Is sevoflurane replacing halothane?

British Journal of Anaesthesia 1998; 80: 123–128

CORRESPONDENCE

Is sevoflurane replacing halothane? Sir,—Mostafa and Atherton describe the successful use of sevoflurane as an induction agent in three adult patients with difficult airways.1 My practice includes anaesthesia for children with difficult airways. I recorded some details of the first 30 such children in whom anaesthesia was induced with sevoflurane when it became available in late 1995. My method is to use 8% sevoflurane in 100% oxygen or in a 2:1 nitrous oxide–oxygen mixture delivered from a T-piece by a bare hand. Sevoflurane is discontinued when the patient loses consciousness, halothane or isoflurane being introduced. Anaesthesia was induced in 30 children (15 boys, 15 girls). Mean age was 3.3 yr (range 1 month to 12 yr); 11 children were aged 3 months or less. The diagnoses are listed in table 1. Table 1 Diagnoses of children with “difficult airways” who received sevoflurane for induction of anaesthesia Cleft lip and palate Laryngeal papillomata Subglottic stenosis Laryngeal web Micrognathia Tracheocutaneous fistula Beckwith’s syndrome Choanal atresia Laryngomalacia Mandibular fracture Total

9 6 5 2 2 2 1 1 1 1 30

Mean time to loss of consciousness was 34.1 (SD 9.23; range 21–62) s. In all of these inductions, the only untoward event was one child who coughed twice. No patient suffered breath-holding or desaturation. My previous calculations2 confirm Mostafa and Atherton’s impression that sevoflurane is not especially expensive when used as an induction agent. Sevoflurane is now my routine inhalation induction agent for children with difficult or “easy” airways. I. BARKER Sheffield Children’s Hospital Western Bank Sheffield 1. Mostafa SM, Atherton AMJ. Sevoflurane for difficult tracheal intubation. British Journal of Anaesthesia 1997; 79: 392–393. 2. Barker I. The cost of volatile anaesthetic agents. British Journal of Anaesthesia 1996; 76: 749.

Propofol in paediatric intensive care Sir,—I was pleased to see another rational study of propofol in a small group of children in intensive care, together with a supportive editorial.1 2 It always seemed to me that propofol was the scapegoat for the shortcomings of paediatric intensive care practice in the UK in the late 1980s. The problem of relative lipid overloading alluded to in many of the case reports and subsequent articles may be circumvented by use of 2% propofol solution, but this was not mentioned in the current article or accompanying editorial. Surely the ethical window has re-opened to allow a properly controlled multicentre evaluation funded by the manufacturer, akin to the adult study, but preferably using the 2% solution. The FDA has recently changed its policy on paediatric studies of new drugs.3 It was interesting that the current evaluation received an exemption from the MCA in the UK and this should encourage paediatric intensivists to collaborate to produce an ethically acceptable and scientifically sound study. N. S. MORTON Department of Anaesthesia Royal Hospital for Sick Children Glasgow 1. Hatch DJ. Propofol in paediatric intensive care. British Journal of Anaesthesia 1997; 79: 274–275.

2. Martin PH, Murthy BVS, Petros AJ. Metabolic, biochemical and haemodynamic effects of infusion of propofol for long-term sedation of children undergoing intensive care. British Journal of Anaesthesia 1997; 79: 276–279. 3. Cote CJ. Unapproved use of approved drugs. Paediatric Anaesthesia 1997; 7: 91–92.

