Prophylaxis of alcohol withdrawal syndrome in alcohol-dependent patients admitted to the intensive care unit after tumour resection

Transcription

1 British Journal of Anaesthesia 1995; 75: Prophylaxis of alcohol withdrawal syndrome in alcohol-dependent patients admitted to the intensive care unit after tumour resection C. D. SPIES, N. DUBISZ, W. FUNK, S. BLUM, C. MÜLLER, H. ROMMELSPACHER, G. BRUMMER, M. SPECHT, L. HANNEMANN, H. W. STRIEBEL AND W. SCHAFFARTZIK Summary P Br. J. Anaesth 75 Key words Alcohol, withdrawal. Intensive care. Intensive care, alcohol withdrawal. Complications, alcohol withdrawal syndrome. Chronic alcohol misuse has been reported in 92 % of patients with oral carcinoma and in 61 % of patients with oesophageal carcinoma [1]. Fifty percent of these patients are alcohol-dependent and develop alcohol withdrawal syndrome (AWS) after surgery which can induce potentially life-threatening complications and prolong their stay in the intensive care unit (ICU) [2, 3]. Prevention of AWS by prophylactic drugs is reported to shorten ICU stay in alcohol-dependent patients [2, 3]. There are many different drugs available to treat alcohol withdrawal [4 19], but the optimum regimen is not known [20]. Most recommendations are based on studies performed in detoxification units [8 15]. The transmitter imbalance in these patients is caused by withdrawal of ethanol [21 23] but after surgery or intubation, stress may aggravate transmitter imbalance, and an increased dose of medication is required in the ICU [4 7]. Ethanol is not administered in detoxification units, while it is used commonly in the ICU [16 18]. In the ICU multiple medications are often preferred as single drug therapy necessitates the use of higher doses compared with detoxification units. The prophylactic regimens used in the present study were chosen from a previous study performed in 672 surgical centres [24]. In 28 %, prophylaxis was not administered because of the difficult preoperative evaluation. As withholding prophylaxis in alcoholdependent patients would lengthen ICU stay [2, 3], this was deemed unethical; 72 % used ethanol or pharmacological treatment for the prophylaxis of AWS. Pharmacological treatment was based, in the majority of cases, on a combination of benzodiazepine, chlormethiazole, and clonidine [4 7]. To our knowledge, no controlled studies exist defining which of these standard regimens are most appropriate in the ICU. We have determined if ICU stay, prevention of AWS and major intercurrent complication rate differed between four different standard ICU regimens. CLAUDIA D. SPIES*, MD, NORMAN DUBISZ, RA, WALTER FUNK, MD, GLENDA BRUMMER, MD, MARTIN SPECHT, MD, LUTZ HANNEMANN, MD, H. WALTER STRIEBEL, DEAA, MD, WALTER SCHAFFARTZIK, MD (Department of Anesthesiology and Operative Intensive Care Medicine); SUSANNE BLUM, MD (Department of Neurology); Benjamin Franklin Medical Centre, Free University Berlin, Germany. CHRISTIAN MÜLLER, PHD (Department of Clinical Chemistry and Biochemistry); HANS ROMMELSPACHER, MD (Department of Neuropsychopharmacology); Rudolf Virchow Medical Centre, Free University Berlin, Germany. Accepted for publication: June 14, *Address for correspondence: Klinik für Anaesthesiologie und operative Intensivmedizin, Universitaetsklinikum Benjamin Franklin, Freie Universitaet Berlin, Hindenburgdamm 30, Berlin, Germany.

