Screening for alcohol problems: what makes a test effective?

Alcohol Research & Health, Wntr, 2004 by Scott H. Stewart, Gerard J. Connors

Screening tests are useful in a variety of settings and contexts, but not all disorders are amenable to screening. Alcohol use disorders (AUDs) and other drinking problems are a major cause of morbidity and mortality and are prevalent in the population; effective treatments are available, and patient outcome can be improved by early detection and intervention. Therefore, the use of screening tests to identify people with or at risk for AUDs can be beneficial. The characteristics of screening tests that influence their usefulness in clinical settings include their validity, sensitivity, and specificity. Appropriately conducted screening tests can help clinicians better predict the probability that individual patients do or do not have a given disorder. This is accomplished by qualitatively or quantitatively estimating variables such as positive and negative predictive values of screening in a population, and by determining the probability that a given person has a certain disorder based on his or her screening results. KEY WORDS: AOD (alcohol and other drug) use screening method; identification and screening for AODD (alcohol and other drug disorders); risk assessment; specificity of measurement; sensitivity of measurement; predictive validity; Alcohol Use Disorders Identification Test (AUDIT)

**********

The term "screening" refers to the application of a test to members of a population (e.g., all patients in a physician's practice) to estimate their probability of having a specific disorder, such as an alcohol use disorder (AUD) (i.e., alcohol abuse or alcohol dependence). (For a definition of AUDs and other alcohol-related diagnoses, see the sidebar "Definitions of Alcohol-Related Disorders.") Screening is not the same as diagnostic testing, which serves to establish a definite diagnosis of a disorder; screening is used to identify people who are likely to have the disorder. These people are often advised to undergo more detailed diagnostic testing to definitively confirm whether or not they have the disorder. When a screening test indicates that a patient may have an AUD or other drinking problem, the clinician might initiate a brief intervention and arrange for clinical followup, which would include a more extensive diagnostic evaluation (Babor and Higgins-Biddle 2001).

Regardless of the context in which screening tests are administered and the subsequent responses, it is important to have an appreciation of the strengths and limitations of screening tests. Accordingly, the main purpose of this article is to review the characteristics of screening tests that influence their usefulness in clinical settings. This includes their validity, sensitivity, and specificity. In addition, the article discusses methods to quantify the likelihood that a patient with a given screening result actually has the disorder (i.e., the postscreen probability). A review of different screening tests, particularly those that can be used in specific settings or with special populations, is beyond the scope of this article. The accompanying table summarizes the features of some of the most commonly used screening instruments. Additional screening tools and their characteristics have been reviewed by Connors and Volk (2003) and are described in the other articles in this issue and the companion issue of Alcohol Research & Health.

WHAT DISORDERS ARE AMENABLE TO SCREENING?

Not all disorders are suitable for screening; in fact, for certain disorders, screening tests may not be helpful or desirable. The main goal of screening is to identify patients at risk for a given disorder or at early stages of the disorder, so that they can begin to receive effective treatment and avoid or ameliorate the morbidity and mortality associated with the disorder. Consequently, disorders should have the following characteristics to be considered suitable for screening:

* They should be a cause of substantial morbidity or mortality.

* Effective treatment should be available that leads to a measurable improvement in morbidity and mortality compared with no treatment.

* Early treatment initiated after a positive screening result should lead to a better outcome than treatment which is initiated later in the disease process, when the disease has produced obvious symptoms that have led to a diagnosis. For example, in a general medical setting, patients should have better outcomes if an intervention is initiated after a screening test, such as the Alcohol Use Disorders Identification Test (AUDIT) (Babor et al. 2001), suggests a pattern of "harmful drinking" than if a diagnosis is made and intervention started after the patient already has developed a more severe condition, such as alcoholic liver disease.

* The disorder should be relatively common because, all else being equal, screening for prevalent disorders is more cost-effective than screening for rare disorders.

AUDs and other drinking problems generally fit these criteria. They are a major cause of morbidity and mortality (NIAAA 2000), are prevalent in the population (NIAAA 2003), and effective treatments are available (Saitz 2005). In addition, because AUDs may have an acute presentation (e.g., alcohol-related trauma or gastrointestinal bleeding) or result in long-term adverse consequences (e.g., liver disease) patients benefit from early detection and intervention. Finally, many people with AUDs never are diagnosed correctly. The next sections therefore will explore the characteristics screening tests must possess in order to be useful and effective.