TOBACCO USE AMONG YOUNG ADOLESCENTS IN URBAN INDIA. Melissa Stigler1, Cheryl Perry1, Monika Arora2, Srinath Reddy2, 1University of Minnesota, Minneapolis, MN United States; 2HRIDAY-SHAN, New Delhi, India
Deaths due to tobacco are expected to rise dramatically in many developing nations of the world in the next two decades. During this time, India will experience the highest rate of increase among all regions in the world – the proportion of all deaths related to tobacco will rise from 1.4% (in 1990) to 13.3% (in 2020), according to the World Health Organization. Tobacco is consumed in a variety of forms throughout India. Currently, other forms of tobacco (e.g., gutkha, bidis) are more prevalent than cigarette smoking. Cigarette smoking is starting to increase, especially in urban India, as income rises and a more Western lifestyle is embraced. As in other parts of the world, tobacco use often begins during adolescence and early adulthood in India. Comparatively little is known, however, about which factors may influence the onset of tobacco use in this population. Using data from a large-scale cross-sectional survey, this presentation will explore the potential relationship between a number of etiologic factors and current tobacco use by young students in two cities in India (Delhi, Chennai). The student survey is currently being conducted as a baseline evaluation tool for Project MYTRI, a group-randomized intervention trial designed to prevent the onset of tobacco use among students in grades 6, 7, 8, and 9 in government (lower SES) and private (higher SES) schools in urban India. Thirty-two schools and approximately 15,000 students in grades 6 and 8 are expected to participate in the survey, which was piloted extensively and administered in different languages (English, Hindi, Tamil). Relevant outcome measures include current tobacco use and intentions to use tobacco, for tobacco that is smoked (e.g., bidis, cigarettes) and tobacco that is chewed (e.g., gutkha). Potential correlates include normative estimates, normative beliefs, normative expectations, outcome expectations, functional meanings, knowledge about health consequences of tobacco, refusal skills efficacy, parent-child communication about tobacco, perceived access, and exposure to tobacco advertising. Relationships will be examined using mixed-effects regression models appropriate for clustered data like these. The focus will be on determining which factors account for the most variability in – and are therefore most strongly associated with – tobacco use by these young adolescents in urban India. Analyses will be stratified by gender and SES to explore whether relationships differ by these important demographic variables. Implications of the results of these analyses for intervention programs will be discussed.
CROSS-CULTURAL DIFFERENCES IN THE DSM-IV CONSTRUCT OF ALCOHOL ABUSE: ASSESSING MEASUREMENT BIAS ACROSS HISPANIC AND CAUCASIAN ADOLESCENTS. Adam Carle1, 1U.S. Census Bureau, Washington, DC United States
The possibility exists that assessment instruments are biased and differentially valid and/or reliable across multiple populations. Measurement bias, a form of non-sampling error, occurs when individuals identical on a construct being measured can be expected to have different observed scores as a function of group membership. Measurement invariance holds when individuals equivalent on the construct, but from different populations, have the same probability of achieving a given observed score on the instrument. When measurement bias is present, it is difficult, if not impossible, to interpret group differences on the construct being measured. Recent years have seen a call for model based, empirical methods to address the validity of measurement instruments across diverse populations. Latent variable models, such as confirmatory factor analysis for ordered-categorical measures (CFA-OCM) and others, are relatively recent entries in the research methods field and are a powerful tool for investigating bias. Recent years have seen increased evidence for the internal validity of the DSM-IV alcohol abuse construct in the general population. However, despite reported differences in the prevalence of alcohol disorders, less attention has been given to cross-cultural differences in the construct across Hispanics and non-Hispanic Caucasians. Likewise, little attention has been paid to the validity of the construct in adolescent populations. Using data from the National Longitudinal Alcohol Epidemiological Survey (NLAES), a nationally representative household survey of 42,692 adults, the current paper addresses these issues and presents analyses exploring violations of measurement invariance on a standardized of alcohol abuse across Hispanic and non-Hispanic Caucasian adolescents. The results of CFA-OCM describe the presence and extent of measurement bias across these groups, discuss its role in clinical diagnoses, address the degree to which group differences on the observed scores can be validly interpreted, and demonstrate the importance of establishing measurement invariance prior to making epidemiological estimates.
