A comparison of lorazepam versus diazepam in the treatment of alcohol withdrawal

Full text

Benzodiazepine loading versus symptom-triggered treatment of alcohol
withdrawal: a prospective, randomized clinical trial
☆
José R. Maldonado, M.D.
a,
⁎, Long H. Nguyen, M.D.
b
, E. Merritt Schader, M.D.
c
,
John O. Brooks III Ph.D., M.D.
d
a
Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA 94305-5718, USA
b
Addiction Psychiatry, San Francisco VA-Medical Center, San Francisco, CA, USA
c
Community Solutions, Morgan Hill, CA, USA
d
Semel Institute at UCLA, Los Angeles, CA, USA
Received 3 March 2012; accepted 23 June 2012
Abstract
Objectives: The objectives were to compare the efficacy of a benzodiazepine loading versus a symptom-triggered protocol in the
management of alcohol withdrawal.
Methods: We conducted a prospective, randomized, controlled trial including 47 consecutive patients admitted to one of two tertiary care
medical centers who developed alcohol withdrawal syndrome. Patients were randomly assigned to either a benzodiazepine loading protocol
or a symptom-triggered treatment protocol. The Clinical Institute Withdrawal Assessment for Alcohol-Revised scale (CIWA-Ar) was
recorded throughout the length of stay, along with measures of autonomic system functioning.
Results: The average rate of change of CIWA-Ar scores was −1.5±1.3 for the symptom-triggered group and −2.3±2.5 for the loading group.
Average rate of change for systolic blood pressure was −2.7±5.3 for the symptom-triggered group and −2.3±6.4 for the loading group. There
was no significant difference between the rates of change for either group on either measure. Similarly, there was no significant difference in
total benzodiazepine use between groups. Within 72 h of treatment, 69.6% of patients in the loading group were free of withdrawal symptoms
versus 41.7% in the symptom-triggered group, a difference not reaching statistical significance.
Conclusions: This study did not reveal clear evidence of a clinical advantage for choosing either treatment method.
© 2012 Elsevier Inc. All rights reserved.
Keywords: Alcohol; Benzodiazepines; Withdrawal; Lorazepam; Diazepam; Symptom triggered; Loading method
1. Introduction
Alcohol use disorder (AUD) is the most serious substance
abuse problem in the United States (US) and worldwide
[1,2]. In 2008 in the US, slightly more than half (56% or 129
million) of Americans reported being current drinkers of
alcohol, while 23.3% participated in binge drinking (i.e., ≥5
on the same occasion on at least 1 day in the 30 days prior to
the survey) and 6.9% of the population reported heavy
drinking (i.e., binge drinking on at least 5 days in the past 30
days) [3]. Similarly, the Epidemiologic Catchment Area
survey found that the lifetime prevalence in the general
population of alcohol abuse or dependence is 13.6% [4].
According to a national survey of the Veterans Affairs
system, 42% of all veteran inpatients required medications to
aid detoxification from alcohol [5]. Alcoholism has been
reported in 20% to 50% of hospitalized medical patients [6].
Most alcohol-dependent patients admitted to the general
medical wards will develop alcohol withdrawal symptoms,
significant enough to require pharmacological intervention
regardless of the cause for admission [7]. Alcohol abuse and
withdrawal are associated with an increased risk for medical
comorbidities (e.g., infections, cardiopulmonary insufficiency,
cardiac arrhythmia, bleeding disorders, need for mechanical
ventilation) and longer, more complicated hospital and
intensive care unit stays [8].
Since their introduction in the 1960s, benzodiazepines
have surpassed all other available agents (e.g., barbiturates)
Available online at www.sciencedirect.com
General Hospital Psychiatry xx (2012) xxx –xxx
☆
Trial Registry: http://www.clinicaltrials.gov/ Id No. NCT00523185.
⁎Corresponding author. Tel.: +1 650 799 1582; fax: + 1 650 724 3144.
E-mail address: jrm@stanford.edu (J.R. Maldonado).
