Appendix: Implementation and Adaptation of the Common Study Protocol
The following case studies describe the process of implementing and adapting the common study protocol used in the AIDS Community Demonstration Projects (ACDPs) for a different population in each city. These case studies summarize project activities that took place from July 1991 through August 1993 and present the following information for each site: a) the defined population or community at which the intervention was directed; b) specific data-collection procedures at each project site; c) demographic and other characteristics of the study community at baseline; d) a description of intervention activities in that community; and e) exposure of community members to the intervention as of the early implementation phase.
Case Study #1: Dallas
Four Dallas County census tracts with high rates of syphilis and gonorrhea served as the intervention and comparison areas for the project. The census tracts were grouped into two pairs based on their physical proximity; within each pair, the census tracts were randomly assigned to intervention or comparison status. The selected communities primarily comprised persons residing in low socioeconomic neighborhoods, with both federally subsidized and public housing.
Based on ethnographic data, three sites in each census tract were selected as data-collection points. Interviewers visited data-collection sites according to a randomly determined schedule; at the sites, interviewers followed a prespecified route and used a list of randomly ordered numbers to determine which persons to approach for interview. Persons who completed either the initial or the full street interview received a coupon for food or beverage from a business within the community.
Demographic data from the baseline phase indicated that respondents in the intervention and comparison areas were similar. In both areas, most respondents were men (78% in the intervention areas and 79% in comparison areas) and most were black (95% in intervention areas and 83% in comparison areas). The average age of respondents was 34 years in the comparison areas and 36 years in the intervention areas.
Mean assessments for behavioral stage of change (SOC) at baseline indicated that the mean response in both intervention and comparison areas was between the precontemplation and contemplation stages for consistent condom use for vaginal intercourse with a main partner but between the contemplation and ready-for-action stages for consistent condom use for vaginal intercourse with nonmain partners. The mean SOC for consistent condom use for anal intercourse with nonmain partners also was between the contemplation and ready-for-action stages in both areas (though slightly higher in the comparison area). For consistent use of bleach to clean injection equipment among injecting drug users (IDUs), the mean SOC was above the contemplation stage in both areas.
The Dallas project, known locally as the "AIDS Prevention Project," employed four full-time workers and one half-time worker who recruited, trained, and maintained networks of community members in both intervention areas. The average size of the adult population (greater than or equal to 18 years of age) in the intervention areas was approximately 5,300 persons. Intervention materials were distributed to any person greater than or equal to 16 years of age encountered in the streets, at storefronts, or at interactor sites. The intervention strategy involved the diffusion of prevention materials into the community through the project's networks and through the social networks within these communities. During the start-up phase, an average of 25 network members per month distributed materials. This number increased to 46 in the early implementation phase. Each network member made an average of 53 contacts per month (range 1-80) during which they distributed the intervention materials and encouraged the use of condoms (and bleach among IDUs) to protect against HIV infection. Several local businesses also distributed the intervention materials.
Within each of the communities receiving the intervention, network members and project outreach workers managed a project storefront. Local residents visited the storefronts to learn about HIV prevention; obtain print materials, condoms, and bleach kits; * and receive other services (e.g., blood-pressure screening and education on other health issues).
Training of new network members and updates for long-term network members were held monthly at the storefronts. Some individual training was conducted to accommodate the network members' schedules. Tangible incentives (e.g., baseball caps and tote bags with the project logo, coupons for local businesses, and certificates of participation) were given periodically to network members.
The print materials consisted of quarterly and later monthly newsletters known as the Neighborhood Voice and West Dallas Today, which contained role-model stories, coupons, and relevant community information. The newsletters were distributed along with condoms and bleach kits by the project networks. Between the start-up and early implementation phases, the average number of materials distributed per month increased from 922 to 2,454.
Qualitative interviews, observation, and exposure data indicated that the intervention was both visible by and popular with community residents. For example, events sponsored by the project (e.g., a commemoration of World AIDS Day and the broadcast of a popular local minority talk show) were attended by up to 100 persons. By the early implementation phase, 31% of respondents in the two intervention areas reported exposure to the intervention.
