Describe the current epidemiological evidence for the prevalence, incidence, and development of adolescent and young adult drug and alcohol use, with a focus on both tobacco and alcohol use among American Indian/Alaska Natives (AI/ANs) compared to non-AI/AN. Use of these two substances are highly prevalent among AI/AN youth. Identify unique risk and protective factors through the developmental period of childhood to young adulthood for AI/ANs compared to non-AI/ANs. Please also describe regional differences and potential cultural differences in use. Discuss why such differences may exist and how any efforts to mitigate risk and exploit protective factors can respond effectively to the environmental context in which these differences arise.
Research Objective and Area of Focus. American Indian/Alaska Native (AI/AN) people suffer from greater rates of health disparities compared to any other ethnicity in the United States (J. P. Gone, Trimble, J.E., 2012; Trends in Indian Health 2014 Edition, 2014). Compared with non-Hispanic Whites (NHWs), AI/ANs death rates related to alcohol are 520% higher, chronic liver disease and cirrhosis 368% higher, and complications related to diabetes is 177% higher (Trends in Indian Health 2014 Edition, 2014). AI/AN adolescents through emerging adulthood also have increased risk for a number of health inequities. American Indian preadolescents residing on reservations for example, are more than 10 times as likely to have initiated marijuana use compared to their NHW counterparts (Stanley LR, 2015). Utilizing an inter-disciplinary perspective I will examine the developmental epidemiology of substance misuse among AI/AN youth through emerging adulthood, with a focus on tobacco and alcohol misuse. After describing the prevalence and incidence between AI/AN youth and their non-AI/AN counterparts, I will address risk and protective factors related to tobacco and alcohol highlighting the unique factors related to AI/AN youth. I will conclude by summarizing why health inequities may persist and what is needed in order to effectively address them.
Statement of the Problem: Prevalence and Incidence of Tobacco and Alcohol Use among
AI/ANs and NHWs Adolescence through Emerging Adulthood
Tobacco use Among NHWS vs. AI/AN Youth. AI/AN people have some of the highest smoking rates out of any ethnic group in the United States. A higher proportion of AI/AN adults smoke (21.6%) compared to NHWs (16.6%)(Current Cigarette Smoking Among Adults—United States, 2005-2015, 2016). Smoking accounts for hundreds of thousands of deaths each year and contributes to the leading causes of death among AI/AN adults, including diseases of the heart (third leading cause of death) and cancer (fifth leading cause of death) (Trends in Indian Health 2014 Edition, 2014). Up until 2013, AI/AN youth had the highest smoking rates of any ethnicity (24.6%) (Beauvais F, 2007). Tobacco use among youth is at a forty-year low for all ethnicities, with the smoking rates among AI/AN youth falling dramatically in 2015 to roughly the equivalent of the prevalence rate among NHW adolescents (12.2%). This is believed to be a direct result of public health efforts targeting perceptions of risk and the negative health effects of smoking, increased knowledge in how commercial tobacco companies exploit young consumers in addition to the Tobacco Master Settlement Agreement of 1998, a decrease in the perceived availability due to increases in taxes and compliance checks among adolescents, and finally more youth not having ever tried tobacco (L. Johnston, O’Malley, PM, Miech, RA, Bachman, JG, & Schulenberg, JE, 2016).
Regional Differences in Tobacco Use. Tribes in the Southwestern United States are less likely to smoke than in the Northern Plains (Tobacco Use Among U.S. Racial/Ethnic Minority Groups—African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon General., 1998). Smoking is more prevalent among men in the Southwest and women in the Northern Plains. Regardless of region, the odds of smoking are strongly associated with alcohol use (Henderson, 2005). Examining smoking patterns between Alaska Natives and tribes in the Southwest, current tobacco use among Alaska Natives is more than three times as high (32% vs. 8%) (Redwood, 2010). Unmarried younger males who only speak English in the home were the most likely to smoke. The authors underscore the potential difference in regional use may be due to the fact that Southwest tribes have a ceremonial and sacred practice with tobacco which may act as a protective factor against commercial tobacco use and smoking, whereas Alaska Natives have no such cultural connection to tobacco (Redwood, 2010).
Alcohol Prevalence among AI/ANs Compared to NHWs. Alcohol consumption poses a public health concern worldwide. Alcohol use disorders (AUD) among men for example, is the highest disabling disease, with a global mortality rate of 4% (Reducing Risks, Promoting Healthy Life, 2002). Internationally alcohol use is the largest avoidable risk factor (Rehm, 2009). Frequency of alcohol use in addition to patterns of drinking (including heavy drinking) is especially problematic (Rehm, 2009). This is also true of AI/AN people. The history of AI/AN communities and alcohol is a tumultuous one. Unlike NHWs who have a cultural history of social drinking, alcohol was initially used as a tool to more efficiently swindle resources away from AI/AN people (E. Duran, Duran B, 1995). Additionally, many reservations remain “dry” with tribal laws banning the sale and consumption of alcohol on tribal lands which underscores continued alcohol related issues in American Indian communities (N. R. Whitesell et al., 2012). Although AI/ANs do not drink more than their NHW counterparts and can have high rates of abstinence (Skewes MC, 2016), they are disproportionally affected by the negative outcomes of serious misuse.
The legal drinking age in the United States is 21 years-old. Despite this, many adolescents continue to engage in underage drinking. In 2015 the Youth Risk Behavior Surveillance reported that by age 13, 17.2% of NHW students had tried alcohol. In the past thirty days, 35.2% of NHW students drank alcohol, with 19.7% of that binge drinking (Youth Risk Behavior Surveillance —United States, 2015, 2016). For adolescents specifically, negative consequences of alcohol use include decreased motivation leading to poor academic performance, alcohol use contributing to mood disorders, increased injury and mortality in addition to higher healthcare costs (Hawkins, 1992; J. Miller, Naimi, TS, Brewer, RD, Jones, SE, 2007; The Surgeon General’s Call to Action To Prevent and Reduce Underage Drinking, 2007). Finally, emerging adulthood is important to consider since alcohol misuse and lifetime prevalence tends to be highest from the ages of 18-20 years old (49%) to 27 years old (72%) (L. Johnston, O’Malley, PM, Bachman, JG, Schulenberg, JE 2009). Since more than 50% of alcohol use disorders are diagnosed before age 25, prevention and intervention efforts should target those younger than 20 years old to thwart alcohol use disorder trajectories (Catalano & Ross, 2012).
Compared to NHWs, AI/ANs adolescents have elevated rates of mental health disorders (J. Beals, Piasecki, J, Nelson, S, Jones, M, Keane, E, Dauphinais, P, Red Shirt, R, Sack W, Manson SM, 1997) in addition to having experienced greater rates of severe trauma including physical and sexual abuse (Deters, 2006; Goodkind, 2012) (Thayer/McDonell). Exposure to traumatic or adverse experiences increases risk of substance use disorders among adolescent AI/ANs (Boyd-Ball, 2006). AI/AN youth suffer from higher rates of alcohol and substance use with earlier age of initiation (Whitbeck, 2008; N. Whitesell, Beals, J, Mitchell, CM, Keane, EM, Spicer, P, Turner, RJ, AI-SUPERPFP Team., 2007). By age 11, more than one-third of AI/AN youth will have tried alcohol (Mail, 1995). Additionally, close to 60% of students attending Bureau of Indian Affairs funded schools reported current alcohol use (“Tobacco, alcohol, and other drug use among high school students in Bureau of Indian Affairs-funded schools–United States, 2001,” 2003) and alcohol related automobile accidents accounts for 50-64% of deaths among AI/ANs 15-24 years-old (Romano, 2010). Alcohol use is also associated with increased co-occurring risk: high school dropout (53-67%), suicide (second leading cause of death for 15-24 year-olds) and unsafe sexual behavior (Clarke, 2002; de Ravello L, 2014; Trends in Indian Health 2014 Edition, 2014).
There is a high prevalence of alcohol use during emerging adulthood. Since alcohol use begins with underage drinking, it is therefore important to address this public health concern from a developmental framework (The Surgeon General’s Call to Action To Prevent and Reduce Underage Drinking, 2007). The transition into adolescence involves many psychological-social-biological experiences connected with puberty. These include changes in neuro-cognitive and emotional processing as well as change in relationships with family, friends and community (Graber, Hill & Saczawa, 2014 *in sloboda book*). Brain development impacts decision making such as impulsivity and sensation seeking, both of which have been connected to risk-taking behavior such as underage drinking (Romer, 2010; The Surgeon General’s Call to Action To Prevent and Reduce Underage Drinking, 2007).
Regional Differences in Alcohol Use among AI/ANs. There are regional differences in the prevalence of alcohol use between the more than 565 federally recognized tribal communities and Alaska Native village corporations in the United States. For example, Northern Plains’ tribes have almost twice the rate of lifetime alcohol dependence as compared to tribes located in the Southwest (16.6% vs. 9.8%) (J. Gone, Trimble, JE, 2012). In Washington State, there are higher rates of substance and alcohol use for both AI/AN adults and adolescents compared with other groups (Atkins et al., 2003; Northwest Portland Area Indian Health Board, 2014). In Alaska, alcohol abuse is the tenth leading cause of death among Alaska Natives. The combined mortality rate is 7.1 times higher than NHWs, with the female Alaska Native mortality rate 15.7 times higher than for their female NHW counterparts (Blake, 2016). Although these prevalence rates are mostly connected with AI/ANs 18 years and older, it further emphasizes the critical need to intervene in alcohol use during adolescence to disrupt potential lifetime alcohol dependence.
Regional differences in alcohol use among adolescents follows a similar pattern as adults, with some notable differences. AI/AN 8th graders residing on reservations were more than twice as likely (36.7%) as the national average (14.8%) to ever have gotten drunk (Stanley, 2014). These youth were also more likely to heavily use and binge drink compared to NHWs, with increased rates of lifetime alcohol dependence (J. Beals, Spicer, P, Mitchell, CM, Novins, DK, Manson, SM, the AI-SUPERPFP Team, 2003; Whitbeck, 2008; N. Whitesell, Beals, J, Mitchell, CM, Keane, EM, Spicer, P, Turner, RJ, AI-SUPERPFP Team., 2007). Unlike adults, young women between the ages of 11-13 attending schools on reservations in the Northern Plains region were significantly more likely to initiate alcohol use than were their male counterparts (N. R. Whitesell et al., 2012). Youth in the Great Lakes and Northern Plains regions were also more likely to use substances then those in Oklahoma and the Southwest (K. Miller, Stanley, LR, Beauvais, F, 2012; Tragesser, 2010).
Contrasting reservation regions to urban areas, urban American Indians residing in Las Angeles County for example, had earlier onset of alcohol use compared with all other groups (Dickerson, 2015). Conversely, Alaska Natives attending high school are almost half as likely to have drank alcohol in the past thirty days (19.2% vs. 35.2%) or to have binged compared to NHW students (11.6% vs. 19.7%), although rates do vary by tribal health region (Alaska Division of Public Health, Alaska Youth Risk Behavior Surveillance System, 2016, 2016). Few studies exist examining specific regional differences between tribes or among younger people, with most comparisons being drawn between AI/ANs and NHWs, making this an important area for continued research.
Risk and Protective Factors for Tobacco and Alcohol Misuse
Definition of Risk. Much like the term indicates, risk factors are variables that are connected to fostering dysfunction whereas protective factors assist individuals in resisting the outcome or improving the disorder. Risk and protective factors are rarely connected to a single disorder and are associated with individual, family and community environments. Risk factors have also been shown to be developmentally related, with specific risk factors causing greater issues at certain periods of development (Coie, 1993). Risk factors also have cumulative effects, in addition to the fact that many risk factors are shared among various types of disorders. Protective factors can buffer against the disorder, interrupt the mediational interaction between the risk factor and the dysfunction, or directly decrease the risk. Ideally, prevention science enhances protective factors and decreases risk factors as early as possible so that they do not stabilize as predictors of the disorder. This is why prevention programs are most effective when they include multiple components or protective factors to address multiple risk factors simultaneously, across stages of development, within an individual’s environment, with an emphasis on impacting proximal and distal outcomes (Coie, 1993).
