Testing the POWER Project: A Neighborhood Intervention Trial

Testing the POWER project: A Neighborhood Intervention Trial PROTOCOL SUMMARY Sheana S. Bull, Ph.D. Principal Investigator Protocol # 02-978

Background The POWER project (Prevention Options for Women Equals Rights) was a social marketing pilot project Targeting African American and Latina women aged 15-25 to promote the "mixing" of female and male condoms conducted in Denver, Colorado in 2000 and 2001. The results of the project show promise for increasing awareness and use of female condoms and use of male condoms (Bull et al., 2002). Sixty-four percent of the women in the follow-up study were able to identify one or more unique aspects of the POWER campaign. There was a significant increase in awareness and use of female condoms and in use of male condoms from baseline to follow-up. Being exposed to the POWER campaign was significantly correlated with having heard of (r = 0.13) and seen (r = 0.20) the female condom, and with having talked about female condoms more frequently with friends, family and acquaintances (r = 0.15). Chi-square analyses using follow-up study data (n=269) showed that women exposed to the campaign had significantly greater positive outcome expectancies toward female condom use (p<0.10) and talked about female condoms more frequently with more people (p<0.01).

The CDC reports increasing prevalence of Chlamydia among women in Oakland/Alameda County, San Diego, Las Vegas and Los Angeles and increasing rates of gonorrhea in Los Angeles and Las Vegas (Centers for Disease Control and Prevention, 2000, 2002). National data on teen births indicate that while teen births have been declining, the number of births to women aged 20-24 are increasing as are births to unmarried women. With the advent of the female condom as an efficacious female controlled barrier method for STD/HIV and pregnancy prevention in the early 1990's (Bounds, 1997; Farr, Gabelnick, Sturgen, \& Dorflinger, 1994; Fontanet et al., 1998; McCabe, Golub S., \& Lee, 1997; Soper et al., 1993; Trussell, Sturgen, Strickler, \& Dominik, 1994), women at risk for these conditions were given a greater range of choices to control fertility and prevent disease.

Social marketing utilizes marketing and advertising concepts, integrated with social science theory for the development of behavior promotion campaigns targeted to specific subgroups. Such targeting is a key element of the social marketing approach. Social marketing approaches have been shown to be effective for changing blood cholesterol levels (Lefebvre, Lasater, Carleton, \& Peterson, 1987), smoking behaviors (Kotler \& Roberto, 1989), and women's and children's health (Hornick, 1991; Rasmuson, Seidel, \& Smith, 1988).

The intent of this project is to replicate the POWER pilot campaign using a randomized controlled trial with neighborhoods as the unit of randomization and analysis. We plan to rigorously test the effects of the POWER campaign on women's knowledge of, attitudes toward and use of female and male condoms for the prevention of sexually transmitted disease (STD) and unintended pregnancy. We will work in 12 non-contiguous neighborhoods in four southwestern U.S. cities (four each in Oakland/Alameda County and Los Angeles and two each in San Diego and Las Vegas). We will complete a baseline assessment in all 12 neighborhoods to document knowledge of, attitudes toward and use of female and male condoms. We will then randomly assign six neighborhoods to intervention status and six neighborhoods to control status. We will adapt the campaign for each city, and implement it in six intervention neighborhoods for nine months, and will complete a follow-up assessment similar to the baseline assessment, with added questions regarding exposure to the POWER campaign. Study partners include Dr. Sheana Bull and research staff from the Colorado Health Outcomes Program (COHO) of the University of Colorado Health Sciences Center, Educational Message Services (EMS), and Centers for Disease Control and Prevention (CDC).

Methods In order to achieve our objectives, we will utilize a randomized controlled trial, pre-post test design with the neighborhood as the unit of randomization and analysis. To avoid introduction of bias, COHO staff will maintain responsibility for all research and evaluation activities during pre and post test assessment. EMS will implement and maintain the POWER social marketing campaign in six intervention neighborhoods. CDC will serve as technical advisors to the project. COHO will first identify "high risk" neighborhoods in each study city that have higher than US average rates (as defined by the CDC, 2000) of Chlamydia and Gonorrhea combined with high concentrations of African American and Latina residents aged 15-25. We will then approach contacts in health departments and community organizations in each city and enlist their help in implementing a venue-based method for sampling hard-to-reach populations (more on this process described in the recruiting methods and selection of study population sections below). We will use this method to recruit participants for a cross-sectional, baseline study to document knowledge of, attitudes toward and use of female and male condoms. We will then randomly assign six neighborhoods to intervention status and six neighborhoods to control status. We will adapt the campaign for each city, and implement it in six intervention neighborhoods for nine months, and will complete a follow-up assessment similar to the baseline assessment, with added questions regarding exposure to the POWER campaign.