Combined spinal–extradural analgesia in labour and post-dural puncture headache Sir,—Jenkins,1 and Elton and May2 have recently commented on the incidence of post-dural puncture headache (PDPH) associated with combined spinal–extradural analgesia in labour. Jenkins suggested that the use of low-dose extradural only analgesia is preferable to a combined spinal–extradural technique because of the high incidence of PDPH (2.3%) reported by Collis and colleagues.3 Elton and May, in reply, cited the study by Norris and colleagues4 to suggest that combined spinal–extradural analgesia in labour may not be associated with an excessive incidence of PDPH. The combined spinal–extradural technique in obstetrics was used initially by Brownridge5 6 for anaesthesia for elective Caesarean section. He used a dual space technique. In an uncontrolled series of 200 women undergoing this technique, there were no cases of PDPH. Brownridge attributed this fact to extradural bolus doses of pethidine which these women received for postoperative analgesia. However, this encouraged the notion that combined spinal–extradural techniques might be associated with lower PDPH rates than either technique alone. The incidence of PDPH cited by Jenkins, namely 2.3% in the study by Collis and colleagues3 from Queen Charlotte’s Hospital, is certainly unacceptable in an obstetric regional analgesia service. However, there were particular reasons why the PDPH rate was so high in that particular study. The study was a prospective observational study of combined spinal–extradural analgesia in labour in 300 women. A single space needle-through-needle technique using 16-gauge Tuohy and 27-gauge atraumatic spinal needles was used. The inadvertent dural tap rate was 1%. There was a 10.6% failure rate in obtaining CSF with the long 27-gauge atraumatic needle. This was partly because all 300 women were in the left lateral position during insertion of the combined spinal–extradural. Norris and colleagues4 found that they were significantly more likely to obtain CSF in the sitting than in the lateral position. Sagging of the midline skin crease and reduced hydrostatic pressure head of CSF in this position may conspire to increase the likelihood of a dry tap. Furthermore, trainees with varying levels of competence and experience in any form of obstetric regional block performed all blocks in the Queen Charlotte’s Hospital study. To increase the chances of obtaining CSF, multiple deliberate dural punctures with the spinal needle were used, accounting for the excess incidence of PDPH over and above the observed inadvertent Tuohy tap rate of 1%. As experience with the combined spinal–extradural technique increased in this unit, Morgan and Kadim7 were able to report that in the next 1565 women who received combined spinal–extradural analgesia, inadvertent dural tap rate had decreased to 0.45% and the incidence of PDPH as a result of deliberate dural puncture with a 27-gauge atraumatic spinal needle had decreased to 0.13% from 1.3%; failure to obtain CSF occurred in only 5.3% of cases. The study by Norris and colleagues4 reported an apparently lower incidence of inadvertent dural puncture in women who had combined spinal–extradural analgesia rather than extradural analgesia in labour (notwithstanding that their overall inadvertent dural tap rate was 2.7%). However, that particular unit reserves extradural analgesia for later in labour and combined spinal– extradural analgesia for those who request regional analgesia earlier in labour. As distressed parturients in later labour are more likely to be moving targets than women in early labour, it is not surprising that the inadvertent dural tap rate is higher in this group. This would presumably be the case even if they received combined spinal–extradural analgesia rather than extradural only analgesia. There is no conclusive evidence that combined spinal– extradural analgesia in labour is associated with an excessive incidence of PDPH. But it would seem preferable to restrict the use of

124

British Journal of Anaesthesia

combined spinal–extradural analgesia to those women who request regional analgesia in later labour (so as to benefit from the rapidity of onset of analgesia); those who require instrumental assisted delivery in the absence of a pre-existing regional block; and those who at any time in labour are excessively distressed (so as to rapidly achieve control of the situation).

L. MCLOUGHLIN Department of Anaesthesia Queen Charlotte’s and Chelsea Hospital London 1. Jenkins JG. Ambulatory extradural analgesia in labour and mode of delivery. British Journal of Anaesthesia 1997; 78: 776– 777. 2. Elton CD, May AE. Ambulatory extradural analgesia in labour and mode of delivery. British Journal of Anaesthesia 1997; 78: 777. 3. Collis RE, Baxandall ML, Srikantharajah ID, Edge GE, Kadim KY, Morgan BM. Combined spinal epidural analgesia: technique, management and outcome of 300 mothers. International Journal of Obstetric Anesthesia 1994; 3: 75–81. 4. Norris MC, Grieco WM, Borkowski M, Leighton BL, Arkoosh VA, Huffnagle HJ, Huffnagle S. Complications of labor analgesia: Epidural versus combined spinal epidural techniques. Anesthesia and Analgesia 1994; 79: 529–537. 5. Brownridge P. Epidural and subarachnoid analgesia for elective Caesarean section. Anaesthesia 1981; 36: 70. 6. Brownridge P. Spinal anaesthesia in obstetrics. British Journal of Anaesthesia 1991; 67: 663. 7. Morgan BM, Kadim MY. Mobile regional analgesia in labour. British Journal of Obstetrics and Gynaecology 1994; 101: 839–841 Sir,—I thank Dr McLoughlin for his interest in my letter. The high rate (2.3%) of post-dural puncture headache (PDPH) in the original series of 300 combined spinal–extradurals for analgesia in labour, reported by Collis and colleagues,1 was the only published data available when we made our decision to use low-dose extradural alone rather than combined spinal–extradural analgesia. Subsequently, Morgan and Kadim2 reported a much lower rate of PDPH after combined spinal–extradural analgesia for labour. Recently, cases of both aseptic3 4 and bacterial4 5 meningitis have occurred after combined spinal–extradural analgesia in obstetrics. I believe combined spinal–extradural techniques should be used with caution for analgesia in labour.

