New assessment tools for risk drinking during pregnancy: T-ACE, TWEAK, and others. (tests for factors that lead to fetal alcohol syndrome; includes definition of risk drinking and bibliography) (Special Issue: Alcohol-Related Birth Defects)

Effective screening of pregnant women for drinking that puts the fetus at risk for developing fetal alcohol syndrome is an essential part of prenatal care. Screening tools such as questionnaires have demonstrated varying degrees of accuracy in identifying risk drinkers among pregnant women.

Although much has been done to educate the public about the dangers of drinking during pregnancy, not everyone has received the message that such behavior puts the fetus at risk for acquiring alcohol-related birth defects. Routine assessments of their alcohol intake give women opportunities to ask questions about drinking during pregnancy and give health care providers opportunities to advise patients whose alcohol use may put their pregnancy outcomes at risk. Even when women have heard public health messages about alcohol-related birth defects, having their health care provider ask about their drinking turns a general message into a specific and personal one, which is more likely to evoke a positive response (Minor and Van Dort 1982). Accurate information about the risk involved and advice about drinking during pregnancy are sufficient to eliminate hazardous drinking in many or perhaps even most pregnant patients (Rosett et al. 1983). However, followup and referral to alcoholism treatment professionals are necessary for heavy-drinking pregnant women who are unable to reduce their alcohol intake. Effective linkage with such professionals is a critical component of any screening program.

It also is useful to assess retrospectively a mother's drinking during pregnancy when making a diagnosis in a child who has developmental deficits. Documenting maternal alcohol abuse during pregnancy can help establish a diagnosis of fetal alcohol syndrome (FAS) or fetal alcohol effects (FAE). Although some health practitioners have questioned the utility of diagnosing alcohol-related birth defects because there is no known cure, such a diagnosis may help to identify why a child has problems learning or displays abnormal behavior. Recent studies indicate that some children exposed to alcohol prenatally have serious behavioral and learning deficits even though their IQ's are normal or borderline (Streissguth et al. 1991). Because the etiology of their problems is not recognized, these children are often thought to be willfully inattentive and disobedient. Societal response is censorious, rather than constructive. Therefore, appropriately diagnosing alcohol-related birth defects in these children can be a great relief to parents and teachers and can serve as a basis for implementing remedial programs to minimize the negative consequences of the developmental deficits.

Several tools exist for assessing a pregnant woman's drinking habits, including tests for biochemical indications of alcohol-related damage to body systems and screening procedures based on questionnaires about drinking behaviors. Of the two, questionnaires currently are the tools most frequently used in prenatal clinics. Screening refers to the mass administration a test for a health problem to people who have not already been identified as being at risk for that problem.

Routine assessment of alcohol intake during pregnancy can be thought of as screening for risk drinking during pregnancy. Some screens for drinking during pregnancy are designed to identify levels of drinking that are unlikely to be harmful to the woman herself but that could harm the fetus. Other screens target levels associated with alcohol abuse or dependence,(1) assuming that women who are alcoholics are risk drinkers. The purpose of screening is to identify health problems or risks in time for intervention to prevent serious consequences such as FAS. This article reviews currently used strategies for assessing risk drinking, focusing on the effectiveness of specific screening procedures.

EVALUATING SCREENING METHODS

Screening methods usually are inexpensive and require little time to administer. In contrast, diagnostic procedures, which are designed to determine a patient's actual condition or disorder, are often time consuming, invasive, and costly. However, because they are more accurate, diagnostic procedures are usually used to evaluate the effectiveness of screening methods. In some cases, a well-validated screening method may be used to evaluate a new screening method.

Table 1 summarizes some results of screening for risk drinking during pregnancy (see sidebar for a discussion of risk drinking). Measures that are useful in evaluating the performances of screening procedures are derived from data in such tables: sensitivity, which is the probability, or likelihood, that a woman who is a risk drinker tests positive; specificity, which is the probability that a woman who is a nonrisk drinker tests negative; positive predictive value, which is the probability that a woman who tests positive is a risk drinker; and efficiency, which is the overall percentage of women correctly identified (Hennekens and Buring 1987).

Sensitivity measures the extent to which the screening procedure is successful in identifying all the risk drinkers in a population, and positive predictive value measures the number of women that must be followed up with diagnostic procedures to identify one true risk drinker. For example, a positive predictive value of 20 percent indicates that diagnostic procedures will verify risk drinking in one of five patients testing positive on a screen.

It should be noted that positive predictive values and efficiency scores are influenced by the …