The imprecise definition of risk and protective factors in research has led to miscommunication within the field, and worse yet, possible inaccurate or invalid conclusions drawn from studies (Kraemer, 1997). This can be dangerous when creating prevention or intervention programs based upon the accurate identification of risk and protective factors. Kraemer (1997) provides alternative terminology/typology to assist researchers in the proper classification of variables, including modifying and moderating variables. This model is important in implementing the study design or methodology necessary to determine the causal pathway in risk and protective factors, and has many implications for efficacy in prevention science. An added layer among AI/AN people is the differences in worldview related to defining risks and defining assets, which may be incongruent with Western definitions (Szlemko, 2006).
With the foundation laid by Kraemer (1997), Rutter (2009) strengthens the argument for the need to conduct studies that measure casual inference. The author discusses how the mechanisms of risk are influenced by a diverse range of variables, both environmental and genetic. To determine causal mechanisms, an approach incorporating many strategies previously underdeveloped up to this point, should be pursued (Rutter, 2009). Mediation analysis for example, is argued to be essential in not only assessing treatment efficacy, but in determining how environmental causes function. A recent example is a study on how ethnic identity is associated with mental health. Future optimism was found to significantly mediate ethnic identity and depression among adolescent Lumbee in North Carolina (Smokowski, 2014), important to consider as both have been identified as risk and protective factors for alcohol use in this population. Studies including moderation are also important. Religious affiliation has shown to moderate the effects of deviant peers on alcohol use among AI/AN youth (Yu & Stiffman, 2007). By identifying the processes and mechanisms related to risk and protective factors, interventions can be developed that effectively target negative and positive health-related outcomes.
Risk Factors for Tobacco use across Ethnicities. Risk of tobacco use among youth include environmental factors such as tobacco outlet density, but tend to be more heavily influenced by individual level factors including reporting lower school support (Hodder 2016), having siblings and friends who smoke, observing smoking in movies, marketing related to tobacco, and sensation seeking (Adachi-Mejia, 2012; O’Loughlin, 2017). Depression was another risk factor for youth as a subset of adolescents with depressive symptoms smoke cigarettes as a way to cope (O’Loughlin, 2017).
General Risk Factors Related to Alcohol. Risk factors related to alcohol use during adolescence primarily falls into four major domains comprised of individual or peer, family, school and community. The following risks include literature on behaviors associated with alcohol use, and not necessarily risk factors as defined by Kramer and colleagues (1997) that have established temporal precedence. This underscores a continued methodological issue and one deserving of further examination, however for the sake of this review, I will include risk factors associated with alcohol use. These risks include rebelliousness, attitudes favorable to use and friends who engage in the behavior. Poor family management, conflict and parental attitudes towards use. Low commitment to school and community norms permissive of youth alcohol use have also been shown to increase risk (Monohan et al., 2014 “THE RELATION BETWEEN RISK AND PROTECTIVE FACTORS FOR PROBLEM BEHAVIORS AND DEPRESSIVE SYMPTOMS, ANTISOCIAL BEHAVIOR, AND ALCOHOL USE IN ADOLESCENCE.”) Risk factors predicting drinking patterns during emerging adulthood are similar to those mentioned above with college attendance being another risk factor for this age group (Stone et al., 2012 “Review of risk and protective factors of substance use and problem use in emerging adulthood”).
Individual and peer. Similarly to NHWs, delinquency and associating with deviant or misbehaving peers (in addition to conflicts within the peer group) are all risk factors for alcohol use (Moncher, et al., 1990). Additionally, incongruence between cultural ideals and actual behavior has been shown to increase the risk of alcohol use (Garrett and Herring, 2001). Social adjustment factors including transitioning to different grades and peer pressure also increases risk (NIAAA, 2002). Early alcohol initiation and heavy use among AI/AN youth is a primary predictor (Schinke, et al., 2000; Spiker et al., 2003; Whitesell et al., 2012).
Family stressors. Family history of substance use and mental health disorders predicts use, along with parental attitudes about drug use and the ability of the youth to easily obtain the substance through family (Kulis et al., 2006). Laissez-faire child rearing (Herring, 1997) and lack of stability in the home are also risk factors (Kulis et al., 2006; Herring 1997).
School. As is the case with NHWs, the amount of drug and alcohol use among young AI/AN people contributes to academic failure and negative school environment increasing risk of substance use dependence (Yu et al., 2010).
Community. Issues surrounding unemployment, socio-economic status, and violent/criminal behavior are all risk factors for adolescent alcohol use (CITE). Among reservation youth, excessive free time is a risk factor (Herring, 1997). For urban AI/AN adolescents, it has been shown that involvement in the foster care system and a feeling of isolation from mainstream society contributes to substance misuse (Flynn et al., 1998). Higher levels of neighborhood disorganization is also connected to higher rates of drinking among AI/AN youth (Drinking among Native American and White Youths: The Role of Perceived Neighborhood and School Environment)
Historical contextual risk factors. The ecological developmental model integrates the contextual as well as transactional effects of environment on development (Stormkshak & Disihon, 2002). This influences environmental and interpersonal factors of the child or individual and is shaped by the community, behavioral settings and relationships. This model has also been placed within a health promotion framework that acknowledges that many behaviors are largely determined by the community setting (Green, Richard, & Potvin, 1996). This model acknowledges that there are systems and mechanisms that influence environmental and biological factors of an individual so that each causal pathway must be disentangled and placed within an ecological context.
This is why the ecological model is important for AI/ANs especially with respect to decreasing substance use as one could argue that environmental factors have played a role not only in the current health status of AI/AN youth, but may contribute to biological risk factors as well. Therefore, historic contextual risk factors may exacerbate the health disparities found in AI/AN youth and could be better understood within Bronfenbrenner’s ecological framework and multiple contexts of risk. Some of these risk factors for example are related to colonization, systemic oppression, inter-generational trauma, forced removal of children, and religious persecution (Brave Heart 1998; Duran & Duran 2001; Gone & Trimble, 2012). One study found unique risk factors for AI/AN adolescent substance use including experiences of forced assimilation and acculturation (Beauvais et al., 1998).
Adversity and trauma. As previously mentioned, AI/AN youth are at an elevated risk of experiencing trauma which is connected to substance abuse (Boyd-Ball et al., 2006). Various forms of trauma increase risk including: witnessing violence, major childhood events such as non-interpersonal/interpersonal (physical and sexual violence vs. surviving a car accident), and hearing traumatic news (Whitesell et al., 2009; Boyd-Ball et al., 2006). Adverse Childhood Experiences (ACEs) and total number of ACEs are also risk factors for substance use dependence (Koss et al., 2003). Examples of these might be physical and emotional neglect or abuse, having attended a boarding school, or as previously mentioned, foster care placement and parental alcoholism (Koss et al., 2003).
Definition of Protective Factors. On the other end of the continuum are protective factors and resilience. Positive youth development (PYD) frames resilience as assets focusing on the strengths, potential and talent of each child, first described in positive psychology as virtues and character strengths common among each culture (Seligman et al 2005). The concept of the fragile child has been replaced with the knowledge that everyone is resilient and that there is a difference between preventing risk behaviors and promoting wellness (Masten, 2001). PYD is framed by the idea that the expectation should remain high for the contribution that each youth can make to society, which is primarily accomplished through facilitating a strong moral identity (Damon, 2004). This framework is also very much in alignment with AI/AN belief and has been incorporated into several programs targeting AI/AN substance use (Kenyon 2012). Studies have also shown that PYD is protective against substance use by enhancing youth self-esteem, connections to family, positive peers and community mentors (Schwartz et al 2010).
Protective Factors Related to AI/AN Tobacco Use. Tobacco among many AI/AN tribes is sacred (Duran, 2006). Qualitative results show that American Indian adolescents are well-versed in the ceremonial uses of tobacco and make the distinction between commercial tobacco and traditional uses. Although some have argued that the current incorporation of commercial tobacco into cultural practices may decrease perception of risk among AI/AN youth, the literature has not supported this (Unger 2006). Other protective factors associated with low tobacco use among AI/AN youth were participation in sports (Whitbeck 2001; Orsilla 2007), having college aspirations, and participation in music (Orsilla 2007). Community mindedness, social support, academic orientation, and ethnic identity were also associated with low tobacco use (Lemaster 2002).
Protective Factors Related to Adolescent Alcohol Use. Protective factors against alcohol use during adolescence mainly involve prosocial behavior. This includes are having social skills, socializing with prosocial peers and belief in the moral order. Family and community protective factors include attachment and rewards for prosocial involvement Monohan et al., 2014 “THE RELATION BETWEEN RISK AND PROTECTIVE FACTORS FOR PROBLEM BEHAVIORS AND DEPRESSIVE SYMPTOMS, ANTISOCIAL BEHAVIOR, AND ALCOHOL USE IN ADOLESCENCE”).
Protective Factors for AI/ANs. Not much is known about the interaction between protective factors and alcohol, tobacco and other drug use among AI/AN youth. Some assets related to decreasing alcohol and other substance use have been identified, but these can be conflicting between studies (Beebe et al., 2008). The following promotive factors discussed represents a sampling of AI/AN specific factors in decreasing alcohol use.
Mentoring. Promotive factors for past thirty-day alcohol and tobacco use for AI/AN young people is the presence of a non-parental role model. These are people in the youth’s school or community who are able to offer advice, listen to their problems and are encouraging. Another way this is thought to fit into the existing literature is through the concept of collectivism. This construct ties into mentoring in that it is about being a part of your community and interacting with extended family. This protective factor means showing respect for elders, interdependence, cooperation and feeling a sense of responsibility to your tribe and community which has been linked to improved decision-making with respect to substance use (Beebe et al., 2008). Collectivism is one argument as to why mentoring may be such promotive factor for AI/AN youth especially when compared to NHWs. Additionally, positive family relationships has also been shown to decrease substance use and dependence among AI/AN youth (Yu et al., 2007).
Cultural Pride and Community Involvement. Enculturation refers to how rooted an individual is in their cultural identity and traditional practices (Whitbeck et al., 2004). Enculturation has been shown to be a protective factor for alcohol misuse among AI/ANs. This includes attending powwows, dances or other intertribal activities and having a connection to spirituality and an overall sense of cultural pride (Yu & Stiffman, 2007). However, the authors caution against focusing on “generic” cultural activities as the primary form of prevention since their qualitative results also indicated that many times activities such as powwows are followed up with parties where alcohol is consumed.
Self-reliance. Among the Keetoowah-Cherokee there is a traditional concept of self-reliance which is believed to strengthen positive youth development. Self-reliance is defined as being responsible, confident, and disciplined. One study found that utilizing this cultural protective factor together with Cherokee Talking Circles produced significantly higher reductions and better results at follow-up in decreasing substance misuse among Cherokee youth (Lowe et al., 2012).
Spirituality. Involvement in, or use of time related to religious groups, attending church, and being involved in traditional spirituality is also linked to reduced underage drinking (Beebe et al., 2008; Yu et al., 2007; Whitbeck et al., 2004). Qualitatively, it has been suggested that among AI/AN youth there is a connection between spirituality and cultural pride that enhances wellness and builds resilience (Beebe et al., 2008).
Current State of the Field, Implications for Research and Future Directions. An epidemiological perspective taken with developmental orientation is most effective in decreasing risk and enhancing protective factors. First the disease or health issue is identified and the etiology of the disease or disorder can be examined through the determinants of health, with an ecological framework. Prevention and intervention strategies are then developed to test the etiological model. All of this culminates in a program that undergoes efficacy and effectiveness testing (Sloboda, Glantz & Tarter, 2012). In order to fully assess the health outcomes of AI/AN communities, an ecological framework and self-determinants perspective should be employed (CITE ALL ARTICLES THAT USE ECO MODEL W/NATIVES). This approach emphasizes social structures that contribute to health inequities such as socio-economic status, education, access to healthcare and social services, food and agriculture (Koh, Piotrowski, Kumanyika & Fielding, 2011).
