Improving Delivery of Preventive Interventions in Clinical Practice:
Practical Implications and Future Research Directions

June 11-12, 2009
Natcher Conference Center, NIH Campus, Bethesda, MD

Workshop Background, Purpose, and Charge
Agenda
Roster
Relevant References

Workshop Report (in-page links)
Background and Objectives
Setting the Stage: Introduction, Background, and Context
Approaches Addressing Patient Adherence
Approaches Addressing Clinician Adoption and Implementation
Approaches Using Teams and Connecting Clinical To Other Settings
Issues, Approaches, and Lessons Learned
Recommendations
Recommendations for Practice
Recommendations for Research
Recommendations for Cross-Cutting Facilitating Approaches

Background and Objectives

Delivery in clinical practice of proven cardiovascular disease (CVD) prevention approaches, and adherence to prescribed prevention regimens by patients, are less than optimal. NHLBI has addressed these issues using two main approaches: (1) developing, disseminating, and promoting implementation of clinical guidelines to foster delivery of evidence-based CVD prevention strategies, and (2) supporting a portfolio of research studies examining approaches to help improve implementation in clinical practice of evidence-based CVD prevention services.

On June 11-12, 2009 the NHLBI sponsored a workshop of invited experts to assess the state of the science regarding implementation approaches to improve clinical care for hypertension, cholesterol, and obesity; these topics were selected because they are the major CVD risk factors for which NHLBI is currently developing updates to its clinical practice guidelines. The purpose of the workshop was to examine the scientific evidence in order to: (1) identify implementation approaches proven effective that could inform practice guidelines; and (2) identify promising approaches and gaps in implementation research that need additional study.

The workshop focused on strategies to enhance the adoption and integration of evidence-based CVD prevention interventions into clinical practice patterns. Although the focus was the clinical practice setting, the workshop also examined approaches to link clinical practice settings with community and home settings. The primary interest was ambulatory care, considering both primary and secondary CVD prevention (i.e., reducing or preventing risk factors in people both without and with existing CVD). The goal was to develop practice and research recommendations that will inform the NHLBI CVD risk factor guideline updates and new integrated guidelines for CVD risk reduction as well as inform future NHLBI research directions.

Workshop participants were leaders and representatives from public and private academic and medical institutions with expertise in CVD prevention and in implementation methods and research. Prior to the workshop, three subgroups of workshop members held conference calls to discuss and review current evidence and to develop preliminary recommendations regarding three separate, but related, topics: (1) approaches addressing patient adherence, including medication and lifestyle; (2); and (3) approaches using teams and connecting clinical to other settings. The workshop process included presentations on relevant topics by experts, presentations by each subgroup, and broad discussions among all attendees. The process did not entail a systematic literature review or rating of the evidence. Rather, workshop participants brought their knowledge and expertise to bear on the issues of interest.

The purpose of this report is to document and convey the deliberations, conclusions, and recommendations from the workshop. The key recommendations are documented in this report; prioritization was not attempted.

Setting the Stage: Introduction, Background, and Context

After welcomes and introductions, Dr. Denise Simons-Morton summarized the history of NHLBI leadership in implementation research. NHLBI-solicited research has included the following topics: improving patient adherence, improving hypertension control among minorities, primary care interventions to reduce obesity and physical inactivity, systems approaches to improve delivery of guideline-based care, and an infrastructure to facilitate CVD research across multiple healthcare organizations. Studies have addressed multiple levels important for evidence-based implementation, including patients, clinicians, and clinical institutions in the context of insurance, government, and society. These studies were intended to provide important evidence to help inform clinical practice.

