 273–278. 5. Greene N. Distribution of local anesthetic solutions within the subarachnoid space. Anesthesia and Analgesia 1985; 64: 715–730.

Aprotinin therapy Sir,—We were interested to read the editorial by Royston [1] in which he considered the evidence for a thrombotic potential for aprotinin in the light of our report of serious arterial thrombosis in a patient participating in a prospective study of aprotinin in knee replacement surgery [2]. As we stated in our report, we think it is unlikely that aprotinin was the cause of the arterial occlusion: the patient underwent technically difficult surgery and the extent of pre-existing arterial disease had been masked by severe limitation from the extensive joint disease. However, the possibility remains that the drug may have contributed and as the complication was serious, even a small risk becomes significant. The author suggests that comprehensive prediction of the clinical effect of aprotinin can be made on the basis of published laboratory and clinical studies to date. Whatever the mechanism of haemostatic action in surgical patients may be, it is clear that under some conditions aprotinin favours haemostasis over fibrinolysis. The risk of a thrombotic effect in otherwise heavily predisposed patients therefore cannot be easily discounted. While the existing data documenting the lack of thrombogenic potential is encouraging, evidence for the role of aprotinin under tourniquet conditions is lacking and can only be inferred. In view of the serious nature of this complication, we consider it appropriate to wait for additional safety and efficacy data for aprotinin in surgical haemostasis. However, we are hopeful that additional data will allow us to consider resuming this study. C. M. THORPE W. MURPHY M. LOGAN Department of Anaesthetics and Department of Medicine Royal Infirmary of Edinburgh Edinburgh 1. Royston D. Aprotinin therapy. British Journal of Anaesthesia 1994; 73: 734–737. 2. Thorpe CM, Murphy WG, Logan M. Use of aprotinin in knee replacement surgery. British Journal of Anaesthesia 1994; 73: 408–410.

Comparison of ropivacaine and bupivacaine for extradural analgesia Sir,—I read with interest the article on ropivacaine and bupivacaine in extradural analgesia for the relief of pain in labour [1]. However, several aspects of this article concern me. Although the authors have stressed the potential for cardiovascular and central nervous system toxicity with ropivacaine and bupivacaine, there was no reference to any means of excluding intravascular placement of the extradural catheter in the method used. Inadvertent intravascular placement or migration of an extradural catheter remains one of the most common complications of extradural anaesthesia, occurring in 3–9 % of obstetric patients [2, 3] and may lead to serious consequences [4]. Although one can argue that there is no specific or sensitive test to exclude intravascular placement or migration of the extradural catheter, one should use a combination of techniques, for example inspection of the extradural catheter for blood, aspiration before

British Journal of Anaesthesia injection, fractionation, injection of adrenaline 15 ␮g and injection of air [5]. A generous volume of bupivacaine or ropivacaine was used in this study (20 mg) to ensure that the extradural catheter was not in the subarachnoid space. Would a dose of 8–10 mg have been sufficient? Was it necessary to use as much as 50 mg of ropivacaine or bupivacaine to establish the block, followed by a 25-mg top-up approximately 1 h later? Perhaps this dose accounts for the fact that 30 % of the patients in this study developed hypotension ranging from mild to severe? I would also be interested to know if the incidence of nausea, vomiting and fetal bradycardia was related to those patients exhibiting hypotension. A. A. AL-KAISY Department of Anaesthesia Southport and Formby National Health Service Trust Southport 1. McCrae AF, Jozwick H, McClure JH. Comparison of ropivacaine and bupivacaine in extradural analgesia for the relief of pain in labour. British Journal of Anaesthesia 1995; 74: 261–265. 2. Verniquet AJW. Vessel puncture with epidural catheters. Anaesthesia 1980; 35: 660–662. 3. Kenepp KB, Gulsche BB. Inadvertent intravascular injections during epidural anesthesia. Anesthesiology 1981; 54: 172–173. 4. Albright GA. Cardiac arrest following regional anesthesia with etidocaine or bupivacaine. Anesthesiology 1979; 51: 285–287. 5. Leighton BL, Norris M, Desimone CA, Rosko T, Gross JB. The air test as a clinically useful indicator of intravenously placed epidural catheters Anesthesiology 1990; 73: 610–613.

Sir,—Intravascular placement of extradural catheters is common, but intravascular migration is rare and both complications should not be considered together. In our study all extradural catheters were inspected for blood and an aspiration test was performed before any local anaesthetic was injected. All patients were asked to report any symptoms of early systemic toxicity such as tinnitus or numbness of the lips or tongue. Fractionation was used and is clearly described. The test dose of 20 mg of ropivacaine or bupivacaine was chosen to exclude both intravascular and intrathecal placement of the catheter, and not only intrathecal placement, as Dr Al-Kaisy implies. A test dose of less than 20 mg is unlikely to produce symptoms of systemic local anaesthetic toxicity [ l ]. The use of adrenaline in a test dose produces a transient tachycardia if i.v. injection occurs, but requires additional monitoring (ECG, pulse oximetry) to detect any transient change in heart rate. Adrenaline may also affect motor block and was thus undesirable in this study. Use of air injected down the extradural catheter and precordial Doppler recording of heart sounds is controversial and is not part of our routine practice. The total loading dose of 50 mg of ropivacaine or bupivacaine was not considered excessive and the top-up dose of 25 mg was given only at maternal request, whenever contractions became painful again. Hypotension occurred in six patients in each group. Mild hypotension was defined as a decrease of 20 % from baseline systolic arterial pressure, moderate 20–30 % decrease and severe .40 % decrease. In the eight patients classified as having moderate or severe hypotension, the majority were associated with a high level of block. Concomitant nausea occurred in 50 % of these patients. In none of the six patients reported was fetal bradycardia associated with maternal hypotension. In three patients bradycardia occurred at the beginning of the second stage of labour. F. MCCRAE H. JOZWIAK J. H. MCCLURE Department of Anaesthetics Royal Infirmary of Edinburgh Edinburgh 1. Scott DB, Lee A, Fagan D, Bowler GMR, Bloomfield P, Lundh R. Acute toxicity of ropivacaine compared with that of bupivacaine. Anesthesia and Analgesia 1989; 69: 563–569.