Case Study #2: Denver
In Denver, one intervention was targeted to IDUs residing in two predominantly black-populated areas on the east side of Denver and one predominantly Hispanic-populated area on the west side. One of the areas is the oldest community of blacks in Denver and included two taverns that are alleged distribution points for heroin and crack cocaine. The other two locations included large public-housing areas.
Because migration among IDUs in Denver precluded the use of a comparison area within the city, a comparison area was identified in Dallas. However, changes in this area during the start-up phase led to a sharp decline in the number of IDUs available for interviews; therefore, the comparison area for the Long Beach intervention targeting IDUs also was used for the comparison area for Denver.
Interviewing teams visited nine data-collection locations following a randomly determined schedule. The interviewers followed prespecified routes and used a list of randomly ordered numbers to determine which persons to approach for interview. Persons completing a full questionnaire were given the equivalent of $10 in nonmonetary incentives. To be eligible for the interview, a person must have injected drugs within the past 30 days and either had vaginal or anal intercourse during the past 30 days or shared injection equipment within the past 60 days.
At baseline, the racial/ethnic composition of Denver respondents was 55% white, 29% black, and 14% Hispanic. For the Long Beach comparison sample, this composition was different: 48% of the respondents were black, 26% Hispanic, and 23% white. In both areas, most respondents were male (89% in Denver; 79% in Long Beach); the mean age of respondents was 35 years for Denver and 39 years in Long Beach.
The mean SOC for consistent condom use for vaginal intercourse with a main partner was between the precontemplation and contemplation stages in both Denver and Long Beach. The mean SOC for consistent condom use for vaginal intercourse with nonmain partners was slightly below the contemplation stage in Denver and slightly above the contemplation stage in Long Beach. In both cities, the mean SOC for consistent use of bleach to clean injection equipment was between the contemplation and ready-for-action stages. The mean SOC for consistent condom use for anal intercourse with nonmain partners was above the contemplation stage in Denver and between the precontemplation and contemplation stages in Long Beach.
The Denver intervention for IDUs was known locally as "Project REACH." Two community networks were formed to distribute intervention materials: a peer network consisting of past or current IDUs and an interactor network comprising persons from local businesses or social service agencies. Three full-time Project REACH staff recruited network members and supplied them with intervention materials. They also regularly monitored project areas and visited taverns, community centers, and other places frequented by IDUs. During the start-up phase of the project, the recruitment of peer-network members was more successful on the west side. To increase the project's visibility on the east side, a storefront was opened at the center of one of the intervention areas in April 1993 (early implementation phase). The location of this facility enhanced the project's ability to recruit peer-network members in this area and provided a convenient place for peer-network training and focus-group meetings.
Peer-network training was held monthly for persons who were recruited to join Project REACH. Project staff contacted peer-network members either weekly or biweekly and paid them $5 for each week during which they distributed materials.
The project newsletter, Reaching Out, was produced biweekly and contained role-model stories and relevant community information. The newsletter was distributed along with condoms and bleach kits. The average number of peer-network members distributing materials per month increased from 12 during the start-up phase to 30 during the early implementation phase. However, to enable project staff to focus on recruiting peer-network members, the interactor network was downsized during this period from 26 to 17. Together, the peer and interactor networks distributed an average of 969 print materials monthly during the start-up phase; this increased to an average of 1,216 materials distributed monthly in the early implementation phase.
By the early implementation phase, 22% of IDUs interviewed in the intervention area reported exposure to the intervention. As the project storefront gained visibility in the community, the peer network grew more rapidly. During the late implementation phase, the peer network included as many as 200 persons who distributed an average of greater than 5,000 materials per month. Preliminary data from the late implementation phase indicate that exposure to the intervention continued to increase among respondents.
Case Study #3: Long Beach
The Long Beach project was targeted to three populations: women who trade sex for money or drugs (female sex traders), IDUs, and female sex partners of male IDUs. This case study focuses primarily on intervention activities for female sex traders.
