Method - Center for Development of Human Services

HIV/AIDS Intervention Literature Review
Method September 2002 through January 2003, A literature search was conducted in
electronic databases ;Elisivier Direct, Psychinfo, PA Research II
(ProQuest), PsyARTICLES, Journals at Ovid full text, JSTOR, AIDSLINE, and
MEDLINE. Search terms used included HIV/AIDS, prevention, intervention,
teens and adolescents. This search resulted in 567 citations. Abstracts
of these articles were reviewed for inclusion using the following criteria: . Published in a peer reviewed journal
. Specific to adolescents between the ages of 11-21, two exceptions
were a study that evaluated the effectiveness of a mass media
campaign that reported results from a sample where 64% of the sample
was less than 18 years old and another mass media study sample
included respondents as old as 24..
. Evaluated an HIV prevention program
The review process resulted in 53 articles. Bibliographies of these
articles were reviewed to identify articles appropriate for inclusion that
were missed in the electronic search, and resulted in the addition of 14
articles.
The final list of comprehensively reviewed studies includes 67 articles.
These articles were read and information provided in each study was
organized in a matrix (see table 1) format to allow coding and analysis.
The first category identifies and describes the intervention by the
article citation, intervention name or category, experimental design, main
outcome goals and theory. The second category delineates the intervention
method, intervention content, setting and intervention level. Intervention
duration, in terms of number of sessions and total contact hours and
duration are categorized together as are facilitator, target sample and
sample characteristics including sample size, gender, age, age range and
ethnicity of the sample. The matrix concludes with a summary of the
comparison condition, outcome measures and findings. If available, contact
information for the study is located in the last column.
Information compiled in the matrix was coded for analysis and a summary of
the results are reported here. Coding Summary
Intervention ( See Table 2)
Condom promotion: Of the 67 articles reviewed 4 (6%) reported on an
intervention where the primary focus was the promotion of condom use,
(Alstead, Campsmith, Halley, Hartfield, Goldblum & Wood , 1999; Kirby,
Harvey, Claussenius & Novar, 1989; Sellers, McGraw & McKinlay , 1994).
Condom promotion programs typically consisted of mass media campaigns or
direct mailings that provided information regarding condom use,
availability and STD prevention. One intervention was specifically
intended to increase condom use self efficacy and communication skills
between partners regarding condom use (Gillmore, Morrison, Bassalone,
Gutierrez & Farris, 1997).
HIV/AIDS education: 48 (72%) of the articles used HIV/AIDS education
as an intervention. Some educational approaches were standardized
interventions such as Becoming a Responsible Teen (BART) (Butts & Hartman,
2002,) or Safer Choices, (Basen-Enquist, Coyle, Parcel, Kirby, Banspach &
Carvajal, 2001) (See table 2). Many approaches did not use specific
curricula e.g., Dunn, Rose, Caines & Howorth (1998), Farley, Pomputius,
Sabella, Helgerson & Hadler (1991), and Jainchill, Yagelka, Hawke & DeLeon
(1999).
Peer education: Five articles (7.5%) utilized peer leaders in the
interventions, (Pearlman, Camberg, Wallace, Symons & Finison, 2002)
evaluated the effect of being a peer leader on those students over a period
of time.( Dunn et al., 1998) compared the effectiveness of peer
leaders/educators to community nurses. Other interventions utilized only
peers and facilitator characteristics were not an independent variable
(Schlapmann & Cass, 2000; Sellers et al., 1999; and Shelton, 2001,.
Individualized intervention: Five articles 7.5% evaluated the
effectivness of individualized assessment and intervention. Two of these,
Boekeloo, Schamus, Simmens, Cheng, O'Connor & DeAngelo, (1999) and
Mansfield, Conroy, Emans & Woods (1993), evaluated the effect of
personalized instruction by MD's.
Video tape: Five articles (7.5%) utilized videotaped interventions
with supportive workbooks or opportunities for question and answer from
intervention participants.
Culturally specific: 14 studies (20%) evaluated interventions
designed to appeal to an adolescent audience with a specific ethnic
cultural identity. Two of these studies utilized the Be Responsible, Be
Proud intervention curriculum, (Jemmott, J. Jemmott, L. & Fong,1998 and
Kennedy, Mizuno, Hoffman, Baume & Strand, 2000)
Others: One study, Blake, Ledsky, Goodenow, Sawyer & Hack (2001)
evaluated effects on gay and lesbian students and two considered the
effectiveness of mass media campaigns (Mizuno et al., 2002; Sellers et
al., (1994))
Experimental Design:
Thirty-four (51%) of the studies utilized a randomized controlled
trial to evaluate intervention effectiveness (see table 3). Six studies
(9%) observed results over 12 months or more. Two studies, St. Lawrence,
Crosby, Brasfield, & O'Bannon (2002), and Shrier, Anchetta, Goodman, Chiou,
Lyden & Emmans (2001), observed outcomes at six and twelve months post
intervention. Smith, Dane, Archer, Deveraux & Katner, (2000) observed
outcomes at four and eight months and O'Donnell et al.( 2002), reported
outcomes 24 months post intervention.
