Screening for Alcohol Use

Surveys have found that screening is far from universal, and that only 13% of primary care clinicians use a validated instrument or tool to do so. When unhealthy use is identified, most primary care clinicians recommend self-help groups, but about a fifth of primary care clinicians offer no formal therapeutic intervention at all. In a review of thousands of records, McGlynn et al. found that about one-half of recommended health services were provided to adult patients. Services for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) alcohol dependence were at the lowest level: only 10% of patients with alcohol dependence documented in the medical record had it addressed in any way.

The US Preventive Services Task Force recommends screening and brief intervention for unhealthy alcohol use for adults and pregnant women. The US Preventive Services Task Force found that screening and brief intervention improves important health outcomes and that benefits outweigh any risks. The recommendation is as strong as that for screening mammography for women 40–50 years of age, osteoporosis screening for women 65 years or older, or cholesterol screening in young adults with other risk factors for coronary artery disease. One clear difference between screening and brief intervention for alcohol use and the other preventive services listed is that screening and brief intervention is a more time-consuming interaction with a patient than is ordering a radiological or blood test. In a revenue-driven health care environment, test ordering may add revenue to an institution. Prevention and reduction of alcohol use consequences, in contrast, may slow down the primary care clinician, decreasing volume-based-revenue, even with the prospect of downstream cost-savings. (This scenario may be mitigated in part by the existence of a billing code for screening and brief-intervention for substance use.) Indeed, screening and brief intervention may be cost-effective and even cost-saving, at least from a societal perspective, if the benefits of self-reported use translate into reduced health consequences and care utilization. Based on randomized trials and simulation modeling, screening and brief intervention for alcohol ranks in the top five of preventive services in terms of cost-effectiveness. [CR] ^,

Unhealthy Alcohol Use: Definitions

In the United States, about 70% of men and 60% of women 18 years of age or older drink alcohol. Although there is evidence that low-level consumption is relatively low risk (although it does increase the risk for cancers, such as breast cancer), higher amounts clearly risk numerous health consequences. Thus one dimension of screening for unhealthy alcohol use is solely based on quantity and frequency of intake. The other dimension that screening can address is alcohol consequences.

The National Institute on Alcohol Abuse and Alcoholism has defined cutoffs for unhealthy use that are empirically based in epidemiological literature. Risky drinking amounts are those above these cutoffs. For men, this level is greater than 14 drinks per week or >4 drinks on an occasion; for women this cutoff is >7 drinks/week or >3 drinks on an occasion. For individuals over 65 years of age, the cutoff is the same as that for women. A “drink” is defined, in the United States, as 12–14 g of ethanol (12 oz of beer, 5–6 oz of wine, or 1.5 oz of 80 proof spirits) ( Table 46.1 ). Drinking risky amounts without associated consequences is risky drinking. If there are consequences, patients may have an alcohol use disorder (as defined in the Diagnostic and Statistical Manual of Mental Disorders , Fifth Edition [DSM-5]). An important exception to using consumption to define unhealthy use is when even low-level use risks consequences. Aside from alcohol being a carcinogen, other examples include women intending pregnancy, pregnant women, people taking medications that interact with alcohol, and people with health conditions worsened by even small amounts (e.g., hepatitis C virus infection). The key distinction among individuals with unhealthy alcohol use is whether a moderate to severe alcohol use disorder (AUD) is present. Identifying such is important because the best advice and management are different, as discussed later in this chapter.

What is the best way to screen for unhealthy alcohol use? Although we usually think about face-to-face discussions or questions in the clinical setting, screening runs the gamut from these to telephone or web-based instruments. These approaches should be considered and adapted as appropriate, with an eye to optimal efficiency and effectiveness, for system-based approaches.

Any interaction between the primary care clinician and patient should serve to build the therapeutic relationship. Therefore history-taking should be conducted in an empathic, nonjudgmental way. Embedding questions about alcohol use among other routine medical history questions may serve to decrease resistance and improve both the tenor of the discussion and the quality of information generated. That said, screening is best done with interview questions that should be asked verbatim, as validated.

The first order of business is to ascertain whether the patient drinks at all. The clearest question is: “Do you sometimes drink beer, wine, or other alcoholic beverages?” If the patient does not drink at all, inquiring into the patient’s rationale for abstaining may reveal that the patient has prior use with consequences. If the patient drinks at all, then quantity and frequency of drinking should be evaluated.

One wants to further characterize alcohol intake. It is important to assess the average number of drinks in a week and whether there are any heavy drinking episodes (i.e., drinking in excess of the single-occasion cut-offs delineated by the National Institute on Alcohol Abuse and Alcoholism). This brief screen can be done with three questions.

• 1.
  

  On average, how many days per week do you have an alcoholic drink?

  
  
• 2.
  

  On a typical drinking day, how many drinks do you have?

  
  
• 3.
  

  What is the maximum number of drinks you had on any given occasion during the past month?

With the first two responses, the number of drinks per week can be calculated, and if weekly cutoffs are exceeded, then there is risky drinking. Similarly, if limits per episode are exceeded, the patient is drinking at a risky level.

