Embedding health equity into preventive science research and recommendations

Melissa Simon, MD, MPH
Northwestern University

ABSTRACT:
Achievement of health equity in preventive care and in prevention science requires critical examination of multiple domains across which structural racism and systems of oppression operate. Design thinking and implementation science are two key areas in which health equity can and should be centered to better design population level recommendation and guideline statements along with operationalizing and implementing such statements across populations. Dr. Simon will discuss these issues and give examples from the US Preventive Service Task Force (USPSTF) Recommendations and her own work.

Automatic text mining methods for systematic, structural, integrative, and meta-analytic reviews: Three examples

Carlos Gallo, Ph.D.
Northwestern University

J.D. Smith, Ph.D.
Northwestern University

Gracelyn Cruden, M.A.
University of North Carolina University-Chapel Hill

Comprehensive specialty care versus usual community care for first-episode psychosis: 2-year outcomes from the NIMH RAISE Early Treatment Program

John M. Kane, M.D.
Northwell Health

ABSTRACT:

The National Institute of Mental Health (NIMH) issued a Request for Proposals entitled “Recovery After an Initial Schizophrenia Episode (RAISE)” in November 2008. The goal of the NIMH initiative is to change the trajectory and prognosis of first episode psychosis (FEP). The premise is that by combining state-of-the-art pharmacologic and psychosocial treatments in a patient-centric fashion and having them delivered by a well-trained and coordinated, multidisciplinary team, the functional outcome and quality of life for first episode patients treated in the community can be significantly improved. The specified aims of RAISE are, first, to develop a comprehensive and integrated intervention designed to: promote symptomatic recovery; minimize disability; maximize social academic and vocational functioning; be capable of being delivered in real world settings utilizing current funding mechanisms, and, second, to assess the overall clinical impact and cost effectiveness of the intervention as compared to currently prevailing treatment approaches and to conduct the comparison in non-academic, real world community treatment settings in the U.S

We developed a treatment model (NAVIGATE) and training program based on extensive literature review and expert consultation. Our primary aim is to compare the experimental intervention to “usual care” on quality of life. Secondary aims include comparisons on remission, recovery and cost effectiveness. Patients 15-40 years old with a first episode of schizophrenia; schizoaffective disorder; schizophreniform disorder, psychotic disorder NOS, or brief psychotic disorder according to DSM IV and no more than six months of antipsychotic medications were eligible. Patients are followed for a minimum of two years, with major assessments conducted by blinded, centralized raters using live, two-way video. We selected 34 clinical sites in 21 states and utilized cluster randomization to assign 17 to the experimental treatment-NAVIGATE and 17 to usual care. Enrollment began in July 2010 and ended in July 2012 with 404 subjects. 

Four hundred and four individuals participated in the controlled trial. The mean age was 23. The median duration of untreated psychosis was 74 weeks. Those patients who received the comprehensive care intervention –NAVIGATE- were more likely to stay in treatment, had significantly greater improvement in quality of life, psychopathology and participation in work or school. Duration of untreated psychosis was a significant mediator

Evolving a society that nurtures the wellbeing of every member

Anthony Biglan, Ph.D.
Oregon Research Institute

ABSTRACT:
Viewed broadly, the ultimate goal of prevention science is to assist societies in ensuring the social, psychological, and physical wellbeing of every member of society.  To fully realize the aspirations of prevention scientists, we need a broad strategy for translating knowledge about human wellbeing into population-wide improvements. Far-reaching changes must occur in the delivery of programs, the making of policies, and values and cultural practices that affect the quality of family, school, workplace, and community environments. In this talk, I will describe a broad cultural movement, not unlike the tobacco control movement, that can make nurturing environments a fundamental priority of public policy and daily life, thereby enhancing human wellbeing far beyond anything achieved thus far.

Unleashing the power of prevention: A plan to advance effective policy and practice to prevent behavioral health problems

David Hawkins, Ph.D.
University of Washington

ABSTRACT:
In the last 35 years, prevention scientists have made remarkable progress in developing and testing effective interventions for preventing behavioral health problems in children and adolescents. The Blueprints for Healthy Youth Development website at University of Colorado lists over 60 efficacious preventive interventions.  Yet effective prevention policies and programs are still not widely used.  Dr. Hawkins will present an overview of Unleashing the Power of Prevention, an action plan for increasing the use of tested and effective preventive interventions and thereby reducing child and adolescent behavioral health problems in the US by 20% in a decade.  The presentation will include a discussion of Encouraging Healthy Parenting through Primary Care, an example of how we can unleash the power of prevention nationwide.  Unleashing the Power of Prevention is published as a discussion paper by the National Academy of Medicine and is available at
http://nam.edu/perspectives-2015-unleashing-the-power-of-prevention

Searching for effect heterogeneity: The baseline target moderated mediator design

George Howe, Ph.D.
George Washington University

ABSTRACT:
We present and discuss a novel approach, the baseline target moderated mediation (BTMM) design, that holds substantial promise for advancing our understanding of how genetic research can inform prevention research.  We first discuss how genetically informed research on developmental psychopathology can be used to identify potential intervention targets.  We then describe the BTMM design, which employs moderated mediation within a longitudinal study to test whether baseline levels of intervention targets moderate the impact of the intervention on change in the target, and whether change in those targets mediates causal impact of preventive or treatment interventions on distal health outcomes.  We next discuss how genetically informed BTMM designs can be applied to both microtrials and full-scale prevention trials.  We use simulated data to illustrate a BTMM, and end with a discussion of some advantages and limitations of this approach.

