Advancing Sexual and Reproductive Health and Rights
Family Planning Perspectives
Volume 32, Number 6, November/December 2000

Pregnancy and Alcohol: Many Obstetrician-Gynecologists Are Unsure About Risks or How to Assess Women's Use

Virtually all obstetrician-gynecologists participating in a national survey ask pregnant women whether they use alcohol; most obtain this information during an initial visit with a patient, a practice recommended by the American College of Obstetricians and Gynecologists (ACOG) and the American Academy of Pediatrics. Nevertheless, only 30% of obstetrician-gynecologists feel very prepared to assess pregnant women's alcohol use, and 83% say that they need information on thresholds at which prenatal drinking poses specific threats to the pregnancy or the fetus. Moreover, although the federal government has for 20 years advised pregnant women to refrain from drinking, about half of doctors surveyed believe that occasional use of alcohol during pregnancy will not increase the risk of several adverse outcomes.1

ACOG conducted the survey in 1998 among a sample consisting of active members of the organization plus the members of an ACOG research network of physicians who volunteer to participate in periodic, topical surveys. The survey asked doctors about their alcohol screening practices, their opinions about the level of use that puts women at risk of particular adverse outcomes, and their counseling and referral practices for pregnant women who drink moderately (i.e., have an average of 3-13 drinks weekly) or heavily (i.e., have 14 or more drinks per week or at least five drinks at any one time).

In all, 604 obstetrician-gynecologists completed the survey. Three-fifths of respondents were men, and nine in 10 were younger than 60. About one-quarter had graduated from medical school before 1973, half had graduated between 1973 and 1989, and one-quarter had graduated later.* Fifty-one percent worked in a group private practice; 21% in a solo private practice; 20% in a managed care organization, university or medical school, or government institution; and 8% in other settings. The questionnaire was mailed to ACOG fellows throughout the country; however, because of the sampling procedure used, the final sample may not be representative of the organization's membership.

Ninety-seven percent of respondents said that they obtain information on alcohol use from all of their obstetric patients; 92% do this at the first prenatal visit. Obstetrician-gynecologists use one or more of a variety of approaches to alcohol screening: Forty-eight percent ask women whether they drink, 41% have a nonphysician staff member ask and 19% include the question on a form that patients fill out. Twenty-three percent use one of the standardized screening tools that are available to help physicians detect alcohol use among pregnant women.

If a pregnant patient says that she drinks, 90% of respondents ask about her level of alcohol consumption. The vast majority also talk with her about the adverse effects of prenatal alcohol use--86% if she is a moderate drinker and 97% if she drinks heavily. Similarly, high proportions advise patients who drink to discontinue alcohol use while pregnant (83% for moderate drinkers and 92% for heavy drinkers) or to reduce their alcohol consumption (79% and 88%, respectively). Far fewer, however, refer women for treatment: Only 21% refer moderate drinkers, and 61% refer heavy drinkers.

Providing education about the effects of drinking during pregnancy is not a routine practice for many physicians. While 50% offer information or advice to all pregnant women, 36% raise the issue only if they know or suspect that a patient uses alcohol. Thirteen percent of respondents offer education or advice about prenatal drinking only to women with risk factors associated with alcohol consumption during pregnancy (a history of drug use or heavy drinking, or current smoking).

Although thresholds have not been established, the survey asked physicians how many drinks per week they think pregnant women can consume without increasing their risk of having a spontaneous abortion or of bearing an infant with central nervous system impairment, birth defects or fetal alcohol syndrome. For each of these outcomes, 26-31% of respondents believe that any alcohol consumption is too much, and 16-28% are unsure. The remainder (46-56%) think that some alcohol consumption poses no risk; the average weekly number of drinks considered safe ranged from 4.6 for spontaneous abortion and central nervous system impairment to 6.6 for fetal alcohol syndrome.

The bulk of obstetrician-gynecologists surveyed (66%) feel somewhat prepared to assess women's alcohol use, but only 30% feel very prepared and 4% feel unprepared. The need for additional training in this area is one of the most frequently cited barriers to the provision of this service (mentioned by 65% of physicians); others are time limitations, concerns about patients' sensitivity to the subject and a lack of referral sources (50-70%). When asked what resources they need to improve their ability to assess patients' alcohol consumption, 83% of respondents said information on thresholds for poor pregnancy outcomes.

Twenty-seven percent of physicians think that medical school did not adequately prepare them to assess pregnant women's alcohol use; 35% consider their training on the subject adequate, and the rest believe that it was very good or outstanding. The more positively physicians view their training, the more likely they are to use a standardized screening tool and to feel prepared to assess patients' alcohol use. The level of satisfaction with training is highest among physicians who graduated after 1989 and falls significantly among those who graduated earlier.

In analyses that adjusted for respondents' age, the researchers found few significant differences between male and female physicians' practices regarding pregnant patients who drink. Male doctors are about 10% less likely than females to advise moderate drinkers to stop drinking (prevalence rate ratio, 0.9), and they are 19% less likely to mention referral sources as a resource that would improve their ability to assess patients' alcohol use. However, men are 15% more likely than women to say that women who are pregnant or trying to conceive should abstain from alcohol use.

The timing of physicians' medical education, however, is associated with several aspects of their alcohol assessment practices. When physicians' gender is taken into account, those who graduated before 1973 are nearly 80% less likely than those who graduated after 1989 to use a standardized screening tool and about 40% less likely to consider themselves very prepared to assess alcohol use (prevalence rate ratios, 0.2 and 0.6, respectively). These doctors are less likely than those who graduated most recently to discuss the adverse effects of alcohol use with moderate drinkers, to advise these women to discontinue or reduce their alcohol use, and to advise heavy drinkers to cut down (ratios, 0.7-0.9). Furthermore, they are more likely than those who graduated after 1989 to be unsure as to thresholds for adverse outcomes, to say they need training in alcohol assessment and counseling, and to say that a lack of insurance reimbursement is a barrier to providing these services (1.4-1.7). They are 10 times as likely to feel that their medical school training in this area was inadequate (10.1).

Fewer, and somewhat less striking, differences exist between those who graduated in 1973-1989 and those who completed school later. Physicians who graduated in the middle period are less likely than those who graduated later to use a screening tool, feel prepared to assess alcohol use, discuss the adverse effects of alcohol with heavy drinkers and advise such women to reduce their alcohol consumption (0.5-0.9). They are about 50% more likely than their colleagues who graduated later to consider a lack of insurance reimbursement an obstacle to alcohol screening or counseling, and they are nearly three times as likely to consider their training inadequate.

The researchers observe that despite two decades of federal advisories on the hazards of drinking during pregnancy, "many physicians are not convinced that total abstinence from alcohol is necessary for pregnant women." To remedy this situation, they suggest that professional organizations and public health agencies participate in "meaningful dialogue" about the issue and that federal, professional and nonprofit agencies collaborate to raise awareness of the potential effects of prenatal alcohol use. Furthermore, they note the importance of physicians' being kept up to date on the effects of drinking during pregnancy, on ways to assess patients' alcohol use and on interventions to use when they see pregnant patients who drink. --D. Hollander


1. Diekman ST et al., A survey of obstetrician-gynecologists on their patients' alcohol use during pregnancy, Obstetrics and Gynecology, 2000, 95(5):756-763.

*These periods reflect the evolution of understanding of and education about alcohol use during pregnancy: Prior to 1973, fetal alcohol syndrome had not yet been documented. In the middle period, the first surgeon general's report on prenatal drinking was issued, and medical school curricula began to include information on the subject. The period since 1990 has been characterized by extensive warning labels on alcohol products and official government health advisories.