Monitoring of neuromuscular block after administration of vecuronium in patients with diabetes mellitus

Background. We studied the supramaximal current for ulnar nervestimulation during electromyographic monitoring of onset andrecovery of neuromuscular block using a neuromuscular transmissionmodule (M-NMT Module, Datex-Ohmeda) in patients with Type 2diabetes undergoing anaesthesia with nitrous oxide, oxygen,isoflurane and fentanyl. Methods. Thirty-six diabetic patients were randomly assignedto a post-tetanic count (PTC) group (n=17) or train-of-four(TOF) group (n=19). In addition, 30 non-diabetic patients weredivided into control PTC (n=15) and TOF groups (n=15). Results. In the diabetic patients (diabetes PTC and diabetesTOF groups), the mean supramaximal stimulating current was significantlyhigher than in the non-diabetic patients (control PTC and TOFgroups) (50.5 (SD 14.1) vs 33.4 (6.1) mA, P<0.01). Onsetof neuromuscular block (time to disappearance of T1) after vecuronium0.1 mg kg^–1 in the diabetic patients did not differ significantlyfrom that in the non-diabetic patients (276 (77) vs 244 (44)s, P=0.055). Time to return of PTC1 did not differ significantlybetween the diabetes and control PTC groups (21.0 (12.1) vs15.7 (5.0) min, P=0.126). Times to return of T1 and T4 in thediabetes TOF group were significantly longer than in the controlTOF group (T1: 37.5 (15.2) vs 25.7 (7.6) min, P=0.01; T4: 61.4(23.7) vs 43.5 (11.4) min, P=0.01). During recovery, PTC andT4/T1 in the diabetes PTC and TOF groups were similar to thosein the control PTC and TOF groups, respectively. T1/T0 in thediabetes TOF group was significantly less than in the controlTOF group, 80–120 min after vecuronium (P<0.05). Conclusions. In diabetic patients, supramaximal current is higherthan in non-diabetic patients. After vecuronium, onset of neuromuscularblock and recovery of PTC or T4/T1 are not altered, but timeto return of T1 or T4, and recovery of T1/T0 are delayed indiabetic patients. Br J Anaesth 2003; 90: 480–6     

Anaesthetist's responses to patients' self-reported drug allergies

Background. Patients with drug allergies are commonplace inanaesthetic practice. We investigated the incidence and natureof drug ‘allergies’ reported by surgical patientsattending a hospital pre-admission clinic, and went on to ascertainto what degree drug allergies recorded in the records influenceddrug prescribing during the patients’ hospital stay anddetermine whether any adverse events occurred in relation todrug prescribing in this population. Methods. Patients attending for anaesthetic assessment at aPre-Admission Clinic over a 30 week period were questioned concerningdrug allergies. Medical records of these patients were thenexamined after their hospitalization to assess medications prescribedduring that period. Results. Of 1260 patients attending the Pre-admission clinicduring the study period 420 (33.4%) claimed to have a totalof 644 individual drug ‘allergies’. The most commonagents implicated were antibiotics (n=272), opioid analgesics(n=118) and NSAIDs (n=62); the most common form of these reactionswere dermatological (n=254) and nausea and vomiting (n=124).There were 41 self-reports specifically of anaphylaxis and afurther 61 where there was significant respiratory system involvement. Conclusions. The majority of the self-reported allergies werein fact simply accepted adverse effects of the drugs concerned.The patients’ reported drug ‘allergy’ historywas generally well respected by anaesthetists and other medicalstaff. There were 13 incidents, mainly involving morphine, wherepatients were given a drug to which they had claimed a specificallergy. There were 101 incidents in 89 patients where drugsof the same pharmacological group as that of their allergicdrug were used. There were no untoward reactions in 84 patientswho had claimed a prior adverse reaction to penicillin who weregiven cephalosporins. There were no sequelae from any otherevents. While anaesthetists generally respected patients self-reported‘allergies', more attention needs to be paid to the accuraterecording of patients’ events and a clear distinctionshould be made both in medical records and to the patient betweentrue drug allergy and simple adverse drug reactions.     

