Alcohol-Screening Instruments for Pregnant Women

Grace Chang, M.D., M.P.H.

According to new studies, even low levels of prenatal alcohol exposure can negatively affect the developing fetus, thereby increasing the importance of identifying women who drink during pregnancy. In response, researchers have developed several simple alcohol-screening instruments for use with pregnant women. These instruments, which can be administered quickly and easily, have been evaluated and found to be effective. Because of the potential adverse consequences of prenatal alcohol exposure, short screening questionnaires are worthwhile preventive measures when combined with appropriate followup. KEY WORDS: prenatal alcohol exposure; prenatal diagnosis; alcohol use test; identification and screening for AOD (alcohol or other drug) use; specificity and sensitivity of measurement; breath alcohol analysis; AODR (alcohol- or other drug-related) biological markers

creening pregnant women for five or more drinks per occasion) by 3.5 drinks per week) demonstrated that alcohol use has become of increasing pregnant women has increased substan­ the women who drank more than 3.0 Simportance, because new research tially, from 0.8 percent in 1991 to 3.5 drinks per week increased significantly indicates that even low levels of prenatal percent in 1995 (Ebrahim et al. 1998; their risk of first-trimester spontaneous alcohol exposure can negatively affect Centers for Disease Control and abortion (Windham et al. 1997). the developing fetus. Adverse effects of Prevention 1997). This rise in the rate Identifying women who drink at prenatal alcohol exposure can range from of alcohol consumption among pregnant risky levels during pregnancy poses spe­ subtle developmental problems, or fetal women coincides with growing evidence cial challenges, however, particularly alcohol effects, to full-blown fetal alcohol of the negative effects of low-to-moderate because the definition of pregnancy risk syndrome. In addition, scientists and alcohol consumption during pregnancy. drinking has been refined over time. In clinicians have found that certain neuro­ Increasingly sophisticated research addition, screening for any alcohol use behavioral outcomes associated with has improved scientific and clinical under- during pregnancy is difficult. This arti­ prenatal alcohol exposure can persist in standing of the adverse consequences cle discusses the difficulties involved in the affected person into adolescence of prenatal alcohol exposure. The term screening pregnant women for alcohol (Sampson et al. 1994) and adulthood “pregnancy risk drinking” (i.e., drink­ use; details some of the questionnaires, (Kelly et al. 2000). ing during pregnancy at levels consid­ or instruments, available to facilitate Because no universally safe level of ered risky to the fetus) was previously alcohol consumption during pregnancy defined as the consumption of 1 ounce has been identified (Stratton et al. 1996), or more of alcohol (i.e., two or more GRACE CHANG, M.D., M.P.H., is an the U.S. Surgeon General and the drinks) per day (Sokol et al. 1989), but associate professor of psychiatry at Harvard Secretary of Health and Human Services more recent findings show that even Medical School, Department of Psychiatry, recommend abstinence both before con­ lower levels of alcohol consumption can Brigham and Women’s Hospital, Boston, ception and throughout pregnancy lead to negative pregnancy outcomes Massachusetts. (Stratton et al. 1996; Ebrahim et al. 1998). (Charness et al. 1994; Wong et al. 1995; However, approximately 20 percent of Ikonomidou et al. 2000; Jacobson and Preparation of this article was supported women drink some alcohol during preg­ Jacobson 1994). A study of more than in part by National Institute on Alcohol nancy, and the rate of frequent drinking 5,000 pregnant women who consumed Abuse and Alcoholism grants K24–AA– (i.e., seven or more drinks per week or alcohol moderately (defined as at least 00289 and R01–AA–12548.