PHYSICAL MEDICINE AND REHABILITATION PHYSICIAN´S ENGAGEMENT IN SUBSTANCE ABUSE ASSESSMENT AND PREVENTION. Steven West1, Richard Luck1, Fred Capps1, 1Virginia Commonwealth University, Richmond, VA United States
The rates of substance misuse and addiction by persons with co-occurring disabilities are extreme across most disability categories, and are particularly high for those persons with traumatically acquired disabilities. For example, in the cases of both spinal cord injury (SCI) and traumatic brain injury (TBI), the rates of co-occurring substance abuse range from 50- to 60% (Corrigan, 1995; Drubach, Kelly, Winslow, & Flynn, 1993; Heinemann, Doll, & Schnoll, 1989; Kolakowsky-Hayner, et al., 2002; McKinley, Kolakowsky, & Kreutzer, 1999; Taylor, Kreutzer, Demm, & Meade, 2003). In many cases, the misuse of alcohol and other drugs pre-dates the disabling condition with as much as 60% of TBI and SCI cases being acquired while the individual is intoxicated (Corrigan, 1995; Radnitz & Tirch, 1995). Intervening to promote reduced use or abstinence has been found to be difficult due to a lack of facilities capable of serving clients with such medically complicated conditions. As such conditions are acquired under circumstances that typically lead to increased contact with healthcare providers (typically those associated with physical medicine and rehabilitation [PM&R] departments), facilitating the ability of such providers to intervene early in the process could aid substantially to the prevention and amelioration of substance use concerns. This research was conducted to determine the current practices of PM&R physicians regarding substance abuse prevention efforts with patients with traumatically acquired disabilities. A survey was sent to 683 Board Certified PM&R physicians and indicated that although awareness of substance misuse was common, the use of standardized assessments, prevention activities, and intervention/referral services were not. Such physicians appear to be an untapped source of prevention information to this historically overlooked population.
EARY LIFE AND CORRENT CORRELATES OF RESILIENT COPING AMONG DRUG USERS IN RECOVERY. Felipe Castro1, Rebeca Rios1, Diana Naranjo1, 1Arizona State University, Tempe, AZ United States
Problem: Resilience involves good outcomes in spite of serious threats to adaptation or development (Masten, 2001). “Ego-resilience,” has been described as involving four important dimensions: (1) confident optimism, (2) productive activity, (3) insight and warmth, and (4) skilled expressiveness (Klohnen, 1996). Ego-resiliency enables individuals to “flexibly modify their impulse expression and thus adaptively respond to environmental contingencies and shape them in accordance with their personal goals and desires,” (Klohnen, 1996). Methods: A sample of 40 heavy drug users of Hispanic background and 40 of mainstream American background have been interviewed with an open-ended “platica” approach entitled: “Your Life´s Journey: Who You Are, and Where You are Going. This interview has two stages, a 20 minute qualitative platica interview, and a 2 hour structured interview that examines to examine the process of personal growth and adaptation across important life milestones. This study will examine the item: the most “Difficult Problem” in the past 5 years. This consists of: (1) identification of this most “Difficult Problem” (2) when this happened, (3) a time sequence of what happened, (4) how stressful this was, and (5) what the client did in temporal sequence analysis of their thoughts and feelings. These narratives will be coded for: (1) thematic content, and (2) emotional/effective tone in a format described by Denne, Castro, & Harris (2001). Results: Levels of resilience as indicated by ratings of capacity to recover from this most difficult problem are being examined and coded for analysis within a regression analyses of scores from qualitatively derived thematic categories, and with other scaled scores including: the CES-D scale scores, levels of illicit drug use including poly-drug use, ways of coping, traditionalism, positive and negative machismo, bicultural orientation, self-efficacy in substance use, and health-related behaviors. Discussion: The correlates of early-life and contemporary factors of resilient responses to the major life problem will be examined to generate new hypotheses for a larger future study.
Denne, R., Castro, F. G., & Harris, T. (2001, May). Antecedents and consequences of relapse in stimulant users: Integrating qualitative and quantitative analyses. Paper presented at the Society for Prevention Research 9th Annual Meeting, Washington, DC.