There are many ways to respond effectively to the environmental context in which these health disparities arise.One example related to tobacco and alcohol use is the need for culturally tailored interventions that incorporate protective factors unique to AI/AN communities including spirituality, mentoring and cultural identity. Integrating culture into treatment is especially vital as AI/AN communities maintain a special political relationship with the United States government guaranteed through the trust responsibility and hundreds of treaties (Gone & Trimble, 2012; Duran & Duran, 1995). The United States government is therefore obligated to provide treatment that is culturally appropriate, feasible, and effective in order to addresses health inequities among tribal members. Unlike other groups, the need for culturally grounded or tailored interventions is also supported by the right of tribal nations to maintain self-determination and sovereignty. Further complicating matters is the relationship between AI/AN people and their healthcare providers, which has not always been positive. Consequently, there may be reticence in receiving services that are not culturally-based due to these historical and cultural factors (Gone & Trimble, 2012).
More behavioral and mental health research and interventions conducted with AI/AN clients to verify treatment efficacy among AI/AN communities is also needed (Gone & Trimble, 2012). Although there are many EBPs that are effective among other populations, few studies have examined evidence of effectiveness for tobacco and alcohol treatments among AI/ANs (Donovan 2015). Providers have voiced ethical concerns over implementing EBPs that have not been tested within the AI/AN communities they treat (Novins et al., 2011). There has been progress over the last several decades in the identification, research and the development of prevention and intervention programs that address risk factors and enhance protective factors positively impacting alcohol and tobacco outcomes among AI/AN youth through emerging adulthood, however more work still needs to be done. Having examined the current research on the epidemiology, risk and protective factors related to tobacco and alcohol among AI/AN vs. NHWs adolescent populations, I have identified several gaps in the literature which I will continue to explore in my response to the specialty question.
A director of a social service program has reached out to you to consult on assessing the success of a national parent-leadership program that was implemented two years ago in a low-income, rural community in Colorado. Assume that this program was new to the community and that the number of people who participated was n>500. The focus of this program is to help preserve, protect and advocate for the ‘total’ health of their children (behavioral, physical, and academic). Outline your overall research design and evaluation, paying particular attention to issues of measurement, and a statistical analysis plan. In addition to potential threats to reliability and validity, describe the strengths and limitations of your design as you would when presenting your plan to the director. Discuss the health outcomes, along with the social and policy implications of this issue, especially as it relates to the designing and implementing of prevention efforts and interventions in low-resource, rural communities.
Conducting program evaluations in a real-world setting is complex (RFL; Mabry, 2010). Many considerations related to how best to design the evaluation may not have been thought of beforehand at the social service agency. There are other factors that hinder certain activities to best assess program impact, such as random assignment. This may not have been feasible due to a variety of factors including the desire to not withhold services from people (http://aea365.org/blog/?s=wait+list&submit=Go). I will make several assumptions throughout this response. First, I will assume that random assignment was not possible, but that some research considerations were made, such as including a control condition. Second, the curriculum that the service organization has implemented is evidence-based with existing reliable and validated measures for a rural low-resourced largely homogenous non-Hispanic White community. My assumption is that the measures assess growth in parenting skills and knowledge, parent leadership and civic engagement, and the potential improvement or changes in behavioral, physical and academic outcomes of the participant’s children.
Third, there are two kinds of evaluations of the program that could occur. Since the program has only been ongoing for a couple of years, I would be hesitant to recommend a summative evaluation looking at linking program activities to outcomes before having examined the process, implementation and fidelity of the project first (Hill & Owens 2015). Too frequently stakeholders jump to an outcome evaluation before determining if the program was even implemented properly. This can negatively impact whether the program was effective at all, as disused at length by RFL. Therefore, I would advocate first conducting an implementation evaluation to ensure that the program is being implemented with fidelity. Assuming that the program implementation was successfully executed, I would then suggest a formative evaluation looking at how the program impacted the health outcomes.
There are many research designs that could be beneficial for examining both the effectiveness of the program as well as adherence to fidelity. Randomized controlled trials (RCT) are the gold standard for determining causal effects of a treatment or intervention and have been implemented in diverse communities (McDonell). However, they are not always practical or possible to implement especially in smaller communities. Some issues that can arise from RCTs are scale and cost, the at-risk population you hope to reach is difficult to access, ethical issues or mistrust of research within the population, individuals are nested within the group, the analysis is at the group level and the intervention is examined as a whole so that the effective program components are not known (Fok, Henry & Allen, 2015; Henry et al., 2016; Collins, Murphy & Strecher, 2007). Therefore, for this particular research, a quasi-experimental design would likely be the most feasible. A number of designs are recommended in prevention science. These include multiphase optimization strategy, sequential multiple assignment randomized trial, regression point displacement design, interrupted time-series design, a stepped wedge design, and dynamic waitlist design (Shaddish et al, 2002). Each design will be discussed with specific strengths and weaknesses before identifying the best research design for this study.
Beginning with multiphase optimization strategy trials (MOST), this design is useful for designing and evaluating interventions out of the active components of an intervention and identifies the optimal dose. There are three phases to this design: screening, refining and confirming. This may be challenging for traditional interventions where there may not be many treatment conditions. However, fractional factorial ANOVA may be able to address this issue since not all conditions need to be included in the analysis. Sequential multiple assignment randomized trials (SMART) allows the tailoring of variables based upon participant needs and characteristics. This includes a set of decision rules that permit variation in the amounts of the intervention to be administered in accordance with the tailoring variables. Additionally, time-varying adaptive interventions strategies afford researchers the opportunity to address questions in both a rigorous and holistic manner (Collins, Murphy & Strecher, 2007). Both designs can be integrated by using the multiphase optimization screening phase to identify which tailored variables to include in the time-varying adaptive intervention. This approach is likely to be appealing to funders but is expensive and complex (Collins, Murphy & Strecher, 2007). However, the degree of expertise and sophistication needed to implement such a design is unlikely to be feasible in a rural low-resource area that and would therefore not be a recommended strategy.
Regression point displacement design (RPDD) examines differences in outcomes among those naturally falling at a predetermined cut score for risk or inclusion. Intervention participants are usually compared to archival data such as a standardized test scores or post hoc analysis of income requirements for participating in a child support care program. Differences are attributed to the intervention through comparing the scores of participants assigned to the cut point to those in the intervention. In certain cases, RPDD can approximate causal inference similar to that found in RCTs (Henry et al. 2016). RPDD is better suited when there is preexisting data on the outcome of interest or in larger population-level research.
Interrupted time-series design is useful for establishing a baseline prior to intervention implementation and then repeating measurement through the period of exposure. Intervention impact on outcomes are then determined through the intercept or slope of the scores related to the time of the exposure (Henry et al. 2016). This increases confidence in causality because it allows for repeated replication over time and makes within subject comparisons. Another strength outlined as essential by Gottfredson and colleagues (2015) is that group comparisons can be made for similar or identical time periods, enhancing causal inference. One study found that this research design resulted in estimates comparable to an RCT (Cook et al. 2008 “three conditions under which experiments and observational studies produce comparable causal estimates”). Everyone in the community can be exposed to the intervention over time, which may make an interrupted time-series design more appealing to communities. Although having multiple time periods addresses threats to validity related to time series designs, there is significant tracking and identification of comparable groups over time, which may be just as challenging to conduct as an RCT in addition to the fact that you must have access to, or collected data prior to the initial implementation of the intervention (Henry et al. 2016).
Stepped wedge design includes multiple observations over time so that there is no main effect because each unit is nested within time. Therefore, effects of the intervention are modeled across time as changes from the initial project implementation. The utility in this approach is that the evaluation of causality can still be investigated even when implementation did not occur at a single time. This is useful for community organizations where services must be provided to everyone meeting the program criteria but is limited by treatment effect heterogeneity, delayed effects and modeling assumptions (Henry et al. 2016; Current Issues in the Design and Analysis of Stepped Wedge Trials).
For the present study, I will assume that the service organization pursued a quasi-experimental, mixed-methods approach. A variation of the non-equivalent switching replication would be the design (Trochim, 2006). Two groups receive the intervention in a sequential process. During phase one both groups are assessed at baseline but only one group begins the program. Both groups are then reassessed at post-intervention. Phase two involves the initial comparison group participating in the intervention while the original cohort serves as the control. This allows for all participants to have received the program. Since exposure is staggered, intervention effects can be tracked with causal inference based upon comparing those that have and those that have not crossed over to the intervention condition (Henry et al. 2016). Each unit is like a mini-RCT and power is gained through multiple comparisons and within comparisons between the intervention and control groups. The strengths of this design are the within-subject sensitivity to detect effects and that it is an effective way to evaluate outcomes with strong internal validity. It is also appealing to many organizations as a way to evaluate outcomes without disrupting the normal functioning and operation of the organization or restricting services to people who ordinarily meet the criteria for enrollment (AEA eval etc).
Since the service agency has reached out to me to conduct an outcome evaluation, my evaluation questions would addresses whether the families in the enrolled parent leadership program gained improvements in parenting self-efficacy, positive parenting (warmth, responsiveness and effective discipline) and community engagement. Health outcomes assessed for children would examine changes in social competence, academic achievement, and reduced internalizing/externalizing behaviors (Long-term Impact of Prevention Programs to Promote Effective Parenting: Lasting Effects but Uncertain Processes). Secondary health outcomes would assess changes in substance use and nutrition.
There are many potential sampling strategies that the service organization may have employed in recruiting participants for the parent leadership program. Probability sampling is one approach including simple and stratified random sampling. However, purposive sampling is more common for quasi-experimental research in the social sciences. This sampling strategy is used when there is a predefined group that the researcher is targeting. Purposive sampling includes modal, expert, quota, heterogeneity and snowball sampling. Participants may have been receiving social services from this organization prior to joining the parenting program and most likely joined the program due to their interest in the services in addition to other recruitment or marketing strategies employed by the organization. Participants therefore are assumed to be lower-resource NHW parents residing in a rural town in Colorado. For the qualitative portion of the mixed-method evaluation, snowball sampling would be utilized. Snowball sampling is useful when members of the target population may be harder to reach, such as rural low-resourced parents interested in improving the total health of their children (Shadish).
Reliability and validity are fundamental to the scientific method. Reliability is how consistent and repeatable the measures in the study are at producing the same results. Threats to reliability that can create bias in the instrument consists of measurement error which can be both random and systematic. This bias can be created by the participant and their understanding or reporting of information as well as by the researcher in their measurement choice such as observations, surveys and instruments used (Zaslow et al, 2006). Measurement accuracy can be maximized through multimethod assessments to insure the measures are valid and reliable and is a process that the researcher has the most power to control. In addition, the instruments used should first be validated among the population it is intended to assess. It is also important that those administering the assessments are well trained so that they do not introduce bias as previously mentioned (Zaslow et al, 2006). Both quantitative and qualitative data and potential social service records will be analyzed for better data triangulation.
Generalized causal inference involves both internal and external validity. Construct validity refers to ensuring that you are operationalizing your constructs so that the construct truly captures what you are measuring as the results of a study are a product of the instruments used. As mentioned above, my assumption for the measures used in the parent leadership training curriculum are that they were validated among a population similar to those served at the social service agency. If the reliability of the measures assessed in the given population have not been validated, this can heavily bias results (Zaslow 2006). Some of the assessments would be administered once, while the others would be assessed at three time points: pre-, post-, and 3 month followup. The parent leadership program would have several questionnaires to assess the parenting and health outcomes of the program with some possible components to include outlined below.