An individually tailored, family-centered intervention for childhood obesity

J.D. Smith, Ph.D.
Northwestern University
Cady Berkel, Ph.D.
Arizona State University

ABSTRACT:
Non-adherence with recommendations for healthy lifestyle behavior change and connecting children and families with needed services in the community are significant barriers to effectively making a dent in pediatric obesity in the United States. Few strategies have been tested to reverse this trend. Family-centered behavioral interventions have been found to significantly impact childhood obesity and have the potential to increase treatment effects when coordinated with services in the primary healthcare environment—which is the front-line treatment setting for the majority of children and families, particularly low-income, disadvantaged families. The proposed study pending review at the CDC will assess the effectiveness of a family-centered behavioral and motivation-enhancing intervention delivered in the home or the clinic in collaboration with three urban primary care agencies serving low-income ethnic minority families in the Phoenix, AZ area. The intervention being tested has significant potential for scale up and addresses many of the barriers to effective family-based obesity care.

Improving propensity score weighting using machine learning

Elizabeth Stuart, Ph.D.
Johns Hopkins University

ABSTRACT:
Machine learning techniques such as classification and regression trees (CART) have been suggested as promising alternatives to logistic regression for the estimation of propensity scores. The authors examined the performance of various CART-based propensity score models using simulated data. Hypothetical studies of varying sample sizes (n=500, 1000, 2000) with a binary exposure, continuous outcome, and ten covariates were simulated under seven scenarios differing by degree of non-linear and non-additive associations between covariates and the exposure. Propensity score weights were estimated using logistic regression (all main effects), CART, pruned CART, and the ensemble methods of bagged CART, random forests, and boosted CART. Performance metrics included covariate balance, standard error, percent absolute bias, and 95% confidence interval coverage. All methods displayed generally acceptable performance under conditions of either non-linearity or non-additivity alone. However, under conditions of both moderate non-additivity and moderate non-linearity, logistic regression had subpar performance, while ensemble methods provided substantially better bias reduction and more consistent 95% CI coverage. The results suggest that ensemble methods, especially boosted CART, may be useful for propensity score weighting.

Introduction to machine learning

Mitsunori Ogihara, Ph.D.
University of Miami

Optimizing depression treatment through behavioral intervention technologies

Stephen Schueller, Ph.D.
Northwestern University

Modeling the practice of medicine

Luís Amaral, Ph.D.
Northwestern University

Implementation of evidence-based practices in a large urban publicly-funded mental health system

Rinad Beidas, Ph.D.
University of Pennsylvania

ABSTRACT:
Importance: Few studies have examined the effects of both clinician and organizational characteristics on the use of evidence-based practices in mental healthcare. Improved understanding of these factors could guide future implementation efforts to ensure effective adoption, implementation, and sustainment of evidence-based practices.  

Objective: To estimate the relative contribution of clinician and organizational factors on clinician self-reported use of cognitive-behavioral, family, and psychodynamic techniques within the context of a large-scale effort to increase use of evidence-based practices in an urban public mental health system serving youth and families.

Design: Observational and cross-sectional. Data collected in 2013.Setting: Twenty-three organizations.Participants: We used purposive sampling to recruit the 29 largest
child-serving agencies, which together serve approximately 80% of youth receiving publically funded mental health care. The final sample included 19 agencies with 23 sites, 130 therapists, 36 supervisors, and 22 executive administrators. 
Main Outcome Measures: Clinician self-reported use of cognitive-behavioral, family, and psychodynamic techniques, as measured by the Therapist Procedures Checklist – Family Revised. 

Results: Linear mixed-effects regression models were used; models included random intercepts for organization to account for nesting of clinicians within organization. Clinician factors accounted for the following percentage of the overall variation: cognitive-behavioral (16%), family (7%), psychodynamic (20%). Organizational factors accounted for the following percentage of the overall variation: cognitive-behavioral (23%), family (19%), psychodynamic (7%).   Older clinicians and clinicians with more open attitudes were more likely to endorse use of cognitive behavioral techniques, as were those in organizations that had spent fewer years participating in evidence-based practice initiatives, had more resistant cultures, and had more functional climates. Female clinicians were more likely to endorse use of family techniques, as were those in organizations employing more fee-for-service staff and with more stressful climates. Clinicians with more divergent attitudes and less knowledge about evidence-based practices were more likely to use psychodynamic techniques.  

