Drug Abuse Research:
Science for Policy and Practice
Summary of PresentationsModerator: Mr. Daniel Schecter
Assistant Deputy Director/Acting Deputy Director for Demand Reduction
Office of National Drug Control Policy, USA
Dr. Donald Goldstone
Director, Office of Applied Studies,
Substance Abuse and
Mental Health Services Administration,
Department of Health and Human Services, USA
Dr. Goldstone first provided an overview of SAMHSA's National Household Survey on Drug Abuse. This survey, he noted, was designed to provide a broad epidemiological picture of drug abuse in the general population of the United States and has been conducted annually since 1971. He briefly discussed the survey's history, limitations, recent findings, and probable future directions.
As perceived risk of drug use diminishes, use tends to increase, and in recent years the perceived risk of cocaine use, for example, has decreased.
As Dr. Goldstone noted, in the beginning, the Survey was conducted rather inconsistently in that the time of the year, the intervals between surveys, and the sizes of the samples all varied. The household survey was designed to provide a national probability sample using actual household interviews in which sensitive data were collected anonymously. The survey also included sociodemographic data on the participating families as well as some limited behavioral data (e.g., a few questions about criminal behavior).
Dr. Goldstone highlighted the Survey's 1996 data (the table has since been updated to reflect the 1997 Survey data). He noted that an estimated 4.7 percent of the U.S. population used marijuana, and another 1.4 percent were using other illicit drugs but not marijuana. Furthermore, drug use declined from 1979 to 1996. Although drug use among youth increased from 1992 to 1995, there was a decrease in use, primarily among those aged 12 to 15, from 1995 to 1996. However, Dr. Goldstone claimed that the picture presented by perceived risk may be less sanguine. As perceived risk diminishes, he noted, use tends to increase and in recent years the perceived risk of cocaine use, for example, has decreased.
Dr. Goldstone stated that several significant changes are now under way to make the survey more sensitive to drug trends. For example, the size of the sample, beginning in January 1999, will increase sharply -- from 19,000 to 69,000 households (i.e., 69,000 respondents). Moreover, samples in each of the eight largest states will include 3,700 respondents; for the rest of the country, the sample will be 900 per state. The total sample in each of the three age categories (12 to 17, 18 to 25 and over 25 years) will be 900 per state. The total sample in each of the three age categories (12 to 17, 18 to 25 and over 25 years) will be 23,000. The larger the samples, the more accurate estimates of overall use.
In the year 2000, he noted, a second major change is planned: a computer-assisted, personal interview will be used, which will result in overnight data transfer and a higher quality data set. An audio computer-assisted technique will also be used to ask questions in sensitive areas and increase the validity of responses. The third major change will be in the content of the questionnaire. According to Dr. Goldstone, information solicited will cover such areas as insurance coverage, treatment, and prevention as well as respondents' antisocial behavior.
Although the household survey is a useful indicator of trends, Dr. Goldstone pointed out that it has some limitations. It is focused, for example, on the general population living in households. Thus, he explained, it does not include institutional populations (e.g., prisoners), children under the age of 12, and those unlikely to be found in household units (e.g., the homeless). Moreover, not everyone tells the truth, making absolute information uncertain. Nevertheless, he claimed that the household survey is a useful indicator of trends and an important part of an increasing effort by government to obtain objective data by which progress in solving the drug problem can be more accurately gauged.
See next page for excerpt form 1997 U.S. Household Survey on Drug Abuse