204 Alcohol Research & Health Screening Instruments for Pregnant Women

alcohol screening in this population; and inquiries about drinking patterns before 1996). Progress toward this goal has briefly describes a few laboratory tests pregnancy confirmation are potentially not yet been reported. used for detecting alcohol use among more accurate measures of first-trimester In response to the need for increased pregnant women. drinking (Day et al. 1993). Women are alcohol screening among pregnant women, also likely to deny or minimize their researchers have developed several alcohol- drinking during pregnancy out of embar­ screening instruments specifically for Complications of rassment (Morrow-Tlucak et al. 1989). use with this population. Screening Pregnant Even moderate drinkers may underreport Women for Alcohol Use alcohol consumption during pregnancy (Verkerk 1992). Data from a sample of Screening Instruments A key complication in screening preg­ 361 mothers suggest that women who nant women for alcohol use arises from report drinking more than 1.3 drinks The screening instruments described in the fact that the traditional alcohol- per week during pregnancy actually may this section were tested in diverse clini­ screening questionnaires—such as the be drinking at levels high enough to cal populations and may help identify Michigan Alcoholism Screening Test incur risk for alcohol-related birth defects women using alcohol during pregnancy. (MAST) (Selzer 1971) and the CAGE1 (Jacobson et al. 1991). For example, 53 These instruments vary in that they (Ewing 1984)—are less effective in percent of the women who reported were designed to detect different levels identifying drinking problems among drinking more than 1.3 drinks per week of alcohol use and, therefore, differ in women than among men. This discrep­ during pregnancy reported higher lev­ how they define pregnancy risk drinking. ancy is attributable to the fact that these els of consumption when interviewed In general, a positive screen does not instruments were developed among men, retrospectively. indicate an alcoholism diagnosis; rather, who have different patterns of alcohol A third complication is that standard it may signal to a physician or other consumption and different thresholds questions about quantity and frequency health care practitioner the need to discuss for problem drinking than women of alcohol consumption are unlikely pregnancy risk drinking with a patient. (Babor et al. 1989). In addition, these to be helpful when screening pregnant Routine use of screening questionnaires instruments were developed to detect women for alcohol use. The widely in clinical practices may reduce the alcohol dependence, which is relatively used American College of Obstetricians stigmatization of asking patients about uncommon among pregnant women and Gynecologists (ACOG) Antepartum alcohol use and result in more accurate (Ebrahim et al. 1998). Because of bio­ Record poses three questions about alco­ and consistent evaluation. logical differences between women and hol use: (1) the amount of alcohol con­ Sensitivity and specificity are two men, the same quantity of alcohol con­ sumed per day before pregnancy, (2) the important properties of every screening sumed over the same time period pro­ amount of alcohol consumed per day instrument. The sensitivity of a screen­ duces higher blood alcohol levels in during pregnancy, and (3) the number ing test refers to the probability that a women than in men (Graham et al. of years of alcohol use. The Antepartum person who should test positive, does 1998). Women are also more sensitive Record has a fill-in-the-boxes format so (i.e., the sensitivity of a screen for than men to alcohol-related organ dam- designed to gather standard clinical pregnancy risk drinking is the probabil­ age, such as cardiomyopathy and myopa­ information on obstetric patients. How- ity that a woman who is a risk drinker thy (Urbano-Marquez et al. 1995; ever, compared with the 13-item Prenatal tests positive). The specificity of a Hanna et al. 1992). Therefore, alcohol- Alcohol Use Interview, the ACOG screening test is the probability that a screening-instrument cutoff scores (i.e., Antepartum Record is less successful person who should test negative, does the values that