Repeated measures: Fifteen investigations (22%) implemented a
repeated measures design. Of these, six (9%) included observations of a
control group for comparison. One study, Damond, Breuer & Pharr (1993)
used randomized sampling but had no control group.
Randomized quasi experimental design was implemented in four
investigations. Main et al., (1994) collected outcome data at six months,
and Kirby, Barth, Leland & Fetro (1991) measured outcomes at six and
twelve months.
Three articles (4.5 %) conducted a pre and post intervention cross
sectional randomized survey to evaluate effectiveness of the intervention.
Outcome Goals,
Outcome goals were wide ranging. Nine primary categories of outcomes were
evaluated in the investigations. These included HIV/STD knowledge,
behavioral skills, attitudes and beliefs, self esteem, perceptions,
behavior, and intentions (see table 4). Three studies measured improvement
in service utilization. In many studies the primary outcomes were divided
the into sub outcomes, e.g., behavioral outcomes include onset of sexual
behavior, frequency of unprotected sex, number of partners with whom the
respondent engaged in unprotected sex and use of condoms in last instance
of intercourse. Other studies measured the impact of facilitator
characteristics, matching facilitator race and gender to the race and
gender of the sample or the use of peer educators.
26 studies, (40%) measured changes in participant's HIV/STD knowledge
including transmission, prevention, and personal risk. 17 interventions,
(25%) measured skill development or utilization including communication
skills, decision making skills, condom use skills, and problem solving
skills. 26 studies measured change in participant behavior without
detailing the behaviors. Eleven studies (16.4%) compared the age at onset
of sexual activity, seven (10.4%) measured differences in the frequency of
unprotected sex and 12 (18%) investigations measured changes in reported
condom use. 14 studies (20.8%) measured change in participant perceptions
including perception of personal risk, self efficacy and condom use. 21
studies or 31.3% considered changes in adolescent attitudes. Among the
attitudes considered, six studies (9%), measured attitudes toward
practicing preventive behaviors. Six investigations (9%) measured changes
in adolescent beliefs. Twelve studies also included exploratory outcomes
(see table 5) e.g., Fang, Stanton, Fiegelman & Baldwin (1998) considered
differences in outcome by gender and Rotherham-Borus, Gwadz, Fernandez &
Srinivasan (1998) who compared the outcome of identical interventions of
equal number of contact hours in a format of longer or shorter duration
(intensity).
Theory:
Twenty one theories informed 72% of the investigations (see table 6)
Twelve articles, (18%) did not mention a theory to support the choice of
intervention. The most frequently cited theories were the health belief
model, (Rosenstock, Strecher & Becker 1994) 19%, social influence 15%
(Fisher, 1988; and McGuire & Papagoreoris,1961) social cognitive/social
learning model, (Bandura, 1986) 20% . A specific social learning theory the
information motivation and behavior (IMB) model ( Fisher & Fisher ,1992)
accounted for 8% of the 20% that were based on social learning theory. 12
studies, 18%, utilized psycho-educational or preventive education models.
Three studies, utilized stages of change model (Prochaska & Diclemente,
1982; Prochaska,J.; DiClemente,R.J.; Norcross,J. 1992 and Prochaska,J.;
Redding,C.; Harlow,L.; Rossi,J.; Velicer,W. 1994) to inform motivational
aspects of the intervention. Method:
35 investigations (52%) utilized didactic instruction in the
intervention (See table 7). Of these four studies, or 6% of the total,
tailored the instruction to reflect the cultural identity of the
participants. 18 (26%) of the interventions incorporated opportunities for
discussion. Videos presenting HIV prevention information were used in 21
(31.3%) of the interventions. Six (9%) of the video presentations were
reflective of the cultural identity of the subjects. Interventions informed
by social cognitive learning theory typically used games, goal setting and
problem solving activities. 22 (33%) of these studies incorporated group
exercises, 26 (39%) role play and 31 (46%) utilized skill building
activities. Six studies (9%) included parent education ac