Correspondence

Animal toxins: Scorpaenidae and stingrays Sir,—I read with interest the review article on animal toxins by Karalliedde [1] and was surprised to find no mention of the venomous bony fishes of the family Scorpaenidae, which are considered the most dangerous in both the number and severity of injuries produced [2], or of the various genera of stingrays. Scorpaenids envenomate victims by erecting spines on their dorsal, anal and pelvic fins that pierce the victim’s flesh and release a heat labile protein that differs in potency amongst genera [2]. Species familiar to the tropical or temperate scuba diver include the beautiful and majestic lionfish (genus Pterois, commonly seen in aquaria), the masterfully camouflaged scorpionfish (genus Scorpaena) and the deadly innocence of the aptly named stonefish (genus Synanceja) indigenous to the Indo-Pacific region only and responsible for several fatalities [3]. Envenomation produces intense local pain, swelling, paraesthesiae and discolouration that may lead to tissue necrosis and subsequent sloughing. Systemic symptoms may be mild (syncope, weakness, nausea, vomiting, sweating, fasciculations, abdominal or chest pain) or may culminate in bradycardia, hypotension, shock, respiratory distress, pulmonary oedema and death, which usually results from stonefish envenomation [4]. Treatment of the affected limb by immersion in hot water up to the highest temperature tolerated for at least 1 h inactivates the venom and relieves pain. This effective and simple treatment is now well known to lifeguards in the United States, and is responsible for the decrease in the number of cases presenting to hospital emergency departments. Irrigation of the wound, tetanus prophylaxis and reassurance is usually sufficient in mild cases [4]. Further treatment with vasopressors and i.v. fluids for circulatory support may be required in severe cases of envenomation when the use of stonefish antivenin (Commonwealth Serum Laboratories of Australia) is recommended [4]. Envenomation by weever fishes (family Trachinidae) which are found off the coast of the United Kingdom cause excruciating pain and a clinical picture resembling Scorpaenidae envenomation. The management, too, is the same [4]. While one can indulge in the amazing experience of stroking, petting, feeding and frolicking with the tame, giant stingrays of the Caribbean island of Grand Cayman in shallow waters, it is important to realize that stingrays are a common cause of morbidity in bathers, snorkellers and divers. There are approximately 750 stingray envenomations reported off the US coast every year [5]. As they commonly lie partly hidden on sandy bottoms, bathers inadvertently step on them causing the tail with its venomous spine and backward-pointing serrations to lash forwards and puncture the skin, envenomating tissues and causing laceration. The venom contains proteins, serotonin, phosphodiesterase, 5-nucleotidase and vasoconstrictors that may cause coronary artery spasm, chest pain and changes in the electrocardiogram [5]. The wound usually needs irrigation, surgical exploration and foreign body removal after initial immersion in hot water, which is effective at relieving pain. An x-ray of the affected part to exclude retained foreign bodies, tetanus prophylaxis, antibiotics and surgical closure of lacerations are recommended [5]. Death from stingray envenomation may occur, but is uncommon [6]. Stepping on stingrays can be avoided by shuffling into the water from the shore instead of “walking” or “stepping” in, especially in tropical areas with sandy beaches. However, the cardinal rule to minimize the likelihood of marine envenomation whether swimming, snorkelling, scuba diving or handling fishes from an aquarium, is “Look, but don’t touch”. K. LALWANI Department of Anaesthesia King’s College Hospital London 1. Karalliedde L. Animal toxins. British Journal of Anaesthesia 1995; 74: 319–327. 2. Kizer KW, McKinney HE, Auerbach PS. Scorpaenidae envenomation. A five year poison center experience. Journal of the American Medical Association 1985; 253: 807–810. 3. Halstead BW. Poisonus and Venomous Marine Animals of the World. Princeton, New Jersey: Darwin Press, 1978; 743–798. 4. Russell FE, Dart RC. Toxic effects of animal toxins. In: Amdur MO, Doull J, Klaassen CD, eds. Casarett and Doull’s Toxicology—The Basic Science of Poisons, 4th Edn. New York: McGraw-Hill, 1991; 753–803.

247 5. Haddad LM, Lee RF. Toxic marine life. In: Haddad LM, Winchester JF, eds. Clinical Management of Poison and Drug Overdose, 2nd Edn. Philadelphia: WB Saunders, 1990; 600–612. 6. Russell FE. Report of two fatal case studies on the mechanism of death from stingray venom. American Journal of Medical Science 1958; 235: 566.

Detection of accidental oesophageal intubation Sir,—Clyburn and Rosen suggest [1] that the oesophageal detector device (ODD) [2] is a tool for “enthusiasts”. Their choice of the word “enthusiast” is understandable but unfortunate, as it gives the impression that extra effort is required on the part of the operator. Lunn has pointed out that if mortality from oesophageal intubation is to be reduced, every intubation must be tested [3]. The ODD is so simple, quick and reliable in use, that testing every intubation is not only practical, but rapidly becomes second nature. Meanwhile, it is surely time to stop advancing capnography as the best method of detecting oesophageal intubation. There are (at least) six problems: (1) the positive is a negative. If no signal is produced, there is a doubt as to whether or not there is a technical fault; (2) it takes a long time to obtain the result (repeated waveforms); (3) it is easy to forget [4]; (4) the equipment is expensive; (5) the 1993 NCEPOD report indicated that it was used only in 6.7 % of cases [5]; and (6) an expired carbon dioxide waveform does not guarantee tracheal intubation [6]. Of course, there is nothing wrong with using a stethoscope, provided the stomach is auscultated [7], but the ODD gives an instantaneous and very positive result (provided a correct Ellik bulb is used and not that illustrated in Clyburn and Rosen’s original article [8]). It is true that the bulb has a slow refill in about 15 % of cases [9], which can lead to an occasional false positive [10], but this is a fail safe, and the true positive appears to be unmistakable. One cannot deny Clyburn and Rosen’s assertion that with widespread use mistakes may occur (no technique is foolproof), but lives (and careers) would certainly be saved if we all adopted the painless habit of using the ODD. I believe that the use of the ODD should be mandatory. The situation is analogous to the wearing of seat belts; compulsion is sometimes required for the good of society. I. CALDER Royal Free Hospital London 1. Clyburn P, Rosen M. Accidental oesophageal intubation. British Journal of Anaesthesia 1995; 74: 242–243. 2. Williams KN, Nunn JF. The oesophageal detector device. A prospective trial on 100 patients. Anaesthesia 1989; 44: 412–414. 3. Lunn JN. Change of clinical practice. Anaesthesia 1990; 45: 1–2. 4. Mackenzie CF, Craig OR, Parr MJ, Horst R. Video analysis of two emergency tracheal intubations identifies flawed decision-making. Anesthesiology 1994; 81: 763–771. 5. Campling EA, Devlin HB, Hoile RW, Lunn J N. The Report of the National Confidential Enquiry into Perioperative Deaths 1991/1992. London: Royal College of Surgeons, 1993. 6. Deluty S, Turndorf H. The failure of capnography to properly assess endotracheal tube location. Anesthesiology 1993; 78: 783–784. 7. Andersen KH, Hald A. Assessing the position of the tracheal tube: the reliability of different methods. Anaesthesia 1989; 44: 984–989. 8. Clyburn P, Rosen M. Accidental oesophageal intubation. British Journal of Anaesthesia 1993; 73: 55–63. 9. Zaleski L, Abello D, Gold MI. The esophageal detector device—does it work? Anesthesiology 1993; 79: 244–247. 10. Calder I, Smith M, Newton M. The oesophageal detector device. Anaesthesia 1989; 44: 705. Sir,—We agree that clinical signs are not totally reliable at conferring tracheal tube placement, and confirmation by a technical test is mandatory. However, capnography is still the most tried and tested technique. It is widely available in the operating theatre and is a minimal monitoring standard recom-

248 mended by the Association of Anaesthetists of Great Britain and Ireland [1]. Its low utilization in anaesthetic rooms reported by NCEPOD [2] undoubtedly reflects over-reliance on clinical signs to detect oesophageal tube placement and the misconception that capnography is not necessary at induction. We also disagree with Dr Calder that false positive tests are fail safe as the unnecessary removal of a correctly placed tracheal tube after difficult intubation can be hazardous. As we originally stated in our article, the placement of a tracheal tube should be followed by ausculation over the trachea axilla and epigastrium, and confirmed with a reliable technical test. Ideally this should be capnography. When this is unavailable, then the oesophageal detector device is a simple and suitable alternative. P. CLYBURN M. ROSEN Department of Anaesthetic University Hospital of Wales Cardiff 1. Adams AP, Baird WLM, Sykes MK, Charlton JE, Rosen M, Burrows MM, MacRae WR, Morris P, Lunn JN. Recommendations for Standards of Monitoring during Anaesthesia and Recovery. London: Association of Anaesthetists, 1988. 2. Campling EA, Devlin HB, Hoile RW, Lunn JN. The Report of the National Confidential Enquiry into Perioperative Deaths 1991/1992. London: Royal College of Surgeons, 1993.

