Alcohol enhances the activation of GABA on inhibitory type A gamma-aminobutyric acid (GABA-A) receptors and inhibits excitatory N-methyl-d-aspartate (NMDA) and non-NMDA glutamate receptors in the brain leading to overall decreased central nervous system (CNS) excitability. (*11) As CNS tolerance develops from prolonged or repetitive exposure to alcohol, the brain adapts by up-regulating glutamate receptors and down-regulating GABA-A receptors. In the absence of alcohol, this combination of neurobiological events results in a new homeostatic state of CNS hyper-excitability that can manifest as anxiety, agitation, tremors, irritability, seizures and delirium tremens (DT). Repeated detoxifications also result in long-term neuronal changes often referred to as “kindling.” Kindling is the establishment of stable neurobehavioral alterations secondary to repeated electrophysiologic stimulation. Neuronal kindling is hypothesized to increase alcohol craving (*12) and worsen subsequent episodes of AWS (*13).

The Diagnostic and Statistical Manual of Mental Disorders IV, Text Revision (DSM-IV) defines criteria for diagnosis of AWS (Table 1). (*14) AWS can occur upon cessation or reduction in alcohol consumption and may appear while the patient has a measurable alcohol level. Symptoms are associated with CNS hyper-excitability and can vary from mild manifestations such as anxiety and tremors, to fatal conditions such as delirium tremens. (*13) Symptoms can begin as early as a few hours after cessation or reduction in alcohol consumption or appear several days afterwards (Table 2). Often, symptoms develop after patients are hospitalized for other reasons. In general, severity of withdrawal symptoms is correlated with amount and duration of alcohol consumption. (*13) Approximately 7% of patients with AWS develop alcoholic hallucinations, 5-10% develop withdrawal seizures, and 5% develop alcohol withdrawal delirium better known as delirium tremens (DT). (*15)

Figure 1

Table 1

Figure 2

Table 2

Withdrawal seizures are more common in patients with a history of multiple episodes of AWS (*13). They are typically grand-mal type and may occur several times daily after alcohol cessation. Alternate causes should be sought if focality is identified. An effort to identify DT is critical as this condition carries a 1 to 5% mortality rate. (*16) Alcoholic hallucinosis consists primarily of visual, auditory, or tactile hallucinations and can be differentiated from DT by the absence of sensorium clouding, disorientation, and fluctuating awareness. Risk factors associated with the development of DT include old age, history of DT or withdrawal seizures, presence of severe withdrawal symptoms on presentation, current acute medical illness, history of heavy daily alcohol use, and decreased hepatic metabolism. (*13)

Withdrawal from benzodiazepines and barbiturates and intoxication of amphetamines, cocaine, or anticholinergic drugs can lead to increased sympathetic drive and mental status changes which can mimic AWS. CNS infection, trauma, and thyrotoxicosis can also present similarly to AWS. These conditions should be ruled out before solidifying a diagnosis of AWS.

The Clinical Institute Withdrawal Assessment for Alcohol-Revised (CIWA-Ar) is the most commonly used validated method of quantifying severity of AWS (Figure 1). (*17, *18). The CIWA-Ar can also be used to monitor the progress of patients and determine treatment plans for AWS. (*17, *18, *19, *20) The CIWA-Ar is a cumulative rating of ten AWS symptoms each scored on a scale of 0 to 7 or 0 to 4. Trained staff can reliably conduct this assessment in less than one minute. A CIWA-Ar score below 8 indicates mild AWS; a score between 8 and 15 signifies moderate AWS; a score above 15 constitutes severe withdrawal and a higher risk for the development of DT. The categories of AWS are associated with treatment recommendations, with patients demonstrating moderate or severe AWS requiring treatment with benzodiazepine (discussed below). While the CIWA-Ar is a helpful tool, clinicians and medical staff should realize that symptoms listed are not specific to AWS. Co-morbid medical (e.g. hyperthyroid) or psychiatric (e.g. psychosis) illness and medications (e.g.  blockers) can falsely elevate or decrease CIWA-Ar scores and therefore interfere with proper treatment if not accounted for by the evaluator.