clinicians use to define in identifying prenatal alcohol use. so (i.e., the probability that a woman a positive result from a screening instru­ Researchers suggest that the difference who is not a risk drinker tests negative) ment) most likely need to be set differ­ in findings between the two instruments (Rosner 1990). ently for men and women and particu­ may be attributable to the format of larly for pregnant women (Bradley et the ACOG Antepartum Record and its The T-ACE al. 1998). lack of guiding questions: the ACOG A second complication faced by instrument requires a skilled interviewer The T-ACE was the first validated sen­ researchers is that many women alter their in order to elicit accurate responses about sitive screen for risk drinking (defined alcohol consumption once they learn drinking during pregnancy (Budd et al. as alcohol consumption of 1 ounce or that they are pregnant. Consequently, 2000). more per day) developed for use in A final complication is that obstetri­ obstetric-gynecologic practices (Sokol cians inconsistently screen their patients et al. 1989). An obstetrician developed 1The CAGE screening instrument (Ewing 1984) consists of four questions: (1) Have you ever felt you should Cut for alcohol use during pregnancy. One the T-ACE after observing that asking down on your drinking?, (2) Have people Annoyed you by goal of Healthy People 2000 was to patients about their tolerance to the criticizing your drinking?, (3) Have you ever felt bad or increase obstetricians’ rate of screening intoxicating effects of alcohol did not Guilty about your drinking?, and (4) Have you ever had a drink first thing in the morning to steady your nerves or to for alcohol use to 75 percent, from the trigger denial. The “socially correct” get rid of a hangover (Eye opener)? 1987 rate of 34 percent (Stratton et al. answer is not known (patients do not

Vol. 25, No. 3, 2001 205 feel stigmatized to answer honestly), and make you feel high?” The T-ACE was the AUDIT and SMAST independently tolerance reflects a pattern of drinking. not administered as an independent as well as reviewed the participant’s The four T-ACE questions (see T­ instrument; instead, both the sensitivity medical record. The three criteria used ACE textbox) take less than 1 minute and specificity of the T-ACE were cal­ to evaluate the T-ACE, AUDIT, SMAST, to ask. The T-ACE is positive with a culated from the subjects’ responses to and medical record were as follows: (1) score of 2 or more points. One point is the tolerance question as well as to the alcohol abuse or dependence diagnoses given for each affirmative answer to the annoyed, cut-down, and eye-opener as defined according to the Diagnostic A, C, or E questions. Two points are questions from the CAGE questionnaire. and Statistical Manual of Mental Disorders, given when a pregnant woman reports The T-ACE proved to be superior to Third Edition, Revised (DSM-III-R) that more than two drinks are necessary both the MAST and CAGE in identi­ (American Psychiatric Association 1987), for her to feel “high” or experience the fying pregnancy risk drinking (i.e., which the subject could meet at any intoxicating effects of alcohol. defined as alcohol consumption of more point in her lifetime; (2) risk drinking, Researchers initially evaluated the than 1 ounce daily). Table 1 summa­ defined as having more than two drinks T-ACE in a sample of 971 African- rizes the study’s findings. per drinking day before pregnancy; and American women attending an inner-city We subsequently tested the T-ACE (3) current drinking (i.e., any alcohol antenatal clinic. The researchers admin­ as a self-administered, independent consumption during pregnancy). istered both the MAST and CAGE as screening tool embedded in a health- Table 2 summarizes the sensitivity well as asked the T-ACE tolerance ques­ habits survey with questions about smok­ and specificity of the T-ACE, AUDIT, tion, “How many drinks does it take to ing, stress, weight, and dietary habits in SMAST, and medical record for the a more socially and ethnically diverse three criteria. In addition, sensitivity obstetric population—350 women ini­ and specificity for varying cut-off scores T-ACE tiating prenatal care at the Brigham and for the T-ACE and AUDIT are listed Women’s Hospital in