Ambulatory extradural analgesia Sir,—I believe Drs Buggy, Hughes and Gardiner [1] may have made a mistake in stating that the Queen Charlotte’s group used an intrathecal dose of bupivacaine 25 mg with fentanyl 25 ␮g before extradural administration. This dose most probably results in an extremely high spinal block which certainly prevents the mother from mobilizing. The dose of bupivacaine is 10 times lower; that is 2.5 mg intrathecally with fentanyl 25 ␮g. While it is true that it is not known what the effect of an intrathecal dose of bupivacaine 2.5 mg and fentanyl 25 ␮g has on the neurological status of mothers, the majority of these mothers could mobilize if they wished [2]. The extradural dose of bupivacaine 15 mg containing fentanyl 30 ␮g was given only after the return of discomfort, which occurred at a mean time of 90 min (range 20–245 min). Therefore, at no time was a total dose of bupivacaine 30 mg given initially, as was the case in their study [3]. I therefore repeat the contention that while their article quite rightly cautions the safety of giving bupivacaine 30 mg and fentanyl 30 ␮g extradurally, they have not demonstrated that repeat doses of 10–15 ml of 0.1 % bupivacaine with fentanyl 2 ␮g ml⫺1 present the same dangers. M. WEE Department of Anaesthesia Poole Hospital Dorset 1. Buggy D, Hughes N, Gardiner J. Posterior column sensory impairment during ambulatory extradural analgesia in labour. British Journal of Anaesthesia 1995; 74: 350. 2. Collis RE, Baxandall ML, Srikantharajah ID, Edge G, Kadim Y, Morgan BM. Combined spinal epidural analgesia with ability to walk throughout labour. Lancet 1993; 341: 767. 3. Buggy D, Hughes N, Gardiner J. Posterior column sensory impairment during ambulatory extradural analgesia in labour. British Journal of Anaesthesia 1994; 73: 540–542.

Sir,—Certainly intrathecal bupivacaine 25 mg is a mistake and not what the Queen Charlotte’s group prescribed [1], nor indeed did we interpret it as such. We suggest that this must surely be a misprint or typographical error, as our original letter of reply correctly noted the intrathecal dose of bupivacaine as 2.5 mg [2]. The return of discomfort during labour may not necessarily correlate with the return of neurological function and, in particular, with recovery of posterior column sensory modalities. Nevertheless, further studies are indeed indicated to determine the effect and duration of both the intrathecal dose of bupivacaine

British Journal of Anaesthesia 2.5 mg and the extradural dose of bupivacaine 15 mg on posterior column sensory function. D. BUGGY N. HUGHES J. GARDINER Department of Anaesthesia Mater Misericordiae Hospital Dublin 1. Collis RE, Baxandall ML, Srikantharajah ID, Edge G, Kadim MY, Morgan BM. Combined spinal epidural analgesia with ability to walk throughout labour. Lancet 1993; 341: 767. 2. Buggy D, Hughes N, Gardiner J. Posterior column, sensory impairment during ambulatory extradural analgesia in labour. British Journal of Anaesthesia 1995; 74: 350.

Gastric mucosal pH and splanchnic blood flow Sir,—I read with interest the two recent articles by Takala’s group [1, 2]. The authors used dopexamine [1] and dobutamine [2] to increase cardiac index and splanchnic blood flow after cardiac surgery in an attempt to improve splanchnic tissue oxygenation. The latter was assessed by tonometric estimation of gastric mucosal pH (pHi) and by regional oxygen consumption data. Many factors are known to affect pHi such that in some circumstances it is not a reliable index of the adequacy of splanchnic perfusion. Unfortunately, the presence or absence of several important factors was not described in these articles. Evolving systemic metabolic acidosis is not an uncommon phenomenon during or after cardiac surgery. In these circumstances, a low pHi would be expected to develop and may not necessarily indicate a state of regional dysoxia. Coexistent systemic alkalosis is another potential source of error as it may mask tonometric detection of low pHi [3]. In many cardiac surgical units, metabolic acidosis is treated with i.v. sodium bicarbonate but it is unclear whether or not patients in these studies received such treatment during or after operation. Administration of sodium bicarbonate is known to introduce artefact into tonometric estimation of pHi [4]. Although Takala’s group found that pHi decreased significantly during infusion of dopexamine, there was a comparable reduction in mean values for this variable in the control group (0.04 vs 0.05 with dopexamine), and the reason for the lack of statistical significance for the decrease in the control group may have been related to the smaller number of patients. A downward trend in pHi after cardiac surgery has been documented previously [5]. “How rapidly should low pHi be expected to recover in response to an effective therapy?” is a question pertaining to study design which remains unanswered at present. Takala’s group assessed the effect on pHi after 90 min of inodilator therapy. It is conceivable that this is an insufficient length of time for restoration of local acid–base balance. Nevertheless, the apparent lack of improvement in pHi with dopexamine in the study of Uusaro, Ruokonen and Takala [1] is in keeping with our own findings in non-cardiac patients (predominantly with multiple trauma) in whom pHi was measured up to 3 h after commencing dopexamine [6]. T. J. TRINDER Department of Clinical Anaesthesia Royal Hospitals Belfast 1. Uusaro A, Ruokonen E, Takala J. Gastric mucosal pH does not reflect changes in splanchnic blood flow after cardiac surgery. British Journal of Anaesthesia 1995; 74: 149–154. 2. Parviaienen I, Ruokonen E, Takala J. Dobutamine-induced dissociation between changes in splanchnic blood flow and gastric intramucosal pH after cardiac surgery. British Journal of Anaesthesia 1995; 74: 277–282. 3. Benjamin E, N-Fonayim JN, Hannon EM, Premus G, Oropello JM, Leibowitz AB, Iberti TJ. Effects of systemic metabolic alkalosis on gastrointestinal tonometry. Critical Care Medicine 1992; 20 (Suppl): S65. 4. Benjamin E, Polokoff, E, Oropello JM, Leibowitz AB, Iberti TJ. Sodium bicarbonate administration affects the diagnostic accuracy of gastrointestinal tonometry in acute mesenteric ischaemia. Critical Care Medicine 1992; 20: 1181–1183. 5. Landow L, Phillips D, Prevost D, Vandersalm TJ, Fink MP. Gastric tonometry and venous oximetry in cardiac surgery patients. Critical Care Medicine 1991; 19: 1226–1233.