Constipation and its implications in the critically ill patient

Background. Motility of the lower gut has been little studiedin intensive care patients. Method. We prospectively studied constipation in an intensivecare unit of a university hospital, and conducted a nationalsurvey to assess the generalizability of our findings. Results. Constipation occurred in 83% of the patients. Moreconstipated patients (42.5%) failed to wean from mechanicalventilation than non-constipated patients (0%), P<0.05. Themedian length of stay in intensive care and the proportion ofpatients who failed to feed enterally were greater in constipatedthan non-constipated patients (10 vs 6.5 days and 27.5 vs 12.5%,respectively (NS)). The survey found similar observations inother units. Delays in weaning from mechanical ventilation andenteral feeding were reported by 28 and 48% of the units surveyed,respectively. Conclusions. Constipation has implications for the criticallyill. Br J Anaesth 2003; 91: 815–19     

Drug patient information leaflets in anaesthesia: effect on anxiety and patient satisfaction

Background. Patient information leaflets are produced for allnew drugs, including anaesthetic drugs that are licensed solelyfor physician administration. The effect of this informationon patients’ satisfaction and anxiety has not been investigatedpreviously. Methods. Eighty-five patients were allocated randomly to receivea standard information leaflet about anaesthesia (Group 1) orthe standard leaflet plus the manufacturers’ patient informationleaflets for propofol and remifentanil (Group 2). Anxiety wasassessed using the state trait anxiety index (STAI) and a visualanalogue scale (VAS) for anxiety before and after this informationhad been read. Patients’ attitudes to this informationwere assessed by a short questionnaire. Results. There was no significant difference in STAI or VASscores for anxiety between the two groups before or after theinformation leaflets. Significantly more patients who receiveddrug patient information leaflets felt that they had receivedtoo much information (0% Group 1 vs 18% Group 2, P=0.003). Morethan 64% of patients in both groups said that they would notwish to receive detailed anaesthetic drug information. Therewas a correlation between the STAI and the VAS scores for anxiety(R=0.8). Conclusions. A minority of patients (up to 36%) wish to receivedetailed anaesthetic drug information before anaesthesia. Manufacturers’drug patient information leaflets do not alter preoperativeanxiety and may be safely issued to patients requesting suchinformation. Br J Anaesth 2004; 92: 854–8     

Double lumen tube location predicts tube malposition and hypoxaemia during one lung ventilation

Background. Poor positioning of an endobronchial double lumentube (DLT) could affect oxygenation during one lung ventilation(OLV). We set out to relate DLT position to hypoxaemia and DLTmisplacement during OLV. Methods. We recruited 152 ASA physical status I–II patientsabout to have elective thoracic surgery. The trachea was intubatedwith a left-sided DLT. Tube position was assessed by fibre-opticscope and correction was made after patient positioning andduring OLV. If Pa_O2 was less than 10.7 kPa, the DLT positionwas checked and then PEEP, continuous positive airway pressure(CPAP), oxygen insufflation, or two lung ventilation (TLV) weretried. Results. The DLT was found to be misplaced in 49 patients (32%)after patient positioning, and in 38 patients (25%) during OLV.PEEP to the dependent lung, CPAP or apneic oxygen insufflationto the non-dependent lung, or brief periods of TLV, were appliedin 46 patients (30%). Patients who had DLT malposition afterplacing the patient in the lateral position had a greater incidenceof DLT malposition during OLV (59 vs 9%) and also required eachintervention more frequently (57 vs 10%). Patients with DLTmalposition during OLV also required interventions more often(84 vs 12%). Conclusions. Patients who have DLT malposition after placingthe patient in the lateral position had more DLT malpositionduring OLV and hypoxaemia during OLV. Br J Anaesth 2004; 92: 195–201     

Preoperative use of herbal medicines: a patient survey

Background. There has been recent concern in the media overthe possible detrimental effects of herbal medicines on theperioperative period. Perceived by the public as ‘natural’and therefore safe, herbal remedies may have led to adverseevents such as myocardial infarction, bleeding, prolonged orinadequate anaesthesia and rejection of transplanted organs.In addition, herbal remedies can interact with many drugs givenin the perioperative period. In this article we summarize thepotential perioperative complications that can occur. Methods. In order to determine the extent of use of herbal medicines,we conducted a survey of patients presenting for anaesthesia.During a 3-month period, patients were directly asked by anaestheticstaff if they were currently self-administering herbal medication. Results. Of 2723 patients, 131 (4.8%) were taking one or moreherbal remedy. In only two cases was this recorded in the patients’notes. Women and patients aged 40–60 yr were mostlikely to be taking a herbal product (P<0.05 and P<0.001respectively). The most commonly used compounds were, in descendingorder, garlic, ginseng, ginkgo, St John’s wort and echinacea. Conclusion. Self-administration of herbal medicines is commonin patients presenting for anaesthesia. Because of the potentialfor side-effects and drug interactions it is important for anaesthetiststo be aware of their use. Br J Anaesth 2002; 89: 792–5     