Boston, Massa­ (e.g., in response to the tolerance question chusetts (Chang et al. 1998). in the T-ACE, “more than 2 drinks” T Tolerance: How many drinks We compared the sensitivity and would be a positive response in one does it take to make you feel specificity of the T-ACE with the sensi­ scoring method and “2 or more drinks” high? tivity and specificity of three other pop­ would be a positive response when ular methods of screening for alcohol using a different scoring method). With A Have people Annoyed you by use in other clinical settings: (1) the “tolerance” defined as “2 or more drinks criticizing your drinking? Alcohol Use Disorders Identification to feel intoxicated,” the T-ACE was the Test (AUDIT) (Babor et al. 1992), (2) most sensitive instrument to detect cur- C Have you ever felt you ought the Short Michigan Alcoholism Screening rent alcohol consumption, risk drink­ to Cut down on your drinking? Test (SMAST) (Selzer et al. 1975), and ing, and lifetime DSM-III-R alcohol (3) a review of the patient’s medical diagnoses. However, it was also the E Eye opener: Have you ever record. Researchers gave each participant least specific. had a drink first thing in the morning to steady your nerves or get rid of a hangover? Table 1 Comparison of the T-ACE, CAGE, and MAST in Identifying Pregnancy Risk Drinking The T-ACE is used to screen for pregnancy risk drinking, defined Screening for Pregnancy Risk Drinking* here as the consumption of 1 ounce or more of alcohol per Positive Test Score Sensitivity Specificity day while pregnant. Scores are Instrument (points accrued) (%) (%) calculated as follows: a reply of “More than two drinks” to ques­ tion T is considered a positive T-ACE (> 2) 69 89 response and scores 2 points, and CAGE (> 2) 38 92 an affirmative answer to question MAST (> 5) 36 96 A, C, or E scores 1 point, respec­ tively. A total score of 2 or more MAST = Michigan Alcoholism Screening Test. points on the T-ACE indicates a *Pregnancy risk drinking is defined as the consumption of 1 ounce or more of alcohol per day during pregnancy. positive outcome for pregnancy NOTE: The sensitivity of a screening test is the probability that a person who should test positive, does so (i.e., risk drinking. the sensitivity of a screen for pregnancy risk drinking is the probability that a woman who is a risk drinker tests positive). The specificity of a screening test is the probability that a person who should test negative, does so (i.e., the probability that a woman who is not a risk drinker tests negative) (Rosner 1990). SOURCE: Sokol et al. 1989. SOURCE: Sokol et al. 1989.

206 Alcohol Research & Health Screening Instruments for Pregnant Women

The ideal screening test would be in diverse obstetric populations. The patients at two primary health care cen­ both highly sensitive and highly spe­ questions are easy to both remember ters; and (3) the general population of cific; however, any given test usually and score and can be asked by an obste­ the Buffalo, New York, metropolitan area has a trade off. Screeners typically give trician or nurse in 1 minute. Women (Chan et al. 1993). Subsequent evaluation priority to sensitivity if it is important waiting for their prenatal appointments, of the TWEAK has revealed its promise to identify a condition, even if more for example, could be asked to complete as a screening tool for identifying preg­ false positives are subsequently identi­ the T-ACE as part of a routine patient nant women who are at-risk drinkers, fied. However, if insufficient resources questionnaire to be reviewed during defined as those consuming 1 ounce of are available to evaluate all patients who the visit. alcohol or more daily (Russell et al. 1994). screen positive, then specificity may be The TWEAK is scored on a 7-point considered more important (Russell The TWEAK scale. On the tolerance question, 2 points 1994). Thus, the T-ACE, with a posi­ are given if a woman reports that she tive response to the tolerance question The TWEAK is a five-item screening can consume more than five drinks defined as “more than 2 drinks,” offers tool that includes questions from the without falling asleep or passing out. A the best balance of sensitivity and MAST, CAGE, and T-ACE (see TWEAK positive response to the worry question