0163-8343/$ –see front matter © 2012 Elsevier Inc. All rights reserved.
doi:10.1016/j.genhosppsych.2012.06.016

and have become the first-line treatment for all phases of
alcohol withdrawal syndrome (AWS), including prevention
of alcohol withdrawal delirium and seizures [9–12]. They
have been shown to be the most efficacious available agents
and are considered the treatment of choice for alcohol
withdrawal [7,9,12–15]. A Cochrane review including 64
trials and a total of 4309 subjects demonstrated that
benzodiazepines are superior to placebo for all alcohol
withdrawal symptoms, particularly seizures [3 studies, 324
participants, relative risk 0·16 (95% confidence interval
0·04–0·69)] [16]. Yet, there is no consensus as to the best
agent from this group to use, as randomized controlled trials
have been limited in number and size [11].
There are at least two schools of thought regarding the
clinical use of benzodiazepines for the treatment of alcohol
withdrawal [10,12]. One such school favors the “loading
method.”This requires the use of a long-acting agent (e.g.,
diazepam, onset of action=1–1.5 h, T1/2=20–100 h, plus
36–200 additional h of active metabolites) which is
administered until there has been significant improvement
in withdrawal symptoms [17]. This approach postulates that
agents with long half-lives will allow for self-tapering of the
drug, translating into ease of administration and avoidance of
breakthrough symptoms due to undersedation [18–20].
Critics of this method suggest that some patients may
receive unnecessary medication, thus highlighting the
possibility of oversedation, which may result in respiratory
depression and prolonged hospitalizations [21,22].
The second is the “symptom-triggered”method, which
promotes the use of short-acting agents (e.g., lorazepam,
onset of action=2–4h,T1/2=10–15 h, no active metabolites)
administered in accordance with regular symptom monitor-
ing [e.g., Clinical Institute Withdrawal Assessment for
Alcohol-Revised scale (CIWA-Ar)] [23,24].Symptom-
triggered protocols address the potential for under- or
overmedicating by assessing symptoms on real-time and
administering benzodiazepine dosages only in response to
symptom severity [25]. Proponents of this method suggest
that it is as safe and effective as loading methods while
preventing oversedation, which translates into faster resolu-
tion of symptoms, a reduction in the duration of treatment
and quantity of medication use, and earlier discharge from
the hospital [20,25–29], thus decreasing medical resource
utilization and improving the efficiency of treatment [20].In
fact, a study using this symptom-triggered model reported a
shorter time to symptom control and a total lower medication
needed when compared to a nonprotocol infusion method
[30]. Critics highlight problems of “breakthrough”with-
drawal, the need for constant monitoring and frequent
medication administration, and a potentially greater risk of
developing benzodiazepine dependence [14,18,19,31].
The study described in this paper sought to conduct a
head-to-head comparison between the two different treat-
ment methods in a “real-life”scenario by comparing the
loading method (using a long-acting benzodiazepine agent)
to the symptom-triggered method (using a short-acting
benzodiazepine) to determine whether there are indeed
differences between these different approaches regarding
effectiveness, safety or side effects. This was done by design
as both the comparison between long- and short-acting
agents using the same method and the comparison between
two methods (i.e., loading vs. symptom triggered) have been
conducted without conclusive results [9,14,16,32–37].
2. Methods
2.1. Study patients
The study was an open, prospective, randomized clinical
trial conducted over a 12-month period at two tertiary care
medical facilities, Stanford University Medical Center
(SUMC) and the Palo Alto Veterans Affairs (PAVA)
Healthcare System, in patients who presented with alcohol
withdrawal symptoms. Patients were enrolled regardless of
the initial reason for hospitalization. Eligible patients were
inpatients with a reported history of alcohol withdrawal or
dependence, age 18 or older, who had consumed alcohol
within 24 h of admission and had the ability to consent to
participate in the study. The criteria for exclusion from the
study included pregnancy, history of dementia, reported
active abuse of other central nervous system (CNS)-
depressant agents (e.g., benzodiazepines, barbiturates, opiates),
acute intoxication with a CNS-activating agent (e.g., cocaine,
amphetamines), severe hepatic dysfunction [e.g., international
normalized ratio (INR)N2·0] or unwillingness to participate in
the study. Upon meeting eligibility criteria and consenting to
participate, baseline characteristics were gathered (Table 1).
Patients were randomized by number draw to either a
symptom-triggered or loading benzodiazepine treatment
protocol (Table 2).