Figure 3

Figure 1: The Clinical Institute Withdrawal Assessment for Alcohol-Revised (CIWA-Ar)

Ref #17: 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). Br J Addict 1989;84:1353-7.

There are three goals of drug and alcohol detoxification as described by the American Society of Addiction Medicine (ASAM): 1) “To provide a safe withdrawal from the drug[s] of dependence and enable the patient to become drug-free” 2) “To provide a withdrawal that is humane and thus protects the patient's dignity” and 3) “To prepare the patient for ongoing treatment of his or her dependence on alcohol or other drugs.” (*16)

All patients who are admitted for acute withdrawal (seizures, DT’s) or other comorbidities related to unhealthy alcohol consumption should be evaluated and treated following a standardized protocol. Basic laboratory tests such as complete blood counts, blood alcohol level, urine drug screen, transaminases, bilirubin, albumin, coagulation tests, and electrolyte levels should be performed on admission. Particular attention should be paid to potassium, bicarbonate, magnesium and phosphate, as these electrolytes are commonly decreased in chronic alcohol use secondary to emesis, volume depletion, and malnutrition.

Patients should be treated with thiamine, folate and multivitamins. Hypomagnesia may predispose patients to seizures (in addition to cardiac arrhythmias). While supplementation with magnesium is appropriate if hypomagnesia is present, routine administration of magnesium has not been proven to improve AWS. (*21) Intravenous fluids are indicated in patients with excessive fluid losses (*13). If glucose-containing fluids are necessary, it is suggested they be administered after thiamine to prevent precipitation of Wernicke's encephalopathy. (*13, *22) Patients at risk for AWS can benefit from placement in quiet environments with reduced light and limited interactions with others. (*23) Based on risk factors for specific populations (e.g. indigent populations) and comorbid medical issues, tuberculosis testing and administration of appropriate vaccines can be performed as well.

In the majority of patients with mild to moderate withdrawal symptoms, outpatient detoxification is appropriate, assuming the patient is motivated and has a supportive home environment. This method has been proven to be effective, safe and less costly than inpatient treatment. (*8, *24, *25, *26) If outpatient treatment is undertaken, only small quantities of medication should be prescribed following a fixed-dosing regimen (described in greater detail below) and close monitoring of the patient by another individual should be confirmed.

Relative indications for inpatient detoxification include a history of severe withdrawal symptoms, seizures, DT, multiple previous detoxifications, co-morbid psychiatric or medical illnesses, and lack of a reliable support network. (*23)

Prompt pharmacological treatment is indicated in all cases of AWS, as nontreatment or undertreatment can be fatal. (*27, *28) Benzodiazepines are safe, effective and the preferred treatment for AWS. (*7) Benzodiazepines are cross-tolerant with alcohol and modulate anxiolysis by stimulating GABA-A receptors. (*29) They are proven to reduce withdrawal severity and incidence of both seizures and DT. (*7, *30, *31, *32, *33) All subclasses of benzodiazepines appear to be equally effective in treating AWS (*7). Therefore, choosing a benzodiazepine depends on selection of preferred pharmacokinetic properties in relation to the patient being treated. (*34) Longer acting benzodiazepines such as diazepam and chlordiazepoxide are most often preferred because they allow for easy tapering (*35), have lower potential for abuse (*7) than short acting benzodiazepines, and cause fewer rebounds and withdrawal seizures upon discontinuation. Short acting agents such as lorazepam or oxazepam do not undergo hepatic metabolism like most long acting benzodiazepines and are more appropriate for patients with significantly decreased liver function. (*36) Short acting agents also avoid the risk of excessive sedation caused by longer acting agents and may be beneficial in the elderly or cognitively impaired. (*37)

Two regimens, fixed-scheduled and symptom-triggered, are commonly employed for benzodiazepine dosing AWS. In fixed-scheduled regimens, medication doses are administered at specific intervals and additional doses are given as needed. In symptom-triggered regimens, medications are dosed based upon the patient’s current manifestations of symptoms, which can be evaluated by the CIWA-Ar. (*38)

