• Alcohol screening should be carried out as an integral part of practice in primary care. Consider screening when:
• Registering a new patient.
• Screening for other conditions.
• Managing a chronic disease (for example diabetes, hypertension, or chronic heart disease).
• Carrying out a medicine review.
• If screening everyone is not feasible or practical, focus on people who have an alcohol-related condition or who are at increased risk of harm from alcohol. This includes people:
• With relevant physical conditions (such as hypertension, gastrointestinal disorders, or liver disorders).
• With relevant mental health problems (such as anxiety, depression, or other mood disorders).
• Who have been assaulted.
• At risk of self-harm.
• Who regularly experience accidents or minor traumas.
• Who regularly attend genito-urinary medicine clinics or repeatedly seek emergency contraception.
• Always screen people who make an active call for help, or in whom an alcohol problem is strongly indicated as an incidental finding (see Additional information).

• An active request for help (either from the person, or through friends or relatives) is more likely in people with harmful drinking or alcohol dependence. People who are hazardous drinkers will not usually seek medical help, even though they may be aware that their drinking is putting them at risk.
• Incidental findings (no request for help) may include:
• Abnormal blood tests such as a raised gamma-glutamyl transferase (GGT) and mean corpuscular volume (MCV). These may alert a healthcare professional to an alcohol problem.
• Signs of an alcohol problem, such as dilated facial capillaries, bloodshot eyes, or hand tremor.
• Symptoms suggestive of a possible alcohol problem — professionals should be aware of the complications and comorbid diseases associated with an alcohol problem, or have a raised suspicion of an alcohol problem based on an individual's behaviour (for example use of illicit drugs, smelling of alcohol in consultations, numerous accidents, or requesting numerous sick notes). Recognizing these cues should initiate further questioning and the opportunistic (targeted) detection of a potential alcohol problem.

Screening for problem drinking

• Recommendations on who to screen are based on the public health guidance Alcohol-use disorders: preventing the development of hazardous and harmful drinking published by the National Institute for Health and Clinical Excellence (NICE) [NICE, 2010c].
• Ideally, alcohol screening should be carried out routinely as an integral part of clinical practice. However, NICE has recognized that this is not always feasible or practical, and has listed groups considered to be at increased risk of alcohol problems (based on observational studies and expert opinion).

Incidental findings

• The recommendation to screen people for alcohol misuse after an incidental finding is based on expert opinion from a review from the NHS National Treatment Agency for Substance Misuse [Raistrick et al, 2006]. It is consistent with recommendations made by NICE [NICE, 2010c].
• Problems with alcohol are underdiagnosed for various reasons, including poor recognition by healthcare professionals, or withholding of information by people because of shame or fear of stigmatization [Enoch and Goldman, 2002].

• Tailor discussions on alcohol and screening according to the person's needs, taking into account their faith and cultural beliefs.
• Complete a validated questionnaire with the person, or, if they are competent, ask them to complete one themselves (consult relevant specialists if it is not appropriate to use an English language-based questionnaire).
• Use AUDIT (Alcohol Use Disorders Identification Test) to decide whether to offer a brief intervention or not, and if so, what type (see Interpreting AUDIT).
• If time is limited, use an abbreviated version. These include AUDIT-C (AUDIT-Consumption), AUDIT-PC (AUDIT-Primary Care), SASQ (Single Alcohol Screening Questionnaire), or FAST (Fast Alcohol Screening Test). Follow this up with the full questionnaire if problem drinking is indicated.
• AUDIT and its associated screening questionnaires can be downloaded from www.ncl.ac.uk/ihs/assets/pdfs/hmitm/screeningtools (pdf).
• Biochemical measures (blood tests) should not be used as a matter of course to detect whether a person has been drinking hazardously or harmfully.

Recommendations on the method of screening are based on the public health guidance Alcohol-use disorders: preventing the development of hazardous and harmful drinking, published by the National Institute for Health and Clinical Excellence (NICE) [NICE, 2010c].

• AUDIT (Alcohol Use Disorders Identification Test) detects 92% of genuinely hazardous and harmful drinkers, and excludes 93% of those who are not. It is regarded by NICE as the 'gold standard' screening questionnaire for detecting hazardous and harmful drinking.

• Use AUDIT (Alcohol Use Disorders Identification Test) to identify the person's risk through drinking:
• Low-risk drinking: score of 1–7.
• Hazardous drinking: score of 8–15.
• Harmful drinking: score of 16–19.
• Possible alcohol dependence: score of 20 or more.
• Use professional judgement to revise AUDIT cut-offs downwards (that is to increase sensitivity) when screening:
• Women, including those who are, or plan to become, pregnant.
• People 65 years of age and older.
• People from some black and minority ethnic groups.
• Manage the person according to their risk. See Scenario: Hazardous/harmful drinking and Scenario: Dependence on alcohol (after confirmation of dependence).