Changes in serum albumin concentration and volume expanding effects following a bolus of albumin 20% in septic patients

Background. Patients with systemic sepsis develop a capillaryleak syndrome, and serum –albumin concentration decreases.Hyperoncotic albumin infusion can be used for volume expansionin these patients, but the degree and duration of effect arenot well described. We assessed volume expansion by albumin20% infusion and compared the retention of infused albumin inseptic patients and healthy controls. Methods. We gave albumin 20%, 200 ml as a rapid infusion to70 patients with septic shock and 26 controls. Blood sampleswere taken before and 1, 5, 15, 30, 60, 120 and 240 min afterthe infusion for measurement of serum albumin concentrationand haematocrit. Haemodilution and the percentage of administeredalbumin remaining intravascularly at each time were calculated. Results. The mean proportion of the increase in albumin remainingat 4 h was 68.5 (SD 10)% in septic patients and 79 (5)% in controls(P<0.001). The albumin 20%, 200 ml caused a secondary fluidresorption and volume expansion maximal at 30 min, equivalentto a 430 ml infusion in septic patients and 500 ml in controls. Conclusions. After giving albumin, serum albumin concentrationsdecrease significantly faster in septic patients than in healthycontrols. Br J Anaesth 2004; 92: 821–6     

Measurement of intra-abdominal pressure in intensive care units in the United Kingdom: a national postal questionnaire study

Background. To explore the attitudes of intensivists in theUK to intra-abdominal pressure (IAP) measurement and abdominalcompartment syndrome (ACS) and to determine current practice. Methods. A postal questionnaire study addressed to the leadclinician in the intensive care unit was sent to hospitals inthe UK with a general surgical service. Results. Completed questionnaires were received from 137 ofthe 207 hospitals surveyed (66.2% response rate). Only 1.5%of the respondents (n=2) had no prior knowledge of intra-abdominalhypertension and ACS. IAP had been measured on some occasionby 75.9% (n=104) of the respondents, always by the intravesicalroute. Among those intensive care units that measured IAP, in93.2% (n=97) it was only measured when there was a suspicionof the development of ACS; 3.8% of units (n=4) measured IAPon all patients who had undergone an emergency laparotomy, and2.9% (n=3) measured IAP only in those who had undergone emergencylaparotomy associated with massive fluid resuscitation. Therewas major disparity in the frequency of IAP measurement andwhen to recommend abdominal decompression. Conclusions. Despite widespread awareness of IAH and the ACS,many intensive care units never measure the IAP. When it ismeasured, the intravesical route is used exclusively. No consensusexists on optimal timing of measurement or when decompressivelaparotomy should be performed.      

Moderate hypothermia for 359 operations to clip cerebral aneurysms

Background. Experimental data have suggested that hypothermia(32–34°C) may improve outcome after cerebral ischaemia,but its efficacy has not yet been established conclusively inhumans. In this study we examined the feasibility and safetyof deliberate moderate perioperative hypothermia during operationsfor subarachnoid aneurysms. Methods. A total of 359 operations for intracranial cerebralaneurysms were included in this prospective study. By usingcold intravenous infusions (4°C) and convective coolingour aim was to reduce the patient's core temperature to morethan 34°C within 1 h before operation. The protocol assessedpostoperative complications such as infections, prolonged mechanicalventilation, pulmonary complications and coagulopathies. Results. During surgery, the body temperature was reduced toa mean of 32.5 (SD 0.4) °C. Cooling was accomplished ata rate of 4.0 (SD 0.4) °C h^–1. All patients were normothermicat 5 (SD 2) h postoperatively. Peri/postoperative complicationsincluded circulatory instability (n=36, 10%), arrhythmias (n=17,5%) coagulation abnormalities and blood transfusion (n=169,47%), infections (n=29, 8%) and pulmonary complications (infiltrateor oedema while on ventilatory support) (n=97, 27%). Eighteenpatients died within 30 days (5%). There was no significantcorrelation between the extent of hypothermia and any of thecomplications. However, there was a strong correlation betweenthe occurrence of complications and the severity of the underlyingneurological disease as assessed by the Hunt and Hess score. Conclusion. Moderate hypothermia accomplished within 1 h ofinduction of anaesthesia and maintained during surgery for subarachnoidaneurysms appears to be a safe method as far as the risks ofperi/postoperative complications such as circulatory instability,coagulation abnormalities and infections are concerned.     