There would be a one-time questionnaire to gather demographic information of the participants. Another questionnaire would assess health behaviors as well as current and past drug, alcohol and tobacco use, with some dichotomous responses. Dietary assessment of family nutrition would also be measured. Children’s academic success and school adjustment would be obtained through parent self-report. A parent leadership instrument would assess things like the level of parent modeling, vision of community change, seeking opportunities to be engaged, and creating collaborations and support networks (http://www.leadershipchallenge.com/UserFiles/English_LPI4eSampleReport.pdf; The Triple P-Positive Parenting Program: A systematic review and meta-analysis of a multi-level system of parenting support).
Additionally, an inventory to assess parenting and family functioning with domains including parenting and parent–child relationship as well as a family scale with domains addressing family functioning and cohesion. Each item would be rated on a 5-point scale from 0 (not true at all) to 5 (very much true). To identify higher levels of family issues, some items would be reverse scored. There would also be a measure on child emotional and behavioral domains to assess healthy functioning of the child. These would relate to child internalizing and externalizing behaviors as well as how well the child interacts and listens to their parents Parenting and Family Adjustment Scales (PAFAS): Validation of a Brief Parent-Report Measure for Use in Assessment of Parenting Skills and Family Relationships). Interitem reliability would be assessed through Cronbach’s α with a score of 0.85 on all measurements. Qualitative data would be obtained through interviews, focus groups and through the creation of participant developed success stories. Questions would focus on feedback of class delivery and implementation, impact of curriculum on individual and family health and well-being, as well as child health and functioning. All qualitative data would be coded by the evaluator and one staff member.
Data Analysis Plan
Considerations of the data analysis plan include sample size and power, missing data, attrition, possible research questions and the statistical analysis, in addition to the strengths and weaknesses of the design. After conducting an a priori G Power analysis to determine a small to medium effect size powered at ninety-five percent, a sample size of at least 100 people would be necessary. Since there is a fairly large sample size for participation in the program, it appears the evaluation would be sufficiently powered. Generally speaking, the initial sample size usually becomes an effective sample size when fewer cases can be utilized in the analysis than originally thought. This can be due to attrition and missing data. Missing data is common in longitudinal studies and there are different types of missing data. These include missing completely at random, missing at random and not missing at random. Compared to listwise deletion and single imputation, when data are missing at random multiple imputation is less likely to bias towards the null or inflate treatment effects. Multiple imputation increases internal and external validity and enhances power by effectively utilizing the available data when some of the time points or variables of interest are missing (McPherson 2012), which may be the case for the data provided by the service agency. This would be the preferred method for handling missing data in the study. Another consideration is attrition, which is especially common in longitudinal evaluations. This is problematic if the final sample is not representative of the original sample. When attrition occurs differentially across conditions, it is a threat the internal validity (Lochman & Steenhoven, 2002). Attrition would be examined through statistical analysis to determine if there were differences between those who stayed in the program and those that dropped out.
Data analysis would include calculating percentages for categorical variables and means with standard deviations for continuous variables between groups. Differences in groups at pretest would be assessed through t-tests for continuous variables and chi square tests for categorical variables. If the two groups were equivalent at pretest, one-way analysis of variance (ANOVA) on posttest scores could be used to determine the differences between the treatment and control groups. Another approach could be conducting an ANOVA on the differences between pre- and post- tests to determine if there were significant differences between groups.
If there were small samples sizes and the data were non-normally disturbed a non-parametric analyses would be most appropriate. A Mann-Whitney-U statistic for comparisons of independent groups could be used. For the repeated measures Friedman’s Two-Way Analysis of Variance by Ranks followed by a post hoc Wilcoxon Signed Ranks Test for repeated measures across time (Donovan 2015). Potential parametric tests consist of multiple linear regression or multiple logistic regression that could be used depending on the scale of the outcome variable. Propensity scoring, blocking, and matching may also be pursued (Katz).
Assumptions for the data analysis plan were that the data met all the assumptions of parametric tests. This includes the data were normally distributed, there were no outliers, all the other data checks met the acceptable thresholds, there was an even split in frequencies, groups were equivalent at pretest and assumptions of homogeneity were not violated. Analysis of covariance (ANCOVA) could be assessed through entering the pretest scores as a covariate, this removes pretest effects so that posttest can be more accurately compared between the treatment and control groups. Additionally, Hopkin, Hoyle and Gottfredson argue that reducing variance other than that which can be attributed to the intervention can be addressed through adding covariates that are strongly related to the outcome and account for individual differences and are person-centered. This is thought to clarify the noise in the outcome variable that is explained by the covariates, making it easier to observe the true effect of the intervention. This approach does not lower power when looking at a single predictor. Another approach recommended by Fok, Henry & Allen (2015) is using multilevel modeling to address nonindependence in the model as well as actor-partner interaction model that pools between-within analysis to model group effects. Mixed modeling can be used to analyze differences between groups, with mixed ANOVA used to compare means across a within-subjects variable and a between-subject variable. For example, the research question would be, is there a difference in parent self-efficacy in the parent leadership course vs. parents who did not take the course? The within factor is the pre vs. post vs. followup and the between is enrollment in the program vs. not taking the program (Nurs 527/528).
There are several limitations of the study design proposed for this evaluation. Primarily, lack of random assignment decreases the ability to determine casual relationships. Threats to internal validity could be related to self-selection, which can create systematic bias due to the differences between those who opted to participate in the program vs. those who did not. In a longitudinal design, history as well as maturation or natural changes over time can confound. Attrition, regression to the mean and practice effects can also threaten internal validity. Another limitation is self-report, which can be biased and is potentially influenced by several factors. Some of these factors may be the question wording and format of the response options, contextual influences related to the accessibility of the memory or information being referenced as well as the temporal nature of the question, how people understand what is being asked of them, social desirability and if there is no independent verification of self-report results (Holder, 2010)(Schwarz, 1999). Three major threats to external validity in this case might be the inability to generalize or make causal inferences to other people, communities or service-agencies.
This response outlined how an independent consultant might approach an evaluation of an on-going parent leadership program’s impact on the total health of families and children residing in a rural community in Colorado. As described throughout the response, the focus was in determining the impact of the program on improving the health outcomes compared to waitlist. The primary health outcomes targeted by the parent leadership program were increases in parental self-efficacy, positive parenting such as warmth responsiveness and effective discipline, and community engagement. Secondary health outcomes were related to alcohol, drug and tobacco use. The health outcomes related to children were reduced internalizing/externalizing behaviors, increased social competence, increases in healthy nutritional choices and academic achievement.
There are many social and policy considerations when designing and implementing prevention and intervention efforts in a low-resource, rural community. For sake of this response, my assumption was that the curriculum was chosen based upon the identification of community need. Through this process, funding may have been obtained via State and Federal sources so that the program implementation would have been in alignment both with the previously identified community need and guided by the interests of policymakers or community leaders. This is one way the program addresses social and policy concerns.
Emphasizing the benefits of parent initiatives to funding agencies and policymakers is also important. Depending on the parenting initiative, cost-benefits of parenting programs can be as high as $10.32 for every dollar spent (http://www.wsipp.wa.gov/ReportFile/1089). Evaluations can be complex especially among different populations and particularly when resources are low. There is always the fear in evaluations of on-going programs that results will indicate that the program had no effect on health and social outcomes. This is especially true in low-resourced areas where this type of finding could result in the loss of funding. One way to address this is through framing the evaluation as an opportunity for continuous quality improvement. Even if results were not significant or totally favorable, one point that could be emphasized is the success of the program recruitment and utilization and how this is an important metric in and of itself (RFL). In a rural community having anything close to 500 participants in a two-year period is a huge success and indicates that the program may be meeting an essential community need that could be explored further.
When evaluating a program it is important to think about how the finding will be presented and disseminated. When presenting my findings to the Director of the service agency, the program staff, community stakeholders and policymakers, I would be cognizant of increasing the accessibility of the results for an audience that may not have a background in research and evaluation. As part of the products generated from the evaluation, I would write a brief report that would be visually appealing to stakeholders and policymakers with short sentences and bullet points so that it could be turned into multiple one-page handouts, if necessary.
Additionally, a PowerPoint would be developed that program staff could use at various meetings and conferences to disseminate results or discuss program activities. The PowerPoint would also come in the form of a Prezi to increase the interactive nature of the presentation. Graphs and charts would be represented utilizing data visualization software such as Tableau, which is a data analytics tool that creates dashboards and is easy to use. Data visualization tools have gained popularity in evaluation for creating a level of approachability of data that is not always present in regular research reports (http://aea365.org/blog/dvr-tig-week-janina-mobach-on-contextual-visualization/; http://aea365.org/blog/category/data-visualization-and-reporting). Finally, if the program had been impactful for participants in supporting health and well-being in their community and lives, digital success stories could be created with volunteer participants. These are quick and easy to make with existing software and free websites so that the cost associated with making a digital story would be participant, staff and/or evaluator time.
Program Development, Implementation & Institutionalization
A tribal health department has decided to apply for a federal grant to implement a home-visiting program. You have been hired to assist with writing the proposal including choosing the evidence-based program that will be used, designing the implementation strategy in a tribal community where the program may not have been previously tested, develop the evaluation plan and propose strategies for sustainability. Please describe how you would go about designing your grant application in a manner that meets the needs of your community with the unique historical and political factors, while simultaneously discussing how you would address the four components listed above. Please also be sure to describe the relative merit and utility of collecting qualitative and quantitative data for this project, and discuss the budget implications of your design considerations.
The grant application process includes planning for and attempting to predict the components necessary to successfully rollout a program in a particular community. If I were hired as a consultant by a tribal health department to assist with writing a grant application I would have several recommendations for administrators related to successful program implementation. Throughout the body of this response I will also highlight general considerations related to prevention science as a whole. Recommendations will include how best to design the grant application and the need for evidence based-practice and why this is important in this context (Yuen). I will identify cultural and implementation considerations in addition to social-determinants of health that contribute to the disparity in outcomes among American Indian women which also effect the tribal home-visiting grant application. These include colonization, historical and political conditions, poverty, substance misuse and mental health symptoms (E. Duran, 2006; E. Duran, Duran B, 1995; J. Gone, Trimble, JE, 2012; J. P. Gone, Trimble, J.E., 2012; Trends in Indian Health 2014 Edition, 2014). In this response I will also outline the importance of including an evaluation plan and carefully crafting a budget. Proposed strategies for sustainability will also be considered. Assumptions related to this question will be integrated throughout the response.
Problem Statement. For this portion of the grant application the maternal and child health inequities in addition to the need for services would be identified and outlined. Differences in the social-determinants of health especially for mothers and pregnant women residing on the reservation would be described in detail. Briefly, American Indian women have higher rates of negative maternal and child health outcomes compared to NHWs. For example, the infant mortality rate among American Indian women is 26% higher than for all other races (Trends in Indian Health 2014 Edition, 2014). Prevalence of maternal smoking is the highest among AI/AN women which contributes to a number of negative birth outcomes (“CDC Pregnancy Risk Assessment,” 2010). Preterm birth in AI/AN is 13% as compared to 11% in NHWs (Raglan, Lannon, Jones, & Schulkin, 2015).
Until the 1980s, the IHS practiced racist and eugenic health services through coerced or unconsented sterilization and termination of pregnancies on thousands of AI/AN patients (Temkin-Greener, Kunitz, Broudy, Haffne, 1981; Torpy, 2000). Distrust between AI/AN women and their healthcare providers echoes today as many AI/AN women choose not to receive prenatal care due to this traumatic relationship. Therefore, a discussion of American Indian maternal health cannot be separated from historical as well as current micoraggression. Due to this history, there is an urgent need to implement culturally appropriate services to address the negative birth outcomes of American Indian women and the delayed development of American Indian children. This remains a top priority for the Healthy People 2020 goals and objectives and is further supported through the authorization of the Affordable Care Act (Koh, Piotrowski, Kumanyika, & Fielding, 2011) HRSA https://www.hrsa.gov/grants/manage/homevisiting/).