Conclusions & Relevance: This study suggests that both clinician and organizational factors are important in explaining clinician behavior and the use of evidence-based practices, but that their relative importance varies by therapeutic technique.

Parenting practices in primary care: A structured review

John Landsverk, Ph.D.
Oregon Social Learning Center

Mapping a community of implementation science: A network analysis

Wynne Norton, Ph.D.
National Cancer Institute
(formerly of University of Alabama-Birmingham)

Evaluating construct equivalence and harmonizing measurement when synthesizing results across multiple studies

George Howe, Ph.D.
George Washington University

Profiles of contextual risk during the prenatal period and adolescent substance misuse in a Finnish birth cohort study

W. Alex Mason, Ph.D.
Boys Town

Gilbert Parra, Ph.D.
University of Nebraska (formerly of Boys Town)

ABSTRACT:
Vulnerability for substance misuse is highest among youth who experience multiple contextual risks, such as birth complications and socioeconomic disadvantage, during early development. Although notable progress has been made in research that derives from the cumulative risk and adverse childhood events traditions, there continues to be a need to better understand how to conceptualize children’s exposure to multiple contextual risk factors. This study examined whether subgroups of youth with unique profiles of multiple contextual risk factors during the prenatal period could be identified. The relation between subgroup membership and adolescent substance misuse during adolescence also was investigated. Data from the 1986 Northern Finland Birth Cohort Study (NFBCS) were used. The NFBCS is a large-scale birth cohort study that includes 9,432 youth who were followed from the prenatal period into emerging adulthood. Eleven measures of contextual risk during the prenatal period were examined. Measures of alcohol, cigarette, and illegal drug misuse were assessed when participants were 16 years old. A latent class analysis (LCA) was conducted using Mplus Version 7.3 to identify subgroups of adolescents with unique profiles of contextual risk. Findings indicated that a five class solution provided the best fit to the data. We described the subgroups as Parent Substance Misuse (n = 385; 6%), Mom High-Risk (n = 167; 2%), Large Family Size (n = 521; 8%), Socioeconomic Disadvantage (n = 352; 5%), and Low Risk (n = 5466; 79%). The Large Family Size and Low Risk groups had the lowest levels of substance-related problems. The Mom High-Risk group, characterized by high probabilities of smoking while pregnant, school dropout, and multiple romantic partners had the highest levels of cigarette use. Implications of the findings for prevention science will be discussed including the importance of cultural considerations.

The Patient Centered Outcomes Research Institute's Clinical Trials Advisory Panel: Perspectives from Year 1

Elizabeth Stuart, Ph.D.
Johns Hopkins University

Variations in impact of prevention programs on adolescent depression: An integrative data analysis approach

Hendricks Brown, Ph.D.
Northwestern University

Tatiana Perrino, Psy.D.
University of Miami

Shi Huang, Ph.D.
Vanderbilt University
(formerly of University of Miami)

George Howe, Ph.D.
George Washington University

Technology in prevention research: Selected opportunities and challenges for human subjects protection

Anthony Pisani, Ph.D.
University of Rochester

A moderated nonlinear factor model for the development of commensurate measures in integrative data analysis

Patrick Curran, Ph.D.
University of North Carolina-Chapel Hill

ABSTRACT:
Integrative data analysis (IDA) is a methodological framework that allows for the fitting of models to data that have been pooled across 2 or more independent sources. IDA offers many potential advantages including increased statistical power, greater subject heterogeneity, higher observed frequencies of low base-rate behaviors, and longer developmental periods of study. However, a core challenge is the estimation of valid and reliable psychometric scores that are based on potentially different items with different response options drawn from different studies. This talk will focus on a five-step procedure for building measurement models in IDA and demonstrate this approach using data drawn from n = 1,972 individuals ranging in age from 11 to 34 years pooled across three independent studies to examine the factor structure of 17 binary items assessing depressive symptomatology. A series of nonlinear factor models are estimated and individual- and time-specific scores are obtained for subsequent analysis. Potential limitations of this approach are discussed, as are future directions for ongoing methodological development.

New weighting methods for causal mediation analysis

Guanglei Hong, Ph.D.
University of Chicago

Comparing risk factors for adolescent drug use between Columbia and the United States

Eric Brown, Ph.D.
University of Miami

Augusto Pérez-Gómez, Ph.D.
Corporación Nuevos Rumbos

Juliana Mejía Trujillo, M.A.
Corporación Nuevos Rumbos

New weighting methods for causal mediation analysis

Guanglei Hong, Ph.D.
University of Chicago

Policy dissemination research

Ross Brownson, Ph.D.
Washington University in St. Louis

Fidelity monitoring using speech analysis

Carlos Gallo, Ph.D.
Northwestern University

Flexible mediation analysis in the presence of non-linear relations using natural effect models

Stijn Vansteelandt, Ph.D.
Ghent University

Genetics and prevention science: Some methodological and conceptual challenges (and solutions) in integrating genetics into prevention research

Karl Hill, Ph.D.
University of Colorado-Boulder
(formerly of University of Washington)