Does the platelet function analyser (PFA-100) predict blood loss after cardiopulmonary bypass?

Background. This study was designed to determine if a new point-of-caretest (PFA-100^® platelet function analyser) that assessesplatelet function predicts blood loss after cardiac surgery. Methods and results. Blood samples from 70 patients were drawnbefore and after cardiopulmonary bypass (CPB) for PFA-100^®measurements. The system consists of a cartridge in which amembrane and an aperture are coated with either collagen/adenosine-5'-diphosphateor collagen/epinephrine. The instrument determines the timerequired for full occlusion of the aperture (closure time).We observed a weak correlation between pre-CPB collagen/epinephrineclosure time and second-hour mediastinal blood loss (r=0.34,P=0.01). The sensitivity and positive predictive value of thePFA-100^® measurements were comparable to platelet countfor predicting excessive bleeding after CPB (75 and 27% vs 100and 25%, respectively). Conclusions. The PFA-100^® is a logical test for detectingpatients who could have excessive bleeding after CPB. However,the PFA-100^® was not able to separate patients at low riskof subsequent bleeding from those who had substantial bleeding. Br J Anaesth 2003; 90: 692–3     

Predictive perioperative factors for developing severe sepsis after major surgery

Background. Early identification of high-risk patients undergoingmajor surgery can result in an aggressive management affectingthe outcome. Methods. We designed a prospective cohort study of 93 adultpatients undergoing major oncological surgery to identify thepredictive risk factors for developing postoperative severesepsis. Results. Nineteen of 93 patients developed a severe sepsis aftersurgery; seven of the septic patients died in intensive careunit. Multivariate analysis discriminated preoperative and postoperative(first and second day after surgery) predictive risk factors.The postoperative severe sepsis was independently associatedwith preoperative factors like male gender (OR 4.7, 95% CI between1.5 and 15.5, P<0.01) and Charlson co-morbidity index (OR1.3, 95% CI between 1.07 and 1.6, P<0.01). After the surgery,the presence of systemic inflammatory response syndrome (OR4.0, 95% CI between 1.02 and 15.7, P<0.05) and a logisticorgan dysfunction score on day 2 (OR 3.3, 95% CI between 1.9and 5.7, P<0.001) were found as independent predictive factors. Conclusion. We have shown that some of the markers that canbe easily collected in the preoperative or postoperative visitscan be used to screen the patients at high risk for developingsevere sepsis after major surgery.     

Antiemetic and analgesic-sparing effects of diphenhydramine added to morphine intravenous patient-controlled analgesia

Background. This study was designed to examine the analgesicand dose-related antiemetic efficacy of diphenhydramine–morphinemixture for intravenous patient-controlled analgesia (PCA). Methods. Healthy women, undergoing abdominal total hysterectomywere recruited to this double-blinded randomized placebo-controlledstudy. Patients were randomly allocated to one of three groups(n=40 each). In group 1, patients received saline at inductionand morphine 1 mg ml^–1 alone for postoperative PCA. Patientsin groups 2 and 3 received diphenhydramine 30 mg i.v. at inductionand were given a 1.2:1 or a 4.8:1 ratio, respectively, of diphenhydramine–morphinemixture for postoperative PCA. Results. A total of 112 patients completed the study. The incidenceof postoperative nausea (31.6% vs 67.6%, P<0.01) and vomiting(15.8% vs 40.5%, <0.05) was significantly lower in group3 than in group 1. Furthermore, the incidence of severe nauseawas significantly lower in group 3 than in group1 (2.6% vs 24.3%,P<0.05). The rescue antiemetic requirements were also significantlyless in group 3 than in group 1 (5.3% vs 24.3%, P<0.05).However, there was no significant difference between group 2and group 1 in any of the comparisons. Pain intensity, 24-hmorphine consumption and diphenhydramine-related side-effects,such as sedation or dry mouth, did not differ among the threegroups. Conclusion. An initial bolus of diphenhydramine 30 mg at anaestheticinduction followed by postoperative PCA with a 4.8:1, but not1.2:1, diphenhydramine–morphine mixture provides an effectiveantiemetic efficacy without morphine-sparing effects.     