Needs Assessment. Ordinarily the target population would be described in the needs assessment section of the grant along with the degree of readiness to change. The target population would be American Indian low-resource first-time mothers and their families. A description of available services and gaps in services in the community would be outlined. A formal Needs Assessment to be conducted during the initial phase of the program implementation would also be suggested. This would assist with determining if there is sufficient capacity and community organization to adequately address maternal and child health and if there is an appropriate climate for implementation of identified evidence-based curriculum, ideally championed by a community leader. Multifaceted community capacity building integrated into the intervention itself has also been correlated with directly impacting health outcomes (Sanigorski, Bell, Kremer, Cuttler, & Swinburn, 2008).
During the Needs Assessment it will be important to assess how an evaluation and dissemination activities will ultimately be most useful to the community. use WSIPP as an example of success for eval of programs (For WSIPP Cite Spoth/Steverman and Spoth 2008 for Nurse-Family Partnership if I end up getting the question about home-visiting). Ideally, the home-visiting program will be connected to another program in the tribal health department where lower-income first-time mothers can be easily recruited and there is an integration of the program into the organizational culture. Further, marketing strategies may need to be employed for broader recruitment reach (Spoth 2008).
Evidence. The National Institutes of Health (NIH) has identified “Blue Highways” along the road from basic research to the translation of this into practice (Brownson, Colditz, & Proctor, 2012). There are three phases of translational research referred to as T1, T2 and T3. T1 includes preclinical studies and basic research transferred to humans which is often referred to as “the bench.” Moving to phase T2, clinical trials or randomized controlled trials are conducted shifting from “bench to bedside” to determine the efficacy of the intervention. T3 is the process of determining effectiveness in addition to the dissemination or integration of an intervention into a particular setting or clinical practice to make a true public health impact (Drolet & Lorenzi, 2011)(Drolet & Lorenzi, 2011; SPR MAPS II TASK Force, 2008; Fishbein & Ridenour, 2013). There are potholes along this “highway” due to a variety of factors, including the funding priorities of NIH so that translational research is only budgeted at 9-25% of total funding. This challenge is compounded by the fact that translating research into accepted clinical practice happens about 14% of the time, over a 17 year period (Brownson, Colditz, & Proctor, 2012). Complicating matters, most EBPs never even reach the effectiveness testing phase among AI/ANs.
Nurse Family Partnership (NFP). Examples of successful translational impact paradigms include NFP which emphasizes parenting skills and is family-centered with evidence for impacting breastfeeding, improving vaccination rates, decreasing preterm births, increased life and parenting skills, maternal health behavior’s, early childhood development and substance use (Thorlan et al. 2017; Thorlan & Currie, 2017; Spoth 2008; Stormshak et al 2009). NFP also has support from the legislature and maternal and child health registries (WSIPP; what works clearing house). The program provides bi-monthly home visits to low-resource first-time mothers, from conception until the infant is two years old. NFP is currently in phase T3 among AI/AN populations (HRSA). Cultural and local adaption would therefore be necessary to implement the program in a tribal community. Due to its current evidence base, support from federal funders and the direction of NFP including tribal communities in the research process, this program would be the recommended EBP for the grant application.
In the Descriptive Epidemiology section of this response, EBPs were briefly touched upon with respect to the need for culturally tailored or grounded interventions. This is important to the tribal home-visiting grant application as well because it relates to successful program implementation. Sources of program mismatch for example include attempting to implement an English written/speaking program among Spanish speakers with no previous translation so that people do not understand the program content or there are serious differences in the values and norms between the curriculum and the consumer, a distinct possibility among an American Indian community (Castro, Barrera & Martinez, 2004). Over the last eight years NFP has targeted tribal communities to see if there are similar positive health outcomes that have been found in other populations. Therefore, the grant application would need to emphasis and balance program fidelity with contextual fit since there is not the same evidence-base among American Indian communities.
Contextual fit refers to the strategies, procedures and elements of an intervention matching with the values, needs, capacity and resources of the setting. This is an undervalued factor affecting quality and more research is needed to better understand this interplay as mentioned in other sections of this response. Horner 2014 argues that policymakers should include contextual fit as part of funding criteria to improve the selection, adoption and implementation of interventions, which the Health Resources Services Administration (HRSA) has attempted to do in their home-visiting initiative. As a result of the lack of resources allocated by funding agencies to support EBPs however, many communities do not have the resources necessary to implement such programs making this a good opportunity for the tribal health organization to apply for funds that would support the cultural tailoring of NFP (Kellam & Langevin, 2003)(Kellam & Langevin, 2003; Neblhof). It is also hypothesized that fidelity and quality are moderated by participant responsiveness and that participant responsiveness also mediates outcomes, therefore adaptation has both direct and indirect effects on participant responsiveness which has been demonstrated among AI/ANs (Berkel, Mauricio, Schoenfelder, & Sandler, 2011) add the citation from the article that discussed issues with participant participation)
Community-based participatory research (CBPR) is one way to address potential program mismatch. CBPR has been shown to be especially effective in AI/AN communities because it empowers stakeholders to take charge and actively participate not only in their health, but in the research process as well (Wallerstein, Yen, & Syme, 2011; Wandersman, 2003). Recently public health strategies have more strongly emphasized community-engaged interventions. Community-engaged research assists in determining whether the identified risk factors are priorities for stakeholders and if the community is amenable to change using an asset-based approach (Wallerstein et al., 2011), all of which will be extremely important to the successful program adoption of NFP in a tribal community that may be experiencing birth outcome inequities. Lastly, CBPR is an important strategy for communities that feel like they have been surveyed to death without any actionable results (“helicopter research”), in an ahistorical manner, with no real assessment of readiness to change (Wallerstein et al., 2011).
Coalition development and support is another important area for successfully adapting an EBP like NFP into a tribal community or organization in a manner in alignment with CBPR. The idea behind community organizing and coalition building is that multi-organizational groups, community stakeholders and leaders can come together with greater potential to create sustainable change (Butterfoss, Coalitions and partnerships in community health.). Coalitions and collaborations focus on community development and participation of stakeholders which increases capacity, competence, social capital and community empowerment (Butterfoss, Coalitions and partnerships in community health.). In order to have an effective coalition, the focus of the group on maternal child health must be very clear and goal-oriented, which has been shown to enhance member satisfaction leading to increased effectiveness of the EBP as well as the coalition. Size and focus of the coalition also supports healthy functioning with smaller groups being more effective (Brown, Feinberg, & Greenberg, 2010). One of the goals of NFP is increased community capacity to address maternal infant health inequities which will assist with integrating the program into the local service ecology.
Two examples of successful coalitions creating change are Communities that Care (CTC) and PRO-moting School-community-university Partnerships to Enhance Resilience (PROSER), both of which are university-community partnerships developed to reduce risky behavior among youth and have been successfully implemented among AI/AN communities. Coalitions assist with diffusion of EBPs and with generating buy-in within a community (Brown et al., 2010). Coalitions have also been linked to increased community readiness, TA, funding and fidelity which all contribute to sustainability. Due to the successful use of coalitions, CTC for example demonstrated significant effects for reducing youth substance use (Brown et al., 2010). Given the criteria for EBP in prevention science and the need to culturally adapt programs to determine effectiveness among multiple populations, NFP program fit will be enhanced thorough the cooperative agreement with HRSA and extensive coalition building.
Adoption, Dissemination and Implementation Strategies. Having discussed the importance of EBPs, cultural fit and identifying NFP for the grant application, I turn to adoption, implementation and dissemination strategies for the project which is generally discussed in the work plan section of the grant application. Diffusion of Innovations was proposed by Rogers and refers to the spread and adoption of new innovations over time. This include the need for perceived attributes of the innovation; reaching early adopters; the social system; the adoption process and the diffusion system. Implementation refers to the set of activities and processes necessary to integrate an intervention into a community setting (Rabin & Brownson, 2012). Dissemination strategies focus on how to communicate the components of the intervention to the target audience as well as the adopters of the program (Bopp, Saunders, & Lattimore, 2013; Rabin & Brownson, 2012).
There are many frameworks and theories in prevention science that support the planning implementation efforts that could be integrated into the grant application for the home-visiting program. A framework commonly used in prevention science is the RE-AIM Framework that focuses on the reach, adoption, implementation and maintenance of multilevel indicators including individual and setting that can be used to evaluate dissemination and implementation efforts (Glasgow, Vogt, & Boles, 1999; Rabin & Brownson, 2012). RE-AIM is an important framework for American Indian tribal communities because it incorporates individual and systems level factors with an eye towards adaptation, adoption and institutionalization in a holistic and systems approach (Glasgow et al., 1999).
As one of the four phases of implementation, it will be important for the tribal organization to have an implementation team to ensure success which would be written into the grant application. One way to support this is through providing technical assistance (TA) and coaching to members of the team delivering services with a focus on continuous quality improvement. TA is generally offered to tribes through the funding agency. This can be useful in assisting with program delivery and collection of program measures, but sometimes does not extend to other issues that can crop up for tribal organizations including tribal politics or capacity issues related to staff turnover (Meyers et al., 2012).
Poor training has also been cited as problematic to successfully implementing EBPs. Training sometimes does not adequately prepare facilitators for program delivery largely due to employee absenteeism and high turnover, which is certainly an important consideration in a tribal health department. The focus of TA provided tends to be less about continuous quality improvement and more about putting out fires. Since many analyses of programs have not been sophisticated enough to drill down on core components of effectiveness especially among American Indian communities, poor training can lead to problematic local adaptations that can windup reducing program dosage, does not account for program theory and throws out entire sections of the curriculum which can negatively impact outcomes (Elliot & Mihalic, 2004). Despite this, findings suggest a TA capacity building approach is beneficial regardless of the target population or desired outcomes in addition to organizational structure and will be an important part of the grant application (Meyers et al., 2012).
Sustainability and Institutionalization. Strategies for sustainability and institutionalization refers to the extent that NFP has been adopted into the culture of the organization through policy and practice giving the program longevity in the tribal community. There are three main stages of institutionalization including passage which may include a change from temporary to permeant funding, “cycle or routine” which encompasses having mechanisms in place that reaffirm the importance of NFP in the community through community identification of the need for evaluation of program outcomes or impact of the intervention, and lastly “niche saturation” which occurs when NFP has been successfully integrated into the subsystems of the organization (Rabin & Brownson, 2012). With more emphasis placed on accountability and technical assistance. Wandersman & Florin (2003) suggest answering 10 questions which is believed to lead to better intervention outcomes and institutionalization. A sampling of the questions include, “what are the needs and resources in your organization/school/community/state? How does the intervention fit with other programs already being offered?”
For a reservation community, sustainability will boil down to program fit, capacity and funding. Sustainability will be strengthened through the coalition with community buy-in, the on-going CBPR approach, TA and training that assists with building capacity, and the cultural adaptation of NFP. Federal funding has become increasingly precarious and the future of the Affordable Care Act uncertain. For tribal partners however, home visiting services may continue to be third party billed through Medicaid, enhancing sustainability of NFP services in the community (http://www.nashp.org/sites/default/files/medicaid.financing.home_.visiting.programs_0.pdf ).
Methodology. A mixed-methods approach would be pursued to maximize the effectiveness of the evaluation and to better track performance measures of the NFP program. There are many considerations when employing a mixed-method approach. As Yoshikawa and colleagues (2008) argue, when working with diverse stakeholders’ numeric or quantitative analysis does not provide a full picture. What can enhance understanding is qualitative research that addresses the epistemological assumptions of the population. Depending on the community, words, narratives and experiences should be given equal status with quantitative methods, which will be especially important in a tribal community with a strong oral tradition. Further, a mixed-method approach is guided by the research question and blends the two types of data, with a particular strength of this method being drawn from the ability to work across disciplines in support of the health and well-being of the community (Katz et al, 2016).