2.2. Outcomes
The primary outcome measure was the baseline scores
and rates of change of the CIWA-Ar [24]. The CIWA-Ar is a
widely used scale that monitors alcohol withdrawal symp-
toms of anxiety; agitation; headache; tremor; diaphoresis;
nausea and vomiting; orientation; and tactile, auditory and
visual hallucinations. Throughout their inpatient stay,
patients in both groups were regularly evaluated by the
nursing staff using the CIWA-Ar. Additional [i.e., as needed
(PRN)] medication could only be administered if the
patient's CIWA score was elevated or if vital signs exceeded
established parameters (Table 2). In addition, the CIWA-Ar
was blindly administered three times per day (at 08:00, 14:00
and 21:00) by one of two medical students trained by the
senior author on the administration of the CIWA-Ar.
Secondary measures included measures and rate of change
of autonomic system functioning (as measured by changes in
vital signs). Blood pressure, pulse, temperature and respira-
tory rate measured by the nursing team were recorded at
corresponding times. To facilitate comparison of the usage of
2J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx

benzodiazepines between the two groups, lorazepam dosage
was converted to diazepam equivalents based on the
equipotency conversion of 1 mg of lorazepam to 5 mg of
diazepam [38].
The study was approved by the Stanford Institutional
Review Board and the Palo Alto VA Institutional Review
Board and Research and Development Committees. Written
informed consent was obtained from all patients. Additional
medication (i.e., PRN) was permitted as described in Table 2.
2.3. Statistical analysis
The average rates of change were computed for CIWA-Ar
and systolic blood pressure. For the purpose of computing
the rates of change, a patient was considered to have ended
his or her withdrawal syndrome after having been asymp-
tomatic for 24 h (i.e., receiving four consecutive CIWA-Ar
scores of zero). The rates of change were compared using a
Student's ttest, and the effect of covariates was assessed
through the use of multiple regressions. Results were
analyzed for patients who completed the study, and
intention-to-treat analysis was performed for all patients
enrolled regardless of study completion. Power calculations,
however, demonstrated a 70% chance of detecting an effect.
3. Results
Over the study period, 47 inpatients at SUMC and PAVA
who presented with alcohol withdrawal symptoms complet-
ed the study. Fig. 1 represents the enrollment and
randomization flow diagram. Twenty-four patients received
the “symptom-triggered”intervention, while 23 patients
received the loading method intervention. The patient
sample comprised 46 males and 1 female whose mean age
was 51.7 years (S.D.=9.6, range: 32–81 years). There were
40 Caucasians, 4 Hispanics, 2 African–Americans and 1
Indian–American. Most patients were enrolled at PAVA
(N=34), and the remainder were studied at SUMC (N=13).
Patients were admitted to a variety of services, including
addiction treatment services (N=24), psychiatry (N= 14),
general medicine (N=3), trauma/surgery (N= 3) and medical
intensive care (N=3). The distribution of patients at hospital
site and service was nearly matched between treatment groups.
As shown in Table 1, the two treatment groups were
similar with respect to age, sex, race, body mass index,
primary reason for hospitalization, hospital site, hospital
service and initial CIWA-Ar score. The two treatment groups
were also similar with respect to histories reported for
previous withdrawal symptoms [i.e., minor withdrawal,
seizures, hallucinosis or delirium tremens (DT)], number of
previous detoxifications, years of alcohol abuse or depen-
dence, longest period of sobriety, comorbid psychiatric
diagnoses, abuse of other substances and prevalence of
homelessness. There was a slight difference between the two
groups in blood alcohol level (BAL) and liver function tests
(i.e., INR) at the time of admission; however, these
characteristics were not statistically significantly different
between groups.
Table 1
Baseline patient characteristics
Characteristics Sx-triggered
Group (n=24)
Loading
Group (n=23)
Age, mean (S.D.), y 51.0 (11.3) 52.5 (8.0)
Female, n(%) 1 (4.2) 0 (0.0)
Race or ethnic group
a
,n(%)
Caucasian 21 (87.5) 19 (82.6)
African–American 1 (4.2) 1 (4.3)
Latino 2 (8.3) 2 (8.7)
Indian–American 0 (0.0) 1 (4.3)
Primary reason for hospitalization, n(%)
Detoxification 16 (66.7) 18 (78.3)
Other psychiatric 4 (16.7) 1 (4.3)
Medical/trauma 4 (16.7) 4 (17.4)
Initial CIWA-Ar score,
mean (S.D.)