The intervention was purposefully assigned to an area comprising seven contiguous census tracts covering 2.3 square miles in the central part of the city. This area was chosen because of its high prevalence of drug abuse and prostitution. Comparison data were collected at sites located in other areas of Long Beach and in two nearby communities.
Data-collection sites in the intervention and comparison areas were matched by using data collected during the formative research; interviews were conducted in 126 sites. Interviewers visited data-collection sites following a randomly constructed schedule. They remained at a given location for a short time, collecting a maximum of six interviews during a visit. Women who completed only the initial screening part of the interview received two food coupons (worth $1 each); those who completed the full street interview received $5 in cash. A woman was considered to be a sex trader if she reported a) having exchanged sex for money or drugs in the past 30 days; b) having had at least one nonmain male sex partner in the past 30 days; c) being greater than or equal to 18 years old; and d) having had vaginal intercourse in the past 30 days.
Baseline data indicated that the average age of female sex traders was 31 years among respondents in the intervention area and 33 years among those in the comparison area. The racial/ethnic distribution of the respondents in the two areas also was similar. In both areas, more than 60% of respondents were black; however, whites comprised a higher proportion of respondents in the intervention area (22%) compared with the comparison area (11%). In addition, 8% of intervention-area respondents were Hispanic, compared with 16% of respondents in the comparison areas.
Among female sex trader respondents in both intervention and comparison areas at baseline, the mean SOC for consistent condom use for vaginal intercourse with a main partner was between the precontemplation and contemplation stages, whereas the mean SOC for consistent condom use for vaginal intercourse with nonmain partners was between the ready-for-action and action stages. Among female sex traders who reported injecting drugs, the mean SOC for consistent use of bleach to clean injection equipment also was higher in the intervention area (above the ready-for-action stage) than in the comparison area (between contemplation and ready-for-action stages).
In Long Beach, the prevention activities were known collectively as "Road Dogs," a term that referred to persons, generally drug users, who associated with each other on the streets. Two different intervention packets were developed for distribution to the three target populations. One packet, designed for high-risk women encountered, contained a For Women Only flyer, five condoms, and instructions for correct condom use. The other packet, directed to both male and female IDUs (many of whom were also sex traders), contained a Road Dogs flyer, a 1-oz. bottle of household bleach (labeled with instructions for its use in disinfecting injection equipment), condoms (and instructions), and three alcohol wipes. Because approximately 37% of the female sex traders reported having injected drugs in the past 6 months, this population was given both packets.
In accordance with the study protocol, each flyer contained factual stories of members of the target population who had successfully overcome personal barriers to lowering their risk for HIV infection. Feedback from the network members and target population members indicated that the stories were relevant to the situations and circumstances in which persons often found themselves. Anecdotal data suggested that the women were interested in the stories, particularly those featuring a person they knew or that portrayed provocative situations. Flyers also contained local health-services referral information, word games, and cartoons.
Materials were distributed by target-population peers and through local businesses (e.g., motels and liquor stores) that served members of the target population. Peer-network members directed the target population's attention to the role-model stories and reinforced acceptance of and interest in the materials. Many of the peer-network members were sex traders themselves. Other network members interacted with members of the target populations through familial, social, or business relationships. Network members received a monthly nonmonetary incentive (e.g., a candy-filled mug, a T-shirt with the project logo, or food coupons).
Records were kept on the number of network members and the number of materials distributed for all three populations. In the start-up phase, an average of 21 peer network members and 10 businesses distributed approximately 2,108 materials per month to the three target populations, including female sex traders. By the early implementation phase, an average of 85 persons and 12 businesses distributed an average of 6,357 flyers per month to the three populations. The number of materials distributed by the peer network members and the participating businesses exceeded the number of materials that the outreach staff could have distributed on their own.
By the early implementation phase, 68% of female sex traders interviewed in the intervention area reported exposure to the intervention. These data suggest that even before the late implementation phase, exposure to the intervention was high among this population.
