Early Studies on Prevention of Alcohol Use in Pregnancy

April 23-24, 1998 • Ramada Inn • Bethesda, Maryland

Abstracts

The Seattle Pregnancy and Health Program: A Demonstration Grant to Prevent Fetal Alcohol Syndrome and Intervene in Female Alcohol Abuse During Pregnancy Ruth E. Little, Sc.D., Ann P. Streissguth, Ph.D., and Gay M. Guzinski, M.D.

From the Department of Psychiatry and Behavioral Sciences and the Department of Obstetrics and Gynecology, University of Washington School of Medicine, Seattle, Washington

The Pregnancy and Health Program (PHP) consisted of three components: public education, professional training, and providing services. Our slogan was "When you are pregnant, the best drink is no drink at all." A staff of 27 employees and 15 volunteers participated. The primary target area for service delivery was King County, Washington.

Public education was achieved through a massive campaign of warning signs and brochures at State liquor stores, on buses, in libraries, and in other public places (220,000 informational brochures were distributed). Nine public service announcements were developed for radio and television, and a dysmorphology manual and a "drinking for two" cookbook were produced. A dedicated hotline (5-HEALTH) was answered at least 12 hours a day. It received 4,306 calls during its 2 years of operation. One in every 44 pregnant women in the target area called 5-HEALTH, almost two-thirds of them in the first trimester of pregnancy. The cost of 5-HEALTH was approximately $14.50 per call, or $429 per call for those who became clients of PHP. A continuing legacy of this program is a 24-hour alcohol and other drug hotline operated by the Washington State Division of Alcohol and Substance Abuse that is answered by an information specialist who responds to callers’ questions, recommends not drinking during pregnancy, and refers callers to physicians who have special training in alcohol and substance abuse.

Professional training was achieved by training trainers for each field who then trained professionals in their own disciplines (physicians, nurses, alcoholism counselors, social workers, psychologists, and teachers). Over 6,300 professionals were trained, and over 7,000 informational brochures and 600 technical packets designed for professionals were distributed. Research on random residents in the target area, obstetricians, other health care professionals, and pregnant women before and after the program showed significant increases in the proportion recommending abstinence or limited alcohol use during pregnancy. The finding that more pregnant women recommended not drinking during pregnancy than did their own obstetricians revealed the need for public education as well as professional training.

A two-stage screening service was developed that involved a short questionnaire administered in the waiting room of obstetricians’ offices and a brief intervention interview. Of the 1,300 Seattle women screened, 11% were considered to be at genuine risk (averaging one or more drinks per day, consuming five or more drinks on at least one occasion, or showing evidence of a past or present alcohol problem). A very brief 2-question replacement questionnaire was found to identify 92% of those originally defined to be at genuine risk. The two questions ask whether the individual ever consumed 5 or more drinks on any occasion and whether she felt that she should cut down on her drinking. This simple screening device identified four times as many risk level drinkers as were documented in the women’s obstetrical records. A brief intervention interview was then administered to those at risk. This provided information on alcohol and pregnancy, recommended abstaining from alcohol during the remainder of the pregnancy, and helped each woman work out an individual plan. Women were also offered several intervention sessions with specially trained alcohol counselors, or were referred to alcohol treatment specialists with special training in pregnancy issues.

Three-fourths of the women who were drinking moderately to heavily were able to either stop or significantly reduce their alcohol intake after a brief intervention; 86% were judged by independent raters to have improved. A significant decreasing linear trend was observed between fetal alcohol effects in infants (abnormalities and neurological findings previously associated with maternal drinking) and decreased length of alcohol exposure. Mothers still drinking at risk levels in the third trimester had infants with three times the number of infant abnormalities as those reducing their drinking in the first trimester, and more pregnancy complications.

Lessons Learned from Early Studies of Alcohol Abuse in Pregnancy
Philip A. May, Ph.D.

From CASAA, University of New Mexico, Albuquerque, New Mexico

My colleagues and I began prevention and research project work on fetal alcohol syndrome (FAS) in the southwestern United States in 1979 and have had contact with or directed a number of programs. Described will be the lessons learned from at least four of these efforts.

The initial pilot project of the Indian Health Service (IHS) was carried out from 1979 through 1983. In this project, we used epidemiology as a base for a variety of our efforts. In this first project, we developed and used active case ascertainment methods to facilitate the referral and identification of FAS cases, provided diagnostic services with a treatment plan for the children, and worked to establish the epidemiologic features of FAS for prevention applications. When we applied this methodology, we were able to establish the prevalence and characteristics of FAS children, the prevalence of mothers and families who were producing the FAS cases, and the characteristics of the community and maternal risk factors in the research area. We used these facts to design public health intervention and prevention plans.

The IHS funded the National Indian FAS Prevention Program from 1993 through 1995. It was a limited model, which was based on primary prevention efforts directed at training trainers in all IHS areas. Over 1,900 trainers in FAS Prevention were trained in 92 different locations throughout the United States. Although testing revealed that those trained gained and retained knowledge, the limited primary prevention (universal prevention) model of this program was lacking in several ways. We learned you could not depend on primary prevention alone; that you need community prevalence data to be able to assess the problem, design adequate prevention approaches, and detect changes in FAS rates. We also learned that special incentives are needed to set aside resources for continuing FAS prevention programs over a number of years, and that local programs are extremely dependent on a small number of key individuals whose enthusiasm and expertise drive and carry the prevention effort.