specificity. textbox). The TWEAK is designed to yields 2 points, and positive responses The T-ACE is a valuable and efficient detect alcoholism or heavy drinking to the last three questions yield 1 point tool for identifying alcohol use among and was first tested in three male and each. A woman who has a total score of pregnant women; in addition, it demon­ female samples randomly selected from 2 or more points is likely to be an at- strates acceptability and accuracy in three groups: (1) alcoholics in treat­ risk drinker. identifying a range of alcohol-use levels ment at a county medical center; (2) Like the T-ACE, the TWEAK asks about tolerance to the effects of alcohol. In one study of 4,743 African-American Table 2 Sensitivity and Specificity of the T-ACE, AUDIT, SMAST, women of low socioeconomic status and Medical Record who were given the MAST, the CAGE, and the T-ACE tolerance question, the Criterion Sensitivity* Specificity** calculated sensitivity and specificity of Standard Instrument (%) (%) the TWEAK were 79 percent and 83 percent, respectively, in contrast to the DSM-III-R T-ACE (tolerance > 2) 87.8 36.6 calculated 70-percent sensitivity and lifetime alcohol T-ACE (tolerance > 2) 60.0 66.4 85-percent specificity of the T-ACE. diagnosis AUDIT (> 11) 7.0 99.6 AUDIT (> 10) 11.0 99.0 Periconceptional risk drinking, defined AUDIT (> 8) 22.6 97.4 as 1 ounce or more of alcohol consump­ SMAST 14.8 97.9 tion per day or 14 drinks per week dur­ Medical record 15.6 93.6 ing a typical week before pregnancy (Russell et al. 1994), was the criterion Risk drinking T-ACE (tolerance > 2) 92.4 37.6 standard (i.e., this was the level of (two drinks per day T-ACE (tolerance > 2) 74.3 71.4 drinking that the instruments were try­ before pregnancy) SMAST 11.4 95.9 ing to detect). The ability to generalize Medical record 6.7 89.4 these findings is limited. This is attribut­ Current alcohol T-ACE (tolerance > 2) 89.2 37.8 able to the homogenous makeup of the consumption T-ACE (tolerance > 2) 60.0 66.9 sample, the fact that neither the T-ACE (while pregnant) AUDIT (> 11) 3.3 97.8 nor the TWEAK were administered AUDIT (> 10) 6.7 96.9 as independent instruments, and the AUDIT (> 8) 15.0 93.9 definition of periconceptional risk SMAST 7.5 94.3 drinking, which other researchers have Medical record 20.0 96.1 subsequently updated to 0.5 ounces of alcohol per day (Hankin and Sokol 1995). SMAST = Short Michigan Alcoholism Screening Test. *Sensitivity is the probability that a person who should test positive, does so (Rosner 1990). The TWEAK does not appear to **Specificity is the probability that a person who should test negative, does so (Rosner 1990). offer any significant advantages over NOTE: The sensitivity and specificity for varying cutoff scores for the T-ACE and AUDIT are listed (e.g., in response to the tolerance question in the T-ACE, “more than two drinks” would be a positive response in one the T-ACE. Most studies investigating scoring method and “two or more drinks” would be a positive response under a different scoring method). With the TWEAK’s performance have relied tolerance defined as two or more drinks to feel intoxicated, the T-ACE was the most sensitive instrument to on a definition of risk drinking that detect current alcohol consumption, risk drinking, and lifetime DSM-III-R alcohol diagnoses. However, it was also the least specific. does not reflect more current research. SOURCE: Chang et al. 1998. Nonetheless, it offers another option for clinicians.

Vol. 25, No. 3, 2001 207 Other Screening of drinking by most pregnant women 2000), brief interventions (i.e., short Questionnaires (i.e., it is unlikely that pregnant women counseling sessions) may be especially will consume alcohol right before their effective in this population. Given the Research has not established the utility obstetric appointment) (Testa and potential adverse consequences of pre- of other screening questionnaires—the Reifman 1996; Lundberg et al. 1997; natal alcohol exposure, short screening CAGE, SMAST, AUDIT, and Prenatal Strano-Rossi 1999). However, recent questionnaires are worthwhile preven­ Alcohol Use Interview—for pregnant research has demonstrated the potential tive measures. women. The CAGE and the SMAST value of maternal blood markers for are popular self-report