Klohnen, E. C. (1996). Conceptual analysis and measurement of the construct of ego-resiliency. Journal of Personality and Social Psychology, 70, 1067-1079.
Masten, A. S. (2001). Ordinary people: Resilience process in development. American Psychologist, 56, 227-238.
CONCURRENT 4, EPIDEMIOLOGY, Grouped papers
IMPLICATIONS FOR PREVENTION RISK AND PROTECTIVE FACTORS BETWEEN URBAN & RURAL ADOLESCENTS, FEMALE FORMER & CURRENT SMOKERS, SUBSTANCE ABUSE BY RURAL YOUTH
Chair: Hanno Petras
A COMPARISON OF RISK AND PROTECTIVE FACTORS BETWEEN URBAN AND RURAL ADOLESCENT FEMALE FORMER AND CURRENT SMOKERS. Peggy Meszaros1, Angela Huebner2, Randy Koch3, Hope Merrick4, 1Virginia Polytechnic Institute and State University, Blacksburg, VA United States; 2Virginia Polytechnic Institute and State University, Falls Church, VA United States; 3Virginia Commonwealth University, Richmond, VA United States; 4Department of Mental Health, Mental Retardation and Substance Abuse Services, Richmond, VA United States
This study examined female adolescent tobacco use based on data collected from 8th, 10th and 12th graders throughout the state using the Virginia Community Youth Survey. The survey assessed lifetime and past 30-day ATOD use, 25 risk factors and 10 protective factors. The hypotheses tested were: 1) Specific ecological factors can be identified which increase the odds of being a current smoker, and 2) These ecological factors will vary by urban versus rural environments. Research suggests environmental factors may be different for these populations (Shillington & Clapp, 2000; Horn et al., 1998). The sampling frame used a three-stage, stratified random sample design in order to provide a representative sample at the state and regional levels. The survey was administered to 147 randomly selected classrooms within 18 randomly selected school divisions in 10 catchment areas. From an initial sample of 3,330, there were 3,166 valid surveys, including a final study sample of 1,621 girls; 62% (n=986) classified as “urban” and 38% (n=612) classified as “rural.”
Logistic regression analyses revealed several factors which influenced the odds of being a current smoker for rural, adolescent females. Current rural adolescent female smokers were less likely than their former smoking counterparts to perceive community laws and norms as favorable toward drug use and to hold favorable attitudes toward antisocial behavior. They were more likely to hold favorable attitudes toward drug use, to have friends who use drugs and to perceive rewards for antisocial involvement than former smokers. These include community, peer and individual level factors.
The same logistic regression analysis was conducted to examine these factors for girls in urban areas. Current urban female adolescent smokers were less likely than former smokers to perceive high family conflict and less likely to perceive school rewards for prosocial involvement. They were more likely to perceive school opportunities for prosocial involvement, to hold favorable attitudes toward drug use and to perceive rewards for antisocial involvement than were their former smoking counterparts. These include school and individual factors.
Both urban and rural current female smokers report more favorable attitudes toward drug use and rewards for antisocial behavior than do their former smoking peers. School factors and family factors appear to be more salient for urban smokers, while community factors and peer factors were more salient for rural female smokers. These results suggest that intervention programs targeting female adolescent smokers may need to include a core set of basic interventions as well as some tailored to the specific environment. Specific recommendations will be elaborated.
THREATS TO EARLY CHILDHOOD HEALTH AND DEVELOPMENT IN RURAL AMERICA. Cathy Grace1, Martha Zaslow2, Brett Brown2, Elizabeth Shores3, 1National Center for Rural Early Childhood Learning Initiatives, Mississippi State, MS United States; 2Child Trends, Washington, DC United States; 3Mississippi State University, Little Rock, AR United States
The health of young children and their families directly affects children´s early learning potential. Health problems and lack of access to prevention and early intervention services can significantly reduce the young child´s prospects for school success as well as the family´s overall wellbeing.
