State of the Art and Future Directions in Multilevel Interventions across the Cancer Control Continuum Erica Breslau, PhD, MPH Allen Dietrich, MD Russell Glasgow, PhD Kurt Stange, MD, PhD Mulitlevel Interventions in Health Care March 4-5, 2011, Las Vegas NV Methods Multidisciplinary team Multiple iterations of literature searches Designed to include the breadth of the field Concentrated on cancer control continuum Included other fields (particular cardiovascular disease studies) Developed a matrix of article types Immersion at OHare Airport hotel Characterize how multilevel research currently is
conceptualized and implemented Described illustrative examples Identified opportunities for MLI to advance cancer control 2 Threads of Shared Thinking & Research Contextualized single-level studies intervene at the person/patient or provider/practice, but consider other contextually-important levels Health care systems research includes individual, practice & system levels, and sometimes considers community or policy factors Community-wide studies public health perspective, often including nested health care system, practice, family, person levels Community heart studies Community cancer control studies 3 Observations from the Literature
Multilevel interventions are context-dependent, but context is reported inadequately in most reports Typically, less than 3 levels are reported Theory, models and interventions are not well integrated in reports/studies Most focused on prevention and screening; seldom on diagnosis, treatment, survivorship Reporting of temporal issues is limited Many glimmers of potential of MLI Thought pieces Theory Empirical 4 Opportunities Design Analysis Dissemination
5 Design Greater attention to: Contextual factors across Levels Time Rapid learning designs that evolve over time Dynamic Adaptive Emergent Interfaces across and within levels Specifying levels affecting the phenomena under study 6 (even if they are not the focus of the study) Analysis Greater attention to: Evaluate both process and outcomes
Capture, but move beyond measures of central tendency Use multi-method approaches that Integrate quantitative modeling across multiple levels where relevant quantitative data can be generated, and Qualitative methods to evaluate levels with small numbers, and to identify key inter-level processes Complex systems and dynamic simulation modeling may provide additional insights where data are sparse 7 Translation Greater attention to: Transparent reporting of contextual factors Moving beyond fidelity to Local adaptation / reinvention / evolution Participatory approaches
How multiple levels interact in context, (vs. in isolation) Move beyond sustainable to evolvable interventions 8 An Example Design: Focus: 9 Observational / intervention Preventive services & health behavior Level Individual and Family Who/What
Low income, pregnant mothers, and their children <2 Practice Primary care and home visits Healthcare System Multilevel advisory committee included Medicaid representatives Community Community organization and town and state government Margolis PA, Stevens R, Bordley WC, et al. From concept to application: the impact of a community-wide intervention to improve the delivery of preventive services to children. Pediatrics. 2001;108(3):E42.
Challenges The future is multiple (conditions, behaviors, interactive modalities) The future is complex (and we ignore complexity at our peril)1 All models (and designs) are wrong 2 tolerance, respect, & creativity are needed We may need to UN-learn much of what we have been taught to answer the tough questions 10 Glasgow RE, Emmons KM. Annual Review of Public Health , 2007 StermanJD. Syst Dynam Rev 2002;18:501-531 1 2 Discussion Questions How do we develop MLI that
Create synergy across multiple levels? Continually pay attention to context? Adapt and learn over time? How do we develop designs and analytic techniques that reflect the complexity of the phenomena we are studying? How do we disseminate to foster Informed adaptation or re-invention? Continued evolution and learning? 11 Extra Slides Not for handout May be used for discussion Integrated Dynamic, Multilevel Research-Practice Partnerships Systems Approach Health Care System Evidence-Tested Program
Program as Tested Clinic(s) Fit Critical Elements Delivery Site(s) Non-critical Packaging Organization Program as Marketed Design Appropriate for Question
Program Delivery Staff Partnership Research Design Team And Adaptive Design Broader Health Policy and Cultural Context 14 Adapted from Estabrooks et. al. AJPM, 2005, 31: S45 Opportunities for Integrating Across Levels Cultural, Historical, Social-Environmental Factors (Fundamental distal Determinants) Tracking, Monitoring Progress
Action Plans Individual Level Malleable Factors: Self-efficacy, activation, problem-solving, etc. ) ile ob (m s la th vice He de m Trait Factors: Preferences, illness Representations, emotions Genetic Counseling Sub-personal/Biological
Genomics 15 Ubiquitous Media and Technology Work, built environment, environmental exposures, family, friends Pe Re rso co na So rd l H cia s ( ea l M PH lth ed R); ia Patient-Provider Encounters Shared Decision-making
Pe Re rso co na rd l H s ( ea PH lth R) F M eed on ba ito ck rin , g Breadth of Impact Health Care System Priorities, Resources, Guidelines, etc.
Malleability in Short Timeframe s ie ol op ss o n s. ce M v Ac en Op s ity s am un s rm gr m m ram No Pro l n Co og pr
cia o So rtati o sp an Tr Social-Community Context Work, built environment, environmental exposures, family, friends Incentives, Coverage e bl s ra R pe PH ro d te an In Rs blic g n
EM Pu orti p Re ity s un g e; ce m in ur r u m , ld t Co Bui ruc reso cco t a , s s b on
n o fra In atio d, t rtati l o o gu fo sp Re on ran t The significant problems we face cannot be solved by the same level of thinking that created them. A. Einstein 16 17 Implementation and Dissemination Research Characteristics Contextual Complex
Multi-component programs and policies Non-linear Transdisciplinary Multi-level 18 THE END 19
A bit twist. The same as above, converting to upper case, however, now we want to map each lower case letter to a random upper case letter. Example: 'abc123' 'SGC123', 'hello' 'SGEBU', '123 456' '123 456'
Divorce Act "Marriage breakdown" after 3 years of separation. Sharp increase in divorce rates. Divorce Act. 1 year of separation before divorce. Single ground for divorce (marriage breakdown); it could be established by proving either separation for at least a...
Drug dosing Important factors concentration of drug in plasma rate of drug elimination rate of drug absorption Toxic level Minimum therapeutic level Cp time Therapeutic window Revision of pharmacokinetic terms 1st order elimination rate of elimination depends on plasma concentration...
Significant room for growth in terms of reporting in the league's mandated format, creative business modeling, personnel modeling for player salaries (utilizing the salary's predictive function), considering the league's changing salary cap, luxury tax, etc. See The Milwaukee Bucks case...
20th Anniversary Symposium of the AQICESH Experience-innovation-inclusion Background In 1976: Charter of Human Rights and Freedoms Prohibits all forms of discrimination based on a handicap or the use of any means to palliate a handicap.
An archetype is a term used to describe universal symbols that evoke deep and sometimes unconscious responses in a reader. In literature, regardless of when or where they live, the following are considered archetypes, because they symbolically embody universal meanings...
A data flow diagram (DFD) shows how data moves through an information system but does not show program logic or processing steps. ... Balancing Examples. Ensures that the input and output data flows of the parent DFD are maintained on...
Ready to download the document? Go ahead and hit continue!