A symptom-triggered regimen is preferred in most cases of AWS because it results in the administration of less total medication and shorter duration of treatment. (*39, *40) This regimen may also reduce the risk of undermedicating or overmedicating a patient since dosing is based upon withdrawal symptoms. (*41) The efficacy of symptom-triggered regimens, however, depends on the validity of patient assessment. Consequently, a fixed-dose regimen may be preferred if CIWA-Ar scores cannot be accurately performed (i.e. lack of training, outpatient care setting, co-morbid medical or psychiatric illnesses or use of medications that may affect COWA-Ar measurements. (*7)

Institutional variations in symptom-triggered regimens often create confusion regarding the standard of care when using the symptom-triggered regimens but guidelines from the ASAM are available. As described earlier, the CIWA-Ar is a well-validated method of assessing severity of AWS using a cumulative rating of ten AWS symptoms. (*7) In general, CIWA-Ar scores below 8 should be treated with supportive care; CIWA-Ar scores between 8 and 15 require treatment with benzodiazepines in addition to supportive care (institutional dependent; e.g. diazepam 10-20 mg q1hour, lorazepam 2-4 mg q1h); CIWA-Ar scores above 15 require hourly monitoring for respiratory failure in addition to pharmacotherapy and supportive care.

Several medications can be used as adjunctive treatments of AWS including haloperidol, atenolol, clonidine and phenytoin. Haloperidol is often used for managing severe agitation and hallucinations; this drug can, however, reduce the seizure threshold. (*13) Atenolol, a -blocker, improves vital signs more rapidly and provides better craving control than benzodiazepines alone. (*42) Atenolol should be considered in patients with coronary artery disease, who may not be able to tolerate the strain that sympathetic withdrawal symptoms can place on the cardiovascular system. (*23). Clonidine has also been shown to improve the autonomic symptoms of withdrawal. (*7) Although phenytoin does not treat withdrawal seizures and does not need to be routinely started, it is an appropriate to continue it as an adjunct in patients who have an underlying seizure disorder. (*13)

Anticonvulsants have been proposed as an alternative to benzodiazepines because of their low abuse potential, minimal effect on cognition, and ability to reduce kindling. (*43, *44) One randomized control trial demonstrated the effectiveness of carbamazepine as an alternative to benzodiazepines in the treatment of mild to moderate AWS. (*45) No statistically significant benefit was demonstrated during systemic review of carbamazepine, however; benzodiazepines remain the mainstay of treatment.

The acute management of DT in the inpatient setting utilizes many of the same medications used in AWS. In the United States, benzodiazepines have been established as the medication of choice for the prevention and treatment of DT. They have been proven to be safe and effective in preventing complications of DT. Additionally, haloperidol may be used as an adjunctive treatment with benzodiazepines to treat psychiatric symptoms of DT, including anxiousness, hallucinations, and combativeness. Phenobarbital has a theoretical benefit due to similar underlying mechanism as benzodiazepines, though literature has shown application of phenobarbital leads to negative outcomes such as respiratory depression and coma more often than benzodiazepines and thus should be administered only in a controlled ICU setting as intubation may become necessary. Propafol has also been shown to be a useful agent for treating refractory DT in the intubated patient, particularly in the inpatient setting. Finally, anti-adrenergics, beta blockers, and clonidine may help treat autonomic hyperactivity symptoms, but do not treat delirium. Carbemazepine has been shown to be effective in treating early AWS, though it has not been proven efficacious in treating DT. (*47)

Psychotherapy complements medical treatment in attaining successful outcomes. (*48) Physicians should be familiar with and capable of scheduling outpatient treatment for all methods of psychotherapy. Additionally, they should be able to perform brief interventional psychotherapy confidently as this can be completed while the patient is hospitalized for withdrawal. The National Institute for Alcohol Abuse and Alcoholism (NIAAA) has developed an outline of brief intervention components and goals. (*49) In general, a brief intervention includes an initial counseling session of ten to fifteen minutes to assess alcohol use, and a session to provide feedback, advice, and goal-setting. Follow-up by office visit or phone conversation is also encouraged. (*4) A meta-analysis of brief intervention demonstrated a mean reduction of four drinks of alcohol per week with significant results achieved in men but not women. (*50)