J. G. JENKINS Department of Anaesthesia Royal Surrey County Hospital Guildford 1. Collis RE, Baxandall ML, Srikantharajah ID, Edge G, Kadim MY, Morgan BM. Combined spinal epidural analgesia with the ability to walk throughout labour. Lancet 1993; 341: 767–768. 2. Morgan BM, Kadim MY. Mobile regional analgesia in labour. British Journal of Obstetrics and Gynaecology 1994; 101: 839–841. 3. Cascio M, Heath G. Meningitis following a combined spinal–epidural technique in a labouring parturient. Canadian Journal of Anaesthesia 1996; 43: 399–402. 4. Harding SA, Collis RE, Morgan BM. Meningitis after combined spinal–extradural anaesthesia in obstetrics. British Journal of Anaesthesia 1994; 73: 545–547. 5. Aldebert S, Sleth JC. Meningite bacterienne apres anesthesie rachidienne et peridurale combinee en obstetrique (Bacterial meningitis after combined spinal and epidural anaesthesia in obstetrics). Annales Francaises d’Anesthesie et de Reanimation 1996; 15: 687–688. Sir,—We agree that the evidence that combined spinal–extradural analgesia increases the incidence of post-dural puncture headache (PDPH) is conflicting. This expands the assertion in our original letter.1 Combined spinal–extradural analgesia has proved extremely popular in centres in which it has been introduced, including Queen Charlotte’s Hospital.2 It has been suggested that conventional extradurals with bolus doses of plain bupivacaine are associated with increased operative3 and instrumental delivery and long-term backache.4 It is possible that combined spinal– extradural analgesia in labour has a different profile of side effects.5 It would be prudent to await the results of large scale, randomized controlled studies with long-term follow-up to establish the role of combined spinal–extradural analgesia in labour. We

agree that where it is necessary to achieve rapid onset of profound analgesia, particularly where pain is of sacral origin, combined spinal–extradural analgesia has advantages.6 However, the fact that anaesthetists at Queen Charlotte’s Hospital experienced a greater incidence of side effects when combined spinal–extradural analgesia was introduced compared with subsequent studies2 would imply that the use of an unfamiliar technique in such fraught situations may not be desirable.

C. D. ELTON A. E. MAY Department of Anaesthesia Leicester Royal Infirmary Leicester 1. Elton CD, May AE. Ambulatory extradural analgesia in labour and mode of delivery. British Journal of Anaesthesia 1997; 78: 777. 2. Morgan BM, Kadim MY. Mobile regional analgesia in labour. British Journal of Obstetrics and Gynaecology 1994; 101: 839–841. 3. Thorp JA, Hu DH, Albin RM, McNitt J, Meyer BA, Cohen GR, Yeast JD. The effect of intrapartum epidural analgesia on nulliparous labor: A randomised, controlled prospective trial. American Journal of Obstetrics and Gynecology 1993; 169: 851–858. 4. MacArthur C, Lewis M, Knox EG, Crawford JS. Epidural anaesthesia and long-term backache after childbirth. British Medical Journal 1990; 301: 9–12. 5. May AE, Elton CD. Ambulatory extradural analgesia in labour reduces the risk of Caesarean section. British Journal of Anaesthesia 1996; 77: 692–693. 6. Stacey RGW, Watt S, Kadim MY, Morgan BM. Single space combined spinal–extradural technique for analgesia in late labour. 1993; 71: 499–502.

Transposition of rotameter tubes Sir,—With the publication of the new checklist for anaesthetic apparatus from the Association of Anaesthetists,1 we feel that the incident reported in this letter will highlight the importance of these checks. However, in addition, having read carefully this document we can find no instruction which would ensure that the fault we describe would be detected in the future. There were two reports of a rotameter and a bobbin swap from the USA in 1974.2 3 However, there has previously been no report in this country of such an incident. But the morning after an anaesthetic machine service in this hospital, a full check of the anaesthetic machines in the anaesthetic room and operating theatre was made according to the guidelines of the Association of Anaesthetists of Great Britain and Ireland (AAGBI)4 by the anaesthetist who was working in that theatre. Nothing untoward was noted at this time. Anaesthesia was carried out safely in the first patient in the anaesthetic room before transfer to theatre. As soon as the anaesthetist turned on the gas flows on the machine in the operating theatre he realized that the flowmeter above the nitrous oxide flow control valve was only calibrated to 2 litre. On closer inspection, this flowmeter was noted to be colour coded and labelled for carbon dioxide. The flowmeter above the carbon dioxide flow control valve was conversely colour coded and labelled for nitrous oxide (fig. 1). The anaesthetist administered 100% oxygen and a volatile agent to the patient while the anaesthetic machine was changed. The patient came to no harm. How did the flowmeters get swapped? It is standard practice for the service engineer to remove, clean and replace each flowmeter in turn. On this occasion one of the flowmeters was broken on cleaning and then the next was removed, cleaned and replaced before a replacement for the broken one was found. This led to the swap. Why was this swap not detected during the anaesthetic machine check? The checklist for anaesthetic machines4 states to “set a flow of approximately 5 litre min91” for the flowmeters. The new trainee anaesthetist knows that this is approximately 50–66% up the oxygen and nitrous oxide flowmeter, without reading the scales. Therefore, the flowmeters were not read accurately to 5 litre during this check. Extra vigilance is needed in checking the anaesthetic machine to exclude a crossover of flowmeters after a machine service. How might such a potentially serious fault be avoided in future? Flowmeter design could be changed so that it would not be possible for them to be interchanged. Clearly the procedures being followed by the service engineer were not foolproof. Perhaps after each anaesthetic machine service the

Correspondence

125 4. Checklist for anaesthetic machines. London: Association of Anaesthetists of Great Britain and Ireland, 1990.