8.0 (5.3) 8.2 (4.7)
Body mass index, mean (S.D.) 24.6 (3.9) 26.0 (4.8)
Blood alcohol level on admission,
mean (S.D.)
0.115 (0.149) 0.068 (0.092)
INR, mean (S.D.) 1.25 (0.97) 1.01 (0.08)
Reported alcohol withdrawal
history, n(%)
Minor withdrawal 19 (79.2) 18 (78.3)
Hallucinations 2 (8.3) 9 (39.1)
Seizures 7 (29.2) 7 (30.4)
DT 3 (12.5) 7 (30.4)
Previous detoxifications,
mean (S.D.)
2.0 (2.0) 2.5 (2.5)
Years of alcohol abuse,
mean (S.D.)
26.9 (12.8) 26.8 (11.9)
Longest period of sobriety
in months, mean (S.D.)
15 (23) 26 (35)
Comorbid psychiatric diagnosis, n(%) 15 (62.5) 15 (65.2)
History of other substance abuse, n(%) 9 (37.5) 7 (30.4)
Homeless, n(%) 8 (33.3) 8 (34.8)
Sx=symptom.
a
Race or ethnic group was self-reported.
Table 2
Alcohol withdrawal treatment protocols
Loading protocol
On day 1: Load with diazepam 20 mg by mouth every 2 h × 3 doses. (For
parenteral treatment, load with diazepam 10 mg intravenously every 1 h ×
6 doses). Hold for excessive sedation or RRb10.
Additional diazepam 10 mg by mouth or intravenously may be administered
every 2 h as needed for residual withdrawal symptoms (e.g., CIWA-ArN8
or vital signs alteration suggestive of a hyperadrenergic state, such as
SBPN140, DBPN90, HRN100). Hold for excessive sedation or RRb10.
Symptom-triggered protocol
Lorazepam 1 to 2 mg, by mouth or intravenously, may be administered
every 2 h as needed for active withdrawal symptoms (e.g., CIWA-ArN8or
vital signs alteration suggestive of a hyperadrenergic state, such as
SBPN140, DBPN90, HRN100). Hold for excessive sedation or RRb10.
RR=respiratory rate; SBP=systolic blood pressure; DBP=diastolic blood
pressure; HR=heart rate.
3J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx

3.1. Outcome results
While intention-to-treat analysis included all randomized
subjects, four postrandomization subjects were excluded
from regression analysis because of failure by the treatment
team to adhere to the treatment protocol, such as usage of
multiple benzodiazepine types (N= 2 in the loading group,
N=2 in the a symptom-triggered group).
Results are summarized in Table 3. Overall, 55.3% of all
study patients were free of withdrawal symptoms within 72 h
of admission: 69.6% in the loading group and 41.7% in the
symptom-triggered group. This difference in proportions,
however, did not reach statistical significance (P=.08). For
the entire duration of symptoms, the average rate of change
of CIWA-Ar score was −2.3 points per day (S.D. 2.5) for the
loading group and −1.5 points per day (S.D. 1.3) for the
symptom-triggered group. Although the rate of change of
CIWA-Ar for the loading group was more rapid than that of
the symptom-triggered group, the difference did not reach
statistical significance (PN.05). The average rate of change
of systolic blood pressure was −2.3 (S.D. 6.4) for the loading
group and −2.7 (S.D. 5.3) for the symptom-triggered group,
which did not significantly differ from one another (PN.05).
Furthermore, there were no statistically significant differ-
ences in other vital signs such as heart rate, diastolic blood
pressure and respiratory rate among either treatment group.
Intention-to-treat analysis did not reveal a significant
difference between groups in any of the above measures.
Comparing total benzodiazepine dosages (as calculated
by converting lorazepam dosage to diazepam equivalents
based on the equipotency conversion of 1 mg of lorazepam
to 5 mg of diazepam [38]), there was no significant
difference in the average total benzodiazepine usage between
the two groups (PN.05), with 103.8 mg (S.D. 71.5) for the
loading group and 92.4 mg equivalents (S.D. 103.5) for the
symptom-triggered group.
