respiration, and no patient experienced oxygen desaturation to,92%. Adequate light sedation before awake insertion of theILMA was achieved with total midazolam and fentanyl dosesranging from 3 to 6 [mean 4.4 (0.8)] and 0.1 to 0.3 [mean 0.16(0.6)] mg, respectively.Nopatient recalledexperiencingdiscom-fort during the procedure when questioned after operation.

We obtained good results with awake insertion of a size 3.5air-QTM ILMA device followed by tracheal intubation usingthe device as a conduit in morbidly obese patients (n20)undergoing bariatric surgery. This ILMA device is designed foreasier insertion. It has a curvature approximate to that of theupper oropharyngeal airway and a wider (anteriorposteriordiameter15 mm) and shorter airway conduit than previousmodels.4 5 It has an easily removable airway adapter with nogrill in the ventilating orifice, which may further facilitate inser-tion and placement (Fig. 1C).

In conclusion, the technique we describe may be a viable al-ternative to mask ventilation and direct laryngoscopy for safeairway management in morbidly obese patients. Furtherstudies and detailed comparison with results of other techni-ques may be warranted.

Declaration of interestNone declared.

T. Shiraishi*Tokyo, Japan*E-mail: shiraishi@mcube.jp

1 Brodsky J, Lemmens H, Brock-Utne J, Vierra M, Saidman LJ. Morbidobesity and tracheal intubation. Anesth Analg 2002; 94: 7326

2 Ferson D, Rosenblatt W, Johnasen M, Osborn I, Ovassapian O. Use ofthe intubating LMA FastrachTM in 254 patients with difficult-to-manage airways. Anesthesiology 2001; 95: 117581

3 Brodsky JB, Lemmens HJ, Brock-Utne JG, Saidman LJ, Levitan R.Anaesthetic considerations for bariatric surgery. Proper positioningis important for laryngoscopy. Anaesth Analg 2003; 96: 18412

4 Product study: air-Q. Neonatal Intensive Care 2011; 24: 3940

5 Liu E, Goy R, Chen F. The LMA CtrachTM, a new laryngeal mask airwayfor endotracheal intubation under vision: evaluation in 100 patients.Br J Anaesth 2006; 96: 396400

doi:10.1093/bja/aet389

Bradycardia after dexamethasone forpostoperative nausea and vomitingprophylaxis during induction of anaesthesiaEditorWe report a case of sinus bradycardia after a singledose of i.v. dexamethasone for postoperative nausea andvomiting (PONV) prophylaxis during anaesthesia induction.

A 51-yr-old woman, ASA II, undergoing elective spinesurgery due to a protruding intervertebral disk was broughtto the anaesthesia induction room. She had mild arterialhypertension and non-active rheumatoid arthritis for whichshe received no therapy. She reported allergies to nickel andformaldehyde. Her BMI was 35 and she was a non-smoker

with a history of PONVafter discectomy in the past. Monitoringof the patient included ECG, non-invasive arterial pressure, andpulse oximetry. Her vital signs were: heart rate 8085 beatsmin21, arterial pressure 170/90 mm Hg, SaO2 95% on breathingroom air. An i.v. line was inserted and infusion with Ringerslactate 500 ml was initiated.

Owing to the high risk of PONV,1 a prophylactic dose of i.v.dexamethasone 4 mg was given at anaesthesia induction. Oneminute later, the patients heart rate decreased to 40 beatsmin21 and she felt drowsy. No dose of benzodiazepine or opioidhad been given at this point. I.V. atropine 0.5 mg resulted in nochange in heart rate. A second dose of i.v. atropine 0.5 mg wasgiven 5 min later. The heart rate slowly increased to 8085beats min21. Anaesthesia was induced with i.v. fentanyl 0.15mg, propofol 2 mg kg21, and rocuronium 0.6 mg kg21 and thetrachea was intubated. Anaesthesia was maintained with a com-bination of the volatile anesthetic sevoflurane and intermittentdoses of i.v. fentanyl 0.050.1 mg. No clinically relevantchanges in heart rate or arterial pressure beyond the 20% rangeof the initial values were observed. The course of anaesthesia, op-eration,andthefollowingrecoverywereuneventful.Echocardiog-raphy examination on the first postoperative day showed nocardiac abnormality. The patient was discharged in goodgeneral condition.