Dr. Ira Ockene, workshop co-chair, presented clinical implementation considerations from a broad perspective. He illustrated how patients and doctors think differently about chronic treatment regimens. He also noted evolutionary influences on human health and health behaviors, the role of education and environments, and the need to systematize preventive actions. Dr. Laura Svetkey, workshop co-chair, presented an example of a sequence of research from efficacy to effectiveness to implementation studies: the DASH and DASH-Sodium feeding trials, the PREMIER trial testing multicomponent interventions for BP control, and the Hypertension Improvement Project testing approaches to improve patient and/or provider adherence in practice.

Dr. Ned Calonge reviewed the U.S. Preventive Services Task Force (USPSTF) guideline development process. This process includes defining a question, retrieving relevant evidence, judging the quality of the studies, judging and synthesizing the overall body of evidence, and linking the magnitude and certainty of benefit to a recommendation statement and letter grade. Dr. Sidney Smith described current efforts to update the NHLBI-led cholesterol, hypertension, and obesity clinical guidelines, as well as to develop integrated CVD prevention guidelines. Dr. Thomas Pearson provided an update on the activities of a working group focusing on developing strategies to enhance implementation of the new NHLBI guidelines; the goal is to assure that the revised guidelines are clear, specific, easy to understand, and easy to follow. Dr. David Goff described theories, approaches, and strategies for implementation research, including various definitions. Although the definitions differ somewhat, each embraces the notion of scientific study of methods to improve the uptake and delivery of evidence-based medical care.

The discussion resulted in a general consensus that implementation research is critical to achieving an understanding of the best ways to translate new discoveries into practice in order to improve health.

Approaches Addressing Patient Adherence

Dr. Martha Hill, subgroup chair, introduced this topic, noting that there are various definitions of “adherence”. Consequently it is difficult to quantify the extent of non-adherence. She noted that multiple co-morbidities are common in U.S. adults, as is insufficient health literacy, both of which contribute to poor adherence.

In an overview on adherence research, Dr. Lisa Cooper noted that adherence is not an “all-or-none” phenomenon; rather, it is a complex issue with many dimensions. Adherence can represent a wide variety of behaviors, from single discrete episodes to behavioral patterns over long periods of time. Patient factors include financial difficulties, lack of family and social support, and competing daily demands. Health system and provider factors include provider knowledge, attitudes, skills, and behaviors. Therapy-related factors include regimen complexity and treatment duration. Many theoretical models of health behavior have been applied to adherence, as have planning and intervention frameworks. Discussants suggested that adherence research could be advanced by developing and testing conceptual theories, incorporating patient and clinician perceptions, and collaborating with other disciplines, patients, and communities.

Drs. Gbenga Ogedegbe and Jacqueline Dunbar-Jacob reviewed interventions to improve patient medication adherence for hypertension and hypercholesterolemia. They stated that characteristics of effective adherence interventions include simplification of regimens, behavioral training, combinations of strategies, and intensified care using monitoring, followup contacts, and reminders. They noted gaps in the current research evidence, which include lack of cost-effectiveness data, limited generalizability of studies, lack of gold standards, unreliable and inconsistent measures of adherence, challenges in interpreting multiple measurements, and failure to measure adherence patterns over time. Discussion points included that adherence studies often have short intervention times, small sample sizes, imprecise adherence measures, and lack of patient tailoring.

Drs. Jamy Ard and David Hyman noted that adherence to healthful behaviors, including heart-healthy dietary and physical activity patterns, are highly relevant to controlling CVD risk factors. Clinicians are in a unique position to promote healthful lifestyle behaviors in their patients because of their contacts with patients and their role as valued and trusted sources of information. They showed evidence that interventions have been demonstrated to improve healthful behaviors in efficacy trials in volunteers; however, they noted there is insufficient research conducted in primary care settings and in generalizable patient populations to inform practice. Challenges for implementation in the primary care setting include lack of clinician training, lack of systematic patient screening, time constraints, low frequency of contacts, and limited resources for prevention.