Length of stay was not calculated given the standard
duration of hospitalizations in addiction treatment programs
(e.g., 14-day programs), which accounted for the majority of
Fig. 1. Enrollment, randomization and completion of study.
Table 3
Summary of results
Outcome measure Lorazepam group
(n=24)
Diazepam group
(n=23)
P
value
CIWA-Ar rate of change,
mean (S.D.)
−1.5 (1.3) −2.3 (2.5) N.05
SBP rate of change,
mean (S.D.)
−2.7 (5.3) −2.3 (6.4) N.05
Total benzodiazepine use,
mean (S.D.)
92.4 (103.5) 103.8 (71.5) N.05
Absence of symptoms
within 72 h, n(%)
10 (41.7) 16 (69.6) N.05
4J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx

patients at Veterans Affairs hospitals. Total cost of treatment
was not calculated for similar reasons.
To determine whether baseline characteristics effected the
rate of change of CIWA-Ar scores, we performed multiple
regression analyses in which we included treatment group
(symptom-triggered vs. loading method); blood alcohol level
at admission; liver function at admission (INR, liver function
tests); number of years drinking; history of hallucinations,
seizures or delirium tremens; and the interactions of the
covariates with the treatment group. None of the covariates
or their interaction terms were statistically significant
(PN.05).
The effects of hospital site (SUMC vs. PAVA), type of
inpatient ward/service (medical vs. psychiatric) and route of
administration (parenteral vs. oral) on outcome measures did
not differ significantly between or within groups (PN.05).
The two groups did not differ with regards to use of other
medications or standard supportive therapies, such as
nutrition, hydration, thiamine supplementation, treatment
for other comorbid medical problems (e.g., use of antihy-
pertensive medication) or treatment of comorbid psychiatric
disorders (e.g., use of tricyclic antidepressants or antipsy-
chotics) (PN.05).
3.2. Adverse events
No patients experienced seizures during their inpatient
stay. Two patients, one in each treatment group, developed
treatment-refractory delirium tremens and required transfer
to the intensive care unit. Another patient treated with
lorazepam was concomitantly agitated and excessively
sedated, resulting in a fall that caused a minor abrasion to
the head. There were no events of respiratory depression.
Otherwise, there were no remarkable adverse events
affecting the patients or outcome measures of this study.
4. Discussion
Clinicians have been divided between the two main
treatment methods. The loading method requires the use of a
long-acting agent (e.g., diazepam, chlordiazepoxide), which
is administered at a predetermined dose or until there has
been significant improvement in withdrawal symptoms, thus
allowing for “loading”of the agents. It is postulated that
agents with long half-lives will allow for self-tapering or
elimination of the drug (linked to its long half-life).
Conversely, the symptom-triggered method advocates for
the use of short-acting agents (e.g., lorazepam) and allows
for multiple doses based on the patient's need (i.e.,
symptoms) as measured by an objective withdrawal scale
or clinical assessment. This prospective, randomized,
controlled trial, which closely followed patients through
their subjective and objective signs and symptoms of alcohol
withdrawal, did not reveal any evidence of a clinical
advantage for a benzodiazepine-based treatment protocol
method (i.e., symptom-triggered versus loading method).
Although there was more rapid resolution of symptoms with
the diazepam loading protocol in that a greater percentage of
patients were free of symptoms within 72 h in the diazepam
group, the difference was not statistically significant.
Past studies comparing long- and short-half-life benzodi-
azepines in the treatment of alcohol withdrawal have been
conflicting. Some studies show an advantage to using long-
half-life drugs such as diazepam or chlordiazepoxide by
providing a smoother detoxification, decreased risk of
seizure, minimization of breakthrough symptoms and
superior cost-effectiveness [13,16,17]. Other studies have
shown an advantage to using symptom-triggered approaches
with short-half-life drugs such as lorazepam in order to
minimize total benzodiazepine usage, length of hospital stay
and hepatic stress [21,22,39]. In this study, adverse events of
falls and treatment-refractory delirium tremens in both
treatment groups demonstrate that neither treatment is
without risk or always effective.