Effect of hydroxyethyl starch on coagulation is difficult to asses in vitro

Figure 1 Flowmeters of the anaesthetic machine. machine should be double checked by an independent technician. A much more comprehensive checklist than the one currently described by the AAGBI is possibly required after a machine service. Currently, we reinforce to our trainees that they must set the bobbin height to exactly 5 litre when doing their anaesthetic machine checks to guarantee that the flowmeter calibrations are read accurately. The second question to be asked is what difference to the outlet gas mixture would a switch of this sort make? The two possibilities are of an hypoxic mixture (picked up by the oxygen analyser) or an oxygen-enriched gas mixture leading to the possibility of awareness. Afterwards, oxygen and nitrous oxide concentrations from the faulty machine were measured, setting both flowmeters to read 2 litre as this was the maximum the “nitrous oxide” flowmeter could read. Equal concentrations of each gas were emerging from the common gas outlet. If the flowmeters are calibrated for individual gases, why was the flow of nitrous oxide through a carbon dioxide flowmeter accurate? With the bobbin type of flowmeter, the physical property determining flow is gas viscosity at low flow rates and gas density at high flow rates. The viscosities of carbon dioxide and nitrous oxide are 13.9 and 13.5 Pa.s, respectively. The density of both is 1.98 kg m93. In summary, although service procedures are set and followed by companies, mistakes can occur. No checklist can ever be foolproof. Extra vigilance in checking anaesthetic machines is strongly recommended, particularly after servicing. The importance of following the AAGBI checklist for anaesthetic apparatus1 cannot be overemphasized. However, a more comprehensive checklist may be required after an anaesthetic machine service.

A. J. WALMSLEY J. HOLLOWAY Eastbourne District General Hospital Eastbourne E. Sussex 1. Checklist for anaesthetic apparatus. London: Association of Anaesthetists of Great Britain and Ireland, 1997. 2. Chadwick DA. Transposition of rotameter tubes. Anesthesiology 1974; 40: 102 3. Slater EM. Transposition of rotameter bobbins. Anesthesiology 1974; 41: 101

Sir,—We have read with interest the in vitro studies carried out by Egli and colleagues on the effect of haemodilution with hydroxyethyl starch (HES), gelatin and albumin on blood coagulation.1 However, we cannot agree completely with the conclusion that HES compromises blood coagulation more than other substances because in vitro studies of the effect of HES on blood coagulation are, unfortunately, of limited value. HES leads to coagulation disorders, particularly through impairment of factor VIII/von Willebrand factor. In 1985, Stump and colleagues2 were able to show that this effect only occurs in vivo, not in vitro. They speculated that the reason for this could be either reduced release or impairment of the synthesis of factor VIII/von Willebrand factor. Our own studies support the hypothesis that accelerated elimination of factor VIII/von Willebrand factor is the cause for the observed coagulation disorders.3 It is difficult, therefore, to simulate through in vitro studies the complex interaction of HES with the clotting system and in particular elimination via the reticuloendothelial system. These difficulties are enhanced by the fact that the chemical composition of HES changes markedly in vivo. In recent studies, we were able to show that HES is metabolized in vivo in complex ways,4 5 and that impairment of factor VIII/von Willebrand factor depends exclusively on the in vivo molecular weight of HES. This explains why haemorrhagic complications have almost exclusively been observed with high and medium molecular weight HES, which is not easily metabolized by the body, resulting in high in vivo molecular weights of the HES molecules. HES with a medium or low molecular weight which is easily metabolized does not lead to a reduction in factor VIII/von Willebrand factor beyond the dilution effect, and rarely leads to haemorrhagic complications.6 In vitro studies do not take into account this new finding, because HES is not metabolized in vitro, resulting in unrealistically high molecular weights of the starch molecules which are not observed in vivo. Nevertheless, we believe that the experimental model of Egli and colleagues is innovative and interesting, particularly as the pathological mechanism of impairment of the clotting system through HES and the mechanism of the reduction in factor VIII/ von Willebrand factor have not yet been elucidated. Future in vitro studies that incorporate the recent results should be able to yield interesting new insights.