The discussion highlighted the need to raise awareness of the importance of adherence and of the concept of “risk.” Multifactorial approaches are likely needed including counseling, physician-patient communication, team approaches, and interventions involving both community and clinic. Group members noted that lifestyle and medication should be thought of as a “package” important for CVD risk reduction. Because patients are influenced by their broader environments, not just through clinical encounters, the point was made that adherence could be enhanced by including approaches outside of the primary care setting as well as better methods to reach target population groups. Group members also noted that incorporating an economic perspective and addressing external validity in studies are important.

Approaches Addressing Clinician Adoption and Implementation

Dr. Laura Petersen, subgroup chair, noted that moving beyond “a science of recommendation” to a “science of implementation” includes the need for bi-directional relationships between practice and evidence. Simply providing clinicians with results of internally valid studies is insufficient to change clinician behaviors because of numerous challenges and the complexity of “real world” settings.

Dr. Paul Estabrooks presented conceptual models of clinician adoption. An often-used linear model of translation goes from basic to clinical research and then from clinical research to clinical practice. System models take into account the complex interrelationships between factors at multiple levels, and thus may be more relevant. More complex models allow for adaptations through feedback and modification; these dynamic systems models incorporate ongoing adjustment of intervention and evaluation strategies into their structure. During the discussion, it was noted that reliance on a traditional linear model may be contributing to the well-recognized lag between evidence and practice.

Systematic reviews of interventions to improve physician clinical performance, reviewed by Drs. Charles Eaton and Jonathan Tobin, have evaluated various approaches including CME lectures, printed materials, audience response systems, discussion groups, demonstrations, use of opinion leaders, academic detailing, physician reminders, audit and feedback, and other approaches. They noted that employing multiple techniques has resulted in the biggest improvements in physician performance. However, there are gaps in understanding how processes lead to outcomes, including the need to understand causal mechanisms by which interventions may lead to changes in clinician behavior. Studies of practice-based approaches can arise from researchers, from clinicians or patients, or can be mixed. It was suggested that more input from clinicians and patients is needed, as is the use of behavioral theory to inform behavior change at multiple levels.

Dr. Judith Walsh reviewed approaches employing changes in the healthcare system to reduce the gap between evidence and practice, including Quality Improvement (QI) strategies. Systems changes can include facilitated relaying of clinical information, auditing patient charts and proving feedback to clinicians, creation of multidisciplinary teams, altering roles of staff, case management approaches, or other approaches. Systematic reviews have shown a positive effect of team changes, such as assigning some patient care responsibilities to a team member other than the physician (e.g., pharmacists, nurses, or physician assistants). Studies of “Pay for Performance” approaches, reviewed by Dr. Laura Petersen, reveal some evidence that physician-level incentives may be useful. However, there are many unanswered questions because there are an insufficient number of controlled studies to draw firm conclusions. Dr. Steven Ornstein presented several issues to be considered, which he gleaned from practice and from implementation research: determining the extent of buy-in to evidence-based medicine by clinicians, clarifying the importance of audit and feedback, the need for communication between Electronic Medical Record developers and users, recognition that the medical clinic is an “independent micro-system” for care delivery, the need to understand relationships between intervention dose and effectiveness, and the need to integrate diverse guidelines.

The discussion addressed many issues, including use of clinical support systems; the need to know more about drivers of clinician behavior; the need to identify tools that may be useful to clinicians; the complexity of busy healthcare practices; and the need to address maintenance of new behaviors within complex organizational structures. Group members noted that implementation research should consider behavioral economics, psychology, and cost-effectiveness; that qualitative research may need to complement quantitative studies; and that financial reimbursement, preventive care systems that bypass the physician, and case management by non-physicians should be studied further.

Approaches Using Teams and Connecting to Other Settings

Dr. George Rust, subgroup chair, introduced this topic by noting that a large number of factors and complex interactions are relevant to preventing complications of obesity, hypertension, and dyslipidemia. He noted that optimal, equitable health outcomes require considering patients within family, neighborhood, and community contexts; this perspective blurs boundaries between inside-the-clinic practice and outside-the-clinic health promotion. These considerations lead to complex, multi-faceted interventions targeting global risk, focused on system performance, and using dynamic adaptive interventions with rapid-cycle feedback loops for continuous improvements.