The National Institute on Alcohol Abuse and Alcoholism Clinician’s Guide recommends a single question to screen people who drink for unhealthy alcohol use :

“How many times in the past year have you had X or more drinks in a day?” where X = 5 for men and 4 for women. A response >1 is considered positive.

This single item is both sensitive and specific for detecting unhealthy alcohol use, memorizable, requires no scoring, and is thus appropriate for most practices. Single-question instruments are brief, valid, and efficient.

The Alcohol Use Disorders Identification Test ( Table 46.2 ) is a 10-item instrument developed by the World Health Organization with good performance characteristics for identifying unhealthy alcohol use. It is scored from 0 to 40 with a score of 8 classically, but more recently scores of 5 (for men) and 4 (for women and those older than 60 years of age) have been considered positive. It requires scoring so may be better suited to a pen-and-paper or automated process than to verbal interview. The main advantage of the longer tool is that it provides more information for discussion and can suggest the presence of a disorder with a score of 15–20 or more.

A briefer validated screening tool is the Alcohol Use Disorders Identification Test-C. It comprises three consumption questions from the Alcohol Use Disorders Identification Test (the first three questions) and also requires scoring. Using a score threshold (or cut-point) of 4 for males and 3 for females will identify unhealthy alcohol use, while a score of 7 or more suggests moderate to severe alcohol use disorder. This test is in wide use and is good for practices with workflows that permit scoring.

If a patient is NOT drinking above recommended cutoffs and has no absolute contraindication to any drinking, then the patient should be congratulated on the healthy pattern, educated about the risks and benefits of “moderate” (low amounts of) drinking (the best evidence for the former being for an increased risk for breast cancer and for the latter, possible decreased cardiovascular risk) and educated about optimal limits. If the screening is “positive,” however, then further assessment is recommended for confirmation and to determine severity.

Screening for Tobacco Use

Evaluating the use of tobacco is by-and-large a simpler process than that for use of alcohol. The healthiest level of tobacco intake is none. Because efforts to make smoking a “vital sign,” smoking status is now frequently recorded routinely at contacts with clinicians. A single question, such as “In the past year have you smoked cigarettes or used any other tobacco product?” should be asked of all patients. Caution should be taken with patients who define smoking as regular or current use; they may report they are nonsmokers despite sporadic or recent but not current regular use. Finally, the presence of past or current smoking may itself raise concern for concomitant unhealthy alcohol or other drug use.

Screening for Other Drug Use

The DAST-10 is a screening questionnaire that asks about drug use and consequences. A score of 3 or more is positive. The length of the questionnaire, and particularly the lack of validation studies in primary care settings, limits its utility. The 10 questions probe for physical and social consequences (e.g., blackouts, withdrawal, relationship problems) and loss of control over use and do not identify the drug used.

The Alcohol Smoking and Substance Involvement Screening Test (ASSIST) is a complex instrument of 80 items yielding independent scores for each of multiple substances. A “positive” screening test can be defined as a score of 2 or greater (indicating any drug or alcohol use in the past 3 months), although the cutoff of 4 or greater is probably more clinically useful (indicating either weekly use or less-frequent recent use accompanied by consequences of use). ASSIST has been validated internationally. Although there are 80 items, if no use of a specific substance is reported, only 10 items need to be answered. If the patient reports any use of a substance, then a series of questions are asked about that substance. Its complexity, length, and need for scoring limit its utility in routine clinical primary care practice, although where computers are integrated into clinical or research settings, the test may be usable. The ASSIST is available online from the National Institute of Drug Abuse. Another major limitation of the ASSIST is that the results do not directly identify use of risky alcohol amounts per se, a critical target of screening in primary care settings because of the prevalence and proven value of brief intervention for such patients.

The best approach to drug use screening in primary care is to ask one or two validated questions. The following single question test has been validated : How many times in the past year have you used an illegal drug or used a prescription medication for non-medical reasons—for example, because of the experience or feeling it caused?

Any nonzero response is considered positive. This has good sensitivity (>70% and specificity (>94%) and is appropriate for primary care settings. Because marijuana is becoming legal in many states, it can also be useful to simply ask about the frequency of any marijuana (cannabis) use and then determine if it is medicinal or whether there are any consequences of use in a further assessment.

Brief Intervention

Brief intervention is an essential part of the primary care clinician’s management of patients with unhealthy behaviors in general, and unhealthy alcohol and other drug use in particular. It is covered in detail in the Chapter 43 . Brief intervention is a brief, patient-centered counseling, a conversation of no more than 45 but usually 10–15 minutes. Informed by motivational interviewing, the primary care clinician provides feedback to the patient (after asking for permission to do so) about their substance use, use-related risks, and any consequences of importance to the patient (e.g., social, occupational, legal, medical, psychological consequences), and sometimes how their use compares to norms ( Table 46.3 ). Along with the assessment should come clear and nonconfrontational advice about change, again after asking for permission to do so. The patient’s desires and understanding about relevant behavior change should be elicited, as should their ability and readiness to change. Then with the patient’s agreement, a menu of options for courses of action should be discussed. Their commitment, including an agreement about the next step and (long- and short-term) goals should be agreed upon and recorded. Finally, arrangement for follow-up should be made. The approach must be empathic and supportive of the patient’s self-efficacy.