2 Prophylaxis of alcohol withdrawal syndrome 735 Patients and methods In this prospective, randomized, clinical study, we studied 197 consecutive patients with carcinoma of the upper digestive tract undergoing tumour resection. All patients gave written informed consent to the study, which was approved by the Ethics Committee of the Benjamin Franklin Medical School, Free University Berlin. Patient age, height, weight, acute physiology and chronic health evaluation score (APACHE II) [25] and multiple organ failure score (MOF) [26] were recorded (table 1). In addition, patient history and the results of alcoholismrelated questionnaires (CAGE questionnaire [27] and the smast [28]) were obtained before operation (table 1). The investigator who documented the alcoholism-related history was unaware of the treatment and of the intercurrent complications during ICU stay. All chronic alcoholics fulfilled the DSM- III-R criteria for alcohol dependence [29]. Alcoholdependent patients were allocated randomly to receive one of the following regimens: flunitrazepam clonidine (flu clon); chlormethiazole (cmz hp); flunitrazepam (flu hp); or ethanol. Patients were excluded if they were less than 18 yr of age, did not undergo surgery, had chronic obstructive lung disease and poor pulmonary function (e.g. chronic hypoxaemia Pa O 8 kpa), pneu- 2 monia, congestive heart failure, bradycardia 45 beat min 1, systolic pressure 95 mm Hg, secondor third-degree atrioventricular node block (AVB), pancreatitis, liver insufficiency or had a history of current abuse of benzodiazepines, barbiturates, clonidine or blockers. After assignment to a group, a bolus of the test agent was administered immediately after operation on admission to the ICU, followed by a continuous infusion (table 2). The regimen was adapted so that the patients were always arousable while a revised clinical institute withdrawal assessment for alcohol scale (CIWA-Ar) 20 was maintained and controlled hourly in the first 24 h and then four times daily [30]. The CIWA-Ar is a shortened 10-item scale for measurement of the degree of AWS, including vegetative symptoms, hallucinations and symptoms of delirium tremens [30]. The CIWA-Ar was obtained by experienced investigators blinded to the drug regimen. As an adjuvant, all patients were treated with vitamin B mg day 1 i.v. Serum electrolyte concentrations were measured at least three times daily and abnormalities were treated appropriately. Table 1 Patient characteristics (mean (SD), median (range) or frequency (%)). smast short Michigan alcoholism screening test; CAGE alcoholism-related questionnaire; CDT carbohydrate-deficient transferrin (cut-off 9 mg litre 1 ); MCV mean corpuscular volume; GGT -glutamyl-transferase; ASAT aspartate aminotransferase; ALAT alanine aminotransferase clonidine (n 48) Chlormethiazole (n 49) Ethanol Age (yr) 52 (26 87) 52 (31 76) 53 (29 77) 53 (36 89) (mean (range)) Sex (M/F) 44/4 45/4 42/8 46/4 Height (cm) 174 (10) 174 (6) 172 (10) 176 (7) Weight (kg) 75 (16) 75 (12) 75 (16) 76 (15) Smokers 37/48 (77 %) 36/49 (73 %) 36/50 (72 %) 38/50 (76 %) Ethanol (g day 1 ) 95 (60 380) 90 (60 660) 90 (60 485) 98 (60 550) Beer 24/48 (50 %) 25/49 (51 %) 28/50 (56 %) 31/50 (62 %) Wine 8/48 (17 %) 6/49 (12 %) 8/50 (16 %) 6/50 (12 %) Spirits 31/48 (65 %) 31/49 (63 %) 35/50 (70 %) 32/50 (64 %) Prior detoxification 14/48 (29 %) 12/49 (24 %) 17/50 (34 %) 15/50 (30 %) CAGE 3.8 (0.2) 3.6 (0.1) 3.6 (0.1) 3.7 (0.2) smast 9.4 (5.8) 9.8 (5.2) 8.9 (5.1) 9.7 (4.9) CDT (mg litre 1 ) 11 (3 46) 11 (2 36) 15 (3 38) 12 (2 48) MCV (fl) 99 (9) 98 (7) 98 (6) 99 (7) GGT (u. litre 1 ) 38 (17 100) 39 (5 187) 35 (8 443) 40 (13 174) ASAT (u. litre 1 ) 12 (6 57) 11 (3 36) 12 (6 53) 16 (6 80) ALAT (u. litre 1 ) 16 (10 62) 12 (6 34) 13 (4 36) 15 (5 59) Table 2 Prophylaxis for alcohol withdrawal syndrome (median (range)) clonidine (n 48) Chlormethiazole (n 49) Ethanol Flunitrazepam Chlormethiazole Flunitrazepam Ethanol Bolus 1.0 (0.5 2) mg 150 (50 500) mg 1.0 (0.5 2) mg 3 (2 4) g Maximum dose 7 (1 34) g kg 1 h ( ) mg kg 1 h 1 6 (1 61) g kg 1 h 1 48 (12 157) mg kg 1 h 1 Total amount 68 (4 603) mg 16 (4 48) g 20 (7 257) mg 316 ( ) g Clonidine Haloperidol Haloperidol Bolus 150 (75 300) g 10 (5 20) mg 10 (5 20) mg Maximum dose 0.83 ( ) g kg 1 h 1 30 (10 115) g kg 1 h 1 29 (9 99) g kg 1 h 1 Total amount 4.1 (1.3 69) mg 209 ( ) mg 168 (26 850) mg