Challenges to prevention and early intervention service delivery are significantly different for rural areas. Specialized services may not be available in small towns and rural communities. In addition, cultural and educational differences may deter rural parents from seeking appropriate prevention and early intervention services for their children. Reliable estimates of the prevalence of key risk factors, including indicators of child and family wellbeing and indicators of access to services and use of services, would allow public agencies to better target high-need rural areas with the most needed services and public health education.
However, there has been a gap in demographic knowledge about the health status, risk factors, and access to prevention and early intervention services for young children and their families in rural America. Recent reports by the Commonwealth Fund and Annie E. Casey Foundation have not distinguished between rural, suburban, and urban families. Moreover, it is not feasible to assess child and family wellbeing in rural areas using most major national datasets. Many datasets do not include variables for rural residence. Others contain rural sub-sets but those samples are too small to permit reliable estimates of risk factors or access to services for rural households.
The National Center for Rural Early Childhood Learning Initiatives and Child Trends analyzed several national datasets to determine the prevalence of key risk factors and the accessibility of prevention and early intervention services in rural areas. The datasets included the Early Childhood Longitudinal Study - Kindergarten Cohort (ECLS-K), and the Early Childhood Longitudinal Study - Infant Cohort (ECLS-B). The indicators included child wellbeing markers such as birthweight status, poverty, and mental health status of mothers. The indicators also included accessibility markers such as children´s enrollment in early childhood programs, food stamp receipt, health care coverage, and prenatal care.
The results have implications for public policy at local, state, and federal levels. For example, should the analysis show that rural kindergarteners have significantly lower expressive language skills, state child care resource and referral agencies might allocate more funds for professional development in fostering pre-literacy skills for rural child care workers.
PATTERNS OF SUBSTANCE USE BY RURAL YOUTH: IMPLICATIONS FOR PREVENTION. Ruth W. Edwards1, Beverly Marquart2, Pamela Jumper-Thurman2, Barbara Plested2, 1Colorado State University, Ft. Collins, CO United States; 2Tri-Ethnic Center for Prevention Research, Ft. Collins, CO United States
A nation-wide, stratified random sample of 54,471 7th-12th grade rural youth completed an adaptation of The American Drug and Alcohol Survey as part of a major study of community-level factors and how they may interact with substance use patterns of youth. In this poster, we will present the results of analyses by gender and grade of patterns of combinations of substances used by rural youth. For inclusion in the patterning analysis, substances must have been reported to be used by 5% or more of 7th-12th grade youth in their lifetime. Getting drunk, marijuana, inhalants, methamphetamines, amphetamines, psychedelics, cocaine, ecstasy, narcotics other than heroin, tranquilizers and tobacco all met this criterion. As expected, the predominant patterns were “no use”, “getting drunk only” and “getting drunk and marijuana”. Other patterns which emerged for males and females and younger and older youth are, however, of interest for prevention and interventions.
CONCURRENT 5, ECONOMIC AND COST-UTILITY ANALYSIS, Grouped papers
STRATEGIES FOR ECONOMIC ANALYSES OF SCHOOL-BASED, CLINICAL, AND COURT PREVENTION AND INTERVENTION PROGRAMS. WHAT ARE THE TRUE COSTS OF PREVENTION?
Chair: Alka Indurkhya
COLLECTION OF UNIFORM COST DATA FROM SCHOOL-BASED INTERVENTIONS. Pinka Chatterji1, Linda Juszczak2, Christine Caffray3, 1Cambridge Health Alliance/Harvard Medical School, Somerville, MA United States; 2Montefiore School Health Program, Bronx, NY United States; 3Children's Board of Hillsborough County, Tampa, FL United States
Although school-based prevention interventions have become increasingly common over the past two decades, we lack good data on how much these programs cost. This problem limits our ability to evaluate the cost-effectiveness of school-based programs relative to alternative methods of providing services to children. The over-arching goal of the session is to encourage and inform future economic evaluations of school-based prevention programs. The objectives are the following: (1) to outline economic cost analysis methods, with emphasis on their application to school-based programs; and (2) to describe an ongoing effort to develop a standardized cost instrument that can systematically collect uniform cost data from programs on a national basis. The cost analysis methods covered in the session include defining the study´s perspective, and collecting, analyzing, and interpreting cost data. By using these methods, an estimate of the true economic cost of a school-based intervention can be calculated. The discussion of the cost instrument project focuses on how the instrument was designed and tested, the challenges involved in this process, and practical pitfalls in assessing the costs of school-based programs.
