This study assessed the effectiveness of culturally sensitive interventions (CSIs) (N = 10) designed to address substance use among minority youths. Study methods consisted of systematic search procedures, quality of study ratings, and recta-analytic techniques to gauge effects and evaluate publication bias. The results, across all measures and time frames, reveal small effects (Hedges's g = .118, 95% confidence interval [CI] = 0.004 to 0.232). For recent alcohol use, the effects were small (Hedges's g = .225, 95% CI = 0.015 to 0.435). For recent marijuana use, the effects failed to achieve significance (Hedges's g = .610, 95% CI = -0.256 to 1.476). Suggestions for enhancing the effectiveness of CSIs are discussed along with the benefits and limitations of using meta-analyses versus narrative reviews to assess intervention effectiveness in emerging areas of research.
KEY WORDS: culturally sensitive interventions; substance use
To address substance use among youths from minority cultures, the importance of culturally sensitive interventions (CSIs) is often emphasized (Gil, Wagner, & Tubman, 2004). CSIs can be understood as interventions that incorporate a target population's culture (norms, beliefs, and values) (Resnicow, Soler, Braithwaite, Ahluwalia, & Butler, 2000). Resnicow et al. distinguish between surface structure and deep structure CSIs. The former involves tailoring interventions to conform to observable social and behavioral characteristics of the population. This entails use of people, language, clothing, settings, music, and exemplars that resonate with the population. The latter involves construction of interventions that reflect the social, historical, psychological, and cultural forces related to health and wellness in that population. With African Americans, for example, such a deep structural intervention might incorporate the values of spirituality, communalism, expressiveness, family, and storytelling.
The prevention program developed by Flay, Graumlich, Segawa, Burns, and Holliday (2004) represents one example of an intervention that incorporates a similar set of Afro-centric values and uses a mix of surface and deep structural strategies. Some content was adapted from exiting substance use prevention programs, whereas other, more Afro-centric content was developed specifically for the CSI. The intervention was pilot tested and then refined to enhance its level of cultural congruence.
Although CSIs have garnered considerable philosophical support, debate exists about their effectiveness (Hall, 2001). Proponents noted that interventions are typically designed, developed, and tested in environments that are predominately European American (Sue & Sue, 2008). Such interventions often reflect majority-culture beliefs and values and, consequently, may have limited utility with minority groups who affirm different cultural norms.
Alternatively, others have questioned the effectiveness of CSIs, arguing that contemporary society is increasingly characterized by a blended, post-ethnic youth culture in which norms, beliefs, and values are widely shared across cultural groups (Elliott & Mihalic, 2004). In addition, basic developmental needs are understood to be similar across cultural groups. Adapting proven evidence-based interventions to reflect more locally based cultural norms may undermine their effectiveness and even engender detrimental outcomes.
Despite such debates, little research on effectiveness exists (Hecht et al., 2008). This represents a significant oversight given the growing size of the minority youth population, which is projected to grow from 44% of the nation's children and adolescents to 62% in 2050 (U.S. Census Bureau, 2008). Accordingly, research is needed to determine the effectiveness of currently operationalized CSIs designed to address substance use.
ASSESSING EFFECTIVENESS: NARRATIVE REVIEWS VERSUS META-ANALYSES
Two types of systematic reviews are commonly used to assess intervention effectiveness: narrative reviews and meta-analyses (Borenstein, 2005; Littell, Corcoran, & Pillai, 2008). Both approaches offer unique advantages. Narrative approaches typically review a broader array of issues. In addition to discussing outcomes, they review concepts such as study theory, methods, and interventions in some degree of depth.
These strengths are perhaps particularly relevant in emerging areas of research, which are often characterized by relatively few studies using various designs. A small body of research lends itself to a narrative summarization by an expert in the field. Similarly, study heterogeneity can often be captured more accurately in a narrative format, relative to a meta-analysis.
Meta-analyses focus more narrowly on outcomes (Lundahl, Yaffe, & Hobson, 2009). Statistical procedures are used to obtain an effect size across studies. In contrast to narrative reviews, in which authors subjectively assign some level of importance to each study, meta-analyses use mathematical criteria to control for differences in sample and effect sizes among individual studies. The end result is an overall quantitative estimate of effectiveness, which can be compared and contrasted with other interventions using standardized conventions to determine if an effect size is small, medium, large, or nonexistent (Hedges & Olkin, 1985).
The aim of the present study was to determine the overall quantitative effectiveness of CSIs in addressing substance use (1), across measures and time frames (2), for recent (that is, 30-day) alcohol use and (3), for recent marijuana use. Given this aim, a meta-analysis was conducted. The trade-offs between meta-analyses and narrative reviews are revisited in the Discussion section, in light of the heterogeneous nature of the studies that emerged from the search criteria.
Studies had to meet four criteria to be included in the present analysis:
In keeping with the contextual nature of CSIs, the study had to be conducted in the United States.
The majority of participants had to be under 18 years of age and racial/ethnic minorities (that is, African American, Latino, or Native American).
The intervention had to be described as culturally sensitive and incorporate either surface or deep structural strategies (as defined by Resnicow et al., 2000).
The research had to measure some degree of change over time in substance use.
Studies were excluded on the basis of the following seven criteria: (1) if CSIs were described but not tested; (2) if the study featured retrospective rather than prospective designs; (3) if the study, reported only qualitative findings; (4) if the study examined substance use among minority youth using non-CSIs (that is, standard interventions); (5) if the study focused on changing cultural beliefs or other moderating constructs apart from examining the effects of CSIs on substance use; (6) if the study failed to include the necessary data to conduct a meta-analysis; or (7) if the research examined attitudes and intentions regarding substance use apart from the behavioral use of substances. One exception was made to this latter criteria in the case of a study that used a composite score from a nine-item alcohol use and intentions scale (five items assessed use and four items assessed intentions) (Komro et al., 2006). Finally, some ethnic minority youths (that is, Asians) have relatively low rates of drug use and were not included in the meta-analysis.
Search for Studies
In October 2009, a computerized literature search of databases indexed...