Anaesthetic management of coronary artery bypass grafting in a patient with central core disease and susceptibility to malignant hyperthermia on statin therapy

Central core disease and malignant hyperthermia (MH) are bothassociated with mutations in the RYR1 gene. We report the anaestheticmanagement of one such patient presenting for coronary arterybypass grafting. Her medication included aspirin 75 mg, atorvastatin20 mg, isosorbide mononitrate 60 mg, atenolol 25 mg and glyceryltrinitrite sublingual spray as required. The use of aprotinin,statins and moderate hypothermia in patients with central coredisease and known susceptibility to MH has not been documented. Br J Anaesth 2003; 91: 744–7     

Endotoxaemia during left ventricular assist device insertion: relationship between risk factors and outcome

Background. Endotoxaemia, caused by splanchnic ischaemia duringsurgery, is believed to trigger systemic inflammation and causepostoperative organ dysfunction. A relationship between theplasma concentration of endotoxin during surgery and known riskfactors for postoperative morbidity and mortality (e.g. age,abnormal gastric tonometric variables) and adverse outcome aftersurgery has not been demonstrated. Methods. In a prospective study, the plasma concentration ofendotoxin was measured in 12 patients undergoing implantationof a left ventricular assist device. Automated air gastric tonometrywas performed in all patients. The relationship between plasmaendotoxin concentration, risk factors, and postoperative outcomewas explored. Results. Carbon dioxide gap increased from 0.7 (0.3) to 3.6(1.6) kPa at the end of surgery. Endotoxin was detected in oneof 12 patients at baseline and in nine of 12 patients at theend of surgery (P=0.003). A high plasma concentration of endotoxinat the end of surgery was associated with a higher carbon dioxidegap (r=0.59, P<0.05), and a higher postoperative multipleorgan dysfunction score (r=0.7, P=0.01). Conclusions. The finding of an association between high intraoperativeplasma concentrations of endotoxin, abnormal gastric tonometricvariables and adverse outcome supports the view that endotoxaemiais caused by gut hypoperfusion during surgery and is associatedwith postoperative organ dysfunction. Br J Anaesth 2004; 92: 131–3     

Survey of cancellation rate of hypertensive patients undergoing anaesthesia and elective surgery

Hypertension is the commonest avoidable medical indication forpostponing anaesthesia and surgery. There are no universallyaccepted guidelines stating the arterial pressure values atwhich anaesthesia should be postponed. The aim of this studywas to determine the extent of variation across the South-Westregion of the UK in the anaesthetic management of patients presentingwith stage 2 or stage 3 hypertension. Each anaesthetist in theregion was sent a questionnaire with five imaginary case historiesof patients with stage 2 or stage 3 hypertension. They wereasked if they would be prepared to provide anaesthesia for eachpatient. The response rate was 58%. We found great variabilitybetween anaesthetists as to which patients would be cancelled.Departmental protocols may aid general practitioners and surgeonsin the preparation of patients for surgery, but such protocolsmay be difficult to agree in the light of such a wide variationin practice. Br J Anaesth 2001; 86: 789–93     

Dysaesthesia associated with sternotomy for heart surgery

Background. Chronic pain occurs in 40–50% patients followingcardiac surgery. Dysaesthesia, either in the form of heightenedor diminished skin sensation, are frequently associated withchronic neuropathic pain. Therefore, dysaesthesia in the earlypostoperative period may predict chronic pain. However, thecharacter and causes of dysaesthesia in the early postoperativeperiod are unknown. The aim of this study was to investigatethe incidence, extent, and causes of dysaesthesia followingcardiac surgery by sternotomy. Methods. In a prospective cohort study, 50 patients undergoingsternotomy for cardiac surgery were admitted to the study: 38underwent coronary artery bypass graft (CABG), nine valve surgery,and three combined surgery. Forty-eight hours postoperatively,acute pain was measured by four-point verbal scale. Manual pinprickand cotton wool brushing was used to detect the areas of dysaesthesia. Results. Some form of dysaesthesia was found in 27 (54%) ofthe patients. Using multivariate regression analysis, the totalarea of dysaesthesia was positively associated with CABG surgeryand the severity of postoperative pain (P<0.001). Conclusion. Dysaesthesia is common in the early postoperativeperiod following cardiac surgery using a sternotomy and is associatedwith CABG surgery. The association with severity of pain mayindicate a neuropathic element that is unrelieved by conventionalopioid analgesia.      