All patients presenting with alcohol-related complaints should be assessed for alcohol abuse and dependence. Physicians should begin by assessing the patient’s history for evidence of alcohol-related problems. The quantity of alcoholic intake, history of alcohol use, time since last drink, presence of co-morbid medical or psychiatric conditions, previous alcohol withdrawals, and abuse of other agents should be elicited from the patient. The NIAAA defines at risk drinking as greater than 14 drinks per week or four drinks per occasion for men and greater than seven drinks per week or three drinks per occasion for women. (*52) A standard drink is defined as 0.5 ounces (oz) of alcohol, which is equivalent to 12 oz of beer, 5 oz of wine and 1.5 oz of 80-proof distilled spirit. (*82)

Specific tools and criteria are available to accurately identify and diagnose alcohol abuse and dependence. The CAGE questionnaire is one of many screening tools that can be used in primary care, hospitalized and pre-operative settings to identify alcohol addiction. (*53, *54) Patients are given a point for each of four screening questions (Table 3) that they identify with. A score of 2 or greater suggests that a more detailed assessment with alcohol use is warranted. (*53).

Figure 4

Table 3

Other screening tools that have been utilized include the Michigan Alcoholism Screening Test (MAST), a modified short MAST (sMAST), Alcohol Use Disorders Identification Test (AUDIT), and several others. The CAGE, MAST, and sMAST are the most widely used, with the MAST being more sensitive than CAGE, except in elderly patients, where CAGE may be more effective. Overall the CAGE, MAST, and sMAST perform comparably (*81). Multiple other screening modalities exist with differing sensitivities and specificities based on different patient populations.

Differentiating alcohol abuse from dependency is confusing to some clinicians. In general, abuse refers to misuse of alcohol that leads to impaired functioning, while dependence refers to an inability to control alcohol use for either physical or psychological reasons. Strict criteria for diagnosis of alcohol abuse (Table 4) and alcohol dependence (Table 5) are defined in DSM-IV. (*14) A diagnosis of alcohol dependence supercedes the diagnosis of alcohol abuse.

Figure 5

Table 4

Figure 6

Table 5

Once a diagnosis of alcohol abuse or dependence has been made and the patient is amenable to treatment, the physician should treat the patient for co-morbid medical and psychiatric illnesses, refer the patient to an appropriate addiction therapy program, and discuss the pharmacotherapy options. These decisions can be made in consultation with addiction and psychiatry specialists. While most treatments are completed as outpatient, the physician or member of the medical staff should initiate them before discharge. Though brief intervention is helpful in patients with alcohol abuse (*55), more intensive treatment is needed to address alcohol dependence. Other psychological pharmacological treatments are available and complement each other in attaining successful outcomes in these patients. (*48)

The COMBINE study is the largest randomized trial comparing placebo, behavioral therapy, drug treatment with naltrexone or acamprosate, a combination of these methods, and no intervention for treatment of alcohol dependence. (*56) Results of this study demonstrated that naltrexone is as effective as behavioral therapy in achieving a high number of days of abstinence from alcohol, but was no more effective when combined with behavioral therapy. Acamprosate alone or combined with naltrexone or Cognitive Behavioral Therapy were all more effective than non-treatment, but not placebo. At one-year follow-up, effects of all treatments were absent with the exception of naltrexone, which demonstrated a reduction in relapse to heavy drinking days. This indicates that alcohol abuse can be managed in the primary care setting with naltrexone or behavioral therapy. All physicians should be familiar with the methods described in this study. These methods are briefly discussed below.