The present report might be the first case to draw attentionto potentially serious side-effects of a single-dose of i.v. dexa-methasone. Dexamethasone has been routinely used forprophylaxis and treatment of PONV2 and the quantitative sys-tematic review of Henzi and colleagues3 showedthat a single ap-plicationdidnotseemtoprovokeanyoftheknownside-effectsofcorticoids. These include cardiac arrhythmias47 and evensudden death.8 9 The authors of the cited review point out,however, that they still do not know if a single bolus dose ofdexamethasone 8 or 10 mg is safe in patients at risk ofcorticosteroid-related adverse effects.3

As described in the literature, most of the patients experien-cing complications after i.v. application of glucocorticoids wereeither adults with autoimmune and rheumatic diseases orpremature infants. Our patient had a history of rheumatoiddisease. Furthermore, some authors suggest that high-dosemethylprednisolone may be contraindicated in patients withknown heart disease.7 There are no data confirming whetherthis suggestion relates to the use of dexamethasone.

The preservatives used in the drug preparation might also beacontributing factor.10 Our patient reported allergic reaction toformaldehyde. Although cross-reactivity between formalde-hyde and preservative substances cannot be excluded, ourpatient showed no symptoms of anaphylactic reaction.

Undoubtedly, serious side-effects of dexamethasone arerare. Our case shows, however, that the possibility of seriousside-effects after a low dose of dexamethasone still existsand that these side-effects can occur in the anaestheticpractice. Avoiding rapid bolus application in patients withknown risk factors and continuous monitoring can help withtimely recognition and treatment of the adverse cardiovascu-lar side-effects that may follow after i.v. application ofdexamethasone.

Correspondence BJA

1025

by guest on February 10, 2015http://bja.oxfordjournals.org/

Dow

nloaded from

Declaration of interestNone declared.

M. Marinov*M.-U. FuesselA. F. UnterrainerSalzburg, Austria*E-mail: m.marinov@salk.at

1 Rusch D, EberhartLHJ, WallenbornJ, Kranke P. Nausea and vomitingafter surgery under general anesthesia.DtschArztebl Int2010;107:73341

2 Apfel CC, Korttila K, Abdalla M,et al.; IMPACT Investigators. A factor-ial trial of six interventions for the prevention of postoperativenausea and vomiting. N Engl J Med 2004; 350: 244151

3 Henzi I, Walder B, Trame`r MR. Dexamethasone for the prevention ofpostoperative nausea and vomiting: a quantitative systematicreview. Anesth Analg 2000; 90: 18694

4 Schmidt GB, Meier MA, Sadove MS. Sudden appearance of cardiacarrhythmias after dexamethasone. J Am Med Assoc 1972; 221:14024

5 Moses RE, McCormick A, Nickey W. Fatal arrhythmia after pulsemethylprednisolone therapy. Ann Intern Med 1981; 95: 7812

6 Barry M. The use of high-dose pulse methylprednisolone in rheuma-toid arthritis. Unproved therapy. Arch InternMed 1985; 145: 14834

7 Tvede N, Nielsen LP, Andersen V. Bradycardia after high-dose intra-venous methylprednisolone therapy. Scand J Rheumatol 1986; 15:3024

8 Gardiner PV, Griffiths ID. Sudden death after treatment with pulsedmethylprednisolone. Br Med J 1990; 300: 125

9 Bocanegra TS, Castaneda MO, Espinoza LR, Vasey FB, Germain BF.Sudden death after methylprednisolone pulse therapy. AnnIntern Med 1981; 95: 122