The combination of quantitative and qualitative data provides both prevalence estimates as well as data related to the goals of the tribal home-visiting program explored through process analysis to uncover mechanisms behind cause and effect utilizing the two sources of data. Taking one of the NFP metrics of enhancing family education and employment for example, a research question could address whether young families participating in the NFP program report higher levels of participation in the labor force compared to the control group, which could be explored both qualitatively and quantitatively. This approach assists with triangulation, increasing the validity and reliability of the evidence by using more than one source (Chapal, n.d. cdc). Finally, qualitative data has the potential to enhance cultural acceptability of NFP. One form of qualitative analysis that may be helpful in gathering the experiences of the parents in the home-visiting program might be qualitative description. This form of analysis allows the research to be guided by the words and experiences of the participant themselves with less theory and interpretation coming from the researcher (Neergaard et al, 2009).
Evaluation Plan. Theory is important to evaluation for a number of reasons. It assists with guiding the goals, the outcomes assessed and the values of the program. It can assist with guiding the methods used and how this will be implemented (Chen, 2005)(Chen, 1990). Theory provides the framework to determine the merits of a program and provides the fundamental values of whether an evaluation is useful and is of good quality. Without proper program theory, the assumptions and the subsequent evaluations may be faulty. One way to represent the program theory of change is through logic models. For an American Indian community implementing a home-visiting program, a medicine wheel logic model might be appropriate. This speaks to the traditional values of harmony and balance in four quadrants: mental, physical, emotional and spiritual that map onto issues related to staff education, workplace safety and participant outcomes (please see Amendment A) (CITE). Although Nurse Family Partnership has an existing logic model, a more conventional or standard grant logic model could also be included in the application (INSERT LOGIC MODEL Figure 1. adapted here from Prevention Science 535 and NFP logic model).
Stevahn & King (2016) recommend an interactive evaluation practice facilitating and guiding project aims, which is a natural process of CBPR. American Indian stakeholders and community members are likely to be more concerned with service delivery and less on program accountability. This could be accomplished through a process evaluation to ensure successful adoption and implementation of NFP. The WK Kellogg Foundation Evaluation Handbook suggests considering the balance between the “call to prove with the need to improve.” Stakeholders inherently understand the importance of developing a program and evaluation that is both defensible and credible (RFL). It is also important however, to balance rigor with stakeholder needs (RFL) and this is also built into the NFP program.
More cost-effectiveness and cost-benefit analysis should be pursued in prevention science and this could be incorporated into the proposed evaluation (Gottfredson et al., 2015; Kellam & Langevin, 2003)(Gottgredson et al, 2015; Kellam et al, 2004; Spoth et al, 2004). For the home-visiting program it would be beneficial to suggest establishing the foundation for a potential economic analysis plan that could be conducted in the future. This would assist in demonstrating to stakeholders, providers and policymakers the benefits and impact of the home-visiting program in the community compared to current practice (Gold et al., 1996). For an accurate assessment, secondary costs would also be evaluated. For example, transportation costs would be included for parents to attend the home-visiting group activities which is beyond the scope of regular cost accounting (Crowley, Hill, Kuklinski, & Jones, 2013).
When assessing the potential impacts of the home visiting program, I would include in the evaluation portion of the grant, outcome measures related to costs of the program and delivery of services at the population-level. If an economic evaluation of the program were pursued in the future, the outcome variables would be reported as the incremental cost-effectiveness ratio (ICER) in dollars per quality-adjusted life-year (QALY) (Murphy et al., 2015). ICER assesses cost-effectiveness by comparing the difference in cost between two interventions and dividing by the difference. QALYs are useful because it incorporates both morbidity and mortality allowing for comparisons of health effects of the intervention which would be important to assess in the home-visiting program (Sanders et al., 2016). When the economic evaluation is conducted, it is important to include two reference cases in the analyses based upon the perspectives of the healthcare sector as well as the societal perspective (Sanders et al., 2016). This could be accomplished by tracking costs to the provider, administrative costs of the program, and cost-offsets that may result from the services provided to the community.
Budget Considerations. Funding sources for the program include HRSA or Administration for Children and Families (ACF). There are also a number of state block grants that focus on maternal and child health that could be pursued (Steverman & Shern, 2014). The FY2017 funding cycle for HRSA included competitive funding of home-visiting applications for 5-year awards up to $1,300,000. This proposed tribal home-visiting project could easily be executed with an annual budget of $360,000 (Health and Human Services Administration, 2011). This includes personnel of two full time registered nurse home-visitors, a program coordinator and a data collection/manager, administrative costs related to being an NFP affiliate, program incentives for participation, and fixed and indirect costs. Examples of additional costs are funding for a program evaluator to assess the impact of the NFP program in the tribal community, ongoing nurse education, mileage for driving around the reservation to make home-visits and a budget for the data and reporting system. As previously mentioned, a plan for cost allocation and analysis would also be identified to ensure that the impacts of the program are tracked as early as possible (Crowley et al., 2013).
Younger AI/ANs are at an elevated risk for substance use. There is not a lot that is known about treatment and interventions aimed at AI/AN young adults, especially evidence-based treatment. Describe the theory, policy and research considerations when addressing prevention or intervention efforts that exist or could be developed for AI/AN adolescents and emerging adults (18-29 years-old) suffering from substance use disorders. Please discuss what, if any, types of interventions offer promising solutions for AI/AN communities and what other types of interventions are unlikely to have significant impact. Describe intervention work that is incorporating indigenous protective factors such as cultural identity, cultural continuity and sense of connectedness in addition to Western approaches to address this public health concern. Importantly, be sure to discuss how such important cultural factors can be maintained while still developing and integrating evidence-based approach to substance use disorder treatment. In addition, what is the evidence supporting these interventions and where and how could this be improved?
Substance Use Disorder Interventions: AI/AN Youth-Emerging Adulthood
There are many theory, policy and research considerations when developing drug and alcohol treatments for younger AI/ANs. Researchers have used both traditional indigenous knowledge as well as Western-based theory generated from psychology (transtheoretical model) and biomedical models when forming and adapting programs. However, there is a disconnect between common Western-based theories and AI/AN cultural worldview. Western approaches tend to focus on the individual and their motivation to change while often overlooking family, community, cultural and spiritual components that are important to tribal community members Allen 2016).
AI/ANs suffering from SUDs are more likely than non-Natives to seek services, but due to the lack of available holistic and culturally tailored treatments, have lower treatment retention rates (Beals 2006 help seeking for substance use problems in two American Indian reservation populations). There is general consensus that a strength-based model is the most appropriate for AI/AN communities (Allen; Whitesell; Beebe; Donovan Kenyon; Smokowski; Cross 2010; etc). Allen and colleagues (2016) advocate for including strengths into the intervention theory that relate to the individual, family, community and spirituality. Some of these include reverence for ancestors and kinship knowledge, powerful women, traditional laws, recognizing the spirit and belief in prayer.
With respect to theory, the Medicine Wheel theory has been incorporated into grassroots programs developed in reservation schools and supported through funding agencies including Robert Woods Johnson Foundation and SAMHSA’s Circles of Care. The Medicine Wheel teachings as described in other sections of this response, is a culturally grounded theory that stresses the importance of balancing mental, physical, emotional and spiritual domains for improved health and well-being (Whitbeck 2012). Variations of this theory have been expanded upon. Cross and colleagues (2010) developed the Relational Worldview approach emphasizing the interconnectedness of people with data gathered across the domains, focusing on discovering patterns between the quadrants. Additionally, heuristic models have been developed by Alaska Native communities outlining the process of moving from risk to protective factors. Community members theorize that community, family and individual characteristics are influenced by social/environmental factors (including trauma) which leads to experimental substance use. After use occurs, youth begin to think it over, with self-reflection culminating in a turning point and sobriety.
Interventionists have also used theories that are Western-based but incorporate environmental factors to better support the treatment needs of AI/AN people. Theories that have been used to develop SUD curriculums include the Primary Socialization Theory and the social determinants of health because they capture the external factors that potentially mediate and moderate treatment response (Whitesell 2012). Bronfenbrenner’s ecological approach has also been incorporated into treatment development for the same reason (Yu 2007; Smokowski 2014). Cultural Historical Activity theory is recommended when culturally tailoring curriculums for AI/AN younger people because it incorporates the aspects of language, culture and community that are central to learning. Cultural Compatibility theory has also been utilized to ensure that human interaction around the SUD curriculum is aligned with the community in which it is being implemented. Cognitive theory has also been identified as the basis for cultural tailoring of SUD curriculums because new knowledge is acquired in reference to prior understanding and experiences (Ringwalt 2006). As previously mentioned, Positive Youth Development (PYD) is another theoretical framework that has been successfully applied to prevention of SUDs among AI/AN adolescents. This is because PYD targets assets to enhance the total health of AI/AN younger people in alignment with many AI/AN cultural beliefs (Kenyon 2012).
There are also policy implications stemming from prevention and interventions targeting SUDs among AI/AN youth through emerging adulthood. In 2010 the U.S. Congress reauthorized the Indian Health Care Improvement Act (IHCI) as part of the Affordable Care Act. Along with expanding health insurance parity for alcohol and other substance use and mental health disorders, this act broadens access to needed services as well as additional funding to tribally run healthcare agencies. Under the current administration the IHCI is in jeopardy, creating uncertainty for tribal communities in their ability to adequately address health inequities with potentially devastating health outcomes (Tribal Organization IHCIA Support Letter, 2016). In addition, the FY2018 American First: A Budget Blueprint to Make America Great Again outlines substantial funding cuts to agencies most likely to support research on substance use disorder prevention and treatment among AI/AN youth through emerging adulthood.
Thoughtfulness in the research approach is also integral to the development of AI/AN interventions. Community-based participatory research (CBPR) has been identified as the most effective way to develop, implement, adapt or enhance the cultural acceptability of treatments among AI/AN youth (Rasmus et al., 2016; Donovan et al., 2015; Goodkind et al., 2012; Novins et al., 2012; Okamoto et al., 2012; Cross et al., 2011; Allen et al., 2006). CBPR assists with assessing community readiness, identifying areas of importance to target, getting stakeholder buy-in and finding a champion for the development or implementation of the intervention. Among AI/AN communities, this usually takes the form of developing a culturally grounded intervention or culturally adapting an existing program. When adapting an intervention there are three main areas to focus on: local adaptation, participant engagement and sustainability (Barrera 2016). This form of adaptation was discussed at length in the implementation question so the focus here will be a brief discussion on the importance of CBPR in research and developing culturally grounded or culturally tailoring interventions.
Research considerations for developing SUD interventions among AI/AN communities emphasizes the relational. This includes meetings with community members to identify goals, objectives and tasks as well as a process for effecting positive change. A qualitative component is essential with the inclusion of tribal elders and leaders and an orientation towards capacity building. Another vital aspect of this is that stakeholders are co-owners of the project throughout the process (Lowe 2011). In addition, researchers must have ties to the community beyond solely conducting research. The outside researcher must be respectful of the history and ways of knowing of the community, as well as subscribe to a practice of reciprocity (Rink). Having described the theory, policy and research implications of designing programs for AI/AN youth I turn to prevention and intervention evidence.
Prevention and Intervention Evidence
As various governmental and funding agenesis move more towards accountability and demonstrating program effectiveness in order to justify costs, the evidence-based practice (EBP) movement has gained momentum to demonstrate causal efficacy, moving from clinical experience to scientific evidence. This is problematic for AI/AN communities however as there is an alarming lack of alcohol EBP treatments that have been tested among AI/AN communities generally, and adolescents specifically (Gone & Trimble, 2012; Nebelkopf et al., 2011). This deficit includes an absence of AI/AN participants in RCT studies, limited resources and research investigating the interplay between culture, mental health and treatment, in addition to a lack of representation of AI/AN researchers on research teams (Nebelkopf et al., 2011). Additionally, mediators and moderators of local context and culture which are heavily influenced by co-morbidities, poverty historical trauma as well as resilience and community assets, are not captured by RCTs, which is yet another reason RCTs lose credibility among tribal communities (Nebelkopf et al., 2011; Gone & Alcantara, 2007)
There is also confusion around what constitutes evidence-based, promising-practice, best-practice, community-defined and practice-based prevention and intervention evidence (Eco-hawk, 2010 Compendium of Best Practices for American Indian/Alaska Native and Pacific Island Populations USF Children’s Mental Health Research & Policy Confer). I will make distinctions between the levels of evidence throughout the body of this response. Briefly, evidence-based generally refers to programs where randomized control trials or quasi-experimental trials have been conducted. Promising-practice refers to interventions that have positively impacted outcomes but have not been empirically tested. Best-practice refers to clinical practice or guidelines that have been reached through consensus and thought of as effective by clinicians. Practice-based interventions have not been rigorously evaluated but are supportive of local society, cultural and traditional approaches adopted by the community (Bartgis & Bigfoot, 2010). And finally, community-defined are those interventions that communities use and treatments they determine produce positive results (Nebelkopf et al., 2011).