Outcome of ASA III patients undergoing day case surgery

Background. Day case surgery is becoming more acceptable, evenfor patients with complex medical conditions. Current recommendationssuggest that patients who are graded as American Society ofAnaesthesiologists physical status (ASA) III may be suitablefor this approach. There is only a small amount of publisheddata available to support this. We present a retrospective reviewof ASA III patients who had undergone day surgical proceduresin our unit. Methods. We carried out a retrospective case controlled reviewof 896 ASA III patients who had undergone day case proceduresbetween January 1998 and June 2002 using the existing computerizedpatient information system. The system records admission rates,unplanned contact with healthcare services and post-operativecomplications in the first 24 h after discharge. Results. We demonstrated no significant differences in unplannedadmission rates, unplanned contact with health care services,or post-operative complications in the first 24 h after dischargebetween ASA III and ASA I or II patients. Conclusion. With good pre-assessment and adequate preparationASA III patients can be treated safely in the day surgery setting. Br J Anaesth 2004; 92: 71–4     

Emergence delirium in adults in the post-anaesthesia care unit

Background. Emergence delirium in the post-anaesthesia careunit (PACU) is poorly understood. The goal of this prospectivestudy was to determine frequency and risk factors of emergencedelirium in adults after general anaesthesia. Methods. In this prospective study, 1359 consecutive patientswere included. Contextual risk factors and occurrence of deliriumaccording to the Riker sedation scale were documented. Groupswere defined for the analysis according to the occurrence ornot of agitation, then after exclusion of patients with preoperativeanxiety and neuroleptics, or both, and antidepressants or benzodiazepinestreatments. Results. Sixty-four (4.7%) patients developed delirium in thePACU, which can go from thrashing to violent behaviour and removalof tubes and catheters. Preoperative anxiety was not found tobe a risk factor. Preoperative medication by benzodiazepines(OR=1.910, 95% CI=1.101–3.315, P=0.021), breast surgery(OR=5.190, 95% CI=1.422–18.947, P=0.013), abdominal surgery(OR=3.206, 95% CI=1.262–8.143, P=0.014), and long durationof surgery increased the risk of delirium (OR=1.005, 95% CI=1.002–1.008,P=0.001), while a previous history of illness and long-termtreatment by antidepressants decreased the risk (respectively,OR=0.544, 95% CI=0.315–0.939, P=0.029 and OR=0.245, 95%CI=0.084–0.710, P=0.010). Conclusions. Preoperative benzodiazepines, breast and abdominalsurgery and surgery of long duration are risk factors for emergencedelirium.     

Physiological abnormalities in early warning scores are related to mortality in adult inpatients

Background. Early warning scores using physiological measurementsmay help identify ward patients who are, or who may become,critically ill. We studied the value of abnormal physiologyscores to identify high-risk hospital patients. Methods. On a single day we recorded the following data from433 adult non-obstetric inpatients: respiratory rate, heartrate, systolic pressure, temperature, oxygen saturation, levelof consciousness, urine output for catheterized patients, ageand inspired oxygen. We also noted the care required and given. Results. Twenty-six patients (6%) died within 30 days. Theywere significantly older than survivors (P<0.001). Theirmedian hospital stay was 26 days (interquartile range 16–39).Mortality increased with the number of physiological abnormalities(P<0.001), being 0.7% with no abnormalities, 4.4% with one,9.2% with two and 21.3% with three or more. Patients receivinga lower level of care than desirable also had an increased mortality(P<0.01). Logistic regression modelling identified levelof consciousness, heart rate, age, systolic pressure and respiratoryrate as important variables in predicting outcome. Conclusions. Simple physiological observations identify high-riskhospital inpatients. Those who die are often inpatients fordays or weeks before death, allowing time for clinicians tointervene and potentially change outcome. Access to criticalcare beds could decrease mortality. Br J Anaesth 2004; 92: 882–4