Case Study #4: New York City
In New York City, female sex partners of male IDUs were targeted for the project intervention. The intervention area was a public-housing development in the Lower East Side of Manhattan. Approximately half of the 4,798 residents were adult or adolescent women. A second public-housing development located within six blocks served as the comparison area. This area was chosen for the intervention because during the formative research, respondents indicated that one of the public-housing areas was more crime-ridden than the other and abutted a major drug-dealing section of the city.
Interviewers visited data-collection locations in the intervention and comparison areas following a computer-generated schedule that was based on the buildings in each housing development, day of the week, and time of day. One of a team of female interviewers who spoke both Spanish and English approached every fifth woman the team encountered in the lobbies, courtyards, or other public spaces of the designated building. Participants received $3 for completing the initial screening interview and $20 for completing the full street interview.
A woman was eligible for interview as a female sex partner of an IDU if she reported a) having had a main male sex partner whom she knew or suspected to have injected drugs during the past 5 years; b) having had an active sexual relationship with this partner within the past 30 days; and c) not having injected drugs or traded sex for money, drugs, or other rewards within the past 30 days.
Among female sex partners of IDUs who were interviewed at baseline in the intervention areas, the average age was 32 years; by race/ethnicity, 78% were Hispanic and 19% were black. Similar demographic information was found for respondents in the comparison areas (average age 33 years, 79% Hispanic, 18% black). The mean SOC for consistent condom use for vaginal intercourse with a main partner was similar in both areas at slightly below the contemplation stage. In both the intervention and comparison areas, few women reported having had vaginal or anal intercourse with nonmain partners.
Women who lived in the housing project were recruited to distribute newsletters (businesses were not recruited). This peer network called themselves "companeras," which translates to mean friend or comrade and signifies solidarity. Each weekday, companeras disseminated the Companera Newsletter in 2-3 shifts; each companera distributed approximately 20 newsletters per shift. The companeras were asked to distribute newsletters during at least one shift each month but could distribute materials as often as one shift per day. The monthly newsletter contained two culturally specific role-model stories of women from the Lower East Side. These stories depicted the factors that inhibited or facilitated consistent condom use with a male partner. Condoms were attached to the newsletters.
Companeras were recruited by veteran companeras while they were distributing brochures, through street recruitment by veteran companeras or project staff, and by referrals from staff of different neighborhood agencies who had clients who lived in the intervention housing project. Women who were recruited attended a storefront orientation party to learn about the program. After completing the training, each woman received a $20 honorarium, a certificate of completion, the offer of ongoing information about job opportunities, and a letter of reference if needed in the future. Companeras were also asked to sign up for 4 hours of outreach and print-material distribution per month; after each of these sessions, they were interviewed about the outcome and paid $5.
Several incentives helped to sustain the companera's ongoing participation in the outreach and distribution activities. Incentives were intended to provide both material and nonmaterial rewards and to minimize practical barriers to companera activities. Incentives included a) weekly support-group meetings; b) weekly debriefing interviews; c) special events such as barbecues and holiday parties for companeras and their families; d) weekly telephone reminders to companeras; e) provision of child care during storefront meetings and interviews; and f) refreshments and donated gifts (for companeras and their families) at holidays.
During the start-up period, an average of 13 companeras distributed approximately 704 newsletters monthly. In the early implementation phase, an average of 19 companeras distributed approximately 1,922 newsletters monthly.
A focus group was conducted with 12 experienced companeras to determine reasons given by women for refusal of the companeras' outreach. Although all of the companeras reported low refusal rates, cold weather and time constraints were the most commonly cited reasons for refusal to accept the newsletter. Other barriers to outreach included discomfort about the offer of condoms, claims of already having received HIV education in the neighborhood, and perceptions of being at low risk for HIV infection because of being in a monogamous relationship. By the early implementation phase, 46% of women interviewed in the intervention areas reported exposure to the invervention.
Case Study #5: Seattle
In Seattle, interventions were targeted to men who have sex with men (MSM) but do not self-identify as homosexual (nonhomosexually identifying MSM). The intervention areas included several indoor and outdoor locations in Seattle where men were known to seek potential sex partners. These included the "back room" video arcades of adult bookstores and erotica shops and areas in public parks.