There are various online registries that list projects with levels of effectiveness as well as what evidence was included to determine the rating. The most well-known online registry is SAMHSA’s National Registry of Effective Programs and Practices (NREPP). Currently, only 2 prevention and intervention programs included on the NREPP registry have been specifically designed for AI/AN youth to address substance misuse. These include: Project Venture and Red Cliff Wellness. Interestingly, Project Venture and American Indian Life Skills is considered evidence-based within the AI/AN literature (Okamoto et al., 2014; Nebelkopf et al., 2011) but according to the NREPP registry, it is a promising practice. This underscores continued barriers between a communities’ ability to provide culturally acceptable and effective treatments and whether these programs will be considered evidence-based by potential funding agencies and non-Natives. I will discuss the current list of alcohol treatments provided to AI/AN youth, evidence of effectiveness and future considerations.
Reasons Interventions are Unlikely to Have a Positive Impact
Themes related to intervention components unlikely to have a positive impact emerged in the literature. First, there are issues around cultural adaptation and research. Second, there are issues around “outsiders” conducting research. A primary argument for culturally grounded and culturally tailored interventions is that not enough prevention science has been conducted among AI/AN youth in order to determine what the “active ingredients” are for successful adaptation of existing programs (Okamota 2014). For example, the Keepin’ it REAL program developed for Latino/a youth significantly reduced substance misuse in this population. However, when the program was adapted for AI/AN youth, it produced iatrogenic effects (Dixon et al 2007).
Another important point is that researchers that do not have a connection to the community and are developing interventions for SUDs, although well intentioned, may unknowingly implement projects that are harmful and erode the cultural values and protective factors that they are seeking to enhance (Whitebeck 2012). An example of this is teaching life skills from an individualistic perspective without realizing the value of being community-minded. Think Smart addressed inhalant use of the individual youth and was successful in decreasing use but only of this substance. Later research determined that the reason inhalant use was the only substance that decreased was not related to the program but to the fact that the community had increased their perception of risk and altered its behavior towards allowing children access to certain products. This intervention had failed to take the community into consideration and account for possible contamination effects that would have been more evident had it involved the community to a great degree (Whitebeck 2012).
Treatments Offering Promising Solutions
Project Venture. This program is an experiential outdoor program for middle school youth to address substance use and other health-risk behaviors. Project Venture is school-based as well as including extended outdoor activities, such as backpacking and wilderness trips that incorporate American Indian values and culture. One study found that it positively impacted mental health variables, but not substance use Carter, S. L., Beadnell, B., & Vanslyke, J. (2015). Evaluation of Project Venture: Findings from an innovative positive youth development program for American Indian youth. National Indian Youth Leadership Project.. It is listed on the NREPP website as promising (NREPP website; Okamoto et al., 2014).
Red Cliff Wellness. Developed in partnership with the Red Cliff Band of Lake Superior Chippewa, this is a school-based intervention for youth grades K-12 targeting substance use disorder and prevention. This program has been implemented in more than 185 schools over the last 30 years. One study found the curriculum significantly decreased the rate of alcohol use during the intervention and at one year follow-up (NREPP website).
GONA. Gathering of Native Americans is categorized as an EBP that was developed in the 1990s through SAMHSA’s Center for Prevention for Substance Abuse as a program to reduce alcohol and substance use in American Indian communities, including a component for youth. The curriculum emphasizes healing trauma and revitalizing traditional values, practices and traditions in a conference format over a four-day period. Although GONA is one of the most widely implemented prevention and interventions in American Indian communities (partially due to past funding support from SAMHSA and IHS), it is not listed on EBP registries due to the lack of published outcome research (Nebelkopf et al., 2011; Wright et al., 2011).
Healing of the Canoe. This intervention focuses on the cultural values and traditions of three Pacific Northwest tribes in addressing adolescent drug and alcohol misuse. The curriculum was adapted to three coastal tribal communities using CBPR and was delivered in-school as well as in a workshop format. Results indicated that the intervention was effective in reducing alcohol use through the school year in addition to increasing cultural protective factors of hope, self-efficacy and optimism. Workshop participation resulted in the increase of the previously mentioned protective factors, as well as higher cultural identity and engagement, knowledge about alcohol and drugs, and lower levels of substance use for those who participated compared with youth who did not. Although the sample size was small and the study was not an RCT, the findings are promising and add to the body of literature demonstrating the effectiveness of CBPR in developing and implementing a culturally grounded substance use intervention for American Indian youth (Donovan et al., 2015).
The Alaska People Awakening. This intervention is a great example of quasi-experimental designs being applied to smaller communities and using hierarchical linear modeling. The cultural intervention targets suicide and co-occurring alcohol and substance use among both adults and adolescents (Allen 2009 suicide prevention as a com dev process; 2016).
Cherokee Talking Circles. Similar to Healing of the Canoe, CTC was developed through CBPR. Initial program development process included identifying cultural protective factors. Self-reliance was chosen as it incorporates being responsible, disciplined and confident, integral to positive youth development among the Cherokee. After completing the multi-phase development process, the program was tested through a quasi-experimental design. CTC was compared to Be a Winner/Drug Abuse Resistance Education among Keetoowah-Cherokee high school students who had already initiated substance use. Students met once a week for 10-weeks and talking circles were led by a counselor and a community cultural leader. Findings indicate that the program reduced substance use and related problems compared to controls. Importantly, this effect persisted at 3-month post-intervention follow-up (Lowe 2012).
Promoting Social Competence and Resilience of Native Hawaiian Youth. Focused on rural Native Hawaiian youth, this program addresses substance use prevention through resistance strategies. The program is still undergoing implementation and evaluation, but initial findings have been promising in gaining community and youth engagement, input, and collaboration in the development of materials. This includes scripts, videos and other program resources identified by stakeholders as important components of alcohol and drug prevention and intervention in their community (Okamoto 2014).
Walking On. The Walking On curriculum was developed through a community-university partnership based in CBPR. Through this collaboration AI/AN adolescent’s with histories of trauma and substance misuse were identified as lacking strength-based services that could be delivered in an out-patient setting, especially after returning from residential treatment. In addition to being grounded in Cherokee culture, the curriculum is delivered once a week by clinicians in a group setting with youth and families, and incorporates evidence-based practices of cognitive-behavioral therapy, contingency management and motivational interviewing (Novins et al., 2012). Although still in the development and testing phase, this program offers another example of the potential success in combining culturally-tailored treatment methods with EBPs.
Practice-based and Community-defined
White Bison/Wellbriety. White Bison was founded in the 1980s to merge cultural spiritual teachings, teachings of the medicine wheel and standard AA. The process includes the sacred hoop as symbolism for healing, hope unity and forgiveness with curricula specifically designed for youth. In addition to talking circles, participants mind map, learn the four laws and the healing forest model, and are encouraged to adapt the program to the local traditions and community (Coyhis & Simonelli, 2008; Kenyon 2012). Much of the evidence of effectiveness comes from the wide-ranging utilization of the Wellbriety practice. This includes the community use of the books, manuals and DVDs developed through White Bison in addition to the participation and training of thousands of individuals (Coyhis 2008). Two treatment facilities are now certified to deliver Wellbriety services: Volunteers of America Northern Rockies and the Native American Rehabilitation Association (NARA), an American Indian owned and operated residential treatment facility located in Portland, Oregon.
Cultural Practices. In many tribal communities there is an agreement that “culture is treatment.” Consistently AI/AN people believe that it is a return to traditional practices and ceremony that will ultimately heal and lead to long-term recovery (Gone, 2011). This approach usually includes talking circles, smudging by burning sage, cedar or sweetgrass, drumming, sweat lodges, ceremony, activities related to connection to the land, and medicines/traditional foods as well as equine therapy (Rowan et al, 2014).Structured talking circles have been successfully used in treatment and prevention efforts in tribal communities (Wright et al., 2011; Coyhis 2008).
Another culturally grounded approach to healing and ceremony are sweat lodges. Sweat lodges have shown to significantly reduce alcohol intake among AI/AN men recovering from alcohol misuse in the prison system (Gossage et al., 2003). Among youth specifically, promising results in treating addiction were observed when sweat lodges, singing, drumming, storytelling, art, elders, cultural teachings about tribal history, fasting, ceremonial feasts, natural and traditional medicines, and equine therapy were incorporated into treatment (Rowan et al 2014; Boyd-Ball 2011; Dell 2011a; Dell 2011b; Boyd-ball 2003).
Maintaining Cultural Protective Factors while Developing EBPs
There is still a lot of research that is needed in order to determine how to successfully maintain cultural protective factors while integrating these into evidence-based interventions. AI/AN scholars have made suggestions on how this process could be improved. Since there is considerable heterogeneity among tribes some have argued for an approach that is developed nation by nation instead of assuming a “pan-Indian” prevention method. The Oregon Tribal Best Practices initiative is a tribal identified list of best practices that emphasizes innovative ways to evaluate interventions with more evidence created through practice-based and community-defined approaches (Whitebeck 2012).
A real tension exits between the appropriateness of adaptation vs. a culturally grounded method (Okamoto 2014). Most researchers argue that a program does not have to be one or the other—evidence-based or culturally grounded and that the two are not mutually exclusive. A collaborative approach is necessary (Whitebeck 2012). As described throughout this response, there are opportunities for implementing Western-based methods which integrate a strengths-based approach and is community-engaged. The literature does seems to agree that the best way forward boils down to community determined need and continued research (CITE).
Adachi-Mejia, A., Carlos, H, Berke, EM, Tanski SE, Sargent, JD. (2012). A comparison of individual versus community influences on youth smoking behaviours: a cross-sectional observational study. BMJ Open. doi:doi:10.1136/bmjopen-2011-000767
Alaska Division of Public Health, Alaska Youth Risk Behavior Surveillance System, 2016. (2016). Retrieved from
Beals, J., Piasecki, J, Nelson, S, Jones, M, Keane, E, Dauphinais, P, Red Shirt, R, Sack W, Manson SM. (1997). Psychiatric disorder among American Indian adolescents: Prevalence in Northern Plains youth. Journal of the American Academy of Child and Adolescent Psychiatry, 36(9), 1252-1259.
Beals, J., Spicer, P, Mitchell, CM, Novins, DK, Manson, SM, the AI-SUPERPFP Team. (2003). Racial Disparities in Alcohol Use: Comparison of 2 American Indian Reservation Populations With National Data. American Journal of Public Health, 93(10), 1683–1685.
Beauvais F, T. P., Burnside M, Plested B. (2007). Prevalence of American Indian adolescent tobacco use: 1993-2004. Substance Use & Misuse, 42(4), 591-601.
Berkel, C., Mauricio, A. M., Schoenfelder, E., & Sandler, I. N. (2011). Putting the pieces together: An integrated model of program implementation. Prevention Science, 12(1), 23-33.
Blake, I., Holck, P, Provost, E. (2016). Alaska Native Mortality Update: 2009-2013. Retrieved from
Bopp, M., Saunders, R. P., & Lattimore, D. (2013). The Tug-of-War: Fidelity versus adaptation throughout the health promotion program life cycle. The Journal of Primary Prevention, 34(3), 193-207.