In a collaborative effort with the Tuba City FAS Prevention Program of the Navajo Area IHS and the Navajo Tribe, a program was evaluated that emphasized secondary (selected) and tertiary (indicated) prevention to reduce the existing prevalence rate of FAS in that community. Since this particular community had participated in the early IHS screening efforts, the rate of FAS was known. Women of childbearing age were routinely screened in IHS clinics and referred to the prevention program if found to be drinking during pregnancy or were known to be mothers of children with FAS or fetal alcohol effects (FAE). The prevention program offered women referral to alcohol treatment and/or birth control. This dual approach doubled the chances for success and was evaluated positively with the highest-risk mothers. The program also taught us that case management can positively impact many problem areas of FAS mothers’ lives and thereby help ensure abstinence from alcohol in the future. Although the rate of FAS was not reduced, it did not increase as measured in periodic screening clinics.

A fourth and final program from which we have learned a great deal was carried out in a larger Indian pueblo in New Mexico. This Pueblo tribe had also participated in the IHS pilot projects, and their prevalence of FAS from 1978 through 1982 was 4.1 per 1,000 children. This tribe utilized prevention information from a variety of sources (including the IHS pilot projects, State and local programs, the National Indian FAS Prevention Program, the Center for Substance Abuse Prevention [CSAP]) to develop a comprehensive FAS prevention coordination effort. The Pueblo network screens high-risk women in a variety of settings (including prenatal clinics), refers those with positive screening results to public health nurses and IHS clinicians for case management, provides alcoholism treatment for women who are pregnant, and conducts primary prevention programs in the schools and in various community agencies. This coordinated effort has resulted in a reduction of the FAS rate to between 1.3 to 1.9 per 1,000 for the years 1984 through 1995. This not yet unpublished case is certainly noteworthy.

The Alcohol Warning Label: Analyzing the Impact of Universal Prevention Effort
Janet R. Hankin, Ph.D.

From the Department of Sociology, Wayne State University, Detroit, Michigan

The Alcoholic Beverage Warning Label, which was implemented on November 18, 1989, was designed to warn pregnant women about the risks of antenatal drinking. The purpose of our study was to examine its impact on the knowledge, attitudes, and drinking behavior of pregnant African-American women.

Between 1986 and 1995, 21,127 gravidas were interviewed at the time of their first prenatal visit to Hutzel Hospital’s prenatal clinic. The clinic is located in the inner city of Detroit and is affiliated with Wayne State University and the Detroit Medical Center. Data were collected on periconceptional drinking (alcohol consumed one week around the time of conception) and antenatal drinking (alcohol consumed two weeks prior to the first prenatal visit). Beginning in May,1989, women were asked whether there was a warning label on alcoholic beverages and about their attitudes about drinking during pregnancy. Thus, data were obtained both pre- and post-implementation of the warning label.

When examining the impact of a universal prevention effort such as the warning label, it was important to control for other factors that may affect antenatal drinking. We considered several factors, including changes in the population seeking care (age, gravidity, weeks of gestation when prenatal care was first received), changes in drinking patterns (periconceptional drinking), and other social conditions that might affect drinking (e.g., the unemployment rate). A second issue was the application of appropriate statistical techniques. We chose Time Series Analysis, a technique that corrects for autocorrelations among the observations; detects seasonal, upward, or downward trends; and determines when changes occurred. A third problem was detecting when the labeled containers actually appeared on the shelves. Given the fact that containers had to be labeled at the point of bottling, we anticipated a lag in the impact of the warning label, since it would take time before old stock would clear the shelves. To analyze the intervention, we had to determine the month when the label law might show an effect. For that reason, our models incorporated a lag period. A fourth problem was determining whether the label was more effective in certain subgroups of pregnant women. We examined the impact for subgroups of women defined by their age, gravidity, and risk drinking status. A fifth issue was controlling for other major universal prevention efforts targeting antenatal drinking. We monitored the media during the study period, and could find no other major universal prevention programs. Finally, the duration of any impact was important, so we collected data for six years after the law was passed.

An examination of 18,888 alcoholic beverage containers between May and June 1990 showed a lag between implementation of the law and appearance of labeled containers on the shelves. While 86% of beer containers were labeled, only 69% of wine coolers, 34% of wine bottles, and 30% of liquor bottles were labeled. Women whose beverage of choice was beer or wine cooler were more likely to know about the labels. A significant increase in label awareness occurred in March 1990, and by 1992, the upper limit of awareness–about 80%–was reached. Using sophisticated statistical techniques, controlling for a large number of confounding factors (including unemployment), and analyzing data on over 21,000 gravidas, we found that the warning label had a small impact of short duration on antenatal drinking. However, older women, those with higher parity, and risk drinkers were more likely to ignore the warning label and drink during pregnancy. It is clear that we need to go beyond the warning label to prevent FAS and alcohol-related effects.