Correspondence

249

6. Trinder TJ, Lavery GG, Fee JPH, Lowry KG. Correction of splanchnic oxygen deficit in the intensive care unit: dopexamine and colloid versus placebo. Anaesthesia and Intensive Care 1995 (in press). Sir,—We thank Dr Trinder for his interest in our two articles [1, 2]. It is true that systemic acid–base status does also affect pHi interpretation and a pH-gap (difference between systemic and gastric mucosal pH) rather than absolute pHi values should be used. We are also aware of the possible hazards (among others sodium bicarbonate administration) concerning gastric mucosal pH (pHi) determinations. However, we do not administer sodium bicarbonate to our cardiac surgery patients during or after operation and none of our patients in the two published studies received sodium bicarbonate. Metabolic acidosis after cardiac surgery is indeed very rare in our department. A likely cause for metabolic acidosis in postoperative cardiac patients is insufficient tissue perfusion, and increasing oxygen delivery would be our choice of therapy, not sodium bicarbonate. Metabolic acidosis or alkalosis was not observed in either of our studies. In the study of Parviainen, Ruokonen and Takala [2], mean arterial pH was 7.42 (sem 0.01) and base deficit ⫺1.4 (0.4) during the baseline measurement and 7.40 (0.01) and ⫺1.4 (0.4) during the vasoactive period, respectively. In the study of Uusaro, Ruokonen and Takala [1], arterial pH was 7.41 (0.01) and base deficit ⫺1.5 (0.4) during the baseline measurement and 7–40 (0.01) and ⫺1.6 (0.6) during the vasoactive period, respectively. In addition to systemic acid–base balance and sodium bicarbonate administration, sample handling, delay in analysis of tonometric saline PCO2 and determination of PCO2 with different blood-gas, analysers affect calculated pHi [3, 4] and an effort was made to minimize these sources of error in both of our studies. It is indeed possible that if the number of patients had been larger in the control group in the dopexamine study, a significant decrease in pHi would have been observed in these patients. Gastric mucosal acidosis or a decreasing gastric mucosal pH after cardiac surgery is certainly not a new finding. However, the main finding in this study was that gastric mucosal pH decreased consistently in response to increased splanchnic oxygen delivery by dopexamine; a situation in which an increase in pHi would logically have been expected. We do not agree with Dr Trinder that the time was too short for gastric mucosal pH to improve in response to vasoactive therapy and there are previous reports to support our view. In an experimental study by Antonsson and colleagues when tonometry was actually validated, tonometrically determined gastric pHi decreased to its lowest value in 40 min during partial superior mesenteric artery occlusion and increased to the pre-occlusion level 80 min after occlusion [5]. After total occlusion there was a ⬎ 0.1 pH unit increase in 40 min. In a recent study in patients with septic syndrome, gastric mucosal pH improved significantly in response to 60 min of infusion with dopexamine at the maximum dose studied [6]. We believe that 90 min of effective therapy should be long enough to improve gastric mucosal

acidosis, if any improvement occurs. A. UUSARO I. PARVIAINEN E. RUOKONEN J. TAKALA Critical Care Research Program Kuopio University Hospital Kuopio, Finland 1. Uusaro A, Ruokonen E, Takala J. Gastric mucosal pH does not reflect changes in splanchnic blood flow after cardiac surgery. British Journal of Anaesthesia 1995; 74: 149–154. 2. Parviainen I, Ruokonen E, Takala J. Dobutamine-induced dissociation between changes in splanchnic blood flow and gastric intramucosal pH after cardiac surgery. British Journal of Anaesthesia 1995; 74: 277–282. 3. Wood PR, Cudworth P, Davies GE, Lawler PG Potential hazards in estimation of gastric intramucosal pH. Lancet 1992; 340: 913. 4. Takala J, Parviainen I, Siloaho M, Ruokonen E, Hämäläinen E. Saline Pco2 is an important source of error in the assessment of gastric intramucosal pH. Critical Care Medicine 1994; 22: 96–100. 5. Antonsson JB, Boyle CC, Kruithoff KL, Wang H, Sacristan E, Rothschild HR, Fink MP. Validation of tonometric measure-

ment of gut intramural pH during endotoxemia and mesenteric occlusion in pigs. American Journal of Physiology 1990; 259: G519–523. 6. Smithies M, Yee TH, Jackson L, Beale R, Bihari D. Protecting the gut and the liver in the critically ill: Effects of dopexamine. Critical Care Medicine 1994; 22: 789–795.

Positive pressure ventilation and the laryngeal mask airway in ophthalmic anaesthesia Sir,—The controversy over whether or not intermittent positive pressure ventilation (IPPV) is safe with the laryngeal mask airway (LMA) rages on, with the publication of the article by Valentine, Stakes and Bellamy [1] and the reply by Brain and colleagues [2]. In ophthalmic anaesthesia, the LMA has a particular advantage. Lamb, James and Janicki [3] demonstrated the beneficial effect of substituting the LMA for a tracheal tube on both intraocular pressure and pressor responses, not only at insertion but also at removal. The LMA has other advantages in ophthalmic anaesthesia, such as lack of a tendency to cough, either if the neuromuscular blocker is wearing off or at removal, resulting in a smooth transfer from IPPV to spontaneous ventilation. There is also markedly less discomfort in the throat after operation. However, because the LMA was not designed to provide a perfect seal [4], there have been various attempts to assess the safety of IPPV and the LMA [5, 6] and in particular the potential dangers of gastric aspiration [7–13]. There have also been several case reports [14–18]. Only Verghese, Smith and Young [19] surveyed a large number of patients to try to show the safety of the LMA, reporting on 2359 patients of whom 41 % underwent ventilation. Only two patients exhibited any regurgitation. Haden, Pinnock and Campbell [20] reported 593 cases with no significant problems. I present a series of 1958 patients given a general anaesthetic for ophthalmic surgery. No patient had any clinical signs of inhalation of gastric contents at any time, before, during surgery, in the recovery room or after operation. A total of 1502 patients received IPPV, of whom 1424 received IPPV by means of an LMA. For 77 patients a tracheal tube was used, and two of 1958 patients had a pre-existing tracheostomy; 455 patients breathed spontaneously by means of an LMA. The numbers for each group of operations is shown in table 1. The high proportion of patients having a tracheal tube for oculoplastics shows the author’s preference because of the (as yet unseen) potential for heavy blood loss with dacryocystorhinostomy. Those patients in whom a tracheal tube was used for any other operative group either gave a clear history of reflux or the author was not satisfied with the seal provided by the LMA. This latter problem is not seen very often because of the introduction of the size 5 LMA [2]. Problem patients are those requiring a high insufflation pressure because of obesity or chronic obstructive airways disease, principally asthma. The author has used an arbitrary insufflation pressure of 20 cm as the upper limit, partly because of the increased likelihood of leak and partly because of the adverse effect on intraocular pressure. Changes in the ventilatory pattern to increase frequency and reduce tidal volume ensure that this level is very rarely reached. Both groups are managed better if adequate ventilation can be achieved using the LMA. Increasingly this is less of a problem in ophthalmic anaesthesia as local anaesthetic techniques are used for this group of patients. The smooth change from IPPV to spontaneous ventilation is an important factor in the success of the LMA in ophthalmic anaesthesia. This is particularly so in vitreoretinal work where the retina may be relatively unsupported and coughing is therefore highly undesirable. When I first started using the LMA for ophthalmic anaesthesia, spontaneous ventilation was used for vitreoretinal anaesthesia. This was found to be unsatisfactory as, during work on encircling the globe on the medial side, stimulation of the nerves in this area caused sudden, unpredictable sneezing even when deep levels of anaesthesia were being maintained. Problems with laryngeal spasm at the time of antagonism of neuromuscular block can usually be overcome with an i.v. dose of doxapram. Provided care is taken to use a tracheal tube: (1) in patients with reflux or with a history suggestive of a full stomach, (2) if there is a high inflation pressure and (3) if the airway is not well