Recommendations on the interpretation of screening are based on the public health guidance Alcohol-use disorders: preventing the development of hazardous and harmful drinking, published by the National Institute for Health and Clinical Excellence (NICE) [NICE, 2010c].

• The Alcohol Use Disorders Identification Test (AUDIT) questionnaire is a 10-item questionnaire taking 2–5 minutes to complete. A comparative study has found that the AUDIT questionnaire [Coulton et al, 2006]:
• Had a positive predictive value of 98% (95% CI 97 to 100) for hazardous drinking, and a negative predictive value of 97% (95% CI 94 to 100) for alcohol dependence.
• Was better at predicting an alcohol problem than biochemical markers.
• Was superior to the CAGE questionnaire for detecting hazardous and harmful drinking, but not necessarily for the detection of dependence.
• NICE recommends that 'professional judgement' is used when interpreting AUDIT scores because some people are more vulnerable to the effects of alcohol than others.
• This includes people with a low body mass index (BMI), women, older people, and some ethnic groups (who may not metabolize alcohol as efficiently).
• Increasing the sensitivity of the questionnaire by reducing the cut-off point will be at the expense of the specificity of the tool (that is it will increase the chance of a false-positive result).

• Blood tests, including measurement of blood alcohol concentration, are not recommended for the routine screening of problem drinking in primary care.

The recommendation to avoid the use of blood tests for screening are based on the public health guidance Alcohol-use disorders: preventing the development of hazardous and harmful drinking, published by the National Institute for Health and Clinical Excellence (NICE) [NICE, 2010c].

• Both NICE and the Scottish Intercollegiate Guidelines Network (SIGN) [SIGN, 2003] recommend blood tests for monitoring an established alcohol-related problem, but not for routine screening.
• Biochemical markers may lead to false-positive results if used for screening, and are neither sensitive nor specific for diagnosing an alcohol problem [Coulton et al, 2006; Raistrick et al, 2006].
• Gamma-glutamyl transferase (GGT) levels are raised in 60–80% of people with severe alcohol dependence and in 20–50% of heavy drinkers. Usually it takes 4–5 weeks of abstinence from alcohol for GGT levels to return to the normal range. GGT levels may be raised with other causes of liver disease, and less commonly with pancreatic disease or following a heart attack.
• Mean corpuscular volume (MCV) will increase within 4–8 weeks of the onset of heavy drinking. An increase in MCV may also occur in people with vitamin B₁₂ deficiency, folic acid deficiency, thyroid disease, or chronic liver disease, and during pregnancy.
• Carbohydrate deficiency transferrin (CDT) levels will increase within 3–4 weeks of the onset of of heavy drinking, and appear slightly more sensitive than GGT at detecting moderate alcohol misuse (especially in people with liver disease). Nevertheless, this test is not routine practice in primary care, and local policy on its availability will vary. False-positive results for CDT are very unusual.

• If alcohol dependence is indicated from the answers from the AUDIT (Alcohol Use Disorders Identification Test) questionnaire, confirm the diagnosis using the International Statistical Classification of Diseases and Related Health Problems, tenth revision (ICD-10), criteria for alcohol dependence. According to these criteria, dependence is diagnosed if three or more of the following have been present together during the previous year:
• A strong desire or sense of compulsion to drink alcohol.
• Difficulty in controlling drinking in terms of its onset, termination, or level of use.
• A physiological withdrawal state when drinking has ceased or reduced, or drinking to relieve or avoid such a withdrawal state. Withdrawal symptoms include tremor, sweating, rapid heart rate, anxiety, insomnia, and less commonly seizures, disorientation, and hallucinations.
• Evidence of tolerance, where increased amounts of alcohol are required in order to achieve effects originally produced by lower amounts.
• Progressive neglect of alternative pleasures or interests because of drinking, and increased amounts of time necessary to obtain or consume alcohol, or to recover from its effects.
• Persisting with alcohol consumption despite awareness of overtly harmful consequences.

Recommendations for the diagnosis of alcohol dependence are based on the International Statistical Classification of Diseases and Related Health Problems, tenth revision (ICD-10) criteria [WHO, 1992] and guidelines published by the Scottish Intercollegiate Guidelines Network [SIGN, 2003].

• The diagnosis of alcohol dependence is beyond the scope of current guidelines from the National Institute for Health and Clinical Excellence (NICE) concerning alcohol misuse [NICE, 2010a; NICE, 2010c]. However, this issue will be addressed in the forthcoming NICE clinical guideline Alcohol-use disorders: diagnosis and clinical management of harmful drinking and alcohol dependence, due to be published in January 2011 [NICE, 2010e].