Drs. Kathy Berra and Barry Carter reviewed team-based care interventions as important Quality Improvement strategies. They described studies testing interventions that employed nurses and community pharmacists, which found significant effects when study results are combined in meta-analyses. However, studies generally did not conduct cost-effectiveness analyses, many were conducted outside the U.S., and most did not describe the training given to the nurses or pharmacists sufficiently for replication. They also reviewed demonstration projects that are team based, include both lifestyle and medication interventions, focus on long-term adherence, and measure patient-centered outcomes like quality of life and satisfaction.

Drs. Hector Balcazar and Penny Feldman reviewed the use of community health workers (CHW) – a team approach that can enhance delivery of health-related services by bridging clinical staff and community residents. Dr. Balcazar described the experience with promotoras In Hispanic communities, noting several benefits to both patients and providers. He described factors can aid successful integration between health-care and community, such as open and frequent communication, organizational acceptance, extensive training, thorough documentation, and management support. Dr. Feldman noted that home-based interventions for the elderly are especially important, and there is some evidence that they can reduce nursing-home admissions and hospital stays, as well as improve other outcomes. The presenters noted that, although studies and some practice models show some evidence of effectiveness, additional evidence is needed to enable firm conclusions about the effectiveness of various CHW models.

Drs. Joseph Finkelstein, Beverly Green, and Hayden Bosworth reviewed studies of health Information Technology (IT) for disease management and prevention. Health IT is a promising tool to facilitate decision support for primary prevention and detection, quality improvement, and chronic disease management across doctors’ offices, hospitals, and patients’ homes. Health IT can be used at patient (e.g. personal health records), clinician (e.g. electronic medical records), and healthcare system (e.g. disease registries) levels, including the interface between those levels (e.g. care coordination tools). Some randomized trials have tested whether IT approaches can improve CVD risk factors, such as diabetes, nutrition, and obesity. However, additional research is warranted to address health IT research needs, including RCTs as well as alternative designs.

Discussion points addressed using the importance of communicating health messages, providing consumers with choices for communication avenues, and using technology (e.g., text messaging, email) to streamline communication. Group members suggested integrating technology and community health workers in a system that provides decision support. Members emphasized the importance of patient-provider interactions for behavior change. The discussion highlighted the need to focus attention on sustainability, reach, and cost-effectiveness of technological approaches as well as the need for a menu of common typologies and key characteristics of successful interventions.

Issues, Approaches, and Lessons Learned

Dr. Brian Mittman presented study designs and methods for implementation research, including an overview of implementation science, methodological challenges to evaluation, and challenges for funding agencies. He noted that implementation research includes studying methods to promote uptake of research findings, improve quality and effectiveness of health services, and examine influences on healthcare professional and organizational behavior. Dr. Mittman placed implementation as a last step in a health-research “pipeline,” which includes basic science, translational pre-clinical research, clinical, health behavior and health services research, effectiveness studies, and implementation research. Dr. Mittman reviewed the QUERI Service-Directed Project (SDP) template for designing, conducting, and reporting implementation trials. The group suggested that NIH should fund projects to fill the pipeline with more scientists and to examine practice changes including mechanisms of change and contextual factors. Getting providers to respond and getting IRB approval are among the challenges to implementation research noted in the discussion.

Dr. Gibbons described a real-world example of the QC are guidelines implementation project in Minnesota, including processes, outcomes, and feasibility issues. This project focused initially on developing clinical practice guidelines and later focused on quality improvement using action groups and systems improvement using healthcare redesign. Dr. Gibbons shared examples of feedback given to the clinicians and showed some positive results of the project for improving CVD risk factors. He emphasized that feedback on patient improvement can result in provider behavior change and that sharing best practices in a learning community can stimulate positive change. The group discussed issues of data collection, underserved populations, and expectations of clinicians and others of these types of projects.