250

British Journal of Anaesthesia Table 1 Numbers of patients undergoing each operation, according to whether intermittent positive pressure ventilation (IPPV) was given by means of a tracheal tube (TT) or laryngeal mask airway. SV : spontaneous ventilation

Cataract or trabeoulectomy Vitreoretinal Squint correction Oculoplastics EUA or probing of ducts Evisceration tracheostomies excluded

Total

IPPV–TT

IPPV–LMA

TT % of IPPV

SV–LMA

722 741 343 74 75 1

23 42 2 10 0 0

659 653 79 31 1 1

3.37 % 6.04 % 2.47 % 24.39 %

40 46 262 33 74 0

maintained for any reason, the LMA is not only safe, but a great improvement for patients undergoing general anaesthesia for ophthalmic surgery. C. WAINWRIGHT Shackleton Department of Anaesthetics Southampton University Hospitals NHS Trust Southampton General Hospital 1. Valentine J, Stakes AF, Bellamy MC. Reflux during positive pressure ventilation through the laryngeal mask. British Journal of Anaesthesia 1994; 73: 543–544. 2. Brain AIJ, Brimacombe JR, Berry AM, Verghese C. Reflux during positive pressure ventilation via the laryngeal mask airway ? British Journal of Anaesthesia 1995; 74; 489–490. 3. Lamb K, James MFM, Janicki PK. The LMA for intraocular surgery: effects on intraocular pressure and stress response. British Journal of Anaesthesia 1992; 69: 143–147. 4. Brain AIJ. The development of the laryngeal mask—a brief history of the invention, early clinical studies and experimental work from which the laryngeal mask evolved. European Journal of Anaesthesiology 1991; 4: 5–17. 5. Devitt JH, Wenstone R, Noel AG, O’Donnell RRT. The laryngeal mask airway and positive-pressure ventilation. Anesthesiology 1994; 80: 550–555. 6. Brimacombe J, Berry A. The incidence of aspiration associated with the laryngeal mask airway—a meta-analysis of published literature. Journal of Clinical Anaesthesia (in press). 7. Warner AM, Warner ME, Webber JG. Clinical significance of pulmonary aspiration during the perioperative period. Anesthesiology 1993; 78: 56–62. 8. Kallar SK. Aspiration pneumonitis: fact or fiction? Problems in Anesthesia 1988; 2: 29–36. 9. Haden RM, Pinnock CA, Scott PV. Incidence of aspiration with the laryngeal mask airway. British Journal of Anaesthesia 1994; 72: 496. 10. Illing L, Duncan PG, Yip R. Gastro-oesophageal reflux during anaesthesia. Canadian Journal of Anaesthesia 1992; 39: 466–470. 11. Owens T, Robertson P, Twomey K, Doyle M, McShane AJ. Incidence of gastroesophageal reflux with the laryngeal mask. Anesthesiology 1993; 79: A1053. 12. Akhtar TM, Street MK. Risk of aspiration with the laryngeal mask. British Journal of Anaesthesia 1994; 72: 447–450. 13. Vanner RG. Regurgitation and the laryngeal mask airway. British Journal of Anaesthesia 1993; 70: 380–381. 14. Brimacombe J, Berry A. Aspiration and the laryngeal mask airway—a survey of Australian intensive care units. Anaesthesia and Intensive Care 1992; 20: 534–535. 15. Maroof M, Khan RM, Siddique MS. Intraoperative aspiration pneumonitis and the laryngeal mask airway. Anesthesia and Analgesia 1993; 77: 409–410. 16. Nanji GM, Maltby JR. Vomiting and aspiration pneumonitis with the laryngeal mask airway. Canadian Journal of Anaesthesia 1992; 39: 69–70. 17. Wilkinson PA, Cyna AM, McLeod DM, Campbell JR, Criswell J, John R. The laryngeal mask: cautionary tales. Anaesthesia 1990; 45: 167–168. 18. Barker P, Langton JA, Murphy PJ, Rowbotham DJ. Regurgitation of gastric contents during general anaesthesia using the laryngeal mask airway. British Journal of Anaesthesia 1992; 69: 314–315. 19. Verghese C, Smith TGC, Young E. Prospective survey of the use of the laryngeal mask airway in 2359 patients. Anaesthesia 1993; 48: 58–60.

20. Haden RM, Pinnock CA, Campbell RL. The laryngeal mask for intraocular surgery. British Journal of Anaesthesia 1993; 71: 772.

Optimum opioid for extradural administration Sir,—We read with interest the editorial by Chrusbasik and Chrusbasik on the optimum choice of opioids for extradural administration [1]. The use of extradural opioids alone for postoperative analgesia is rare in British practice. A recently completed survey of 251 United Kingdom hospitals [Cook TM, Eaton JM, Goodwin APL, unpublished observations] showed that only 4 % of departments use opioids alone via this route after major surgery whereas 96 % of the departments use a combination of opioids with local anaesthetic agents. This practice is supported by both laboratory [2] and clinical evidence [3, 4]. The technique of extradural analgesia is widespread: 97 % of departments use extradural analgesia after laparotomy and 81 % after thoracotomy. The choice of opioid used in this way varies: some departments use several opioids, with fentanyl being used by 61 % of departments, diamorphine by 52 % and morphine by 12 %. The authors emphasize the side effects of various opioids via the extradural route but the critical issue of efficacy is less well addressed. McQuay and colleagues have demonstrated the inverse relationship between efficacy and lipophilicity of opioids used spinally [5]. In this respect diamorphine (which is not mentioned in the editorial but is used by more than 50 % of departments) is particularly effective via the extradural route. Fentanyl as a single agent produces poor results [6]; this may be a result in part of the fact that it is readily absorbed in extradural fat. The other opioids mentioned in the editorial may have theoretical advantages over fentanyl, diamorphine and morphine but are rarely used in practice in the United Kingdom and these advantages have not been supported in clinical trials [7]. However, conclusions on clinical utility cannot be drawn without considering procedural details, most notably site of insertion of the extradural catheter, which is known to influence efficacy and side effects of extradural opioids [8]. The survey suggests that 86 % of departments preferentially site extradural catheters at a spinal level which corresponds with the surgical incision. It would appear that at least for British practice the question should not be which is the best extradural opioid but which is the best opioid to use in combination with local anaesthetic agents via the extradural route and at what site. T. M. COOK G. F. STANNARD Royal United Hospital Bath 1. Chrusbasik S, Chrusbasik J. Selection of the optimum opioid for extradural adminstration in the treatment of postoperative pain. British Journal of Anaesthesia 1995; 74: 121–122. 2. Maves TJ, Gebhart GF. Antinociceptive synergy between intrathecal morphine and lidocaine during visceral and somatic nociception in the rat. Anesthesiology 1992; 76: 91–99. 3. George KA, Chisakuta AM, Gamble JAS, Browne GA. Thoracic extradural infusion for post-operative pain relief following abdominal aortic surgery: bupivacaine, fentanyl or a mixture of both? Anaesthesia 1992; 47: 388–394. 4. Dahl JB, Rosenberg J, Hansen BL, Hjortso NC, Kehlet H. Differential analgesic effects of low dose extradural morphine and morphine-bupivacaine at rest and during mobilisation

