Computer-based Prescription Opioid Abuse Prevention for Adolescents

This description will focus on the randomized, controlled trial to evaluate the effectiveness of the web-based prevention intervention the investigators are developing.

Recruitment. Youth will be recruited for study participation via the social networking website, Facebook. This site allows advertisers to choose the demographics of the audience that ads target. This will lead to significantly fewer clicks on the ad from advertising site users who do not meet inclusionary criteria for the study. The investigators plan to post banner ads that promote the study to 12-17 year olds living in the U.S. To increase credibility, the investigators will ensure the ad references the organizations involved in this research project and the funding source. The investigators anticipate running ads for about 3 months to obtain an adequate sample size of approximately 400 for the study.

Screening and Enrollment. Upon clicking on a study ad, individuals will be taken to the study website and presented with a brief introduction to the study and an anonymous screener. Individuals who are eligible but who decline participation will be asked to indicate reasons for non-interest. Persons who click on the ad but are ineligible or who decline participation will be provided with the link to the Drug Enforcement Administration's website (JustThinkTwice.gov).

The investigators will employ several procedures to ensure that prospective participants are indeed youth within the targeted sample. All prospective participants will be asked to complete an electronic form with participant parents' name, address and phone numbers. An information packet will then be mailed to parents, which will include a letter describing the study and a brief summary of why parents are receiving the packet, 2 copies of a consent/assent form, investigators' contact information (including a toll-free number), a request for the participant's email address and phone number, and a self-addressed, stamped envelope for returning 1 signed copy of the consent/assent. To verify parental consent once received, a research team member will call parents to confirm the child's participation. A research team member will also visually inspect parents' and child's signatures on consent/assent forms, and any that are suspiciously similar will lead to non-enrollment.

Youth who provide appropriate consents/assents will be emailed a unique study identification (ID) and temporary password to use on the study's secure website. During the initial login, participants will be required to change their password. Participants will be invited to complete baseline assessments online. As a further validity check, participants will be asked for their Date of Birth twice at baseline. If an individual provides discrepant dates, s/he will not be invited to move forward with randomization but will be referred to the Drug Enforcement Administration's website (JustThinkTwice.gov).

Randomization and Study Conditions. Minimum likelihood allocation will be used to randomize participants, with stratification on race (Caucasian vs. Other), given the disproportionately high rates of prescription opioid misuse among Caucasian youth. Youth in the control group will be informed that they will be prompted when the next online survey is available to complete. Participants assigned to the web-based intervention will immediately (post-group assignment) be asked to start completing modules. Youth will be encouraged to complete 1-2 modules per login, 2x/week (about 30 mins. minimum per login). The core module content for the web-based intervention is as follows: What are Prescription Opioids? Misconception that Prescription Opioids are Safe and Non-addictive, Misconception that Using Prescription Opioids Without a Prescription is Not Illegal, Risks of Prescription Opioid Misuse, Non-medication Alternatives for Pain Management, Refusing Offers to Misuse Prescription Opioids, Refusing Requests From Others for a Prescription Opioid Prescribed to You, How to Know if You or Someone You Know May be Addicted.

Data to be collected. Assessment data will be obtained at: Baseline, post-intervention, and at 3- and 6-month post- intervention follow-ups (or comparable time-points for controls). If results are promising, longer follow-ups may be incorporated into future research projects that allow for longer evaluation periods than the Phase II grant mechanism. Participants in the experimental arm will be asked to complete the post-intervention assessment immediately after they complete all modules. Consistent with an intent-to-treat analysis, participants who do not complete all modules within a 4-week period will still be provided with the opportunity to complete a post- intervention assessment at 4-weeks post-group assignment and to complete follow-ups. The investigators plan to collect the following data in addition to the data captured by the primary and secondary outcome measures (described in that section): basic demographic information; other substance use information; feedback survey data (post-intervention time point only); assessment of usage data (e.g., number of times each module is accessed, duration of usage of modules, time to quiz mastery within modules, and extent of module completion). Participants will complete all assessments online (in a secure manner; see protection against risk).