M. T. GRAUER J. TREIB Neurologische Universitatsklinik Homburg/Saar Germany 1. Egli GA, Zollinger A, Seifert B, Popovic D, Pasch T, Spahn DR. Effect of progressive haemodilution with hydroxyethyl starch, gelatin and albumin on blood coagulation. British Journal of Anaesthesia 1997; 78: 684–689. 2. Stump DC, Strauss RG, Henriksen RA, Petersen RE, Saunders R. Effect of hydroxyethyl starch on blood coagulation, particularly factor VIII. Transfusion 1985; 25: 349–354. 3. Treib J, Haass A, Pindur G, Miyachita C, Grauer MT, Jung F, Wenzel E, Schimrigk K. Highly substituted hydroxyethyl starch (HES 200/0.62) leads to a type I von Willebrand syndrome after repeated administration. Haemostasis 1996; 26: 210–213. 4. Treib J, Haass A, Pindur G, Grauer MT, Wenzel E, Schimrigk K. All medium starches are not the same: Influence of degree of substitution of hydroxyethyl starch on volume effect, hemorheologic conditions and coagulation. Transfusion 1996; 36: 450–455. 5. Treib J, Haass A, Pindur G, Seyfert UT, Treib W, Grauer MT, Jung F, Wenzel E, Schimrigk K. HES 200/0,5 is not HES 200/ 0,5. Influence of the C2/C6 hydroxyethylation ratio of hydroxyethyl starch (HES) on haemorheology, coagulation and elimination kinetics. Thrombosis and Haemostasis 1995; 74: 1452–1456. 6. Treib J, Haass A, Pindur G. Coagulation disorders caused by hydroxyethyl starch. Thrombosis and Haemostasis 1997; 78: 974–983.

126

British Journal of Anaesthesia

Sir,—Drs Grauer and Treib were concerned that the in vitro finding of an exaggerated effect on blood coagulation of hydroxyethyl starch (200 000/0.5) (HES) compared with gelatin and albumin solutions1 would not be completely applicable to the in vivo situation. This concern was based on two findings: first, an apparent lack of a decrease of factor VIII/von Willebrand in vitro with impairment of blood coagulation2; and second, on the observation that the mean in vivo molecular weight of HES decreased during long-term haemodilution in the treatment of cerebrovascular disease with an associated decrease in blood coagulation.3 Stump and colleagues described stable VIII/von Willebrand factors in their in vitro experiments assessing mixtures of HES and plasma.2 These results were achieved in citrated plasma to prevent clotting. Interestingly, when blood coagulation is not inhibited, as in our study using thrombelastography, HES and dextran solutions may precipitate coagulation factors such as factor VIII/von Willebrand in vitro also.4 The effect of HES in decreasing factor VIII/von Willebrand thus would not have been missed in our in vitro studies. In addition, if we were only able to partially detect the effect of a decrease in factor VIII/von Willebrand in our in vitro study, we would have underestimated the in vivo difference between the blood coagulation effect of HES compared with albumin and gelatin solutions. Thus we cannot agree with the concerns of Grauer and Treib that our results would not substantiate a more exaggerated effect on blood coagulation of HES than albumin or gelatin solutions. Treib, Haass and Pindur have recently suggested an interesting new concept to predict the effect of long-term haemodilution therapy using various types of HES on blood coagulation.3 In patients treated for cerebrovascular disease, they found that the mean molecular weight of HES solutions decreased after infusion in vivo and that the effect on blood coagulation of these different HES solutions was better related to the mean in vivo molecular weight measured after 5–10 days than to the mean in vitro molecular weight (i.e. the mean molecular weight of the HES solution in the bottle before infusion).3 During surgery, however, HES is administered for relatively short periods of time according to surgical blood loss and the volume replacement requirements of the patient. Thus additional HES solution of the original in vitro molecular weight is infused continually and therefore HES molecules with a molecular weight similar to in vitro molecular weight exert their effect on blood coagulation. The concept of the (long-term) in vivo molecular weight of a particular HES solution affecting blood coagulation is of limited clinical relevance for the intraoperative and early postoperative period. In vitro testing of the effect of various volume expanders on blood coagulation using thrombelastography is considered adequate. Nevertheless, in vitro findings have to be confirmed in vivo before final conclusions regarding clinical management can be deduced.

G. A. EGLI A. ZOLLINGER T. PASCH D. R. SPAHN Institute of Anaesthesiology University Hospital Zürich Switzerland 1. Egli GA, Zollinger A, Seifert B, Popovic D, Pasch T, Spahn DR. Effect of progressive haemodilution with hydroxyethyl starch, gelatin and albumin on blood coagulation. British Journal of Anaesthesia 1997; 78: 684–689. 2. Stump DC, Strauss RG, Henriksen RA, Petersen RE, Saunders R. Effects of hydroxyethyl starch on blood coagulation, particularly factor VIII. Transfusion 1985; 25: 349–354. 3. Treib J, Haass A, Pindur G. Coagulation disorders caused by hydroxyethyl starch. Thrombosis and Haemostasis 1997; 78: 974–983. 4. Alexander B. Effects of plasma expanders on coagulation and hemostasis: dextran, hydroxyethyl starch, and other macromolecules revisited. Progress in Clinical and Biological Research 1978; 19: 293–330.