Correspondence

5. 6.

7.

8.

after major abdominal surgery. Anesthesia and Analgesia 1992; 74: 362–365. McQuay HJ, Sullivan AF, Smallwood K, Dickenson AH. Intrathecal opioids, potency and lipophilicity. Pain 1989; 36: 111–115. Sandler AN, Stringer DG, Panos L, Badner N, Freidlander M, Koren G, Katz J, Klein J. A randomised, double-blinded comparison of lumbar extradural and intravenous fentanyl infusions for postthoracotomy pain relief. Analgesic, pharmacokinetic and respiratory effects. Anesthesiology 1992; 77: 626–634. Vercauteren MP. The role of the perispinal route for postsurgical pain relief. In: van Aken, ed. Clinical Anaesthesiology. New Developments in Extradural and Spinal Drugs Administration. London: Balliere Tindall, 1993; 769–792. Chisakuta AM, George KA, Hawthorne CT. Post-operative extradural infusion of a mixture of bupivacaine 0.2 % with fentanyl for upper abdominal surgery. A comparison of thoracic and lumbar routes. Anaesthesia 1995; 50: 72–75.

Sir,—We congratulate Drs Cook and Stannard on the results of their survey. It is undoubtedly very interesting to know what is actually occurring in at least one part of Europe. However, it is equally interesting that what becomes popular at any time or place is not necessarily supported by evidence that it is the best in all circumstances. Whereas there are undeniable situations in which combining local anaesthesia with extradural opioids has improved effectiveness, there are also undeniable reports of situations in which no significant benefit has been demonstrated [1]. The failures to show benefit are partly, but not entirely, explicable in terms of concentration of local anaesthetic, volume and rate of infusion, type of opioid used in the combination, type of surgery or small numbers in studies that failed to show effect. The almost universal use in Britain of local anaesthetic–opioid combinations, rather than extradural opioids alone may not be accompanied by universal improvement in effectiveness, but is probably excusable nonetheless on the grounds that there is some safety in standardizing the regimens for postoperative analgesia to avoid confusing those who have to put them into effect, and that the users are well enough attuned to the side effects of local anaesthetics to be able to avoid the worst of the potential harm. The single most useful practical feature of opioid-only infusions is that extradural opioids do not need to be given in the relatively large volumes that are required to ensure spread of local anaesthetics in the extradural space, and 24 or 48 h supplies of opioid can be contained in small infusion devices that are easily portable. On the theoretical side, attempts to identify the best opioid for extradural use (if it is possible) had to begin with the study of opioids alone, without the potential complications of having to decide the type, concentration and infusion rate of local anaesthetic with which to combine them, and the type of surgery in which the combination might be more effective than the opioid alone. Any study to determine the best opioid for extradural use in combination with local anaesthetic would have to be based on a postulate that the rank order of extradural use of opioids in combination differs from the rank order of opioids alone, which implies that there might be some specific interactions between some opioids and some local anaesthetics. This is certainly not impossible, but is it likely enough to warrant what would clearly be a huge investigative effort? We suspect that the popularity of fentanyl and diamorphine in Britain is largely pragmatic, because of their ready availability in preservative-free form. We have argued elsewhere [2] that the prime criterion by which opioids should be ranked for use by extradural infusion or extradural patient-controlled analgesia (PCA) is the ratio of i.v. to extradural dose requirements. By this, morphine should be the best, with a ratio of about 9:1, and sufentanil, buprenorphine and methadone should be the worst with a ratio that is not convincingly different from 1. This is reasonably consistent with the inverse relationship between lipophilicity and effectiveness demonstrated by McQuay and colleagues [3] for the discharge of nociceptive neurones in the exposed spinal cord of the rat. However, there are major inconsistencies in the relationship in the middle order of lipophilicity, which probably reflect the added complexities of extradural administration in clinical situations [2].

251 A second criterion which favours the use of more lipophilic opioids is speed of onset. This may not be important for the start of an extradural infusion if it is commenced during operation or if the first hour or so is covered by the intraoperative local anaesthetic effect persisting into the early postoperative period. It may be more important if the opioid is to be given by extradural PCA, for which the patient needs to be able to perceive an analgesic effect within a reasonable time from making a demand. We have tried combining fentanyl with morphine in PCA, to attempt to produce a rapid onset of effect with a low ongoing requirement for opioid but have found that the demand rate produces concentrations of fentanyl that increase into the systemic analgesic range [4]. Diamorphine may be unique in being a lipophilic pro-drug that acts through less lipophilic metabolites (monoacetylmorphine and morphine itself). It is not available for therapeutic use in Germany, although we have studied it in collaboration with colleagues in Bristol. We found systemic concentrations of morphine during extradural PCA that were well below the range for systemic analgesia (as for extradural PCA with morphine; 1), and derived an informal estimate of between 2 and 3:1 for the ratio of i.v. to extradural dose requirements. A recent randomized study after Caesarean section produced an estimate close to 6:1 [5] (although it should be noted that the extradural doses were given on-demand by a nurse rather than self-administered by PCA). Accurate placement of the tip of the extradural catheter is logically and practically important if the best effect is to be obtained from the local anaesthetic component of an extradural infusion (large volumes can compensate to some extent for inaccurate placement whether the injectate contains local anaesthetic, opioid or both). We all wish to see convincing evidence that accurate placement is important for extradural infusions of opioid alone. We could then be reassured that the opioid effect was as localized as we should like to the target segments of the spinal cord. The evidence thus far has not been very convincing. S. CHRUBASIK Department of Internal Medicine J. CHRUBASIK Department of Anesthesiology Ruprecht-Karls-University Heidelberg Heidelberg, Germany 1. Chrubasik J, Chrubasik S, Mather L. Postoperative Epidural Opioids. Heidelberg, New-York: Springer-Verlag Berlin, 1993. 2. Chrubasik J, Chrubasik S, Martin E. The ideal epidural opioid—fact or fantasy. European Journal of Anaesthesiology 1993; 10: 79–100. 3. McQuay HJ, Sullivan AF, Smallwood K, Dickenson AH. Intrathecal opioids, potency and lipophilicity. Pain 1989; 36: 111–115. 4. Chrubasik S, Chrubasik J, Pfisterer M, Hage R, Martin E. Pharmacokinetics during epidural morphine vs fentanyl/ morphine for the management of postoperative pain. Anesthesiology 1994; 81 (Suppl): A1051. 5. Stoddart P, Cooper A, Russell R, Reynolds F. A comparison of epidural diamorphine with intravenous patient-controlled analgesia using the Baxter infusor following Caesarean section. Anaesthesia 1993; 48: 1086–1100.