Analysis. Although random assignment is designed to account for baseline differences, groups will be compared for differences on demographics and other characteristics using analysis of variance for continuous variables, Wilcoxon Rank Sum Tests for non-normal variables and chi-square tests for categorical measures. If a specific characteristic is predictive of outcome, it will be considered as a potential covariate in subsequent analyses using mixed-effects analyses of covariance. Primary analyses will include all participants randomized to a study condition independent of early dropout, consistent with an intent-to-treat approach to clinical trials. The investigators also will compare demographics and baseline characteristics of those who do and do not complete the study. Analyses will be performed using 'SAS' statistical software (type I error =.05).

Given the pervasive misconceptions that many youth have in their understanding of the risks associated with prescription opioid misuse, the investigators will primarily focus analyses on knowledge about prescription opioid misuse as well as measures that are highly predictive of future use (attitudes, intentions, skills and perceived risk). Actual rates of use will be a secondary focus, because of the planned evaluation window and because this is a primary prevention trial and is not primarily focused on recruiting and promoting behavior change among a select sample of youth who are misusing prescription opioids at baseline. The investigators will be able to assess if the web-based program results in lower rates of use up to 6 months post-intervention.

Outcome measures will be evaluated in separate analyses. Data from primary outcomes will be evaluated using mixed-effects analyses '(SAS, PROC MIXED),' which allows for incomplete data and covariance structures other than compound symmetry. In addition to the study condition indicator, these models will include, in the fixed effects portion of the model, a main effect for time as well as a two-way interaction between study condition and time. The key estimate from this model is the two-way interaction effect, which indicates whether changes in outcome (e.g., knowledge) over time are different for the two conditions.

Protection Against Risk. During the trial, both the web-based intervention and the online assessment tool will be accessible to participants by password protection by an encrypted Internet connection via 128-bit (256-bit if available) Secure Sockets Layer (SSL), the e-commerce industry standard for securing communications on the World Wide Web. The server that houses the web-based prevention program will be protected behind two firewalls and physically secured in a locked cabinet. Administrative access (e.g., by the Server Administrator) to the machines will only occur using secure shell (SSH), a secure, encrypted protocol for remotely connecting to a machine. Servers will be backed up nightly to external hard drives using an encrypted file system. Additionally, the offsite back-up process encrypts critical data (using a 'FIPS' 140-2 certified encryption algorithm) and uploads that data to a private, secure Amazon S3 bucket (a generic "cloud computing" storage service offered by Amazon).

All data provided by participants when using the interactive prevention program (e.g., when completing quizzes to test learning) or completing online assessments will be transmitted via encrypted 'SSL' to secure servers and will not be accessible to anyone not affiliated with the research project.

Consistent with the Health Insurance Portability and Accountability Act (HIPAA), the investigators will store personally identifiable data separately from all other data obtained from participants. To accomplish this, two applications will be created. The first application will store the (limited) personally identifiable data that is planned to be obtained from participants, including participant email address, participant username/password, participant parent name/parent contact info (for purposes of Informed Consent) and participant date of birth (as part of planned age verification checks). This "personal" application will also store a unique ID string consisting of 32 randomly generated characters, which would be passed back and forth between this application and the application storing all non-personally identifiable data ("anonymous data"). The "anonymous data" application will store the bulk of data collected from participants (data from online assessments as well as data of participants' usage of the web-based intervention) but will not have any knowledge of users' personal information and can only communicate with the personal data application through a strict set of Application Programming Interfaces (APIs). This approach will ensure that personally identifiable data and other data collected from participants are never stored together.

Emails sent to youth to prompt them to access the program or complete assessments will not contain any identifiable information or any references to substance abuse or other confidential or sensitive issues. Participants and participant parent/legal guardian will be explicitly informed about how the participant's sensitive data will be handled in the Informed Consent form. To further protect participant confidentiality, a Certificate of Confidentiality from the National Institute of Drug Abuse will be obtained.