Haemostatic changes during total knee arthroplasty with different anaesthetics Sir,—The study by Sharrock and colleagues1 on haemostatic changes in total knee arthroplasty with different anaesthetic techniques reflects the continuing search for a better understanding of

the mechanisms of postoperative deep vein thrombosis in major joint arthroplasty. There are three comments I would like to make. While they are partially correct in their discussion that in total hip arthroplasty we did not confirm using spinal anaesthesia Modig and colleagues’ findings of suppressed fibrinolysis with extradural block,2 this was largely a reflection of haematological methodology. There were demonstrable differences between general anaesthesia and spinal anaesthesia in some measured haemostatic variables.3 When comparing spinal anaesthesia with general anaesthesia in the immediate perioperative period, platelet count did not increase to the same extent, thrombin generation index (a measure of activation of coagulation) and factor VIIIRAg did not increase and there was less activation of fibrinolysis (as demonstrated by euglobulin clot lysis time). These changes were consistent with suppression of the neuroendocrine response to surgery with spinal anaesthesia, as demonstrated by changes in plasma concentrations of cortisol in our patients4 and were similar in the general anaesthesia group to those seen with exercise and exogenous infusions of epinephrine.5 6 In our study, no exogenous catecholamines were used and it is possible that any differences between general and extradural anaesthesia in the study of Sharrock and colleagues were hidden by the use of epinephrine in the extradural local anaesthetic solution and i.v. ephedrine for cardiovascular control. With respect to their comments on limb blood flow, based on our impedance plethysmography study during hip arthroplasty,7 the enhanced blood flow seen in the surgical limb after relocation of the prosthesis and into the postoperative period with spinal anaesthesia compared with general anaesthesia may not necessarily apply to knee arthroplasty. In the latter case, maximal vasodilatation would be expected in the exsanguinated limb after tourniquet release, at least initially, irrespective of the anaesthetic technique used. It will be difficult to follow through this hypothesis with flow studies on patients undergoing knee arthroplasty. Given the New York group’s very large clinical experience in knee arthroplasty, I would be interested in their views on correct exsanguination of the limb before tourniquet inflation and its importance for the occurrence of deep vein thrombosis. Not all orthopaedic surgeons use a compression Esmarch bandage, and when they do so it is not always applied efficiently. This could well influence coagulation–fibrinolysis changes in the surgical limb.

M. DAVIS Tai Tapu New Zealand 1. Sharrock NE, Go G, Williams-Russo P, Haas SB, Harpel PC. Comparison of extradural and general anaesthesia on the fibrinolytic response to total knee arthroplasty. British Journal of Anaesthesia 1997; 79: 29–34. 2. Modig J, Borg T, Bagge L, Saldeen T. Role of extradural and of general anaesthesia in fibrinolysis and coagulation after total hip replacement. British Journal of Anaesthesia 1983; 55: 625–629. 3. Davis FM, McDermott E, Hickton C, Wells E, Heaton DC, Laurenson VG, Gillespie WJ, Foate J. Influence of spinal and general anaesthesia on haemostasis during total hip arthroplasty. British Journal of Anaesthesia 1987; 59: 561–571. 4. Davis FM, Laurenson VG, Lewis J, Wells JE, Gillespie WJ. Metabolic response to total hip arthroplasty under hypobaric subarachnoid or general anaesthesia. British Journal of Anaesthesia 1987; 59: 725–729. 5. Cash JD, Allan AGE. The fibrinolytic response to moderate exercise and intravenous adrenaline in the same subjects. British Journal of Haematology 1967; 13: 376–383. 6. Small M, Twedell AC, Rankin AC, Lowe GD, Prentice CR, Forbes CD. Blood coagulation and platelet function following maximal exercise: effects of beta-adrenoceptor blockade. Haemostasis 1984; 14: 262–268. 7. Davis FM, Laurenson VG, Gillespie WJ, Foate J, Seagar AD. Leg blood flow during total hip replacement under spinal or general anaesthesia. Anaesthesia and Intensive Care 1989; 17: 136–142. Sir,—We appreciate Dr Davis’ comments on our article, as he was perhaps the first person to highlight the role of regional anaesthesia in the genesis of venous thrombosis.1 The first question relates to his observation of increased thrombin generation during operation and decreased fibrinolysis after operation in patients receiving general anaesthesia during total hip replacement (THR).2 We noted no increase in thrombin generation (measured by thrombin–antithrombin complexes) or in

Correspondence

127

the effect of thrombin formed (fibrinopeptide A), nor did we note any difference in fibrinolysis, although our study did not extend into the postoperative period. These differences most likely reflect the different operations and possibly the differences in blood loss between spinal and general anaesthesia during THR. There is no effect of anaesthetic technique on blood loss with total knee replacement (TKR). The issue of the use of catecholamines during anaesthesia is an important point. Although epinephrine and exercise increase fibrinolysis in young people, we were unable to detect any augmentation during THR in elderly patients.3 In contrast, we have noted that infusions of epinephrine are associated with less thrombogenicity during TKR performed with a tourniquet.4 These are preliminary observations and whether or not this represents a suppressant effect of epinephrine on coagulation or a vascular effect is unclear. With a tourniquet, there is little opportunity for extradural anaesthesia to augment flow in the operative limb and thereby alter venous thrombotic tendencies. This was one of the reasons we chose to study TKR, as flow should not be a confounding variable. As we were unable to see any difference in fibrinolysis or markers of thrombin generation in our patients, we hypothesized that augmentation of flow after surgery may account for the decreased rate of deep vein thrombosis (DVT) noted with extradural anaesthesia after TKR. Whether or not this is true is unknown. The final point relates to how well the leg is exsanguinated before tourniquet inflation. I suspect this is a significant determinant of the degree of venous thrombosis. Any blood remaining in the leg will be static and a potential nidus for DVT. For this reason, our surgeons assiduously attempt to exsanguinate the leg with an Esmarch bandage before inflating the tourniquet. However, to the best of my knowledge, this has not been studied. I thank Dr Davis for his comments, as a fuller understanding of factors contributing to intraoperative thrombosis will help reduce morbidity after surgery.