Cannabis as a medicine Sir,—There is currently much interest in the rehabilitation of cannabis as a medicine. From an academic point of view, Doyle and Spence [1] are correct in their call for further research into its use. However, as a clinician, I am very concerned that its reintroduction into clinical use should not await the results of multiple, prolonged and complex clinical trials, taking years to complete. Cannabis is probably unique in being both a new drug that we want to study and also an old one that has been available for several thousand years. It is a drug with which society has huge experience but doctors very much less. It has dangers and side effects but is also probably far safer than aspirin. There is a large body of laboratory research on it, although paradoxically little clinical research into its analgesic properties. Its study is going to be a challenge to clinicians. Many patients already take cannabis to relieve pain and muscle spasm. Their experience is in contrast with the doubts of Doyle

252 and Spence who comment “…cannabinoids may be useful in some patients with chronic pain”. Recently, I have found an increasing number of patients “outing” about cannabis and telling me that they find that it is more effective in relieving pain than prescribed drugs (including opioids). All of these patients are desperate enough to break the law to obtain analgesia for better symptom control. Some do not like having to smoke it nor do they like some of the side effects. Others find its use pleasant. However, is this any different to the experience of patients taking morphine for pain? Morphine is now used commonly by pain specialists in the management of most types of pain, both acute and chronic. Over the last few years it has been liberated, without evidence of an associated increase in the number of drug misusers. Is cannabis likely to be any different? Pain specialists are well used to using drugs in situations for which they have not been formally evaluated and are not licensed. For the majority of their patients, they conduct single-patient trials of treatment. To stop this practice would stop pain relief services in their tracks. Therefore while we continue to use amitriptyline, Depo-Medrone, sodium valproate, clobazam, flecainide and many other drugs, unlicensed for use in pain relief, there is little justification for not also having cannabis available. Doyle and Spence see little use for cannabis in acute pain although they cite no evidence. It may have a use in the management of acute back pain with severe muscle spasm for example, as an alternative to diazepam and opioids. Could there also be a place for its use for analgesia and sedation in the ITU? Clearly from a humanitarian point of view, there is little logic in denying patients a drug that we know will relieve some or all of their pain, particularly when there is no effective alternative. This is different from Doyle and Spence’s final comment implying that cannabinoids should be studied “…in comparison with optimal conventional treatment”. To await the results of comparative trials before releasing cannabis for use in chronic pain would fail many seriously ill and dying patients. The drug will remain unavailable until the Department of Health agrees to re-classify it to schedule 2 and until a pharmaceutical company produces cannabis as a medicine in preparations that satisfy the Medicines Control Agency. Until then many patients are being denied treatment for pain that is untreatable by other methods. W. G. NOTCUTT James Paget Hospital Great Yarmouth 1. Doyle E, Spencer AA. Cannabis as a medicine. British Journal of Anaesthesia 1994; 74: 359–361.

Cardiac arrest after unrecognized dynamic inflation Sir,—I read with interest Drs Myles, Madder and Morgan’s article on cardiac arrest after unrecognized dynamic hyperinflation [1] and describe a similar case. A 59-yr-old woman with asthma of 20 years’ duration was admitted with a severe infective exacerbation leading to type 2 respiratory failure. This necessitated paralysis, tracheal intubation and ventilation of her lungs in the casualty unit, with consequent transfer to intensive care. Her arterial pressure was 150/85 mm Hg. Initial ventilation was pressure controlled to 25 cm H2O with a rate of 18 b.p.m., primarily to treat hypercapnia. Oxygenation was not a problem. By the next morning expiration was continuing up to inspiratory triggering and so expiratory time was increased. However, over the course of the following night the patient gradually and inexplicably became hypotensive. Systolic pressure, measured directly, had decreased from 110 mm Hg to 65 mm Hg while the ECG was unchanged showing sinus rhythm. It was only on disconnection from the ventilator with manual decompression of the chest and expiration of trapped gas that her arterial pressure improved dramatically. Ventilation was resumed at the lowest rate the machine would allow (6 b.p.m.) with an I : E ratio of 1 to 4. Hypercapnia was tolerated but oxygenation was still satisfactory. Similar episodes of cardiovascular compromise recurred that responded similarly such that it became necessary to instigate hourly ventilatory disconnection with manual chest decompression. At no stage did the chest x-ray show pneumothorax or sinus rhythm alter. By day 4 it was possible to dispense

British Journal of Anaesthesia with this regimen and increase ventilation to combat hypercapnia. The problem did not recur. The patient spent a further 6 weeks in the ICU and 5 weeks in the medical ward before being discharged home. Cardiovascular collapse caused by dynamic lung hyperinflation is a well documented phenomenon [2–5]. I agree with Myles, Madder and Morgan that this should be included as a cause of electromechanical dissociation in resuscitation guidelines. M. MERGER Department of Anaesthesia Bristol Royal Infirmary Bristol 1. Myles PS, Madder H, Morgan EB. Intraoperative cardiac arrest after unrecognized dynamic hyperinflation. British Journal of Anaesthesia 1995; 74: 340–342. 2. Rosengarten PL, Tuxen DV, Dziukas L, Scheinkestel C, Merret K, Bowes G. Circulatory arrest induced by intermittent positive pressure ventilation in a patient with severe asthma. Anaesthesia and Intensive Care 1991; 19: 118–121. 3. Myles PS, Weeks AM. Alpha 1-antitrypsin deficiency: circulatory arrest following induction of anaesthesia. Anaesthesia and Intensive Care 1992; 20: 358–362. 4. Martens P, Vandekerckhove Y, Mullie A. Restoration of spontaneous circulation after cessation of cardiopulmonary resuscitation. Lancet 1993; 341: 841. 5. Sprung J, Hunter K, Barnas GM, Bourke DL. Abdominal distension is not always a sign of esophageal intubation: Cardiac arrest due to “Auto-PEEP”. Anesthesia and Analgesia 1994; 78: 801–804.