More specifically, the clinician should determine the patient’s perception of their use and need for change (e.g., “Do you think your drug use is a problem?”). For patients who are not ready to change, the goals are to increase problem awareness, express concern, and agree to disagree. Sometimes a trial of abstinence or cutting down can be useful. For patients considering change, the goal is to tip the balance toward change by eliciting positive and negative aspects of drinking and not drinking, to demonstrate discrepancies between patients’ values and actions. Once the patient has decided to change, reviewing options for the next steps is recommended. The patient will need support and encouragement, and a reminder that the therapeutic relationship will continue regardless of continued unhealthy use or success in cutting down.

Management of Risky Alcohol and Drug Use

Brief intervention has been demonstrated to significantly improve self-reported drinking outcomes when delivered in many clinical settings and by varied clinical personnel, although the best controlled trial evidence is for screening and brief intervention in primary care settings by primary care clinicians. Although they are less well-studied, brief interventions can also be written, phone, or computer/web-based, and can be single or multiple contacts. The best evidence for efficacy is for multicontact interventions in people without a moderate to severe disorder. Implementation depends on the particular practice setting. The evidence for efficacy on health outcomes and utilization is inconclusive, although some individual studies are suggestive of benefit.

Fleming and colleagues demonstrated in a randomized controlled trial of a multicontact brief intervention, that significant effects on self-reported drinking and on health care utilization and expenditures can be detected for up to 4 years. One meta-analysis demonstrated that brief intervention for alcohol decreases mortality. Meta-analyses predict that on average alcohol intake will decline by 38 g per week (a 15% decrease) and that the proportion of people drinking risky amounts decreases to 57% in brief intervention groups and 69% in controls.

The evidence supporting brief intervention for drugs other than alcohol and tobacco is more limited. There are many high-quality randomized controlled studies demonstrating no significant effect of brief intervention on nonalcohol drug use. However, a small but growing number of controlled trials have successfully tested brief intervention for drugs after screening in outpatient settings. Bernstein and colleagues studied patients who used cocaine and or heroin who presented for care in outpatient (not primary care) settings. A single motivational brief intervention delivered by a trained health promotion advocate reduced cocaine and heroin use at 6 months more than written advice alone. For example, more cocaine users who received brief intervention were abstinent than were controls (22% vs. 17%). The World Health Organization Alcohol Smoking and Substance Involvement Screening Test phase 3 trial found that brief intervention had efficacy with respect to intermediate outcomes, but there is doubt that the outcomes are clinically significant. Blow and colleagues conducted a randomized controlled trial of brief intervention in urban emergency departments and found that certain methods (e.g., delivery by a trained therapist or by computer) resulted in significant decreases in drug use up to 12 months later.

Brief interventions can decrease substance use, but even under the best circumstances (e.g., the evidence for alcohol brief intervention in primary care) many people continue to have unhealthy use. The effectiveness of repeated brief interventions is unknown, but primary care settings do provide the opportunity to assess alcohol use and consequences over time, and to address behavior change. Over time, with repeated conversations, the case for change can build in breadth and depth. As the clinician learns more about the patient and their substance use, the advice and rationale may become more personally salient and effective. The key for clinicians delivering longitudinal care is to maintain an empathic alliance and continue to address substance use. Drinking despite known negative consequences is a DSM-5 criterion for a disorder, so clinicians should remain alert to any consequences, and be clear in the education of the patient about the connection of the drinking and the consequence. Continued drinking in this case likely signals greater severity of the AUD.

If there is an important person in the patient’s life, it may be effective to bring this person into the discussion, with the patient’s consent. This person should be invited to a face-to-face visit with the patient. They may offer an important perspective, for instance by furnishing information about consequences of drinking or by assisting in the process of change. It may also be necessary for the behavior change to be a goal shared by the two people for success to take hold.

Once a patient has been able to reduce drinking to healthier levels, the clinician should recognize and affirm this success and monitor for any recurrence.

This discussion of management of drinking may be applied to use of other drugs. It should be noted that in many cases in the United States, the illegal activities involved in procuring illicit drugs is an obvious risk that can be used in motivational brief interventions, although it is becoming less an issue for cannabis/marijuana. Several drugs, depending on route of administration, induce physiological dependence at high rates, so controlled use will be rare; for instance, smoking freebase or crack cocaine is rarely a casual behavior over which there is robust control.

Management of Alcohol Use Disorder

Several approaches to moderate to severe alcohol use disorder (AUD) can be effective in primary care settings, namely, pharmacotherapy and counseling. Referrals to specialist care can also be helpful. Bear in mind that it is unlikely that a brief intervention alone will result in significant decrease in drinking of people with moderate to severe AUD.