12

Springer

Dear Author: Please find attached the final pdf file of your contribution, which can be viewed using the Acrobat Reader, version 3.0 or higher. We would kindly like to draw your attention to the fact that copyright law is also valid for electronic products. This means especially that: •

You may print the file and distribute it amongst your colleagues in the scientific community for scientific and/or personal use.

•

You may make your article published by Springer-Verlag available on your personal home page provided the source of the published article is cited and Springer-Verlag and/or other owner is mentioned as copyright holder. You are requested to create a link to the published article in Springer's internet service. The link must be accompanied by the following text: "The original publication is available at springerlink.com". Please use the appropriate DOI for the article. Articles disseminated via SpringerLink are indexed, abstracted and referenced by many abstracting and information services, bibliographic networks, subscription agencies, library networks and consortia.

•

Without having asked Springer-Verlag for a separate permission your institute/your company is not allowed to place this file on its homepage.

•

You may not alter the pdf file, as changes to the published contribution are prohibited by copyright law.

•

Please address any queries to the production editor of the journal in question, giving your name, the journal title, volume and first page number.

Yours sincerely,

Springer-Verlag

Intensive Care Med (2003) 29:2230–2238 DOI 10.1007/s00134-003-2033-3

Claudia D. Spies Hilke E. Otter Bernd Hüske Pranav Sinha Tim Neumann Jordan Rettig Erika Lenzenhuber Wolfgang J. Kox

Received: 10 July 2002 Accepted: 5 August 2003 Published online: 14 October 2003 © Springer-Verlag 2003 This study was sponsored in part by the German Research Society (DFG-SP 432/1-1 and 1-2) Electronic Supplementary Material Supplementary material is available in the online version of this article http://dx.doi.org/10.1007/s00134-2033-3.

C. D. Spies (✉) · H. E. Otter · B. Hüske T. Neumann · W. J. Kox Department of Anesthesiology and Intensive Care Medicine, Charité-Universitätsmedizin Berlin, Campus Charité Mitte, Schumannstrasse 20/21, 10117 Berlin, Germany e-mail: [email protected] Tel.: +49-30-450-531052 Fax: +49-30-450-531911 P. Sinha Institute of Clinical Chemistry, Charité-Universitätsmedizin Berlin, Berlin, Germany J. Rettig School of Medicine, University of Connecticut, Conn., USA E. Lenzenhuber Department of Anesthesiology and Intensive Care Medicine, Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin, Berlin, Germany E. M. Sellers Departments of Pharmacology, Medicine, and Psychiatry, University of Toronto, Toronto, Ont., Canada

ORIGINAL

Edward M. Sellers

Alcohol withdrawal severity is decreased by symptom-orientated adjusted bolus therapy in the ICU

Abstract Objective: To examine the effect of bolus vs. continuous infusion adjustment on severity and duration of alcohol withdrawal syndrome (AWS), the medication requirements for AWS treatment, and the effect on ICU stay in surgical intensive care unit (ICU) patients. Design and setting: Prospective randomized, double-blind controlled trial in a surgical ICU. Patients: 44 patients who developed AWS after admission to the ICU. Interventions: Patients were randomized to either (a) a continuous infusion course of intravenous flunitrazepam (agitation), intravenous clonidine (sympathetic hyperactivity), and intravenous haloperidol (productive psychotic symptoms) if needed (infusion-titrated group), or (b) the same medication (flunitrazepam, clonidine, or haloperidol) bolus adjusted in response to the development of the signs and symptoms of AWS (bolus-titrated group). Measurements and results: The administration of “as-needed” medication was determined using a validated measure of the severity of AWS (Clinical Institute of Withdrawal Assessment). Although the severity of AWS did not differ between groups initially, it significantly worsened over time in the infusion-titrated group. This required a higher amount of flunitrazepam, clonidine, and haloperidol. ICU treatment was significantly shorter in the bolus-titrated group (median dif-

ference 6 days) due to a lower incidence of pneumonia (26% vs. 43%). Conclusions: We conclude that symptom-orientated bolus-titrated therapy decreases the severity and duration of AWS and of medication requirements, with clinically relevant benefits such as fewer days of ventilation, lower incidence of pneumonia, and shorter ICU stay. Keywords Alcohol withdrawal syndrome · Symptom-orientated therapy · Surgical intensive care unit · Clinical Institute of Withdrawal Assessment · Critically ill patients · Pneumonia

