This study was conducted from January 14, 2005, to February 28, 2005, drawing on a total undergraduate population of 20 138 full-time students (10 339 women and 9799 men) attending a large US public research university. After the study received institutional review board approval, a random sample of 5389 full-time undergraduate students was drawn from the total undergraduate population. The entire sample was mailed $2.00 along with a prenotification letter that described the study and invited students to self-administer a Web survey by using a URL address and unique password. Informed consent was obtained online from each participant. Nonrespondents were sent up to 4 reminder e-mails. The Web survey was maintained on an Internet site running with the secure socket layer protocol to ensure privacy and security. By participating in the survey, students became eligible for a sweepstakes that included cash prizes, travel vouchers, tickets to athletic events, and iPods. The final response rate was 68%, and potential nonresponse bias was assessed by administering a short form of the questionnaire via telephone to a randomly selected sample of 159 students who did not respond to the original Web survey.
The sample consisted of 3639 undergraduate students (53.6% women and 46.4% men). The mean (SD) age of students in the sample was 19.9 (2.0) years. The racial/ethnic distribution of the sample was 67.4% white, 12.1% Asian, 6.0% African American, 4.2% Hispanic, and 10.2% other ethnic categories. The sample was made up of 28.5% freshmen, 23.4% sophomores, 23.1% juniors, and 25.0% seniors. The demographic characteristics of the sample closely resembled the overall student population at this university. The family income distribution for the sample was as follows: 12.4% less than $50 000, 23.0% from $50 000 to $99 999, 17.9% from $100 000 to $149 000, 11.8% from $150 000 to $249 999, 9.2% from $250 000 or more, and 25.8% did not know. A total of 46.5% of the sample lived in university residence halls, 43.7% resided in a house or apartment, 4.4% lived in a fraternity or sorority house, and 5.4% resided in some other living location; 13.1% of respondents were active members of a social fraternity or sorority.
Lifetime medical use of prescription medication was measured using the following question: “Based on a doctor’s prescription, on how many occasions in your lifetime have you used the following types of drugs?” A separate question was asked for each of the following 4 classes of prescription drugs: (1) sleeping medication (eg, Ambien [zolpidem], Halcion [triazolam], Restoril [temazepam], temazepam, triazolam); (2) sedative or anxiety medication (eg, Ativan [lorazepam], Xanax [alprazolam], Valium [diazepam], Klonopin [clonazepam], diazepam, lorazepam); (3) stimulant medication (eg, Ritalin [methylphenidate], Dexedrine [dextroamphetamine], Adder-all [dextroamphetamine and amphetamine], Concerta [methylphenidate], methlyphenidate); and (4) pain medication (ie, opioids such as Vicodin [hydrocodone and acetaminophen], OxyContin [oxycodone], Tylenol 3 [acetaminophen] with codeine, Percocet [oxycodone and acetaminophen], Darvocet [propoxyphene and acetaminophen], morphine, hydrocodone, oxycodone). The response scale for each question ranged from (1) never to (7) on 40 or more occasions. Similar variables were used to assess past-year medical use of prescription medication.
Lifetime nonmedical use of prescription medication was assessed by asking the following question: “Sometimes people use prescription drugs that were meant for other people, even when their own doctor has not prescribed it for them. On how many occasions in your lifetime have you used the following types of drugs, not prescribed to you?” There were separate questions for the same 4 classes of prescription drugs as medical use with identical response scales and wording. Similar variables were used to assess past-year nonmedical use of prescription medication.
Lifetime prescription drug use status was assessed by creating 4 distinct groups of lifetime prescription medication use: (1) never used 1 or more of the 4 classes of prescription medication (nonuse); (2) used only 1 or more of the 4 classes of prescription medication as prescribed by their physicians (medical use only); (3) used 1 or more of the 4 classes of prescription medication as prescribed by their physicians and prescription medication that was not prescribed to them (both medical and nonmedical use); and (4) used only 1 or more of the 4 classes of prescription medication that was not prescribed to them (nonmedical use only). Similar 4-level variables were developed for each specific drug class and past-year prescription medication use.
Screening for probable drug abuse was assessed using the Drug Abuse Screening Test, Short Form (DAST-10), which is a self-report instrument that can be used in clinical and non-clinical settings to screen for probable drug abuse or dependence on a wide variety of substances other than alcohol.30
Respondents who used drugs other than alcohol in the past 12 months were asked whether they had experienced 10 drug-related problems (eg, inability to stop using drugs, simultaneous polydrug use, illegal activities to obtain drugs, blackouts as a result of drug use, medical problems as a result of drug use, withdrawal symptoms, feeling bad or guilty about drug use, family complaints about drug use, and family avoidance because of drug use). The DAST-10 items were not drug specific, and respondents were informed that drug
refers to use of prescription drugs not prescribed to you or in a manner not intended by the prescribing physician or use of other drugs such as marijuana, cocaine, lysergic acid diethylamide, or Ecstasy. On the basis of previous research, if a respondent positively endorsed 3 or more DAST-10 items, this was considered a positive screening result, denoting risk for probable drug abuse or dependence.31–33
The DAST-10 has been shown to have good reliability (Cronbach α=0.86) and temporal stability (test-retest intraclass correlation coefficient=0.71) and identifies individuals who need more intensive assessment for substance abuse problems.31
Evidence for concurrent validity comes from previous work based on these data, which showed that the DAST-10 was positively correlated with frequency of illicit drug use and negatively correlated with age at onset of illicit drug use.34
Maisto et al33
evaluated the DAST-10 using the Diagnostic and Statistical Manual of Mental Disorders
(Fourth Edition) (DSM-IV
drug use disorder diagnosis as the criterion and found levels of sensitivity and specificity of 0.70 and 0.80, respectively, when using a cutoff point of 3. The Cronbach α in the present study was 0.69 for the DAST-10 items.
Data analyses included 3639 respondents, and all statistical analyses were performed using SPSS statistical software, version 13.0 (SPSS Inc, Chicago, Illinois). We used χ2 tests to compare the prevalence of medical use and nonmedical use according to student characteristics. We used χ2 tests, analysis of variance tests, and multiple logistic regression models to compare DAST-10 scores across the following 4 mutually exclusive groups of lifetime and past-year prescription medication users: (1) non-use, (2) medical use only, (3) both medical and nonmedical use, and (4) nonmedical use only. Multiple logistic regression models were conducted with nonusers serving as the reference group and were adjusted for the following covariates: sex, race/ethnicity, class year, family income level, living arrangement, social fraternity or sorority membership, and age at onset for use of alcohol, tobacco, and marijuana. Adjusted odds ratios and 95% confidence intervals were reported.