Recruiting methods Participants will be recruited from the selected study neighborhoods for baseline and follow-up studies using a venue-based application of time-space sampling (Muhib et al., 2001). The method involves first identifying venues, places at which members of the target population may be found, then identifying days and times during which members of the target population are likely to be found at the selected venues. After venue day-time (VDT) units are identified, forming a sampling frame, a random sample of VDTs is selected and eligible members of the target population are systematically sampled within selected VDTs.

Throughout the nine month baseline data collection period we will maintain a sampling frame of VDTs from which to recruit participants. Sampling frames will be updated monthly, with the addition of sporadically occurring VDTs and the removal of VDTs that are no longer viable (have failed to yield sufficient eligible participants or are infeasible due to practical constraints). Each month, after the sampling frame is updated, VDTs will be randomly selected and placed on the sampling event calendar. At the time of each sampling event, trained staff will systematically approach (e.g. every third woman appearing to be 15 years or older) potential participants who enter a predefined sampling area of the venue. Potential participants will be screened to determine eligibility. Members of the target population who meet all of the eligibility criteria and who agree to participate in the study will then complete informed consents prior to enrollment. They will self-administer the study questionnaire, and will be offered an incentive valued at $10 for their time.

This process for venue-based sampling will be repeated for the follow-up data collection process. We will revisit the sampling frame prior to beginning data collection and will remove no longer viable VDTs while adding new VDTs. Subsequent follow-up sampling activities will mirror those conducted at baseline.

Consent procedures Training of staff to explain study process. Participants will self-administer the questionnaire. However, recruitment of participants, explanation of consent process and study protocols, and distribution of incentives will require trained staff. We will have a core recruitment director hired by COHO trained in study protocols and COMIRB 101. Any additional persons hired to assist with data collection will be required to attend an IRB training in their area, and take the IRB 201 test online.

Setting in which consent will be obtained. Consent will be obtained in multiple community settings, e.g. bars, community organizations, community businesses (such as beauty salons, clothing stores, grocery stores), recreation centers. The settings will be selected at random using the procedure described above. Potential participants will be approached and eligibility will be determined by asking potential participants a few short eligibility questions.

Assessment of subject comprehension and autonomy. Staff will ask the participant to explain the study in their own words prior to handing them a questionnaire to complete.

Consent form copies will be made available to the participant. Staff will ask participants to sign two copies of the informed consent. They will offer the participant one copy and will keep one copy. Documentation of the consent process in a medical source document is not applicable for this study, as the study will take place in community rather than medical settings.

Intervention The POWER campaign will be implemented in six neighborhoods chosen at random (see methods section above); two neighborhoods in Oakland/Alameda County, two in Los Angeles and one each in San Diego and Las Vegas. The campaign will last nine months, and will be implemented by EMS. We will have up to 100 display points in 25 sites in each neighborhood. Some settings will be larger than others, and will require many more display points (e.g., 10) while others may be very small and will require fewer display points (e.g., two). Six control neighborhoods (those not chosen at random to be intervention neighborhoods) will be involved in the study. No POWER campaign materials will be placed in these neighborhoods. All study participants at baseline will be asked to complete an assessment to document knowledge of, attitudes toward and use of female and male condoms. All study participants at follow-up will complete an assessment similar to the baseline assessment, with added questions regarding exposure to the POWER campaign. None of these assessments are part of regular clinical care for participants.

Inclusion and exclusion criteria Persons who meet eligibility criteria include women aged 15-25 who speak English. Children recruited for the study include those women aged 15-17. Although women who are pregnant will not be excluded from the research, they are not targeted for recruitment.

We are requesting a waiver of parental permission from COMIRB to include women under age 18 in the study. Per California and Nevada Statutes, minors are allowed to consent for contraceptive services, STD testing and pregnancy testing without parental consent. Detailed justification for this request is included in the full protocol. Minors (15-17 years old-determined at the time of recruitment) will be encouraged to discuss their participation with a parent or guardian prior to enrollment in the study.

Patient accrual We will enroll 300 women in each of the 12 study neighborhoods at the baseline study and 300 women in each of the 12 study neighborhoods at the follow-up study, for a total of 3600 at baseline, 3600 at follow-up and a total of 7200 for the entire study. This number includes the total number of women who will need to sign a consent form in order to reach study end points.

This study will employ the neighborhood as the unit of analysis. Thus, six neighborhoods will be assigned to intervention status. Baseline data will be collected prior to campaign implementation. The POWER campaign will be implemented in these neighborhoods for a nine month period. Follow-up data collection will take place in all 12 study neighborhoods. Thus, individuals per se are not assigned to intervention or control status, rather the neighborhoods where people live will be assigned to intervention or control status.