2231

Introduction

Patients and methods

Every fifth patient admitted to hospital proves to be positive for alcoholism [1]. Following major surgery chronic alcoholics have an increased morbidity and mortality rate [2, 3, 4]. In addition to infections, cardiac, and bleeding complications, chronic alcoholics are at risk of developing alcohol withdrawal syndrome (AWS) [3]. If not treated prophylactically, one-half of chronic alcoholics develop AWS [3, 5]. However, despite preventive treatment approximately one-fourth of these patients develop AWS postoperatively [3, 5]. Moreover, those who develop AWS have a prolonged ICU stay [6]. The severity of AWS in ICU patients following major surgery is higher than that in psychiatric patients as evaluated by the validated Clinical Withdrawal Assessment for Alcohol, revised scale (CIWA-Ar) initially developed for psychiatric patients [6, 7, 8] but which has also been used in ICU settings [9]. The medication requirements for the treatment of AWS in ICU patients are usually higher [3]. The first-line of therapy for AWS is the administration of benzodiazepines [10]. When used for AWS, benzodiazepines are generally administered on predetermined dosage schedules for several days, often in gradually tapering doses. While monotherapy with benzodiazepines is usually administered for AWS in psychiatric patients, 64% of ICU patients with AWS receive drug combinations [3, 11, 12]. The most common adjunctives to benzodiazepines are clonidine (used for autonomic signs) and haloperidol (for productive psychotic states such as hallucinations) [10]. Intravenous benzodiazepines and intravenous clonidine are usually administered continuously, while intravenous haloperidol is added either according to a predetermined dosing schedule or at a constant infusion rate [3]. In contrast to these conventional ICU infusion regimes, bolus therapy is currently available for detoxification to decrease the amount of benzodiazepines and the treatment period [7]. It is unclear whether bolus therapy would also be beneficial for ICU patients. Oversedation may prolong ICU treatment [13] whereas lack of sedation may put the patient at risk for, for example, dislocation of tubes and prolonged weaning [14]. In AWS the adjustment of continuous infusions is usually delayed, resulting either in comatose patients with the consequence of prolonged ICU treatment or in recurrent symptoms of alcohol withdrawal. There is a lack of randomized studies on sedation in the ICU [15]; to our knowledge, only two randomized controlled trials have examined the treatment of AWS in ICU patients [11, 16]. The aims of this study were to investigate bolus and continuous infusion therapy for AWS with respect to: (a) the severity and duration of alcohol withdrawal syndrome (primary outcome measure), (b) benzodiazepine, clonidine, and haloperidol requirements (secondary outcome measure), and (c) duration of ICU treatment (tertiary outcome measure).

The study protocol was approved by the institutional ethics committee. All patients fulfilled the criteria of the CIWA-Ar (>20) for the requirement of pharmacotherapy [8]. The CIWA-Ar is a validated, reliable measure of the current severity of alcohol withdrawal composed of ten items (see Electronic Supplementary Material). Patients were included within 4 h after the onset of alcohol withdrawal. Exclusion criteria were age under 18 years, pregnancy, intubation and ventilation at the onset of AWS, continuous sedation at the onset of AWS, concurrent acute medical diagnosis, i.e., infections and hypoxia, hemoglobin lower than 8 g/dl, bleeding at the onset of AWS, acute angina, metabolic or endocrinological disorders, head injury, intoxications or psychiatric illness, a history of seizures of any cause, corticosteroid use or chemotherapy, current use or withdrawal from clonidine, β-blockers, or haloperidol and withdrawal from or known misuse of opiates, benzodiazepines, or barbiturates. Patients who were diagnosed as being alcohol-dependent before the onset of AWS were excluded (Fig. 1), receiving prophylactic treatment [3, 17]. After inclusion in the study patients with AWS treatment failure defined as a persisting AWS (CIWA-Ar >20) for longer than 8 h despite administration of rescue medication were also excluded. Protocol In this prospective randomized (per envelope) controlled study 300 patients with trauma or gastrointestinal surgery, fulfilling criteria for possible alcohol abuse (not dependence) as defined by the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) [18], and with an alcohol consumption higher than 60 g/day, were included. In the case of elective surgery patients gave their written informed consent before surgery; in patients with severe trauma relatives or their legal representatives gave consent on admission to the ICU. Physical examination and liver function tests were performed, and complete cell count, mean corpuscular volume, and carbohydrate-deficient transferrin were assessed to establish any diagnosis of alcohol abuse [3, 17, 19, 20]. The alcoholism-related questionnaire CAGE (see Electronic Supplementary Material) was administered to patients (surgery cases) or their relatives (trauma cases) and readministered after successful treatment of AWS from the patients upon reorientation [3]. The maximal value of the two CAGE assessments was taken. Prior to surgery and on admission to the ICU all patients had only admitted alcohol abuse and not dependence [3]. In the case of alcohol dependence (according to DSM-IV) known before onset of AWS patients were given prophylactic treatment and were excluded. Of the original 300 patients 107 developed AWS following admission to the ICU (Fig. 1); 59 of these were not eligible due to exclusion criteria. In this respect intubation was a crucial aspect of this study because before being included patients had to be extubated in order to exclude all differential diagnoses other than AWS. Therefore a total of 48 patients were finally included and randomized in either an infusion-titrated group (ITG) or a bolustitrated group (BTG; Fig. 1). A further four patients were later excluded, two due to contradictory study related histories (one ITG, one BTG), one (ITG) for additional requirement of rescue medication within 8 h of treatment, and one (ITG) who died from pulmonary embolism 8 h after enrollment. Finally, data from 44 patients were evaluated (Fig. 1). Basic patient characteristics and alcoholism-orientated diagnoses did not differ between groups (Table 1). There were no significant differences in baseline AWS characteristics between the treatment groups.