measures of detecting levels of alcohol use during alcoholism and are well studied in alco­ pregnancy that may result in overt holic and nonalcoholic subjects and alcohol-related deficits in newborns. among males (Bradley et al. 1998). The However, the most significant and most TWEAK AUDIT is a 10-item questionnaire that common result of prenatal alcohol expo- identifies harmful and hazardous drink­ sure, neurobehavioral dysfunction, is Tolerance: How many ing during the past year and has been not an outcome recognized in the new- T drinks can you hold? validated in six countries (Cherpitel 1995). born period. Therefore, research has yet The Prenatal Alcohol Use Interview is to establish the relevance of these blood Have close friends or rela­ a 13-item questionnaire that has been markers to the more common fetal alco­ W tives Worried or com­ tested in a sample of 56 women thus far hol effects (Jones and Chambers 1998; plained about your drink­ and requires further evaluation (Budd Stoller et al. 1998). (See the article by ing in the past year? et al. 2000). Bearer on pp. xx-xx of this issue for more Two large studies of disadvantaged, information on potential biomarkers to Eye Opener: Do you minority, obstetric patients (Hankin and detect alcohol use during pregnancy.) E sometimes take a drink in Sokol 1995; Russell et al. 1996) reported the morning when you that the calculated sensitivity and speci­ get up? ficity of the T-ACE and TWEAK were Summary superior to the CAGE in identifying Amnesia: Has a friend or risk drinking (defined as 1 ounce or Simple screening questionnaires, such A family member ever told more of alcohol consumption per day). as the T-ACE, provide valuable tools you about things you said In another study, we gave the SMAST, for identifying women who are using or did while you were AUDIT, and T-ACE questions inde­ alcohol during pregnancy. The T-ACE drinking that you could pendently to 350 pregnant women has been shown to identify any alcohol not remember? (Chang et al. 1998) and calculated how consumption during pregnancy as well well each of the three instruments could as higher amounts of drinking. Research Do you sometimes feel predict lifetime DSM-III-R alcohol diag­ has demonstrated that any alcohol con­ K(C) the need to Cut down on noses and any drinking during pregnancy. sumption during pregnancy increases your drinking? The SMAST did not perform better the risk of continued drinking during than chance as a predictor for either of pregnancy (Chang et al. 1999). The TWEAK is used to screen the two drinking categories. Although The T-ACE is administered easily. for pregnancy risk drinking, the AUDIT had good predictive abil­ A clinician may either ask the T-ACE defined here as the consumption ity, the definition of a “positive” score questions directly or request that the of 1 ounce or more of alcohol on the AUDIT for drinking pregnant patient complete the questionnaire while per day while pregnant. Scores women remains to be identified and waiting for her appointment. The T­ are calculated as follows: A positive confirmed through further research. ACE has been tested and demonstrated response to question T on to be acceptable and effective in both Tolerance (i.e., consumption of formats. more than five drinks) or question Laboratory Tests for A positive screen is not an indict­ W on Worry yields 2 points Detecting Alcohol Use ment. Rather, it is an opportunity for each; an affirmative reply to the clinician and patient to discuss pre- question E, A, or K scores 1 point Although the central focus of this arti­ natal alcohol exposure. The discussion each. A total score of 2 or more cle is on screening questionnaires, other may lead the clinician to refer the patient points on the TWEAK indicates methods of detecting alcohol use dur­ for a diagnostic assessment. Or the a positive outcome for pregnancy ing pregnancy deserve some comment. clinician may offer a brief intervention risk drinking. Use of breath analysis or urinalysis in if the patient does not have a severe pregnant patients is not likely to be fea­ alcohol problem. Because most preg­ SOURCE: Chan et al. 1993. sible or acceptable, given the rapid nant women are highly motivated to metabolism of alcohol and the pattern change their behaviors (Hankin et al.