IS THE DRUG COURT MODEL EXPORTABLE? AN EXAMINATION OF THE. Rosalie Pacula1, Christine Eibner2, John MacDonald1, Andrew Morral2, 1RAND Corporation, Santa Monica, CA United States; 2RAND Corporation, Arlington, VA United States
Despite significant declines in alcohol-related fatalities over the past 20 years, alcohol continues to be a contributing factor in 40% of U.S. traffic fatalities (NHTSA, 2005), the leading cause of death among U.S. residents between the ages of 4 and 33 (Subramanian, 2005). A disproportionate number of these fatalities are caused by drivers previously convicted of driving under the influence (DUI) of intoxicants (Fell, 1990). In California (Peck et al., 1994), as elsewhere in the U.S. (Fell, 1994; Yu & Williford, 1991), approximately 35% of all DUI convictions are for drivers with prior DUI convictions in the past 5-7 years. Prior driving record and severity of alcohol problems constitute the two most important dimensions underlying DUI recidivism (Peck et al 1994). There is an expectation within many criminal justice communities that specialized therapeutic DUI courts may be effective at reducing DUI recidivism because court supervised abstinence and rehabilitation program participation may reduce alcohol problems and therefore reduce future drinking-driving incidents, DUI offenses, and alcohol-involved accidents. In this paper we examine the effectiveness and cost-effectiveness of an experimental DUI court at reducing problem drinking and DUI recidivism among a sample of repeat offenders in Los Angeles County. Approximately 285 volunteers who plead guilty to a 2nd or 3rd time DUI offence between March 2000 and December 2002 were randomly assigned to either specialized therapeutic DUI court supervision or standard sanctions, usually administered by the same judge. Baseline and follow-up interviews conducted two years after assignment are used to evaluate the impact of the experimental condition on self-reported drinking and driving behaviors. Official records are also examined. Preliminary findings suggest that although significant differences in problem drinking behaviors and recidivism do not exist across the two conditions, the DUI court intervention was still more cost effective than standard sanctions in accomplishing these two objectives.
COST-EFFECTIVENESS OF AN INTERVENTION TO PREVENT DEPRESSION IN AT-RISK TEENS. Frances Lynch1, Mark Hornbrook1, Greg Clarke1, Michael Polen1, Elizabeth O'Connor1, John Dickerson1, Nancy Perrin2, 1Kaiser Permanente Division of Research, Portland, OR United States; 2Oregon Health & Science University, Portland, OR United States
Background: Depression is common in adolescent offspring of depressed parents and can be prevented, but adoption of prevention programs is dependent on the balance of incremental costs and benefits associated with these interventions. We examine the incremental cost-effectiveness of a group cognitive behavioral intervention to prevent depression in adolescent offspring of depressed parents. Methods: We used data from a recent randomized controlled trial conducted to prevent depressive episodes in at-risk offspring (13 to 18 years old) of adults treated for depression in a health maintenance organization. Teens were randomized to either usual care (n=49) or usual care plus a 15-session group cognitive therapy prevention program (n=45). Clinical outcomes were converted to measures of depression-free-days and quality adjusted life years. Costs were measured in an ongoing basis during the trial. Total health maintenance organization costs, costs of services received in other sectors (e.g., schools), and family costs were combined with clinical outcomes to conduct an incremental cost-effectiveness analysis comparing the intervention with usual care for one year following the intervention.
Results: Average cost of the intervention was $1632, total direct and indirect costs increased $610 in the intervention group, however the result was not statistically significant, suggesting a possible cost-offset. Estimated incremental cost per depression free day in the base-case analysis was $10 (95% CI: -13 to 52) or $9275 per quality adjusted life year (95% CI: -12,148 to 45,641).
Conclusion: Societal cost-effectiveness of a brief prevention program to reduce the risk of depression in offspring of depressed parents is comparable with that of accepted depression treatments, and cost-effective in comparison to other health interventions commonly covered in most insurance contracts.