2232

Fig. 1 Decision tree of exclusion/inclusion procedure. DSM-IV Diagnostic and Statistical Manual of Mental Disorders, fourth edition; ICU intensive care unit; AWS alcohol withdrawal syndrome

Table 1 Baseline patient characteristics and alcoholismrelated diagnosis in the bolustitrated (BTG) and infusiontitrated (ITG) groups; absolute number (parentheses frequency) or median (parentheses interquartile range) (APACHE II Acute Physiology and Chronic Health Evaluation II, adm. on admission to ICU)

Age (years) Sex: M/F Body mass index Carcinoma Trauma Smokers Coronary artery disease Chronic obstructive lung disease Duration of anesthesia (h) APACHE II Ethanol (g/d) CAGE (mg)a Blood alcohol, adm. (mg/l) Carbohydrate-deficient transferrin (fl)b Mean corpuscular volume, adm. (U/l)c γ-Glutamyl transferase, adm. (U/l)d

BTG (n=23)

ITG (n=21)

p

49 (41–58) 18/5 25 (23–26) 13/23 (57%) 10/23 (44%) 19/23 (83%) 2/23 (9%) 12/23 (52%) 11 (6–14) 9 (5–14) 90 (65–280) 3 (3–4) 0 (0–0.2) 9.4 (5.8–11.7) 93.1 (87.7–96.8) 28 (23–62)

41 (33–68) 18/3 25 (24–30) 10/21 (48%) 11/21 (53%) 18/21 (86%) 1/21 (5%) 6/21 (29%) 10 (6–17) 12 (7–16) 90 (80–240) 3 (2–4) 0 (0–0.3) 7.6 (4.4–16.7) 92.8 (89.0–96.4) 20 (10–75)

0.29 0.53 0.37 ≥0.99 ≥0.99 0.72 0.76 0.37 0.60 0.21 0.64 0.46 0.93 0.92 0.77 0.31

a Alcoholism-related

questionnaire (see Electronic Supplementary Material), administered before and after AWS therapy, maximum value determined b Cutoff 9 mg/l c Cutoff 92 fl d Cutoff 28 U/l Assessment and data monitoring Basic patient characteristics and the Acute Physiology and Chronic Health Evaluation II were recorded. In addition to the CIWA-Ar, the Glasgow Coma Scale (GCS) and the Ramsay Sedation Scale (RSS; see Electronic Supplementary Material) [13, 21] were recorded. The CIWA-Ar was administered by medical research staff trained by videotapes from the Addiction Research Foundation, Clinical Institute in Toronto, Ontario [8]. To monitor the subjects’ response to therapy the vital signs, RSS, and CIWA-Ar were assessed at baseline every 10 min in the first 60 min of AWS therapy, followed by every hour for the first 8 h and thereafter every 4 h. Each item of the CIWA-Ar was set at 0 if determination was not possible following intubation. Patients were monitored prospectively for the development of hallucinations, seizures, and delirium tremens as soon as the diagnosis of alcohol abuse was established. These patients usually do not re-