Dr. Howard Leventhal presented models of health psychology relevant to healthcare systems. He reviewed studies that demonstrated the importance of the physician-patient relationship for successful implementation of guidelines. He showed examples illustrating the effect of patient perceptions and beliefs about symptoms on adherence and techniques for personalized, behavioral patient management. He noted that implementation science needs a process model to identify factors essential for efficient and effective behavioral interventions for self-management. Dr. Leventhal also recommended that guidelines focus on patients’ individual experiences because of the differences in perceptions, motivations, and needs.

The discussion emphasized physicians reinforcing the empowerment of patients to appraise their symptoms, the importance of affective processes, and the importance of shared language and shared responsibility for individuals’ health. The group discussed the importance of what happens with patients in their life environments both before and after the office visit. Group members indicated that cultural differences should be considered, as they relate to belief of causal and cure frameworks and framing of symptoms. It was noted that there is a large disconnect between theory and practical application of theories.

Recommendations

Discussion of Recommendations

The workshop attendees considered and discussed the broad range of topics and issues presented to develop recommendations for practice and research. In addition to recommendations for each topic category, several recommendations arose for facilitating approaches that involve both practice and research. The next section lists the key recommendations made during the workshop.

Recommendations for Practice

General recommendation:
NHLBI clinical guidelines should include recommendations for implementation strategies at multiple levels (patient, clinician, and system) that are based on scientific evidence from implementation research.

The following specific recommendations are based on expert opinion; systematic reviews of the evidence are needed.

In general, implementation approaches to be used in practice should:

Be feasible for delivery in community practice settings, scalable, adoptable, sustainable, equitable, and integratable with technology available to the setting
Be affordable and based on evidence of cost-effectiveness
Have effects of sufficient magnitude to warrant implementation
Be modifiable to align with local variability in practice settings

To improve patient adherence, clinicians should:

Regularly ask patients about their lifestyle and medication adherence, treating lifestyle and medication as a package, not separately
Inquire about and discuss patients’ conceptual models, beliefs, cultural issues, and barriers; validate and “hear” patients’ concerns and issues
Use patient-centered behavioral counseling approaches, including barriers identification and problem solving; receive training in patient-centered counseling approaches if needed
Simplify medication regimens whenever possible and encourage use of medication-taking reminder systems (like pill boxes and machines)

Primary care practices should:

Take a patient-centered approach by addressing multiple CVD risk factors comprehensively, including hypertension, cholesterol, obesity, smoking, and dietary and physical activity behaviors
Identify and manage CVD risk factors by providing consistent screening, raising patient awareness of risk, and initiating and titrating medications based on evidence-based clinical guidelines
Train clinicians in treatment and counseling approaches (e.g., patient-provider communication approaches, patient-centered counseling)
Implement relevant measures for screening and followup, including measurement frequency, based on evidence-based guidelines, and train clinicians in recommended measurement methods (e.g., BP measurement, BMI calculation)
Establish and employ clinical teams (see below)

Clinical teams should:

Comprise multiple disciplines (e.g., MD, RN, NP, PA, nutritionist, behavioral counselor, pharmacist, community health worker, and others as available)
Employ approaches that assign some patient care responsibilities to team members other than the physician
Use effective team communication about patient and system issues
Use patient-centered approaches, including case management or disease management systems, working with patients to recognize health risks and enhance awareness, employing intervention strategies that personalize care in the context of patients’ capabilities, environments, and lifestyles, and providing systematic monitoring and follow up
Enhance ties between clinical and community settings, referring patients to structured community programs for lifestyle issues as needed, with followup and interaction between the program and medical team

Provider systems should:

Devote resources to enhance systematic screening and implementation of evidence-based interventions for lifestyle and medication addressing CVD risk factor control
Use technology to leverage care (e.g., e-mail communication with patients, decision-support IT, EMRs, metrics on quality of care, feedback to clinicians)
Improve clinician care delivery through educational approaches tailored to clinicians that employ multiple approaches such as enhancing CME with other approaches like academic detailing
Provide information back to clinicians such as by establishing and using reminder systems, providing feedback on patient risk-factor goal achievement, and providing information on patient medication prescription filling
Integrate practice-based research within the clinical setting to better quantify the extent of problems and define possible feasible intervention opportunities

Guidelines writers should:

Write guidelines so they are implementable, including being simple, explicit, clear, understandable, and actionable by using explicit language for specific clinical behaviors
Write guidelines so they are explicit about evidence quality and strength
Include guideline implementers on expert panels to ensure development of an evidence-based implementation plan for the guidelines
Recognize barriers to implementation, considering beliefs, venue, format, and channel

Recommendations for Research

General recommendation:
The NHLBI should expand its current research portfolio of dissemination and implementation research studies, including studies at all levels (patient, clinician, and system) and within a community context.

The following recommendations are intended to be broad areas or approaches; specific research topics and questions, too numerous to list here, are encompassed within these areas.

Many different types of implementation research studies should be funded, including studies that:

Are based on identification of implementation issues through dialogue with administrators, clinicians, researchers, and patients
Improve understanding of how patient, provider, systems, and community approaches are useful for improved CVD prevention and outcomes
Examine a variety of different types of study questions, e.g., social, cognitive, and affective factors that underlie implementation at various levels (patient, provider, system); community-based-participatory approaches considering the clinical practice as the community; mediators by which interventions may work
Employ various methods and designs as suitable to the specific research question; designs can include RCTs, GRTs, quasi-experimental studies, observational studies, natural experiments, qualitative research, adaptive designs, systems modeling, hybrid efficacy-effectiveness and effectiveness-implementation studies, and others
Develop and use new methodologies for studying multi-component and multi-level interventions
Employ conceptual, theoretical, and logic models and use relevant theories to design interventions at multiple levels and to examine causal mechanisms linking interventions to outcomes
Are high-risk, innovative, and potentially high-impact, e.g., testing dramatically different care models and systems
Are cross-disciplinary, including non-traditional disciplines (e.g., anthropology, industrial/organizational psychology, systems engineering, health economics, health policy, and others)
Examine equity of diffusion and reduction of disparities
Examine issues related to implementing multiple guidelines for primary care, not just those for CVD prevention

Implementation intervention research that is practice-based and pragmatic should be funded that:

Has high external generalizability through broad eligibility criteria of patients and clinical sites
Determines how efficacious interventions tested in research settings can be translated into feasible and effective interventions delivered by practice staff
Tests intervention “packages” or strategies designed for dissemination and implementation after completion of the research
Examines how to enhance the “dose” of intervention able to be delivered in clinical practice settings, or determines the “minimal effective dose” of interventions to achieve an effect
That informs sustainability, e.g., studies of long duration testing interventions designed for incorporation and maintenance within a practice, with delayed outcomes assessment after intervention completion
Employs two phases, e.g., qualitative or pilot research in phase 1 to inform interventions tested in phase 2
Tests interventions that use feedback loops to modify interventions over time such as dynamic, locally adaptive intervention approaches
Uses EMR and other technology for data collection to minimize participant burden, enable more representative enrollment, and enhance generalizability
Evaluates economic and patient-centered outcomes, including cost-effectiveness, utilization, health related QOL, and patient, provider, and community satisfaction
Uses common measurements, training, and intervention methods and other elements of study design across studies, when possible, to allow cross-study comparisons
Compares different implementation approaches (Comparative Implementation Research), e.g., targeting clinicians vs. targeting systems, comparing various QI strategies, allowing local intervention adaptation vs. high levels of fidelity, comparing single component vs. multi-component interventions, comparing alternative time allocations for clinic visits