Discussion

The Seattle Pregnancy and Health Program began five years after FAS was first identified. Although the community’s awareness of the risks of drinking during pregnancy was high, there was little understanding of what qualified as risky drinking. Respondents in Seattle thought three to four drinks of distilled spirits per day were safe, although beer and wine were generally considered safer. Dr. Streissguth and her colleagues focused much of their public education work on the equivalent amounts of alcohol in different drinks and only recommended abstinence to the women in their program after much careful consideration. Their approach to public education was upbeat. Its slogans–"mothering begins before birth," "when pregnant, the best drink is no drink at all"–focused on the goal of bearing healthy babies rather than castigating mothers for their risk drinking. Cookbooks, videos, radio spots, and bus posters reinforced these messages.

Professional training was another important feature of the project. Although obstetricians proved the most resistant to change, by the program’s end, 90% were asking their pregnant patients about their drinking practices. The program staff also learned that of the mothers receiving services from the program, those who began earlier in their pregnancies had the healthiest babies. The biggest unresolved problem for the program was how to detoxify pregnant women, a problem that remains today.

With hindsight, Dr. Streissguth would conduct the Seattle Pregnancy and Health Program in much the same way, but with one important difference. From day one, she would work toward integrating the program within its sponsoring institution so that services would endure when government funding ended.

Dr. May’s early experiences in responding to FAS in Indian country taught him the importance of obtaining baseline prevalence as a foundation for program development and of building programs within a cultural and community context. Women in these early programs were almost 30 years old and had multiple FAS children. Experience in working with them led to the development of a second effort designed to identify high-risk women as early in their pregnancy as possible. A training-of-trainers approach was used with local advocates who conducted culturally appropriate prevention programs in their respective communities. Since there was no baseline knowledge of prevalence, however, it was difficult to determine whether these efforts were having any impact. A later effort in Tuba City, Arizona, rectified that problem. A small staff established a screening program in a medical setting to target specifically women with one FAS baby. Abstinence was the goal, but reductions in use were considered positive outcomes. Since approximately a third of the mothers had lost custody of their children, winning them back became a key motivator for participating in the program. Although family planning is a concern for many Native Americans, a dual approach of contraception and abstinence worked well with this subpopulation. T-shirts, posters, and other educational materials featuring native art were well received by the target women and the larger community. Those pregnant women who maintained contact with the program achieved an 81% abstinence rate.

In the last of these early programs in Indian country, having healthy babies was linked to the Zuni tradition. Screening was extended to WIC programs, and child protective services among other locations, and results of the screens were used as wake-up calls. Both men and women were involved in the intervention, which was seen as protecting the Zuni culture. As a result of the intervention, the FAS rate was cut in half.

The longitudinal study of beverage warning labels directed by Dr. Hankin considered the impact of the warning labels on low-income, high-risk, pregnant African-American women in terms of their increased knowledge about risks and changes in attitude and subsequent drinking practices. The study accounted for social conditions (e.g., unemployment) that might influence positive and negative changes in drinking by using a time-series technique that could detect seasonal changes and determine when change occurred and whether it was real. Results of the study showed that knowledge increased and that warning labels had an impact on antenatal drinking although it was of a short duration. The study also showed that drinking upturns mimicked the climb in unemployment and that women who ignored the labels were risk drinkers of older age and higher parity. Most important, the study showed that universal prevention efforts alone do not have a significant impact on those women at highest risk for bearing an FAS baby; multiple, targeted initiatives are needed.

Workgroup members raised concerns about the perceived drop in public interest in FAS. Although specific groups (notably adoptive parents) are eager for information, Dr. Kaskutas observed that there has been a decline in the proportion of women who report exposure to public service announcements on the topic. Dr. Coles and Robert Schacht attributed some of the decline to reduced funding for FAS-related activities from such Federal agencies as the Indian Health Service and changes in the health care environment, which no longer funds substance abuse treatment services at the same level. While not disagreeing with the thrust of those arguments, Dr. May also noted that within the substance abuse treatment system, relatively few programs treat pregnant women.

Workgroup members believed that continued broad exposure for FAS messages was vitally important but agreed that universal approaches must be balanced by comprehensive community-based programs comprised of targeted interventions. Fulton Crews cited the example of grandparents as a potential target for future interventions. Dr. May concurred and urged a larger role for the extended family and more emphasis on spouse engagement. He did, however, issue a caution. In his experience, prevention/intervention programs frequently become diverted and find themselves more focused on treating FAS-affected children than preventing/intervening with the behavior of risk-drinking mothers and women of childbearing age. Dr. Streissguth reinforced the importance of focus. In her programs, she involves parents of FAS-affected children as exemplars who function as powerful prevention agents within their respective communities.

Summary

• To make inroads on FAS, universal prevention approaches must be balanced with indicated and selected programs.
• Sound data (especially prevalence data) are a prerequisite for effective FAS prevention/intervention programs.
• If FAS interventions are not institutionalized, they will not survive.