Sir,—The case described by Dr Mercer is not uncommon in the intensive care unit, although it usually remains unrecognized. Modern ventilator management of airway obstruction and adult respiratory distress syndrome (ARDS) highlight the potentially adverse effects of positive pressure ventilation, such that many intensivists now accept a degree of permissive hypercapnia [1]. In fact, adverse haemodynamic effects and barotrauma cause more morbidity than a mild to moderate degree of respiratory acidosis. If respiratory acidosis is thought to be causing complications such as arrhythmias or pulmonary hypertension, then a slow infusion of i.v. bicarbonate solution can be used. We have had a large amount of experience in managing single and bilateral sequential lung transplantation in patients with severe end-stage obstructive lung disease, with less thin 20 % requiring cardiopulmonary bypass. One of the main reasons for our success has been the use of permissive hypercapnia and appreciation of the existence of dynamic hyperinflation. P. S. MYLES Alfred Hospital Melbourne Australia 1. Tuxen DV. Permissive hypercapneic ventilation. American Journal of Respiratory Critical Care Medicine 1994; 150: 870–874.

General anaesthesia for day-case cataract surgery Sir,—I read with interest the paper by Moffat and Cullen [1] in which two techniques of general anaesthesia for day-case cataract surgery were compared. They stated that “In the UK the majority of cataract operations are performed under general anaesthesia on an inpatient basis” [2], but that “It is estimated that between 20 % and 40 % of cataract patients in the UK are acceptable for day-case surgery” [3]. They also felt that, in the short term, there was little likelihood of an increasing use of local anaesthesia techniques for such procedures. Recent experience in my own unit, however, is in marked contrast with these observations. Cataract surgery has been performed routinely under local anaesthesia for the past 15 months. The local blocks are carried out by two different anaesthetists using an identical technique consisting of careful explanation, local anaesthetic eye drops, and intracone and medial

Correspondence pericone injections of 0.375 % bupivicaine, 2 % lignocaine, hyaluronidase 7.5 ␮ml⫺1 and adrenaline 1: 200,000. In the past 6 months, 121 of 124 cataract operations were carried out under local anaesthesia. The indications for general anaesthesia were patient confusion in one case and two patients requested general anaesthesia. There were no complications with any block and patient satisfaction was very high, with excellent surgical operating conditions being provided. Almost all of these patients would have been suitable for daycase surgery and it is anticipated that this will soon become the norm. The only reason why many stayed in hospital overnight appears to be that we are still in a “transition period” between general and local anaesthesia. There has been overwhelming acceptance of this technique by patients, surgeons and nursing staff, and we feel that with greater motivation from anaesthetists and surgeons, the vast majority of cataract surgery is suitable for local anaesthetic as day-cases. P. N. WHITEHEAD Department of Anaesthesia Noble’s Hospital Isle of Man 1. Moffat A, Cullen PM. Comparison of two standard techniques of general anaesthesia for day-case cataract surgery. British Journal of Anaesthesia 1995; 74: 145–148. 2. College of Ophthalmologists Audit Commission. National Cataract Study. London: College of Ophthalmologists, 1991. 3. Audit Commission for Local Authorities and NHS in England and Wales. A Short Cut to Better Services: Day Surgery in England and Wales. London: HMSO, 1990.

Design of nitric oxide delivery systems Sir,—We were interested to read Young’s article [1] on the development of a universal nitric oxide (NO) delivery system. The aim of any such delivery system is two-fold. First, to ensure a constant concentration of NO throughout the whole inspiratory phase (FINO), as high concentrations of NO are toxic. Second, to minimize the formation of highly toxic nitrogen dioxide (NO2), which is produced by the rapid reaction between NO and oxygen. The rate of this reaction is proportional to the square of the NO concentration [2], so that the concentration of NO2 formed by the delivery system is closely related to the square of the peak FINO. We suggest that in order to fully evaluate an NO delivery system it is important to both calculate the peak FINO using a carbon dioxide tracer technique [3] and to directly measure the formation of NO2. Unfortunately, Young has not presented these data for his “universal NO delivery system”. Although the “area ratio” published offers a crude measurement of the ability of NO flow to “track” the total gas flow during inspiration, it bears little relation to the peak FINO. Furthermore, the accuracy of the calculated “area ratio” must be questioned, because it is based on the measurement of NO flow by a pneumotachograph at flow rates of 0.2–0.4 litre min91. It is doubtful if the pneumotachograph can accurately measure such low flows, particularly if they vary rapidly with time. We agree that, theoretically injection of NO after the ventilator should minimize NO2 formation in an ideal system. Unfortunately, if the flow detector and valves have an inadequate dynamic response time, this results in a high peak FINO and in paradoxically high NO2 concentrations.

253 Until such systems are fully evaluated it may be safer, at present, to opt for a constant flow NO blender, such as the Nomius (Sweden) [4], and to accept the limitations of using a Siemens “Servo” ventilator with its low pressure input. The Manley ventilator, although much less sophisticated, is a true minute volume divider; this minimizes the residual volume and the gas transit time, leading to less NO2 formation than with the Servo [5]. The Nomius has been evaluated and shown to be safe with both of these ventilators. Whatever the NO delivery system, it is mandatory to monitor continuously the inspiratory concentrations of both NO and NO2. A.V. V. POWROZNYK R. D. LATIMER A. ODURO Department of Anaesthesia Papworth Hospital Cambridge 1. Young JD. A universal nitric oxide delivery system. British Journal of Anaesthesia 1994; 73: 700–702. 2. Foubert L, Fleming B, Latimer RD, Jonas M, Oduro A, Borland C, Higenbottam T. Safety guidelines for the use of nitric oxide. Lancet 1992; 339: 1615–1616. 3. Moors AH, Pickett JA, Mahmood N, Latimer RD. Evaluation of the potential dangers of nitric oxide administration when given using an intermittent flow ventilator. Journal of Cardiothoracic and Vascular Anesthesia 1994; 8 (Suppl. 3): 21. 4. Stenqvist O, Kjelltoft B, Lundin S. Evaluation of a new system for ventilatory administration of nitric oxide. Acta Anaesthesiologica Scandinavica 1993; 37: 687–691. 5. Powroznyk AW, Booth JV, Stevens JJWM, Latimer RD. Another universal nitric oxide delivery system. Association of Cardiothoracic Anaesthetists, 1994. British Journal of Intensive Care (in press). Sir,—The nitric oxide delivery system described was tested with two pneumotachographs, one to sense flow from the mass flow controller and one to sense flow from the ventilator. The pneumotachograph that Powroznyk, Latimer and Oduro are concerned about (measuring flow from the mass flow controller) has a resistance of 0.034 cm H2O ml⫺1 s, which implies that using a Validyne transducer with a 2-0-2 cm H2O sensitivity, the maximum range of flow measurement is about 0–3 litre min⫺1. Thus flows of 0.2–0.4 litre min⫺1 are measured easily, and the measurement range of the pneumotachograph/Validyne transducer combination matched the full control range of the mass flow controller. The “area” method I used does not therefore suffer from inaccuracies caused by inappropriate measuring techniques. As explained in the article and highlighted in subsequent correspondence [1], the mass flow controller I used does not have an ideal response time; it is too slow. As a result the mass flow controller tends to deliver less nitric oxide than desired, not more, and high peaks (which are characteristic of systems with a fast response) do not occur. With the capnograph in the position indicated, any “peaks” would be detected during inspiration, but none was seen. J. D. YOUNG Nuffield Department of Anaesthetics Radcliffe Infirmary Oxford 1. Gilly H, Baum M. Nitric oxide delivery system. British Journal of Anaesthesia 1995; 74: 630–631.