3 736 British Journal of Anaesthesia We evaluated 393 patients with carcinoma of the upper digestive tract. After diagnostic procedures, 109 patients did not undergo surgery and therefore were not included in data analysis. Another 64 patients were excluded because of underlying diseases being a contraindication to one of the four regimens: 31 with respiratory disorders, 23 with cardiac disorders, six with gastrointestinal disorders such as pancreatitis and liver insufficiency, and four for other reasons. Thus we studied 220 patients but 23 were excluded: seven patients in the flu clon group (one because of the onset of second-degree AVB, two after repetitive bradycardia associated with sustained hypotension, three developed AWS and continued to hallucinate and so required additional therapy with (CIWA-Ar 20), and one withdrew consent); six patients in the cmz hp group (four because of hypersecretion which required tracheal suctioning more than every 5 min, one because of QT c prolongation over 520 ms and one who died before admission to the ICU); five patients in the flu hp group (two patients because of QT c prolongation ( 30% of baseline) which was not caused by hypokalaemia, hypomagnesaemia or bradycardia, one required additional clonidine to treat hypertension, one patient withdrew consent and one was transferred to another hospital); five patients in the ethanol group (two developed hallucinations and delirium tremens which required a change in drug regimen and three who received additional benzodiazepines at night during ICU stay). Blood was obtained on admission to hospital (table 1). Serum carbohydrate-deficient transferrin (CDT) was measured by microanion exchange chromatography and subsequent turbidimetry [31]. CDT is a new marker of chronic alcohol misuse [32] which increases after a daily load of 60 g and normalizes within days after ethanol withdrawal. A CDT greater than 9 mg litre 1 was defined as pathologically elevated [3]. All other laboratory variables were measured according to standard clinical routines. The primary outcome measures were duration of ICU stay and prevention of AWS. Secondary outcomes included the overall incidence of major complications. Pneumonia requiring mechanical ventilation, sepsis, cardiac complications, bleeding disorders and death were considered major intercurrent complications. All patients received perioperative antibiotic prophylaxis with cefotiam on admission to the ICU. The antimicrobial regimen was adapted according to the susceptibility of the isolated organisms. If the patient developed pneumonia and no organism was isolated, cefotaxime and gentamicin were administered. When AWS developed, the diagnosis was made according to an accepted algorithm [33]. Diagnosis was confirmed by a neurologist and symptoms were documented. If the diagnosis of AWS was established and the CIWA-Ar exceeded 20 [30], a bolus of medication was administered. A CIWA-Ar 20 was required after bolus administration in each group. Patients with persistent CIWA-Ar scores 20 that could not be reversed by increased doses of the drug regimen received additional appropriate medication and were withdrawn from the study. Infections were determined with respect to CDC criteria [34]. Tracheobronchitis was diagnosed if the patient had at least three of the following five signs or symptoms: cough, rhonchi, wheezing or other auscultatory findings without evidence of pulmonary consolidation, purulent sputum production, body temperature 38 C, leucocystosis or organisms isolated from culture obtained by tracheal aspirate or bronchoscopy. A diagnosis of pneumonia was made if consolidation was evident