Information collected during the formative research indicated that the mobility of this population precluded having separate intervention and comparison areas within Seattle. Beginning in August 1991, comparison data were gathered in Long Beach, primarily in public parks and other outdoor soliciting areas.
To collect a sufficient number of responses from the comparison area, the baseline data-collection period was extended through September 1991 for the intervention and comparison areas. Otherwise, the start-up and early implementation phases followed the same schedule as in the other interventions.
Men were interviewed at 17 locations in three geographic areas of Seattle and in 17 locations in Long Beach. In both cities, these locations included public parks and outdoor cruising areas; in Seattle, interviews were also conducted in adult bookstores and video arcades.
In Seattle, interviewers made an approximately equal number of visits to each of these locations during a data-collection period. Interviewers used a list of random numbers to select men to approach; when few men were at a location, each man was approached. Men not looking for sex (e.g., police, employees of bookstores and arcades, men with their families) were not interviewed. Those completing the survey received $10.
Interviewers in Long Beach visited each location several times during a data-collection period; the most frequently visited locations were those with more activity among nonhomosexually identifying MSM. Because of the small number of potential respondents, interviewers approached all men who appeared to be greater than or equal to 18 years of age. Participants received $5 for completing the initial screening portion of the interview and another $5 for completing the full street interview.
For outcome analyses, a man was included if he reported all three of the following criteria: a) oral or anal intercourse with a man in the past year; b) anal intercourse with a man or vaginal/anal intercourse with a woman in the past 30 days; and c) self-identification as heterosexual or bisexual.
Baseline data indicated that the average age of men interviewed in Seattle was 36 years and, in Long Beach, 32 years. In Seattle, 82% of respondents were white compared with 46% in Long Beach, where there was a higher proportion of black (18%) and Hispanic (27%) respondents. For consistent condom use for anal intercourse with nonmain partners, the mean SOC for respondents in the Long Beach comparison area was above the action stage, whereas the mean was between the ready-for-action and the action stages for respondents in Seattle. In Seattle, the mean SOC for consistent condom use for vaginal intercourse with a main partner was between the precontemplation and contemplation stages, whereas in Long Beach, the mean SOC was between the contemplation and ready-for-action stages. The mean SOC for consistent condom use for vaginal intercourse with nonmain partners was between the contemplation and ready-for-action stages in both the intervention and comparison areas.
The Seattle intervention project was known in the community as the "Shiftin' Gears Project." One outreach worker recruited a network of men (composed primarily of homosexual men) and supplied them with intervention materials. Both these men and the interactors (local merchants, social service providers) were recruited to distribute condoms and brochures known as Shiftin' Gears, which contained targeted role-model stories.
The network members distributed intervention materials in locations where non-homosexually identifying MSM were known to frequent. They engaged other men in brief outreach interactions, directing the men to the particular topic or cognitive element highlighted in each role-model story and praising and reinforcing reported behavior changes. Actual stories gathered through interviews with local men were abstracted for the brochures. Both the men who distributed materials and the interactors received an incentive of $20 per month as well as occasional thank-you cards.
During the start-up phase, an average of three men and 13 interactors distributed an average of 1,004 brochures per month. During the early implementation phase, approximately four men and 12 interactors distributed an average of 888 brochures each month. Maintaining interactors was difficult because managers at some of the distribution sites complained about increased litter (e.g., used condoms and discarded project brochures) and decreased condom sales. Maintaining participation of peer network members also complicated the distribution process because many men reported feeling uncomfortable distributing materials in sexually charged environments. Therefore, there were fewer men participating and lower distribution levels than had been anticipated. By the early implementation phase, 22% of men interviewed in the intervention area reported having been exposed to the intervention.
Disclaimer All MMWR HTML versions of articles are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the electronic PDF version and/or the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to firstname.lastname@example.org.
Page converted: 09/19/98
This page last reviewed 5/2/01