10 Puntis JWL, Morgan MEI, Durbin GM. Dexamethasone-inducedbradycardia. Lancet 1988; 2: 1372

doi:10.1093/bja/aet392

Long-stay patients with cancer on theintensive care unit: characteristics, riskfactors, and clinical outcomesEditorPublished literature on outcomes of patients with pro-longed admissions to general intensive care units (ICUs) havedemonstrated increased mortality, morbidity, and resourcerequirements.16 However, studies have yet to specifically

consider long-stay ICU patients with cancer. Critically ill cancerpatients pose a challenge to healthcare systems, as the impactof critical care on cancer progression is largely unknown and crit-ical illness itself often precludes aggressive cancer therapy inthe ICU setting. These considerations have contributed tospeculation that long-stay critically ill cancer patients maysurvive their ICU stay only to succumb to their cancer soonafter discharge, leading some healthcare funders to questionthe value of prolonged ICU care in this group of patients.

The aim of this study was to determine the clinical charac-teristics and outcomes of long-stay ICU patients with cancerand also to identify prognostic risk factors of outcome in thisgroup. Retrospective data on cancer diagnosis, pre-admissionchemotherapy, APACHE II score, laboratory tests, organ sup-port, and reason for ICU admission were collected on all cancerpatients admitted to the Royal Marsden Hospital ICU, a tertiaryreferral centre for cancer in the UK, with an ICU stay .16 daysduring a 6 yr period (January 20062012). The definition oflong stay was based on 2 standard deviations from the meanlength of stay in our unit.

Two hundred and three patients met the criteria for inclu-sion in the study. Long-stay patients accounted for 2.6% oftotal ICU admissions, but 24.0% of ICU budget. The mostpreva-lent cancer diagnoses for long-stay ICU patients were haem-atological (65 patients; 32.0%) and upper gastrointestinal (51patients; 25.1%) malignancies. Common reasons for ICU ad-mission were elective surgery (88 patients; 43.3%); respiratoryfailure (37 patients; 18.2%), and sepsis (36 patients; 17.7%).ICU, in-hospital, and 12 month mortality for all long-stay ICUpatients with cancer were 25.6% (52 patients), 32.5% (66patients), and 48.3% (98 patients), respectively. Risk factorsassociated with outcome in long-stay patients were investi-gated using the univariate logistic and Cox proportionalhazards regression. Foreach outcome, clinical risk factors asso-ciated with ICU, in-hospital, 12 month mortality, and time todeath (P,0.05) were chosen from the corresponding univari-ate tests for inclusion in the multivariate model. Interestingly,age, type of cancer, disease status, and APACHE II score on ad-mission were not found to be significantly associated withoutcome in long-stay cancer patients. For time to death afteradmission to ICU, respiratory failure [present vs absent, hazardratio (HR) 2.1, 95% confidence interval (CI) 1.23.7, P0.01],steroid use (present vs absent, HR 2.2, 1.43.4, P,0.001), andchemotherapy before ICU admission (yes vs no, HR 2.7, 1.64.7, P,0.001) were significantly associated with outcome.

Table1. Multivariate regression analysis: significant clinical predictors associated with ICU, in-hospital, and 12 month mortality, and time to deathafter admission to ICU for long-stay critically ill cancer patients. For categorical variables, ORs are provided. HRs are provided for time to death fromICU admission. P-values ,0.05 are shown in bold

Variable ICU mortality In-hospital mortality 12-month mortality Time to death

OR (95% CI) P-value OR (95% CI) P-value OR (95% CI) P-value HR (95% CI) P-value

Respiratory failure 1.3 (1.01.3) 0.02 1.2 (0.991.5) 0.06 1.2 (0.941.5) 0.15 2.1 (1.23.7) 0.01

Renal replacement 1.2 (1.01.3) 0.03 1.2 (1.11.4) 0.002 1.1 (0.971.3) 0.13 1.3 (0.821.9) 0.3

Steroids 1.4 (1.21.6) /JPEG2000GrayImageDict > /AntiAliasMonoImages true /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 175 /MonoImageDepth 4 /MonoImageDownsampleThreshold 1.50286 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects true /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/Description >>> setdistillerparams> setpagedevice