Limitations of this study include the unblinded design and
the small sample size. However, the majority of items on the
CIWA-Ar, as well as objective measurements, were not
subject to experimenter interpretation, so effects from lack of
blinding would be limited. Generalizability may be limited
due to the majority of patients being Caucasian (85.1%) and
male (95.8%). The heterogeneous medical services included
in the study could contribute to the large variance; however,
the effects of hospital site, inpatient service and route of
administration on outcome measures did not differ signifi-
cantly between or within groups (PN.05). The population
under study had relatively low BALs on admission, which
are consistent with the reason for admission (e.g., treatment
for alcohol withdrawal). In addition, our patients required
relatively small amounts of benzodiazepines, suggesting that
timely interventions were associated with lower required
doses. Our sample also had levels of premorbid DTs or
seizures that were relatively low, but consistent with reported
statistics among alcohol withdrawal population (e.g.,
incidence of withdrawal seizures up to 10%, DTs occurring
in about 5%) [40]. Moreover, the prevalence of comorbid
medical problems also appeared small as many were
primarily admitted for detoxification.
We did not include a placebo group because of the ethics
involved in withholding treatment given the high morbidity
and mortality. With respect to medication administration, we
used simultaneous comparisons —treatment method
(loading vs. symptom triggered) and agent choice (i.e.,
long acting vs. short acting) —because of clinical necessity.
By virtue of its pharmacokinetics, if a long-acting agent is
used in the symptom-triggered protocol, its long half-life in
effect would have turned it into a loading protocol. Similarly,
part of the comparison was, in fact, whether a disadvantage
of the symptom-triggered method using a short-acting agent
will allow for a greater incidence of breakthrough symptoms
(including autonomic instability or seizures).
In addition, many alternative agents have been used,
mostly as adjuvant to rather than as alternative to
5J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx

benzodiazepines. These include barbiturates (e.g., pheno-
barbital), anesthetic agents (e.g., propofol), muscle relaxants
(e.g., baclofen), β-blockers and α-agonists, other nonbenzo-
diazepine GABA-agonist agents (e.g., chlormethiazole),
gamma-hydroxybutyric acid and, of course, ethanol itself
[41–45]. The use of these agents is beyond the scope of this
report and has been extensively discussed elsewhere
[10,12,40,41,46–48]. Yet, to date, experts continue to
agree that benzodiazepines are the treatment of choice
[10,12,15,36]. In fact, available data suggest that benzodi-
azepines are the only pharmacological agents known to
provide “protective benefit against alcohol withdrawal
symptoms, in particular seizures”and DT [14,36,49].
Despite the conflicting opinions on choice of agent, it is
clear that aggressive treatment with benzodiazepines for
alcohol withdrawal syndromes is useful in minimizing both
morbidity and mortality. As reported by others, our results
confirm that provided individual differences in metabolism
and hepatic function, both loading and symptom-triggered
protocols are effective in the treatment of all phases of AWS.
Furthermore, this study suggests that early, aggressive use of
a loading protocol with a long-acting benzodiazepine may
improve the initial symptoms of withdrawal more rapidly
than the symptom-triggered use of a short-acting agent, but
that neither method produces a significant difference in
overall rate of improvement and duration of symptoms. Our
findings, based on a head-to-head comparison of two
commonly used alcohol withdrawal protocols, suggest that
the choice of protocol (i.e., loading vs. symptom triggered)
in the management of alcohol withdrawal may not be as
critical as anecdotally assumed for the reduction of
symptoms or for the minimization of total benzodiazepine
use. However, it confirms previous studies’findings
suggesting that the success of “symptom-triggered regimens
is based on their protocol-driven, individualized nature and
are distinct from purely ad hoc PRN orders”[11,27].
References
[1] Lieber CS. Medical disorders of alcoholism. N Engl J Med
1995;333(16):1058–65.
[2] Williams GD, et al. In: N.I.o.A.A.a. Alcoholism, editor. Apparent per
capita alcohol consumption: national, state and regional trends, 1977–
94; 1996, pp. 3–6.
[3] SAMHSA. Results from the 2008 National Survey on Drug Use and
Health: national findings. Rockville (MD): Substance Abuse and
Mental Health Services Administration (SAMHSA); 2009.
[4] Robins LN, et al. Lifetime prevalence of specific psychiatric disorders
in three sites. Arch Gen Psychiatry 1984;41(10):949–58.