N. E. SHARROCK Department of Anesthesiology Hospital for Special Surgery New York NY USA 1. Davis FM, Quince M, Laurenson VG. Deep vein thrombosis and anaesthetic technique in emergency hip surgery. British Medical Journal 1980; 281: 1528–1529. 2. Davis FM, McDermott E, Hickton C, Wells E, Heaton DC, Laurenson VG, Gillespie WJ, Foate J. Influence of spinal and general anaesthesia on haemostasis during total hip arthroplasty. British Journal of Anaesthesia 1987; 59: 561–571. 3. Sharrock NE, Go G, Mineo R, Harpel PC. The haemodynamic and fibrinolytic response to low dose epinephrine and phenylephrine infusions during total hip replacement under epidural anaesthesia. Thrombosis and Haemostasis 1992; 68: 436–441. 4. Sharrock NE, Go G, Sculco TP, Harpel PC. The effect of duration of tourniquet inflation on markers of thrombosis during total knee arthroplasty. Thrombosis and Haemostatis 1997; (Suppl. 608): PS-2478.

Extradural anaesthesia in elective Caesarean section Sir,—We read with interest the article by Morton and colleagues1 and would like to raise several points. First, we were concerned that the results of nine women from a study containing only 38 women were not valid for analysis of efficacy or pharmacokinetics, or both. This is a drop out rate of approximately 25%. Second, one patient who apparently had sensory block to T6 after receiving 0.75% ropivacaine 20 ml and increments of 2% lidocaine had spinal anaesthesia for Caesarean section because surgery was delayed. Analysis of table 3 shows that in this patient, surgery started 114 min after extradural injection. Analysis of table 2 shows that the median duration of sensory block at T6 was 3.4 h (range 1.3–5.8 h). Although the results of this patient were not included in table 2, this patient should still have had a significant sensory block. Why was this patient then given a spinal anaesthetic? If the height of the sensory block had receded it could easily have been extended by injecting more local anaesthetic via the extradural catheter. Furthermore, there are several reports in the literature where very high or total spinal blocks have occurred when spinal anaesthesia has been performed after extradural anaesthesia.2–4

Third, two mothers received accidental i.v. injection of ropivacaine 75 mg and 150 mg, respectively. The second of these patients developed mild signs of i.v. injection after a dose of 75 mg but symptoms resolved, and therefore the authors gave another 75 mg at which time the symptoms recurred. Surely the authors should have stopped injecting after the initial 75 mg, aspirated the extradural catheter and waited a few minutes to see if any signs of sympathetic, sensory or motor block developed. If no signs of extradural block had developed the catheter should have been assumed to have been placed i.v. Furthermore, were there any signs of i.v. catheterization at the time of siting and were the catheters aspirated before injecting the test dose? It is fortunate that the study was not double-blind with bupivacaine! Fourth, the authors failed to mention which form of anaesthetic was administered to the mother who received ropivacaine 150 mg i.v. The authors did not give the results of neonatal outcome in the two mothers who had accidental i.v. injection. Were their NACS scores normal? Finally, as the authors mentioned in their discussion, the open design of the study had limitations, namely it prevented true comparisons with other agents or even with 0.5% ropivacaine. The conclusions of the study were somewhat predictable in that 0.75% ropivacaine should provide better analgesia compared with 0.5% ropivacaine or bupivacaine.5 However, what is interesting is that 21 of 31 neonates in this study had a 2-h NACS score of less than 35 while in the study of Griffin and Reynolds,5 where 0.5% ropivacaine was used, only five of 31 neonates had a 2-h NACS score of less than 35.