quire pharmacological treatment to prevent AWS [6] as they had denied alcohol dependence prior to surgery and admission to ICU [18]. Treatment regimens ITG or BTG therapy (Fig. 1) was initiated according to the median of previous studies in surgical patients [3, 10, 11] and clinical routine [22]. An initial cumulative dose of flunitrazepam boluses was given to control AWS. After 10 min depending on symptoms additional boluses of adjunctive agents were administered, haloperidol for hallucinations and clonidine for autonomic signs, and another bolus of flunitrazepam in the case of agitation. This medication regime was repeated in both groups until a CIWA-Ar score lower than 20 was achieved. As we could not exclude seizures and kindling mechanisms with a subsequently severe course [3, 10], a non-

2233

Fig. 2 Treatment decision tree. CIWA-Ar Clinical Withdrawal Assessment for Alcohol, revised scale; RSS Ramsay Sedation Score; asterisk respective drug for safety reasons

blinded continuous flunitrazepam infusion was then started for both groups (Fig. 2). Afterwards, the two different treatment procedures were carried out. All infusions were started in a dose depending on the initial necessary cumulative dose of boluses of the respective drug (Fig. 2). In the ITG infusions contained the respective drug, while the syringes for the bolus administration in this group were labeled with the drug but contained placebo (Fig. 2). According to CIWAAr and RSS, the following modifications in the continuous infusion were carried out: (a) the continuous infusion of each drug was increased by 20% if symptoms persisted longer than 10 min, (b) flunitrazepam infusion rates were decreased by 20% if the CIWAAr was lower than 10 and RSS lower than 4, (c) clonidine was decreased by 20% in the event of bradycardia (fewer than 60

beats/min) or hypotension (MAP lower than 60 mmHg), and (d) all drugs were reduced by 10% per day if the CIWA-Ar remained below 10 for 24 h. In the BTG the continuous infusions were titrated in the same way as for the ITG, containing placebo, except for the continuous flunitrazepam infusion containing the respective drug for the safety reasons mentioned above. Additional boluses of flunitrazepam, haloperidol, or clonidine (dose according to the initial cumulative bolus and clinical signs) were administered immediately after a CIWA-Ar higher than 20 was observed. The syringes for bolus administration in the BTG were labeled with flunitrazepam, haloperidol, or clonidine and contained the respective drug. The infusion rate of the placebo infusions was modified in the same way as described for the ITG (Fig. 2).

2234

In both groups maintenance therapy was adapted to achieve a CIWA-Ar lower than 10 (first priority) and a RSS between 2 and 4 (second priority). If attaining a CIWA-Ar below 10 was not possible without the GCS being 8 or lower, the patient was intubated. If any symptoms were present, additional boluses were administered. After each bolus the infusion rate of haloperidol, clonidine, and flunitrazepam was modified (Fig. 2). Rescue medication As rescue medication propofol was allowed in all patients. This medication was not blinded and was titrated as needed (50–400 mg) to manage any symptoms that may have threatened the patient (Fig. 2). Treatment failure was determined if AWS persisted despite treatment and propofol rescue medication, i.e., if the CIWA-Ar was not lower than 20 within 8 h. These patients were excluded from the study. Outcomes The primary outcome measures were severity and duration AWS. Severity was determined according to the CIWA-Ar score and duration as the number of days on which the maximal CIWA-Ar exceeded 20. Secondary outcomes included benzodiazepine, clonidine, and haloperidol requirements. Tertiary outcomes were duration of ICU treatment. Infections were determined by Centers for Disease Control criteria [23]. In addition, all patients with pneumonia had a new pulmonary infiltrate as revealed by radiography of the chest. Sepsis was defined according to the Consensus Conference of the American College of Chest Physicians/Society of Critical Care Medicine [24]. Bleeding was diagnosed if the patient required blood transfusions or surgical intervention for persistent bleeding. Cardiac complications included arrhythmias, congestive heart failure, and myocardial ischemia. Drug-related complications due to clonidine (a–d) and haloperidol (d) included: (a) bradycardia (heart rate