[5] Tracy SW, Trafton JA, Humphreys K. The Department of Veterans
Affairs Substance Abuse Treatment System: results of the 2003 Drug
and Alcohol Program Survey. Palo Alto (CA): U.S. Department of
Veterans Affairs, Veterans Affairs Health Care System, Program
Evaluation and Resource Center; 2004.
[6] Moore RD, et al. Prevalence, detection, and treatment of alcoholism in
hospitalized patients. JAMA 1989;261(3):403–7.
[7] Saitz R, O'Malley SS. Pharmacotherapies for alcohol abuse.
Withdrawal and treatment. Med Clin North Am 1997;81(4):881–907.
[8] O'Brien Jr JM, et al. Alcohol dependence is independently associated
with sepsis, septic shock, and hospital mortality among adult intensive
care unit patients. Crit Care Med 2007;35(2):345–50.
[9] Holbrook AM, et al. Meta-analysis of benzodiazepine use in the
treatment of acute alcohol withdrawal. Cmaj 1999;160(5):649–55.
[10] Maldonado J, DiMartini A, Owen J. Psychopharmacological treatment
of substance use disorders in the medically ill. In: & Ferrando S, et al,
editor. Clinical handbook of psychopharmacology in the medically ill.
VA: APPI; 2010, pp. 537–55.
[11] McKeon A, Frye MA, Delanty N. The alcohol withdrawal syndrome.
J Neurol Neurosurg Psychiatry 2008;79(8):854–62.
[12] Maldonado J. An approach to the patient with substance use and abuse.
Med Clin North Am 2010:94(6):1169-205, x-i.
[13] Kosten TR, O'Connor PG. Management of drug and alcohol
withdrawal. N Engl J Med 2003;348(18):1786–95.
[14] Mayo-Smith MF. Pharmacological management of alcohol withdraw-
al. A meta-analysis and evidence-based practice guideline. American
Society of Addiction Medicine Working Group on Pharmacological
Management of Alcohol Withdrawal. JAMA 1997;278(2):144–51.
[15] Stern TA, et al. Current approaches to the recognition and treatment of
alcohol withdrawal and delirium tremens: "old wine in new bottles" or "new
wine in old bottles". Prim Care Companion J Clin Psychiat 2010;12(3).
[16] Ntais C, et al. Benzodiazepines for alcohol withdrawal. Cochrane
Database Syst Rev 2005(3):CD005063.
[17] Sellers EM, et al. Diazepam loading: simplified treatment of alcohol
withdrawal. Clin Pharmacol Ther 1983;34(6):822–6.
[18] Miller NS. Pharmacotherapy in alcoholism. J Addict Dis 1995;14(1):
23–46.
[19] Ritson B, Chick J. Comparison of two benzodiazepines in the
treatment of alcohol withdrawal: effects on symptoms and cognitive
recovery. Drug Alcohol Depend 1986;18(4):329–34.
[20] Saitz R, et al. Individualized treatment for alcohol withdrawal. A
randomized double-blind controlled trial. JAMA 1994;272(7):519–23.
[21] Miller Jr WC, McCurdy L. A double-blind comparison of the efficacy
and safety of lorazepam and diazepam in the treatment of the acute
alcohol withdrawal syndrome. Clin Ther 1984;6(3):364–71.
[22] Solomon J, Rouck LA, Koepke HH. Double-blind comparison of
lorazepam and chlordiazepoxide in the treatment of the acute alcohol
abstinence syndrome. Clin Ther 1983;6(1):52–8.
[23] McKay A, Koranda A, Axen D. Using a symptom-triggered approach
to manage patients in acute alcohol withdrawal. Medsurg Nurs
2004:13(1):15-20, 31; quiz 21.
[24] Sullivan JT, et al. Assessment of alcohol withdrawal: the revised
clinical institute withdrawal assessment for alcohol scale (CIWA-Ar).
Br J Addict 1989;84(11):1353–7.
[25] Jaeger TM, Lohr RH, Pankratz VS. Symptom-triggered therapy for
alcohol withdrawal syndrome in medical inpatients. Mayo Clin Proc
2001;76(7):695–701.
[26] Daeppen JB, et al. Symptom-triggered vs fixed-schedule doses of
benzodiazepine for alcohol withdrawal: a randomized treatment trial.
Arch Intern Med 2002;162(10):1117–21.