on a chest x-ray, with the onset of purulent sputum or a change in the character of the sputum, and rales or dullness to percussion on examination were present in the area corresponding to the consolidation. Organisms isolated from specimens obtained by tracheal suctioning or bronchial alveolar lavage are shown in table 4. Sepsis was defined according to the Society of Critical Care Medicine Consensus Conference [35]. Cardiac complications included bradycardia ( 40 beat min 1 ) associated with hypotension, hypotension not reversible with fluid administration, conduction abnormalities, arrhythmias (e.g. three consecutive premature ventricular complexes (PVC), frequent multiform PVC, R-on-T PVC, atrial fibrillation), QT c prolongation 20 % over baseline which was not caused by hypokalaemia, hypomagnesaemia or bradycardia, congestive heart failure with pulmonary or peripheral oedema and angina. Cardiac complications caused by sepsis with sustained hypotension requiring positive inotropes and vasopressors were not included under this category. A bleeding disorder was diagnosed if the patient required a blood transfusion or surgery because of persistent bleeding. Complications caused by other disease processes such as sepsis requiring fluids and blood transfusion were not included under this category. All data are expressed as mean (SD) (normally distributed), median (range) (not normally distributed) or frequency. Statistical analysis was performed by the Kruskal-Wallis test for variables not normally distributed or analysis of variance (ANOVA) if variables were normally distributed. If a significant difference was detected by the Kruskal Wallis test or ANOVA, the Wilcoxon signed rank sum test, the t test for intergroup differences or the chi-square test to compare dichotomous variables were used to determine which groups differed significantly. P 0.05 was considered significant. Results Patient characteristics did not differ significantly between groups (table 1) and there were no significant differences between alcoholism-related history (table 1), laboratory markers (table 1) and severity of the physiological derangement on admission to the ICU (table 3). The doses of medication required are given in table 2. The regimens did not differ significantly with respect to the prevention of AWS (table 3),

4 Prophylaxis of alcohol withdrawal syndrome 737 Table 3 Scores and complications during ICU stay (mean (SD), median (range) or frequency (%)). Max maximum; min minimum; OA ICU on admission to the ICU; FA ICU following admission to the ICU; PP prior to prophylaxis; ND/OR neck dissection/oesophageal resection; APACHE acute physiology and chronic health evaluation score; MOF multiple organ failure score; CIWA-Ar revised clinical institute withdrawal assessment of alcohol scale; AWS alcohol withdrawal syndrome; V/H/D vegetative withdrawal symptoms/hallucinations/delirium tremens; Pa O2 / F I O (kpa) arterial oxygen 2 tension/inspired oxygen fraction ratio; PA Pseudomonas aeruginosa; SA Staphylococcus aureus; VO2 (ml min 1 m 2 ) oxygen consumption; Do 2 (ml min 1 m 2 ) oxygen delivery; lactate (mmol litre 1 ); noradrenaline ( g kg 1 min 1 ); Hb (g d 1 ) haemoglobin; CHF congestive heart failure; BH bradycardia hypotension; AVB new onset of atrioventricular node block (first-degree); PVC premature ventricular complex ( 3 consecutive PVC, frequent multiform PVC, R-on-T PVC) clonidine (n 48) Chlormethiazole (n 49) Ethanol P ND/OR 30/18 34/15 31/19 29/ APACHE II (OA ICU) 11 (4) 11 (3) 12 (3) 12 (3) MOF (OA ICU) 3 (3) 3 (2) 3 (2) 3 (2) CIWA-Ar 5 (3) 4 (3) 5 (3) 5 (4) AWS 6/48 4/49 5/50 2/ Highest CIWA-Ar 22 (21 39) 24 (22 35) 24 (22 35) 21 and 24 V/H/D 0/2/4 1/1/2 2/1/2 2/0/0 Begin (FA ICU) (days) 1.5 (0 3) 1.5 (0 3) 1.0 (0 2) 0 and 3 AWS duration (days) 2.1 (0.9 8) 0.8 (0.5 4) 1.5 (0.8 4) 0.5 and 1 Tracheobronchitis 12/48 33/49 22/50 20/ Pneumonia 8/48 11/49 9/50 10/ Worst Pa O2 / F I O ( ) 28.3 ( ) 32.5 ( ) 38.4 ( ) Organisms isolated 3/8 4/11 3/9 2/10 PA/SA 1/1 2/1 