Estimated duration of the study The baseline data collection will take place over nine months, starting in month six of the project. The campaign will last the subsequent nine months, and the follow-up data collection will take place in the next subsequent nine months, for a total of 27 months.

Persons who enroll in the baseline and/or follow-up studies will spend a total of 20 minutes in completing the consent process, self-administering the questionnaire, and in receiving their incentive.

Study instruments We will ask all participants at baseline to complete a questionnaire to document knowledge of, attitudes toward and use of female and male condoms. Participants in the follow-up study will be asked to complete a questionnaire identical to the baseline assessment, with added questions regarding exposure to the POWER campaign.

Data Safety and Monitoring Plan Analyses for this study will be led by Dr. Sheana Bull, with input from CDC Colleagues Drs. Sam Posner, Lillian Lin and Sherri Varnell. Three analysis strategies will be employed: (a) comparison of neighborhood-level (unadjusted) means, (b) comparison of neighborhood-level adjusted means, and (c) hierarchical modeling.

We anticipate no adverse events associated with participation in this study beyond possible embarrassment and reluctance to answer sensitive questions. That said, any serious and unexpected adverse event will be reported to the IRB immediately. Dr. Sheana Bull will oversee this aspect of the study. The informed consent indicates that should medical care be necessary for any person responding to the questionnaire, we can facilitate access to care by providing referrals. It also indicates that we will not be able to pay for that care. The consent also offers two telephone numbers for persons who wish to obtain more information on the study, referrals or information on their rights as a research subject.

Plans for interim analyses. We regularly monitor recruitment and risk behaviors disclosed by participants on the assessment. Because we will need to continue enrollment up to 7200 persons to determine if the intervention is effective, we do not plan to stop the study prior to achieving this enrollment.

DSMB. Because this is a behavioral study, and offers low risk outside of some embarrassment over answering personal questions, we do not plan to convene a data safety and monitoring board independently of monitoring this work.

Study conduct is monitored each day by Dr. Sheana Bull, who regularly tracks enrollment and recruitment processes. She is responsible for oversight of all protocol implementation for the project. Specific monitoring duties include regular contact with persons recruiting participants to ensure that protocols for recruitment are adhered to.

Procedures for protecting the privacy of subjects and maintaining confidentiality of data. All participants in the study are asked to sign a consent form and a receipt acknowledging they received an incentive for the study. These two forms are the only documents that have any identifiers on them. We will not scan these materials or keep electronic copies of them. We will keep signed consent forms separate from completed questionnaires. Persons collecting data will place signed consent forms and receipts in a separate locked file from completed questionnaires. All questionnaire data will be entered into a computer on the COHO server. This is a COHO dedicated server, with firewall, login and encryption security, none of which are accessible from outside the COHO local area network.

Statistical analyses planned for data include development of a database using ACCESS (Microsoft Corporation, 2002) to store and manage study data, and to import data into SAS (version 8.1) for quantitative analyses. Analyses of data from the baseline and follow-up studies will concentrate on establishing knowledge of, attitudes toward and use of female and male condoms. Analyses of the follow-up study data will also include documentation of exposure to project material, including open-ended questions on the ability to name a relevant feature or message related to the material.

We have planned a nested cross-sectional design with baseline and follow-up observations, implementing a randomized controlled community-level trial with neighborhood as the unit of randomization and analysis. Neighborhoods will be block-randomized within cities to achieve balance between intervention and control conditions on city-level factors. Assuming an intraclass correlation of 0.02 with a sample size of 300 per neighborhood this study will have 80% power to detect a difference in proportions between intervention and control communities of 20% (45% vs. 36%).

Risks-Subjects There are few anticipated risks related to this research. We believe that participants may experience some embarrassment and or psychological discomfort related to the subject content. There are no anticipated physical or legal risks faced by participants.

Investigators/Institutions There are no anticipated risks to investigators and institutions related to this research.

Benefits It is anticipated that any persons enrolled in the study who live in intervention neighborhoods will be exposed to project materials that will contain useful information about how to prevent pregnancy and STD infection. We will make campaign materials available in control neighborhoods after collecting follow-up data, so this information will be available for all study participants. If shown effective, this campaign will be an extremely cost effective way to deliver important information and one that can reach a much larger audience than that typically seen in clinic settings. If the program is shown not to be effective, we will have useful data on the limitations of the POWER campaign for promotion of increased condom use.

Funding All funding for this study is provided by the Association of Teachers of Preventive Medicine through the Centers for Disease Control and Prevention.

Participants in the baseline study will receive $10 for completing a questionnaire. Participants in the follow-up study will receive $10 for completing a questionnaire. These monies will be prorated, and women who partially complete a questionnaire will receive $5.

Special consent issues We seek a waiver of parental consent for those women aged 15-17 participating in the study. Justification for said waiver is included in the section on inclusion and exclusion criteria above.