Cognitive Behavioral Therapy (CBT), Twelve Step Facilitation programs such as Alcoholics Anonymous (AA), and Motivational Enhancement Therapy (MET) are the main psychotherapy methods used in treating alcohol dependence. CBT is a structured, goal-directed intervention that aims to help patients identify how their thoughts contribute to maladaptive drinking behaviors. The goal is to develop new and healthy behaviors to control one’s thoughts. Though CBT is proven to be effective alone in the treatment of alcohol dependence, controversy exists regarding whether CBT is synergistic with drug treatment. (*57, *56).

Twelve Step facilitations programs such as Alcoholics Anonymous (AA) are familiar to most physicians. AA seeks to help patients with alcohol addiction through voluntary fellowship and a spiritual basis for recovery. In addition to wider availability, AA is beneficial for patients because it is more cost-effective (*58) and equally efficacious (*59) as other outpatient treatment options. In patients who are uncomfortable with the spiritually-based approach employed by AA, alternative treatment groups are available. There is, however, insufficient data to determine the effectiveness of these programs.

MET involves patient-centered, non-confrontational counseling aimed at resolving ambivalence about behavioral changes that contribute to alcohol abuse and dependence. (*60, *61) MET has equivalent three year outcomes when compared to CBT and 12 step facilitation programs. (*62)

Pharmacotherapy is of proven benefit (*63) and can reduce relapse in patients who have been treated for alcohol dependence (*64, *65). Thus far, naltrexone, acamprosate and disulfiram are approved for treatment of alcohol dependence in the US. (*70)

Naltrexone is an opioid receptor antagonist that reduces craving for alcohol by inhibiting the actions of endogenous opioids responsible for producing the pleasurable effects of drinking. (*53, *66) Prior to publication of the COMBINE trial, naltrexone was proven to reduce relapse and number of drinking days (*66) and increase participation in behavioral therapy. (*67, *68). While previous trials suggested that naltrexone should only be used as an adjunct to behavioral therapy (*48) because of their proposed synergistic effect (*69), the randomized COMBINE demonstrated that naltrexone can be used without behavioral therapy. (*56)

Treatment with naltrexone is not completely benign. Relapse is common after the drug is discontinued (*58), and naltrexone can induce opioid withdrawal in opioid users. Naltrexone can also cause hepatocellular damage and is contraindicated in patients with liver failure or acute hepatitis, both of which are common in persons with alcohol dependence. Furthermore, most of the studies described above had short-term follow up periods. The long-term efficacy of naltrexone is relative weak and controversial (*70, *71, *56) Long-acting intramuscular depots of naltrexone have been developed to increase compliance (*72) and newer opioid antagonist such as nalmefene with more favorable side effect profiles are currently being investigated. (*73)

Acamprosate is a structural analog of gamma-aminobutyric acid that decreases excitatory glutamergic neurotransmission. Prior to publication of the COMBINE trial, multiple studies demonstrated to the effectiveness of acamprosate in maintaining abstinence from alcohol (*74), leading to its approval for use in 2004. However, after the publication of the COMBINE trial, the effectiveness of acamprosate has largely been associated with the placebo effect. (*56)

Disulfiram is an inhibitor of aldehyde dehydrogenase and is used as aversive therapy for maintenance of abstinence. Patients who consume alcohol while taking disulfiram experience symptoms such as hypotension, nausea, vomiting, flushing, headache, and palpitations that can be severe enough to cause convulsions or death. (*75) Disulfiram may reduce drinking days in dependent patients (*65) but the results are controversial (*76, *77) Given its hazardous side effects, patients must be cognizant of the consequences of alcohol consumption while on disulfiram and must champion its use. (*78) Even under these circumstances, disulfiram is rarely used in the US.

Other drugs that have shown promise for the treatment of alcoholism are still under investigation. Baclofen, topiramate, ondansetron have all been studied and found to be helpful, however, sufficient clinical data is not yet available presently to recommend their use.

Once abstinence from alcohol is achieved, it is important for physicians to monitor for and treat relapse as well as other co-morbidities associated with alcohol use. (*79) Physicians should be supportive of patients and encourage them to participate in treatment groups. Physicians should understand that a minority of patients with alcohol dependence may eventually be able to consume alcohol safely. (*80)