M. PATEL A. BAXTER Department of Anaesthetics Royal Women’s Hospital Melbourne, Australia 1. Morton CPJ, Bloomfield S, Magnusson A, Jozwiak H, McClure JH. Ropivacaine 0.75% for extradural anaesthesia in elective Caesarean section: an open clinical and pharmacokinetic study in mother and neonate. British Journal of Anaesthesia 1997; 79: 3–8. 2. Beck GN, Griffiths AG. Failed extradural anaesthesia for Caesarean section. Complications of subsequent spinal block. Anaesthesia 1992; 47: 690–692. 3. Bets B, Broccoli E, Brown AR. Is spinal anesthesia after failed epidural anesthesia contraindicated for Cesarean section? Anesthesia and Analgesia 1993; 77: 629–631. 4. Stone PA, Thorburn JA, Lamb KSR. Complications of spinal anaesthesia following extradural block for Caesarean section. British Journal of Anaesthesia 1989; 62: 335–337. 5. Griffin RP, Reynolds F. Extradural anaesthesia for Caesarean section: double blind comparison of 0.5% ropivacaine with 0.5% bupivacaine. British Journal of Anaesthesia 1995; 74: 512– 516. Sir,—We thank Drs Patel and Baxter for their interest in our article. The reasons for the results from seven women being invalid for efficacy analysis were technical difficulty in siting the extradural catheter or failure to establish sensory block. Blood samples from another two women (with satisfactory extradural blocks) were not available for analysis for reasons stated in the article. Many authors do not report the fate of patients who are excluded from a study after recruitment (i.e. after giving informed consent) but we believe it is good scientific practice to do so. Furthermore, because ropivacaine was then an unlicensed drug, we felt it important to report the results from all patients who received it, including four of the above seven patients. Drs Patel and Baxter question our choice of subarachnoid anaesthesia for the woman whose Caesarean section was delayed, inferring from the pooled data that she should either have had adequate block or block could have been extended. First, a median duration of sensory block at T6 of 3.4 h does not imply that this patient’s block would have been adequate at the time of surgery. Second, we do not agree that her block could have easily been extended. The reason that she had received additional local anaesthetic was because of inadequate block. In addition, we were concerned about the possibility of systemic local anaesthetic toxicity. We are aware of the reports of high block when subarachnoid anaesthesia is used after inadequate extradural block but felt that the balance of risks still favoured this technique. The second of the two mothers who received accidental i.v. injections of ropivacaine underwent Caesarean section under subarachnoid block. She received 0.75% ropivacaine 20 ml because

128

British Journal of Anaesthesia

her symptoms (not signs) were inconclusive. The injection was given slowly, which we believe is vital for safety, and therefore even if bupivacaine had been used, warning symptoms should have occurred before serious systemic toxicity developed. All extradural catheters were placed and all clinical decisions made by the consultant in charge of the patient’s anaesthetic care. This person was not the investigator. As stated in the article, all catheters were aspirated. NAC scores at 2 h and 24 h for the two women who received accidental i.v. injections were 30, 31 and 33, 36, respectively. None of the neonates in the study gave clinical cause for concern except that, as discussed in the article, one developed transient tachypnoea of the newborn.

C. P. J. MORTON J. H. MCCLURE S. BLOOMFIELD Department of Anaesthetics Royal Infirmary of Edinburgh NHS Trust Edinburgh

The expanding role of simulators in risk management Sir,—We read with interest the recent correspondence by Dr Byrne and Professor Jones1 in response to Professor Spences’ editorial2 on the expanding role of simulators in risk management. We are also aware of the excellent work that is being done with the ACCESS simulator. We admit that we are fortunate in Bristol to have acquired one of the two commercially available American high-fidelity simulators. The METI version that we purchased cost approximately $200 000. The rest of the project money was used to house the simulator in a purpose-built education centre with additional computer-assisted learning (CAL) and conference facilities. The METI high-fidelity simulator can be used by single-handed operators for individual or group tutorial style teaching. It can also be used for the more glamorous scenario-based “anaesthesia crisis

resource management” training. It is true that this type of training is time consuming, requires many trainers and benefits only a few trainees at a time. In Bristol, we are encouraging the involvement of committed medical teachers (not just anaesthetists) to use our centre. Medical students and trainee anaesthetists attend the centre for tuition on a regular basis. Geographical location does mean that those in the South West of England are more likely to benefit from informal teaching programmes but we hope the commercial courses will appeal to many around the country. The manikin is immobile at present but it is our intent that teaching and training by simulation be available to a wider audience in the future. Pilot projects in the USA involve moving the simulator from one hospital base to another. This is potentially damaging to expensive equipment but mobile units may be available in the future. Alternatively, telemedicine links may provide opportunities for simulation teaching to a wider audience. If high-fidelity simulators are to be used for trainee assessment and continuing medical education in the future, several points need to be considered. First, access to simulator facilities will need to be readily available. Second, trainers using simulators will need to cooperate and develop teaching and assessment programmes together, and third, the benefits of simulation as a teaching tool or in improving clinical performance need to be validated. We are sure that the future success and developments of simulation lies in the collaborative efforts of those working in this exciting field of education.

F. FORREST S. MATHER M. TOOLEY Bristol Medical Simulation Centre Management Board Bristol 1. Byrne AJ, Jones JG. The expanding role of simulators in risk management. British Journal of Anaesthesia 1997; 79: 411. 2. Spence AA. The expanding role of simulators in risk management. British Journal of Anaesthesia 1997; 78: 633–634.