208 Alcohol Research & Health Screening Instruments for Pregnant Women

REFERENCES United States during 1988–1995. Obstetrics and RUSSELL, M.; MARTIER, S.S.; SOKOL, R.J.; ET AL. Gynecology 92:187–192, 1998. Screening for pregnancy risk-drinking. Alcoholism: American Psychiatric Association. Diagnostic and Clinical and Experimental Research 18:1156–1161, Statistical Manual of Mental Disorders, Third Edition, EWING, J.A. Detecting alcoholism: The CAGE 1994. Revised. Washington, DC: the Association, 1987. questionnaire. Journal of the American Medical Association 252(14):1905–1907, 1984. RUSSELL, M.; MARTIER, S.S.; SOKOL, R.J.; ET AL. BABOR; T.F.; KRANZLER, H.R.; AND LAUERMAN, Detecting risk drinking during pregnancy: A com­ R.J. Early detection of harmful alcohol consump­ GRAHAM, K.; WILSNACK, R.; DAWSON, D.; AND parison of four screening questionnaires. American tion: Comparison of clinical, laboratory, and self- VOGELTANZ, N. Should alcohol consumption mea­ Journal of Public Health 86:1435–1439, 1996. report screening questionnaires. Addictive Behaviors sures be adjusted for gender differences? Addiction 93:1137–1147, 1998. SAMPSON, P.D.; BOOKSTEIN, F.L.; BARR, H.M.; 14:139–157, 1989. AND STREISSGUTH, A.P. Prenatal alcohol exposure, BABOR, T.F.; DE LA FUENTE, J.R.; SAUNDERS, J.; HANKIN, J.R., AND SOKOL, R.J. Identification and care birthweight, and measures of child size from birth to 14 years. American Journal of Public Health AND GRANT, M. AUDIT. The Alcohol Use Disorders of problems associated with alcohol ingestion in preg­ Identification Test. Guidelines for Use in Primary nancy. Seminars in Perinatology 19:286–292, 1995. 84:1421–1428, 1994. Health Care. Geneva, Switzerland: World Health HANKIN, J.; MCCAUL, M.E.; AND HEUSSNER, J. SELZER, M.L. The Michigan Alcoholism Screening Organization, 1992. Pregnant, alcohol-abusing women. Alcoholism: Clinical Test: The quest for a new diagnostic instrument. American Journal of Psychiatry 127(12):89–94, 1971. BRADLEY, K.A.; BOYD-WICKIZER, J.; POWELL, S.H.; and Experimental Research 24:1276–1286, 2000. AND BERMAN, M.L. Alcohol screening question­ ELZER INOKUR AND VAN OOIJEN HANNA, E.; DUFOUR, M.C.; ELLIOT, S.; STINSON, S , M.L.; V A.; R , L. naires in women: A critical review. Journal of the A self-administered Short Michigan Alcoholism F.; AND HARFORD, T.C. Dying to be equal: American Medical Association 280:166–171, 1998. Screening Test (SMAST). Journal of Studies on Women, alcohol, and cardiovascular disease. British Alcohol 36(1):117–126, 1975. BUDD, K.W.; ROSS-ALAOLMOLKI, K.; AND ZELLER, Journal on Addiction 87:1593–1597, 1992. R.A. Two prenatal alcohol use screening instruments SOKOL, R.J.; MARTIER, S.S.; AND AGER, J.W. The IKONOMIDOU, C.; BITTIGAU, P.; ISHIMAUR, M.; ET compared with a physiologic measure. Journal of T-ACE questions: Practical prenatal detection of AL. Ethanol-induced apoptotic neurodegeneration Obstetric, Gynecologic, and Neonatal Nursing 29: risk-drinking. American Journal of Obstetrics and and fetal alcohol syndrome. Science 287:1056– 129–136, 2000. Gynecology 160:863–871, 1989. 