Patient adherence research should be conducted that:

Employs basic-behavioral sciences to help determine why patients behave the way they do, including developing and test conceptual bases and theories, incorporating patient and clinician perceptions and behavioral contexts, and collaborating with other relevant disciplines, with patients, and with communities
Examines various aspects of adherence related to both lifestyle and medication, including healthcare seeking, appointment making and keeping, attrition and retention, and satisfaction
Examines attitudes, beliefs, concerns, motivations, barriers, cultural issues, behaviors, and illness/treatment models as they may affect lifestyle and medication adherence, including barriers that may be structural (costs/medication access) or psychosocial (depression/anxiety)
Tests interventions tailored to patients’ situations, preferences, and beliefs, and determines key characteristics of effective tailoring
Focuses on patients with need, e.g., those nonadherent or at high risk for nonadherence in general and with respect to impact on disparities
Tests interventions that combine lifestyle and medication for risk factor control, uses a “stepped-care” approach, and/or combines words and images
Addresses the issues above with respect to reducing health disparities

Studies of clinicians and medical systems delivery should:

Determine the degree of use and acceptance of evidence-based medicine and guidelines; use qualitative or survey research methods to assess attitudes, beliefs, motivations, facilitators, cultural issues, context, barriers, and behaviors for clinician implementation of evidence-based care
Examine;natural experiments” and study “best practices” by examining existing approaches (e.g., embedding evidence-based guidelines into existing EMR and decision-support tools), and examining outcomes associated with existing approaches, including impacts on disparities
Evaluate pay-for-performance approaches, including who should receive incentives; the optimal magnitude, frequency, and duration of incentives for quality; the use of incentives to prevent over-use of services; persistence of effects after intervention; types of performance targets; cost-effectiveness; and proportion of health care payments that should be based on performance
Examine what type of information would be useful to incorporate into clinical-support tools and what types of tools are preferred, including various IT approaches

Relationships between clinical and community settings should be examined, such as:

Access to care, including healthcare seeking behaviors and care availability, interventions to improve access, and effects of those interventions on unintended consequences, such as barriers to access
Effects of various approaches linking clinical to community settings, including effects on health disparities by age, SES, race/ethnicity, and geography
Effectiveness of approaches using community and lay health workers, including cost-benefit and cost-effectiveness outcomes
Approaches examining patient-provider-computer interactions such as those in patients homes, including telemedicine
Various infrastructure approaches to linking clinical to community settings (e.g., departments of public health, community health centers, community organizations, extension service) and various types of approaches such as referrals, forms of health communications strategies, and social marketing

Recommendations for Cross-Cutting Facilitating Approaches

General Recommendation:
The NHLBI should lead or collaborate with other groups on approaches to facilitate an implementation focus in both practice and research.

The following specific recommendations for facilitating approaches were made:

Develop a taxonomy of intervention approaches, process and impact metrics, and outcomes for use in implementation research and practice
Work with IRBs and NIH review groups to support more boundary-crossing implementation studies that address real-world needs
Develop a comprehensive guide to implementation research methods (similar to the QUERI SDP template; a precedent is the NHLBI guidance on multicenter RCTs)
Augment training opportunities to enhance the current generation’s skills in implementation research and develop the next generation of implementation scientists
Use funding mechanisms and infrastructures to facilitate rapid progression across different phases of research, and that require collaboration across institutions (e.g., networks, centers, the NIH CTSAs)
Develop a better approach for communicating research processes and results, including data sharing, common data elements, study details, and intervention details to enable replication in practice, such as making intervention manuals and materials widely available
Develop a “public utility” decision-support system to enable better delivery of guidelines-based care (e.g., the VA EMR system)
Develop a catalog of guidelines across NIH institutes and other federal agencies for use by primary care practices