Alcohol accounted for an estimated 3668 deaths and 95 917 hospital separations in Australia in the 1996–97 financial year.1 Alcohol-related deaths in Australia declined from 460 per million population in 1990 to 369 per million in 1997.2 The net economic cost of alcohol to the economy in 1992 was estimated to be $4.5 billion (this estimate includes increased healthcare expenditure and costs to industry from impaired productivity, increased accidents and absenteeism).3

Prevention and treatment of alcohol-related problems has been improving in recent decades. In the 1980s, brief interventions4 were developed for problem drinkers who reject abstinence or are unsuitable for this treatment goal. Brief interventions involve a combination of techniques, including motivational interviewing, feedback to patients of likely adverse consequences of current drinking, self-monitoring of drinking, developing a contract for future drinking, providing strategies to cut down drinking, and regular follow-up. Most clinicians try to discourage patients with life-threatening complications from alcohol from pursuing brief interventions.

Recently, more effective pharmacological treatments have been developed for alcohol dependence. The aetiology, natural history, compliance with and response to treatment of alcohol dependence are similar to those for other common, chronic, relapsing–remitting conditions readily accepted by the medical profession as worthy of treatment.5 Treatment of alcohol dependence has been shown to substantially reduce healthcare costs in the period after compared with the period before treatment.6

The main drugs used to treat alcohol dependence are acamprosate and naltrexone. Their profiles are shown in Box 1.

This review is restricted to major studies providing the highest-quality evidence. Studies were preferred if they were larger, had a longer study duration and were more recent. Studies or reviews with a more rigorous design were preferred. Only one study directly compared acamprosate and naltrexone.10

Formulating guidelines for pharmacotherapy of alcohol dependence is difficult because of the paucity of data from direct comparison of acamprosate and naltrexone. However, acamprosate should probably be considered the first-line treatment for patients with moderate to severe alcohol dependence, because of the larger body of supporting evidence and the benefits extending after treatment. Naltrexone is indicated for alcohol-dependent patients in whom acamprosate has not proved effective or has not been well tolerated, or for individuals whose lifestyle or past history indicates that compliance with taking medication is poor. The use of naltrexone for managing heroin dependence is controversial. However, there may be a place for prescribing naltrexone to alcohol-dependent patients who are also dependent on heroin.

Defining clear guidelines for use of the two main pharmacotherapies (acamprosate and naltrexone) is difficult in the present state of knowledge. This difficulty stems from the fact that the various studies have examined the use of these drugs over varying groups of outcome measures and study durations. Studies of acamprosate have generally used absolute-abstinence-based measures as the primary outcomes, while naltrexone studies have also measured more relative "harm-reduction" measures, such as relapse to heavy drinking or total amount of alcohol consumed. Acamprosate appears to have a prolonged action for up to a year after therapy has ceased, but compliance with a medication requiring thrice-daily administration is often difficult. On the other hand, naltrexone has well documented efficacy, at least in the initial three months of treatment, and is easier to take on a once-daily basis. Consideration of the outcomes desired by the patient, compliance history, other drug therapy and medical conditions may all influence the choice made by the prescriber.

In patients with alcohol dependence who also suffer from chronic pain, naltrexone will cause some not insurmountable problems. In this situation, or in acute pain, analgesia can be provided by non-opioid drugs such as non-steroidal anti-inflammatory drugs (including parenteral ketorolac). Other approaches could include local or regional anaesthesia. Attempting to overcome blockade with high doses of opioids is dangerous and not recommended outside an intensive care unit. Opioid withdrawal may be precipitated in alcohol-dependent patients treated with naltrexone if they have also been taking heroin recently. Naltrexone has been used (with uncertain benefit) for treating heroin dependence, but to avoid the problem of precipitating heroin withdrawal naltrexone should only be introduced in patients who have abstained from opioids for seven to 10 days.

The optimal duration of acamprosate or naltrexone treatment has not been established, but six months should be considered a minimum, and 12 months a more desirable duration of treatment.

It is generally recommended that naltrexone should be avoided in patients with advanced liver disease or elevated results on liver function tests. Higher than recommended doses of naltrexone may elevate liver function test results, but these dangers may be overstated. It is not clear how severe liver damage has to be before naltrexone administration becomes dangerous. Avoiding naltrexone in patients with decompensated liver disease is prudent; naltrexone should also be used with caution in patients with less severe forms of hepatic impairment. Some authors recommend that naltrexone be avoided in patients with aspartate transaminase levels three times greater than normal, while others set the limit at five-times normal.

Few studies have included many patients with the combination of severe alcohol dependence and mental illness. It is therefore unclear what effect mental illness has on the efficacy of acamprosate or naltrexone. However, one study included patients with stable mental illness and showed that they could be treated safely.20

There are no existing studies on the effectiveness and safety of the combination of acamprosate and naltrexone, but there is no theoretical reason preventing the combined use of these drugs.

Acamprosate and naltrexone have been approved under the Pharmaceutical Benefits Scheme (PBS) for use in treating alcohol dependence, provided that the patient is in a comprehensive treatment program for alcohol dependence with the goal of maintaining abstinence. Each authority prescription lasts for two months. There is no stated limit to the duration of treatment, but further extension of PBS authority requires an additional application each time. The PBS subsidy reduces the price (30 days' supply) from $170.10 (acamprosate) or $167.28 (naltrexone) to $22.40 (both drugs). Product information for naltrexone states that treatment duration is up to 12 weeks. However, the length of treatment is at the discretion of the prescriber.

Disulfiram inhibits acetaldehyde dehydrogenase, so that alcohol consumption results in a build up of acetaldehyde, causing extremely distressing symptoms, including flushing, syncope, nausea, vomiting and diarrhoea. It is available in 200-mg tablets; the usual starting dose is 100 mg daily increasing to 300 mg maximum, with 200 mg being the usual dose. It has been available for many years for treating alcohol dependence, but is prescribed rarely as compliance is often poor. Evidence of efficacy is limited, although this may be partly due to difficulties in trial design24 (E1).

Nalmefene has similar properties and a proposed similar mechanism of action to naltrexone26 (E2). It is not currently used to treat alcohol dependence other than in research settings.

Ondansetron, a selective 5-HT3-receptor antagonist, has been shown in one study to have a beneficial effect on early-onset alcohol dependence27 (E2), presumably by modulating dopamine release in mesocorticolimbic dopamine pathways. There is insufficient evidence to justify its routine use at present.

Other psychoactive drugs, such as lithium and some selective serotonin-reuptake inhibitor antidepressants, have been suggested, but no positive effect on alcohol dependence has been demonstrated in addition to the documented benefits in treating depression and other psychiatric conditions24 (E1).

When pharmacotherapy is included in the management of alcohol dependence, doctors should ensure that patients are also followed up closely and regularly, and should draw up a comprehensive treatment plan with each patient. This should include attempts to resolve any psychosocial issues. Prescribers are required to obtain a Health Insurance Commission (HIC) authority before prescribing acamprosate. The comprehensive plan can be provided by general practitioners using, where required, Expanded Primary Care (EPC) items. Engaging allied health professionals, such as alcohol and drug counsellors, may help patient management and also helps fulfil HIC requirements. A comprehensive treatment plan should be tailored to the needs of the individual patient and may need to involve combinations of modalities, including detoxification, counselling, referral to self-help groups or group therapy. The diagram in Box 3 may help guide doctors through the stages of managing patients who drink hazardous or harmful quantities of alcohol.

Box 4 provides advice for patients, as well as telephone numbers for help-lines throughout Australia.

3: Managing patients who drink hazardous or harmful quantities of alcohol