1060, 2000. Centers for Disease Control and Prevention. Alcohol STOLLER, J.M.; HUNTINGTON, K.S.; PETERSON, consumption among pregnant and childbearing-aged JACOBSON, J.L., AND JACOBSON, S.E. Prenatal alco­ C.M.; ET AL. The prenatal detection of significant women—United States, 1991 and 1995. Morbidity hol exposure and neurobehavioral development. alcohol exposure with maternal blood markers. and Mortality Weekly Report 46:346–350, 1997. Alcohol Health & Research World 18:30–36, 1994. Journal of Pediatrics 133:346–352, 1998. JACOBSON, S.W.; JACOBSON, J.L.; SOKOL, R.J.; ET CHAN, A.K.; PRISTACH, E.A.; WELTE, J.W.; AND STRANO-ROSSI, S. Methods used to detect drug AL. Maternal recall of alcohol, cocaine, and mari­ RUSSELL, M. The TWEAK test in screening for abuse in pregnancy: A brief review. Drug and alcoholism/heavy drinking in three populations. juana use during pregnancy. Neurotoxicology and Alcohol Dependence 53:257–271, 1999. Teratology 13:535–540, 1991. Alcoholism: Clinical and Experimental Research STRATTON, K.; HOWE, C.; AND BATTAGLIA, F., EDS. 6:1188–1192, 1993. JONES, K.L., AND CHAMBERS, C. Biomarkers of fetal Institute of Medicine Summary: Fetal Alcohol Syndrome. Washington, DC: National Academy Press, 1996. CHANG, G.; WILKINS-HAUG, L.; BERMAN, S.; ET exposure to alcohol: Identification of at-risk preg­ nancies. Journal of Pediatrics 133:316–318, 1998. AL. Alcohol use and pregnancy: Improving identifi­ TESTA, M., AND REIFMAN, A. Individual differences cation. Obstetrics and Gynecology 91:892–898, 1998. KELLY, S.J.; DAY, N.; AND STREISSGUTH, A.P. in perceived riskiness of drinking during pregnancy: Antecedents and consequences. Journal of Studies on CHANG, G.; WILKINS-HAUG, L.; BERMAN, S.; AND Effects of prenatal alcohol exposure on social behav­ Alcohol 57:360–367, 1996. GOETZ, M.A. A brief intervention for alcohol use ior in humans and other species. Neurotoxicology and during pregnancy: Results from a randomized trial. Teratology 22(2):143–149, 2000. URBANO-MARQUEZ, A.; ESTRUCH, R.; FERNANDEZ­ Addiction 94:1499–1508, 1999. SOLA, J.; ET AL. The greater risk of alcoholic car­ LUNDBERG, L.S.; BRACKEN, M.B.; AND SAFTLAS, A.F. Low to moderate gestational alcohol use and diomyopathy and myopathy in women compared CHARNESS, M.E.; SAFRAN, R.M.; AND PERIDES, G. with men. Journal of the American Medical Association intrauterine growth retardation, low birthweight, Ethanol inhibits neural cell-adhesion. Journal of 274:149–154, 1995. Biological Chemistry 269:9304–9309, 1994. and preterm delivery. Annals of Epidemiology 7:498–508, 1997. VERKERK, P.H. The impact of alcohol misclassifica­ CHERPITEL, C.J. Screening for alcohol problems in tion on the relationship between alcohol and preg­ the emergency department. Annals of Emergency MORROW-TLUCAK, M.; ERNHART, C.B.; SOKOL, nancy outcome. International Journal of Epide­ Medicine 26:158–166, 1995. R.J.; MARTIER, S.; AND AGER, J. Underreporting of miology 21(suppl.):S33–S37, 1992. alcohol use in pregnancy: Relationship to alcohol. DAY, N.L.; COTTREAU, C.M.; AND RICHARDSON, Alcoholism: Clinical and Experimental Research WINDHAM, G.C.; VON BEHREN, J.; FENSTER, L.; G.A. Epidemiology of alcohol, marijuana, and 13:399–401, 1989. SCHAEFER, C.; AND SWAN, S.H. Moderate maternal cocaine use among women of childbearing age and alcohol consumption and the risk of spontaneous pregnant women. Clinical Obstetrics and Gynecology ROSNER, B. Fundamentals of Biostatistics. Belmont, abortion. Epidemiology 8:509–514, 1997. 36:237–245, 1993. CA: Duxbury Press, 1990. p. 55. WONG, E.V.; KENWRICK, S.; WILLEMS, P.; AND EBRAHIM, S.H.; LUMAN, E.T.; FLOYD, R.L.; RUSSELL, M. New assessment tools for risk drinking LEMMON, V. Mutations in cell adhesion molecule MURPHY, C.C.; BENNETT, E.M.; AND BOYLE, C.A. during pregnancy. Alcohol Health & Research World L1 cause mental retardation. Trends in Neuroscience Alcohol consumption by pregnant women in the 18:55–61, 1994. 18:168–172, 1995.

Vol. 25, No. 3, 2001 209