Boyd-Ball, A., Manson, SM, Noonan, C, Beals, J. (2006). Traumatic events and alcohol use disorders among American Indian adolescents and young adults. Journal of Traumatic Stress, 19(6), 937-947.
Brown, L. D., Feinberg, M. E., & Greenberg, M. T. (2010). Determinants of community coalition ability to support evidence-based programs. Prevention Science, 11(3), 11, 287-297.
Butterfoss, F. D. (Coalitions and partnerships in community health.). 2007. San Francisco, CA: John Wiley & Sons, Inc.
Catalano, R., Fagan, AA, Gavin, LE, Greenberg, MT, Irwin Jr, CE, & Ross, D., Shek, DT. (2012). Worldwide application of prevention science in adolescent health. The Lancet, 379(9826), 1653-1664.
Chen, H. T. (2005). Practical program evaluation: Assessing and improving planning, implementations, and effectiveness. In: Sage.
Clarke, A. (2002). Social and emotional distress among American Indian and Alaska Native students: Research findings. Retrieved from Charlston WV:
Coie, J., Watt, NF, West, SG, Hawkins, JD, Asarnow, JR, Markman, HJ, Ramey, SL, Shure, MB, Long, B. (1993). The science of prevention: A conceptual framework and some directions for a national research program. American Psychologist, 48, 1013-1022.
Crowley, D. M., Hill, L. G., Kuklinski, M. R., & Jones, D. E. (2013). Research priorities for economic analyses of prevention: Current issues and future directions. Prevention Science, 1-10.
Current Cigarette Smoking Among Adults—United States, 2005-2015. (2016). Retrieved from
de Ravello L, E. J., S, Tulloch, S, Taylor, M, Doshi, S. (2014). Substance use and sexual risk behaviors among american Indian and alaska native high school students. Journal of School Health, 84(1), 25-32.
Deters, P., Novins, DK, Fickenscher, A, Beals, J. (2006). Trauma and posttraumatic stress disorder symptomatology: patterns among American Indian adolescents in substance abuse treatment. American Journal of Orthopsychiatry, 76(3), 335-345.
Dickerson, D., Brown, RA, Johnson, CL, Schweigman, K, D’Amico, E.J. (2015). Integrating motivational interviewing and traditional practices to address alcohol and drug use among urban American Indian/Alaska Native youth. Journal of Substance Abuse Treatment, 65, 26-35.
Drolet, B. C., & Lorenzi, N. M. (2011). Translational research: understanding the continuum from bench to beside. Translational Research, 157(1), 1-5.
Duran, E. (2006). Historical Trauma: Treating the Soul Wound, The Spirit of Alcohol: Treating Addiction, Community Intervention. In Healing the Soul Wound: Counseling with American Indian and other Native Peoples.: Teachers College Press.
Duran, E., Duran B. (1995). Native American Postcolonial Psychology. Albany, New York: State Univeristy of New York Press.
Elliot, D. S., & Mihalic, S. (2004). Issues in disseminating and replicating effective prevention programs. Prevention Science., 5(1), 47-52.
Glasgow, R. R., Vogt, T. M., & Boles, S. M. (1999). Evaluating the public health impact of health promotion interventions: The RE-AIM framework. American Journal of Public Health, 89(9), 1322-1327.
Gone, J., Trimble, JE. (2012). American Indian and Alaska Native mental health: Diverse perspectives on enduring disparities. Annual Review of Clinical Psychology., 8, 131-160.
Gone, J. P., Trimble, J.E. (2012). American Indian and Alaska Native mental health: diverse perspectives on enduring disparities. Annual Review Clinical Psychology, 8, 131-160.
Goodkind, J., LaNoue, M, Lee, C, Freeland, L, Freund, R. (2012). Feasibility, acceptability, and initial findings from a community-based cultural mental health intervention for American Indian youth and their families. Journal of Community Psychology, 40(4), 381-405.
Gottfredson, D. C., Cook, T. D., Gardner, F. E. M., Gorman-Smith, D., Howe, G. W., Sandler, I. N., & Zafft, K. M. (2015). Standards of evidence for efficacy, effectiveness, and scale-up research in prevention science: Next generation. Prevention Science, 16(7), 893-926.
Hawkins, J., Catalano, RF, Miller, JY. (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, 112(1), 64-105.
Henderson, P., Jacobsen, C, Beals, J., & the AI-SUPERPFP Team. (2005). Correlates of cigarette smoking among selected Southwest and Northern Plains Tribal Groups:
The AI-SUPERPFP study. American Journal of Public Health, 95(867-872).
Holder, H. (2010). Prevention programs in the 21st century: What we do not discuss in public. Addiction, 105, 578-581.
Johnston, L., O’Malley, PM, Bachman, JG, Schulenberg, JE (2009). Monitoring the Future National Survey Results on Drug Use, 1975–2008. . Retrieved from
Johnston, L., O’Malley, PM, Miech, RA, Bachman, JG, & Schulenberg, JE. (2016). Monitoring the Future national survey results on drug use, 1975-2015: Overview, key findings on adolescent drug use. . Retrieved from
Kellam, S. G., & Langevin, D. J. (2003). A framework for understanding “evidence” in prevention research and programs. Prevention Science, 4(3), 137-153.
Koh, H. K., Piotrowski, J. J., Kumanyika, S., & Fielding, J. E. (2011). Healthy People: A 2020 vision for the social determinants approach. Health Education & Behavior, 38(6), 551-557.
Kraemer, H., Kazdin, AE, Offord, DR, Kessler, RC, Jensen, PS, Kupfer, DJ. (1997). Coming to terms with the terms of risk. Archives of General Psychiatry, 54(4), 337-343.
Mail, P. (1995). Early modeling of drinking behavior by Native American elementary school children playing International Journal of Addiction, 30(9), 1187-1197.
Meyers, D. C., Durlak, J. A., & Wandersman, A. (2012). The quality implementation framework: A synthesis of critical steps in the implementation process. American Journal of Community Psychology, 50(3-4), 462-480.
Miller, J., Naimi, TS, Brewer, RD, Jones, SE. (2007). Binge drinking and associated health risk behaviors among high school students. Pediatrics, 119(1), 76-85.
Miller, K., Stanley, LR, Beauvais, F. (2012). Regional differences in drug use rates among American Indian youth. Drug and Alcohol Dependence, 126(1-2), 35–41.
Murphy, S. M., McDonell, M. G., Mcpherson, S., Srebnik, D., Angelo, F., Roll, J. M., & Ries, R. K. (2015). An economic evaluation of a contingency-management interventions for stimulant use among community mental health patients with serious mental illness. Drug and Alcohol Dependence, 153, 293-299.
O’Loughlin, J., O’Loughlin, EK, Wellman, JR, Sylvestre, MP, Dugas, EN, Chagnon, M, Dutczak, H, Laguë, J, McGrath, JJ. (2017). Predictors of cigarette smoking initiation in early, middle, and late adolescence. Journal of Adolescent Health 1-8.
Rabin, B. A., & Brownson, R. C. (2012). Developing the terminology for dissemination and implementation research. In R. C. Brownson, G. A. Colditz, & E. K. Proctor (Eds.), Dissemination and implementation research in health. New York, NY: Oxford.
Reducing Risks, Promoting Healthy Life. (2002). Retrieved from
Redwood, D., Lanier, AP, Renner, C, Smith, J, Tom-Orme, L, & Slattery, ML. (2010). Differences in cigarette and smokeless tobacco use among American Indian and Alaska Native people living in Alaska and the Southwest United States. Nicotine & Tobacco Research, 12(7), 791-796.
Rehm, J., Mathers C, Popova S, Thavorncharoensap M, Teerawattananon Y, Patra J. (2009). Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. Lancet, 9682(373), 2223-2233.
Romano, E., Lacey, RC. (2010). Alcohol and Highway Safety: Special Report on Race/Ethnicity and Impaired Driving Retrieved from
Romer, D. (2010). Adolescent risk taking, impulsivity, and brain development: Implications for prevention. Developmental Psychobiology, 52(3), 263-276.
Rutter, M. (2009). Understanding and testing risk mechanisms for mental disorders. Child Psychology and Psychiatry, 50(1-2), 44-52.
Sanigorski, A. M., Bell, A. C., Kremer, P. J., Cuttler, R., & Swinburn, B. A. (2008). Reducing unhealthy weight gain in children through community capacity-building: results of a quasi-experimental intervention program. Be Active Eat Well. International Journal of Obesity., 32(1060-1067).
Skewes MC, L. J. (2016). Sobriety and alcohol use among rural Alaska Native elders. International Journal Circumpolar Health, 4.
Smokowski, P., Evans, CB, Cotter, KL, Webber, KC. (2014). Ethnic identity and mental health in American Indian youth: examining mediation pathways through self-esteem, and future optimism. Journal of Youth and Adolesence, 43(3), 343-355.
Stanley, L., Randall, SDH, Swaim, C, Beauvais, F. (2014). Rates of substance use of American Indian students in 8th, 10th, and 12th grades living on or near reservations: Update, 2009-2012. Public Health Reports, 129, 156-163.
Stanley LR, S. R. (2015). Initiation of Alcohol, Marijuana, and Inhalant Use by American-Indian and White Youth Living On or Near Reservations. Drug and Alcohol Dependence, 155, 90-96.
Steverman, S., & Shern, D. (2014). Prevention mental, emotional, and behavioral disorders: Financing and implementation strategies. Alexandria, VA: Mental Health America.
Szlemko, W., Wood, JW, Thurman, PJ. (2006). Native Americans and alcohol: past, present, and future. Journal of General Psychology, 133(4), 435-451.
The Surgeon General’s Call to Action To Prevent and Reduce Underage Drinking. (2007). Retrieved from
Tobacco Use Among U.S. Racial/Ethnic Minority Groups—African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon General. (1998). Retrieved from Atlanta, Georgia: U.S.:
Tobacco, alcohol, and other drug use among high school students in Bureau of Indian Affairs-funded schools–United States, 2001. (2003). MMWR Morb Mortal Wkly Rep, 52(44), 1070-1072.
Tragesser, S., Beauvais, F, Burnside, M, Jumper-Thurman, P. (2010). Differences in illicit drug-use rates among Oklahoma and non-Oklahoma Indian youth. Substance Use & Misuse, 45(13), 2323-2339.
Trends in Indian Health 2014 Edition. (2014). Retrieved from
Wallerstein, N. B., Yen, I. H., & Syme, S. L. (2011). Integration of social epidemiology and community-engaged interventions to improve health quality. American Journal of Public Health, 101(5), 822-830.
Wandersman, A. (2003). Community science: Bridging the gap between science and practice with community-centered models. American Journal of Community Psychology, 31(3-4), 227-242.
Whitbeck, L., Yu, M, Johnson, KD, Hoyt, DR, Walls, ML. (2008). Diagnostic prevalence rates from early to mid-adolescence among indigenous adolescents: first results from a longitudinal study. Journal of the American Academy of Child and Adolescent Psychiatry, 47(8), 890-900.
Whitesell, N., Beals, J, Mitchell, CM, Keane, EM, Spicer, P, Turner, RJ, AI-SUPERPFP Team. (2007). The relationship of cumulative and proximal adversity to onset of substance dependence symptoms in two American Indian communities. Drug and Alcohol Dependence, 91(2-3), 279-288.
Whitesell, N. R., Kaufman, c. E., Keane, E. M., Crow, C. B., Shangreau, C., & Mitchell, C. M. (2012). Patterns of substance use initiation among young adolescents in a Northern Plains American Indian tribe. American Journal of Drug and Alcohol Abuse, 38(5), 383-388.
Youth Risk Behavior Surveillance —United States, 2015. (2016). Retrieved from
Yu, M., & Stiffman, A. R. (2007). culture and environment as predictors of alcohol abuse/dependence symptoms in American Indian youths. Addictive Behaviors, 32, 2253-2259.