1/2 1/0 Sepsis 2/48 4/49 3/50 3/ VO2 (min) 113 and (58 152) 112 (89 145) 123 (78 135) Do 2 (max) 706 and ( ) 735 ( ) 778 ( ) Lactate (max) 1.3 and ( ) 1.4 ( ) 1.6 ( ) Noradrenaline (max) 0.45 and (0 1.40) 0.14 (0 0.97) 0.02 (0 0.47) Bleeding disorder 4/48 9/49 8/50 8/ Blood transfusion (ml) ( ) ( ) ( ) ( ) Hb (min) 8.0 ( ) 7.7 ( ) 8.0 ( ) 7.5 ( ) Cardiac disorder 9/48 5/49 4/50 6/ CHF/BH/AVB/PVC 0/6/2/1 1/0/0/4 0/1/0/4 5/0/0/1 Death 3/48 4/49 3/50 3/ (flu clon 5 %; cmz hp 64 %; flu hp 10 %; ethanol 8 %; P ). The frequency of mechanical ventilation on admission to the ICU (P ) and duration of mechanical ventilation during ICU stay did not differ between groups (flu clon median 2 (range 0 28) days; cmz hp 2.5 (0 40) days; flu hp 2 (0 18) days; ethanol 2 (0 15) days; P ). Severe cardiac complications which led to withdrawal from the study were observed with the three regimens: three patients receiving flu clon, one patient receiving cmz hp and two patients receiving flu hp; there was no significant difference between groups (P ). Figure 1 Intensive care unit (ICU) stay with respect to the four different regimens. The median ( ) is given for each of the four regimens: flunitrazepam clonidine (flu clon); chlormethiazole (cmz hp); flunitrazepam (flu hp); and ethanol (P ). ICU stay (fig. 1) and the incidence of major intercurrent complications between groups. The incidence of tracheobronchitis was increased significantly in the cmz hp group (table 3). Hypersecretion was also more frequent in this group Discussion In this study the efficacy of each prophylactic regimen was controlled by the CIWA-Ar [30]. One of the problems in the interpretation of clinical trials of various drug regimens for prophylaxis or treatment of AWS is that they do not use validated measures of the dependent variables. The CIWA-Ar relies on a constellation of clinical signs and subjective symptoms and so the presence of other acute illnesses can lead to raised CIWA-Ar values unrelated to alcohol withdrawal. However, the study was performed by experienced investigators who were specifically trained to use the CIWA-Ar in this ICU.

5 738 British Journal of Anaesthesia We found that the incidence of AWS did not differ significantly between groups. In most patients AWS was controlled by increased doses. Five patients, three in the flu clon and two in the ethanol group, were withdrawn because they continued to hallucinate and required. The incidence of AWS in patients receiving prophylaxis is difficult to state. Drug regimens are complicated in surgical ICU and they differ from other ward settings. This may account for the increased doses required to prevent the symptoms of alcohol withdrawal and the use of combination therapy [24]. As there are, to our knowledge, no controlled studies in surgical ICU, it is difficult to compare the true incidence of AWS in patients receiving prophylaxis with other studies performed in detoxification units. Benzodiazepines, chlormethiazole, and clonidine are commonly used in the treatment of AWS [4 15]. In contrast, i.v. ethanol as a prophylaxis for AWS is also commonly used but is controversial [16 18], probably because most studies of the treatment of AWS are performed in psychiatric wards that aim to detoxify patients. In surgical patients the main goal is to prevent AWS in alcoholdependent patients and to shorten hospital stay [2, 3]. In the ICU, only a single study by Hansbrough and colleagues [16] administered i.v. ethanol to alcoholic burns patients. They prevented AWS with continuous i.v. infusions of ethanol at rates of mg kg 1 h 1, the same range as used in the present study. Ethanol administration in this previous study [16], however, has not been compared with any other standard regimen such as benzodiazepines. The investigator who performed the CIWA-Ar was unaware of the regimen administered and this reduced any bias that could have occurred in the assessment of prevention and severity of AWS. We did not double-blind the study because no controlled