[27] Fitzgerald FT. As the occasion arises: PRN sedative orders in alcohol
withdrawal treatment. Arch Intern Med 2002;162(10):1093–4.
[28] Hardern R, Page AV. An audit of symptom triggered chlordiazepoxide
treatment of alcohol withdrawal on a medical admissions unit. Emerg
Med J 2005;22(11):805–6.
[29] Wartenberg AA, et al. Detoxification of alcoholics: improving care by
symptom-triggered sedation. Alcohol Clin Exp Res 1990;14(1):71–5.
[30] DeCarolis DD, et al. Symptom-driven lorazepam protocol for
treatment of severe alcohol withdrawal delirium in the intensive care
unit. Pharmacotherapy 2007;27(4):510–8.
[31] Shaw GK. Detoxification: the use of benzodiazepines. Alcohol
Alcohol 1995;30(6):765–70.
[32] Bird RD, Makela EH. Alcohol withdrawal: what is the benzodiazepine
of choice? Ann Pharmacother 1994;28(1):67–71.
[33] Erstad BL, Cotugno CL. Management of alcohol withdrawal. Am J
Health Syst Pharm 1995;52(7):697–709.
6J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx

[34] Lejoyeux M, Solomon J, Ades J. Benzodiazepine treatment for
alcohol-dependent patients. Alcohol Alcohol 1998;33(6):563–75.
[35] Liskow BI, Goodwin DW. Pharmacological treatment of alcohol
intoxication, withdrawal and dependence: a critical review. J Stud
Alcohol 1987;48(4):356–70.
[36] Mayo-Smith MF, et al. Management of alcohol withdrawal delirium.
An evidence-based practice guideline. Arch Intern Med 2004;164(13):
1405–12.
[37] Saitz R, Friedman LS, Mayo-Smith MF. Alcohol withdrawal: a
nationwide survey of inpatient treatment practices. J Gen Intern Med
1995;10(9):479–87.
[38] Colon EA, Popkin MK. Anxiety and panic. In: Rundell JR, & Wise
MG, editors. Textbook of consultation–liason psychiatry. Washington,
DC: American Psychiatric Association Press; 1996, pp. 402–25.
[39] Ahmed S, Chadwick D, Walker RJ. The management of alcohol-
related seizures: an overview. Hosp Med 2000;61(11):793–6.
[40] Sarff M, Gold JA. Alcohol withdrawal syndromes in the intensive care
unit. Crit Care Med 2010;38(9 Suppl):S494–501.
[41] Amato L, Minozzi S, Davoli M. Efficacy and safety of pharmacolo-
gical interventions for the treatment of the alcohol withdrawal
syndrome. Cochrane Database Syst Rev 2011(6):CD008537.
[42] Caputo F, Bernardi M. Medications acting on the GABA system in
the treatment of alcoholic patients. Curr Pharm Des 2010;16(19):
2118–25.
[43] Caputo F, et al. Gamma hydroxybutyric acid (GHB) for the treatment
of alcohol dependence: a review. Int J Environ Res Public Health
2009;6(6):1917–29.
[44] Leone MA, et al. Gamma-hydroxybutyrate (GHB) for treatment of
alcohol withdrawal and prevention of relapses. Cochrane Database
Syst Rev 2010(2):CD006266.
[45] Liu J, Wang L. Baclofen for alcohol withdrawal. Cochrane database of
systematic reviews 2011(1):CD008502.
[46] Ait-Daoud N, Malcolm Jr RJ, Johnson BA. An overview of
medications for the treatment of alcohol withdrawal and alcohol
dependence with an emphasis on the use of older and newer
anticonvulsants. Addict Behav 2006;31(9):1628–49.
[47] Minozzi S, et al. Anticonvulsants for alcohol withdrawal. Cochrane
database of systematic reviews. 2010(3):CD005064.
[48] Muzyk AJ, et al. Role of {alpha}2-agonists in the treatment of acute
alcohol withdrawal. Ann Pharmacother 2011;45(5):649–57.
[49] Amato L, et al. Benzodiazepines for alcohol withdrawal. Cochrane
Database Syst Rev 2010;3:CD005063.
7J.R. Maldonado et al. / General Hospital Psychiatry xx (2012) xxx–xxx