studies in ICU patients and no recommended dosage schedules are available. ICU stay did not differ between groups. This may result from the low power of the study. But even if two regimens only had been compared, more than 1000 patients would have been required to detect a significant difference, assuming a power of 0.8 and an of Such investigations can only be performed as multicentre studies. The major side effect of chlormethiazole in our patients was bronchial hypersecretion. Respiratory depression and therefore the necessity for artificial ventilation has been reported with the higher doses required for treatment of AWS [11]. Tracheal intubation, mechanical ventilation and the severity of pre-existing diseases are strongly associated with the risk of developing nosocomial pneumonia, the most common hospital-acquired infection in the ICU [36 38]. In the present study, the groups did not differ with respect to their pre-existing illnesses and the frequency and duration of mechanical ventilation. The latter may result from vigorous attempts to wean patients from ventilatory support. There were no differences in major intercurrent complications. However, the study was not designed to have enough power to detect small differences in the rates of these complications. Severe cardiac complications which have been reported for clonidine and [5, 39] resulting in withdrawal from the study were distributed equally between the regimens. Acknowledgement This project was supported in part by the German Research Society (DFG HE 916/7-2). References 1. Seitz W, Simanowski UA. Ethanol and carcinogenesis of the alimentary tract. Alcoholism, Clinical and Experimental Research 1986; 10 (Suppl.): 33S 40S. 2. Spies C, Neumann T, Rommelspacher H, Müller C, Schaffartzik W. Sensitivity and specificity of carbohydratedeficient transferrin (CDT) to detect chronic alcohol abuse in polytraumatized patients. Anesthesiology 1994; 81: A Heil T, Spies C, Bullmann C, Neumann T, Eyrich K, Müller C, Rommelspacher H. The relevance of CDT (carbohydratedeficient transferrin) in the preoperative diagnosis of chronic alcohol abuse in intensive care patients after elective tumour resection. Anaesthesist 1994; 43: Nolop KB, Natow A. Unprecedented sedative requirements during delirium tremens. Critical Care Medicine 1985; 13: Metzger E, Friedman R. Prolongation of the corrected QT and torsades de pointes cardiac arrhythmia associated with intravenous in the medically ill. Journal of Clinical Psychopharmacology 1993; 13: Schinzel H, Weilemann LS, Swars H, Kelbel C, Meyer J. Experiences in treatment of acute alcohol withdrawal syndrome with clonidine in an intensive care unit. Intensivmedizin 1993; 30: Adams F. Neuropsychiatric evaluation and treatment of delirium in the critically ill cancer patient. Cancer Bulletin 1984; 36: Saitz R, Mayo-Smith MF, Roberts MS, Redmond HA, Bernard DR, Clakins DR. Individualized treatment for alcohol withdrawal. A randomized double-blind controlled trial. Journal of the American Medical Association 1994; 272: Pycha R, Miller C, Barnas C, Hummer M, Stuppaeck C, Whitworth A, Fleischhacker WW. Intravenous flunitrazepam in the treatment of alcohol withdrawal delirium. Alcoholism, Clinical and Experimental Research 1993; 17: Nutt D, Adinoff B, Linnoila M. Benzodiazepines in the treatment of alcoholism. In: Galanter M, ed. Recent Developments in Alcoholism: Treatment Research, vol. 7. New York: Plenum Press, 1989; Athen D. Comparative investigation of chlormethiazole and neuroleptic agents in the treatment of alcoholic delirium. Acta Psychiatrica Scandinavica 1986; 73: Robinson BJ, Robinson GM, Maling TJB, Johnson RJ. Is clonidine useful in the treatment of alcohol withdrawal? Alcoholism, Clinical and Experimental Research 1989; 13: Baumgartner GR, Rowen RC. Clonidine vs chlordiazepoxide in the management of acute alcohol withdrawal syndrome. Archives of Internal Medicine 1987; 147: Nutt D, Glue P, Molyneux S, Clark E. 2-Adrenoreceptor function in alcohol withdrawal: a pilot study of the effects of iv clonidine in alcoholics and normals. Alcoholism, Clinical and Experimental Research 1988; 12: Wadstein J, Manhem P, Nilsson LH, Moberg AL, Hökfeld B. Clonidine versus clomethiazole in alcohol withdrawal. Acta Psychiatrica Scandinavica 1986; 73: Hansbrough JF, Zapata-Sirvent RL, Carroll WJ, Johnson R, Saunders CE, Barton CA. Administration of intravenous alcohol for the prevention of withdrawal in alcoholic burn patients. American Journal of Surgery 1984; 148: Glickman L, Herbsman H. Delirium tremens in surgical patients. Surgery 1968; 64:

6 Prophylaxis of alcohol withdrawal syndrome Gower WE, Kersten H. Prevention of alcohol withdrawal symptoms in surgical patients. Surgery, Gynecology and Obstetrics 1980; 151: Yam I, Forbes A, Kox WJ. Clonidine in the treatment of alcohol withdrawal in the intensive care unit. British Journal of Anaesthesia 1992; 68: Fuller RK, Gordis E. Refining the treatment of alcohol withdrawal. Journal of the American Medical Association 1994; 272: Linnoila M, Mefford I, Nutt D, Adinoff B. Alcohol withdrawal and noradrenergic function. Annals of Internal Medicine 1987; 107: Hoffman PL, Rabe CS, Grant KA, Valverius P, Hudspith M, Tabakoff B. Ethanol and the NMDA receptor. Alcohol 1990; 7: Schulteis G, Koob G. Dark side of drug dependence. Nature (London) 1994; 371: Imdahl H, Imdahl A. Prophylaxis and therapy of alcoholic delirium tremens in surgery Analysis of questionnaire inquiry. Aktuelle Chirurgie 1992; 27: Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: A severity of disease classification. Critical Care Medicine 1985; 13: Goris RJA, TeBoekhorst TPA, Nuytinek JKS, Gimbrére JSF. Multiple-organ failure. Archives of Surgery 1985; 120: Ewing JA. Detecting alcoholism, the CAGE questionnaire. Journal of the American Medical Association 1984; 252: Selzer ML, Vinokur A, Van Rooijen L. A self administered short Michigan Alcoholism Screening Test (smast). Journal of Studies on Alcohol 1975; 36: American Psychiatric Association, Committee on Nomenclature and Statistics. Diagnostic and Statistical Manual of Mental Disorders, 3rd rev Edn (DSM-III-R). Washington, DC: American Psychiatric Association, 1987; Sullivan JT, Sykora K, Schneiderman J, Naranjo CA, Sellers EM. Assessment of alcohol withdrawal: the revised clinical institute withdrawal assessment for alcohol scale (CIWA-Ar). British Journal of Addiction 1989; 84: Müller C, Bräutigam K, Moritz R, Rüggeberg J, Köttgen E. A simplified test procedure for measuring carbohydratedeficient transferrin in serum. European Journal of Clinical Chemistry and Clinical Biochemistry 1993; 31: A Stibler H. Carbohydrate-deficient transferrin serum: A new marker of harmful alcohol consumption reviewed. Clinical Chemistry 1991; 37: Cassem NH. Psychiatric problems of the critically ill patient. In: Shoemaker WC, Ayres S, Grenvik A, eds. Textbook of Critical Care. The Society of Critical Care Medicine. Philadelphia: Saunders, 1989; Garner JS, Jarvis WR, Emori TG. CDC definitions for nosocomial infections, American Journal of Infection Control 1988; 16: Members of the American College of Chest Physicians/ Society of Critical Care Medicine Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Critical Care Medicine 1992; 20: Celis R, Torres A, Gatell JM, Almela M, Rodriguez-Roisin R, Agusti-Vidal A. Nosocomial pneumonia: a multivariate analysis of risk and prognosis. Chest 1988; 93: Craven DE, Kunches LM, Kilinsky V, Lichtenberg DA, Muke BJ, McCabe WR. Risk factors for pneumonia and fatality in patients receiving continuous mechanical ventilation. American Review of Respiratory Disease 1986; 133: Malangoni MA, Crafton R, Mocek FC. Pneumonia in the surgical intensive care unit: factors determining successful outcome. American Journal of Surgery 1994; 167: Rettmar K, Stierle U, Muhl E, Gehring H, Sheikzadeh A, Diederich KW. Sinus bradycardia, prolonged QT-interval, and ventricular defibrillation under - and clonidine-therapy in alcohol withdrawal syndrome. Intensivmedizin 1992; 29: