Vol. 37 No. 1 January/February 2015

41

A Framework to Guide Implementation Research for Care Transitions Interventions Sydney M. Dy, Mahima Ashok, Roberta C. Wines, Lucia Rojas Smith

Introduction Deficits in quality of care for care transitions from the hospital to the ambulatory setting result in approximately 20% rates of readmissions (Jencks et al., 2009) and preventable adverse events (Forster et al., 2003). Although various interventions to improve these transitions have been implemented, a recent systematic review found few controlled studies and low strength of evidence. In addition, most studies did not describe the intervention context or implementation process (Rennke et al., 2013). Interventions to improve care transitions are “a set of actions designed to ensure the coordination and continuity of health care as patients transfer between different locations” (Coleman and Boult, 2003). They are complex in involving multiple settings and organizations and a bundle of components often selected from various models and adapted to specific circumstances (Geary and Schumacher, 2012). These may involve predischarge interventions in the hospital, such as patient/ caregiver education, and postdischarge interventions, such as outreach to patients (Rennke et al., 2013). The Centers for Medicare and Medicaid Services (CMS) increasingly emphasizes interventions that cross settings or are centered in community organizations (Centers for Medicare and Medicaid Services, 2013), which require new approaches for evaluation. Complex systems interventions, such as those for care transitions, require many components that often interact and involve multiple aspects of systems, including different provider types and settings (Geary and Schumacher, 2012). Implementation research focuses on describing interventions and how they are implemented into routine

Abstract: Evaluating implementation of complex interventions to improve care transitions and comparison across studies is challenging due to issues such as variation in methods and lack of reporting key evaluation elements. This article describes a framework for evaluating implementation of hospital to ambulatory care transitions interventions and application to a case study. We searched published and gray literature for relevant frameworks. We adapted the general Consolidated Framework for Implementation Research, adding elements relevant to other complex interventions. We refined these adaptations through structured expert input and application to case studies. Key adaptations included conceptualization around organizations, not just settings, and around patientand caregiver-centeredness. Although these interventions are often oriented toward institutional outcomes such as readmissions, tailoring interventions to specific patient needs strengthens effectiveness. Coordination and communication are important between organizations and providers and with patients and caregivers. Roles of those involved in the intervention—providers, administrators, and facilitators from different organizations—are also key constructs. Finally, as these interventions often are tailored to specific settings and adapt over time, assessing intervention design—which components are implemented as part of the bundle, how they are actually implemented, and their differential impact on effectiveness—is critical.

practice, as well as evaluating why they succeed or fail (Implementation Science, 2013), and is critical for successfully disseminating interventions. Complex systems interventions may be more challenging to evaluate because the interventions and their implementation vary more than with interventions limited to one setting and because of interactions between intervention, system, and patient characteristics. A framework organizing these elements and guidance for how to apply the framework based on implementation theories and customized to unique issues in complex systems and care transitions could guide evaluations

Keywords care transitions transitional care hospitalization patient readmission coordination quality improvement implementation intervention studies framework Journal for Healthcare Quality Vol. 37, No. 1, pp. 41–54 © 2015 National Association for Healthcare Quality

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

42

Journal for Healthcare Quality

both for developing high-quality evidence to disseminate interventions and for facilitating comparability across studies. The Consolidated Framework for Implementation Research (CFIR) (Damschroder et al., 2009) was developed for researchers to assess effectiveness of implementation within a specific context and promote dissemination. The goal of the CFIR was to synthesize various terminologies, definitions, and constructs described in the literature into a consolidated framework and common taxonomy for implementation research on health service delivery. Drawing on 19 different theories, the CFIR includes four domains that describe the internal and external context of implementation: Intervention Characteristics, Outer Setting (e.g., external policy), Inner Setting (characteristics of setting where the intervention is based), and Process (elements of implementation). Each has a set of defined constructs (or subcategories) addressing key elements (Damschroder et al., 2009). The structure of the CFIR and the synthesis of a variety of preexisting frameworks provided an efficient starting point. However, CFIR, as a general model of implementation within bounded organizational settings, does not address many of the distinct attributes of complex system interventions, such as involvement of teams and multiple provider roles, and does not address issues of measures of implementation and outcomes. Specifically for care transitions, CFIR does not address issues such as crossing settings and involving community organizations in interventions, patient- and caregivercenteredness, and adapting intervention components over time. The objectives of this project were to (1) describe a framework adapted from CFIR for implementation research for interventions for care transitions and (2) apply the framework to develop an evaluation approach for case studies. The purpose of the framework is to guide research and evaluation of care transitions implementation to address how, why, and where these interventions succeed or fail to achieve intended outcomes and how their components can be disseminated.

Methods This project is part of a larger report building frameworks for implementation research for three complex systems interventions (Rojas Smith et al., 2014), including patient-centered medical homes and process redesign, which are described in extensive detail in the full report (Rojas Smith et al., 2014). This article summarizes the portion of the report on interventions for care transitions. We first developed frameworks for these interventions using similar methods, including literature reviews, framework development, and expert panel input (Figure 1). We synthesized the results of the framework development to modify the CFIR taxonomy for complex interventions generally, with specific frameworks tailored to unique aspects of each intervention. We then developed flowcharts for applying the frameworks and applied them to two case studies with input from the expert panel.

Literature Scan and Framework Adaptation The framework development for care transitions interventions used a series of steps (Figure 1), described in the full report. In brief, we first conducted a literature scan to identify frameworks for care transitions intervention models (which include multiple components) and initiatives (which may involve multiple models) and then abstracted resources for relevant elements that we used to modify both the graphical and tabular forms of the CFIR adapted for complex interventions (with input from the patient-centered medical home and process redesign processes). We conducted MEDLINE searches from 2005 to 2012 including keyword terms such as “care transition*” and “transitional care” and identified key gray literature Web sites, including Web sites for key care transitions intervention models and initiatives, e.g., Project RED—Re-engineered Discharge (Agency for Healthcare Research and Quality, 2013), IHI STAAR—Institute for Healthcare Improvement STate Action on Avoidable Rehospitalizations (Institute for Healthcare Improvement, 2013), the Care Transitions Intervention (Coleman, 2013),

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

43

Figure 1. Care transitions framework development.

the Transitional Care Model (Naylor, 2013), and Project BOOST—Better Outcomes for Older adults through Safe Transitions (Society of Hospital Medicine, 2013) and reports from the study team and experts. Two reviewers evaluated articles and web resources, which we included based on consensus if they contained a relevant framework or model and were specific to the hospital to ambulatory transition and to medical illnesses. We reviewed full-text articles using the same procedures. Sources (detailed in the full report; Rojas Smith et al., 2014) included frameworks designed for care transitions generally (Moore et al., 2007), specific interventions (Coleman et al., 2006; Naylor and Sochalski, 2010; Naylor, 2013), classifying types of interventions (McCarthy et al., 2013; Rennke et al., 2013), quality measurement (American Medical Association, 2009; National Transitions of Care Coalition, 2013), and application of theoretical models to care transitions (Geary and Schumacher, 2012; Graham et al., 2009). The guiding questions for the data abstraction from included resources, conducted by one investigator and

reviewed by a second investigator, included identifying key framework themes and domains and constructs and relationships among the domains and constructs. We used the selected articles to develop an adapted version of the CFIR graphical representation and draft table listing domains and constructs, as well as a set of questions on key issues on modification of the framework to address with the technical expert panel.

Technical Expert Panel Input on Content and Application of the Adapted Framework We recruited a technical expert panel to provide input on the adapted framework for care transitions interventions implementation research. Individuals included researchers with extensive experience in the topic area, individuals with relevant policy, clinical, and management background, and experts in general implementation research. Most major care transitions intervention initiatives (e.g., RED, STAAR, BOOST) were represented on the expert panel. The first expert panel telephone meetings focused on the applicability of the initial adaptation of the

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

44

Journal for Healthcare Quality

CFIR framework, graphic representation, and domains and constructs, with the goal of addressing whether the adapted CFIR was an appropriate guide for care transitions implementation research and comments on the adapted framework, including the revised figure, domains, any additional constructs that should be added, and how best to frame the issues of patient-centered care, caregivers, links between settings, and accountability. For the second expert panel meeting, we summarized the expert input on the adaptation of the framework and developed a set of recommendations and questions for additional clarification. We also developed two case studies (one describing a collaborative and one from the patient/caregiver perspective) to evaluate usability of the framework for implementation evaluations. In the meeting, we asked the panel to address these additional questions and how the framework could be made more useful for their own implementation research. The framework was then adapted based on this panel input, underwent peer and public review, and was revised accordingly.

Results The domains (overarching categories) and key included constructs (subcategories) of the Care Transitions Framework are described below. The constructs for each domain are shown in Table 1, with detailed rationale for each construct provided in the Appendix, (Supplemental Digital Content, http://links.lww.com/JHQ/A1). Below, we describe selected constructs specific to implementation research for care transitions interventions that were adapted or added through the literature scan and expert panel process by domain and application to one of the case studies (Box 1). Key overarching adaptations included broadening the focus of the CFIR and emphasizing interactions between organizations, reframing constructs to include different organizations within an intervention, and adding a construct for community resources, as these are key for care transitions and an increasing focus of interventions.

The technical expert panel also stressed emphasizing patient-centered care and engagement and adding and separating out caregiver issues. Figure 2 graphically represents the relationships between these domains. On the left side, five domains are shown in a circle that includes the attributes of the intervention itself and the individuals and organizations or systems participating in the intervention. The outer ring of the circle represents the External Context, outside the domains of intervention, organizations, and individuals. An arrow to the right of the circle points to the Measures of Implementation, which influence Outcomes. Intervention Characteristics are the characteristics of the intervention being implemented in a particular organization, including core activities or components (the essential and indispensable elements of the intervention itself). These may be fixed or mutable attributes; they are considered and assessed before implementation and influence adoption decisions. A key construct is adaptability, which reflects the frequent need in intervention design to choose different components and combine them into a “bundled” intervention in care transitions interventions. Since organizations using named models (e.g., Project RED) frequently customize them, researchers should precisely describe the intervention elements rather than using the name of an existing package. External Context includes the economic, political, and social context within which an organization(s) resides and that may affect the implementation process. Given the need for interaction with multiple providers and community components, these are key issues to consider in evaluating care transitions interventions. Organizational Characteristics includes structural characteristics, networks and communications, culture, climate, and readiness, which all interrelate and influence implementation. The domain can include hospital and ambulatory organizations, if both are integrally involved in the intervention, and any other core organizations (e.g., community-based organizations such as coalitions, agencies, and

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

45

Table 1. Domains and Constructs of the Care Transitions Framework Intervention characteristics A. Vision and change strategy B. Targeted groups for intervention C. Intervention source D. Evidence strength and quality supporting the intervention E. Relative advantage of chosen intervention elements F. Feasibility G. Adaptability (to meet local needs) H. Trialability (testable on a small scale) I. Complexity J. Compatibility with characteristics of the organization(s) K. Radicalness of the envisioned change; disruption L. User control: end users’ authority to fix problem on their own M. Location of intervention activity: need for external services, supplemental providers, or new roles N. Design quality and packaging: intervention components within a bundle or pattern O. Workflows: need for change with the intervention P. Task/process standardization with the intervention Q. History of similar interventions within the organization(s) External context A. External networks, partnerships, and systems B. External pressure (e.g., competing organizations implementing interventions) C. External policy incentives and disincentives D. Technological environment E. Population needs and resources F. Community resources Organizational characteristics A. Structural characteristics B. Team and network characteristics C. Culture D. Implementation climate: tension for change, mandate, accountability, relative priority, organizational incentives, and learning climate E. Readiness for implementation: leadership engagement and staff commitment F. Access to information, training, education G. Information Technology and Health Information Technology resources: systems and accessibility H. Physical space and presence of organizations I. Staff time J. Other resources: for example, grant or other funding for the intervention K. Patient self-management infrastructure L. Continuity: includes relationships and exchange of information M. Patient/caregiver-centeredness Characteristics and roles of providers, patients, and caregivers A. Knowledge and beliefs related to the intervention B. Skills and competencies related to implementation C. Role in implementation D. Authority E. Self-efficacy F. Collective efficacy: belief that implementation can be carried out G. Stage of change toward implementation H. Identification with organization(s) I. Socioeconomic demographics

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

46

Journal for Healthcare Quality

Table 1. (Continued) J. Patient needs and resources K. Caregiver needs and resources L. Other personal attributes Process of implementation A. Planning: assessing, goal-setting, feedback, contingency planning B. Acquiring and allocating resources C. Process ownership D. Transition roles: organizational leaders, opinion leaders, formally appointed implementation leaders, champions, external change agents, frontline staff, integrators (building relationships and bridging between organizations), patients, caregivers, and other stakeholders E. Engaging: organizations, external context F. Executing: decision making, staging, and iteration (implementation in stages) G. Reflecting and evaluating: measurement capability and data availability Measures of implementation A. Acceptability B. Adoption/abandonment C. Appropriateness: degree of fit and relevance D. Intervention cost E. Fidelity F. Reach: within the population and organization G. Penetration: depth of integration H. Replicability I. Sustainability J. Evolvability: capability of being sustained through adaptation and refinement Outcomes A. Patient- and caregiver-centered outcomes B. Patient/caregiver experience C. Provider experience D. Processes of care: patient-centered, coordinated, comprehensive, accessible, quality, and safety E. Effectiveness F. Timeliness G. Clinical outcomes H. Healthcare utilization I. Cost effects/impact J. Value K. Unintended consequences

collaboratives). The construct Accountability reflects the shared accountability for implementation and success across and within organizations that is part of the intervention. Within an organization, certain disciplines or units may be more involved than others and some may not be involved (e.g., nursing leadership, hospitalists, and the emergency department). External organizations and networks may be categorized in the outer setting but must be involved and accountable to some extent for successful implementation. The

construct Measurement capability and data availability is critical, as lack of data or of quality data often leads to failure of the intervention. Common issues go beyond information technology–related barriers and include measurement differences between organizations; lack of availability or sharing; accountability for collection, documentation, and analysis; and timeliness. The construct Continuity addresses the importance of relationships and exchange of information, both between organizations and with patients/caregivers.

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

47

Box 1. Applying the Care Transitions Framework: A Case Study Below is an adapted version of a case study from the Robert Wood Johnson Foundation’s Aligning Forces for Quality program Web site (Robert Wood Johnson Foundation, 2013). RTI International and other authors of this framework document are not authors of this case study and do not possess any copyright to this work. We use this modified example below to illustrate how the Care Transitions Framework may be used.

Reducing Readmissions and Integrating Care in Cincinnati Cincinnati is home to several hospitals and physician groups that eagerly compete with each other for market share. The Health Improvement Collaborative of Greater Cincinnati Alliance, sponsored by the Robert Wood Johnson Foundation Aligning Forces for Care Quality (AF4Q), has found that its clinicians need to work together to coordinate care. This effort (the Collaborative) partnered with the local hospital association and the Greater Cincinnati Health Council to reduce heart failure readmissions under a program called Accountable Care Transformation, or ACT. Nineteen hospitals and health systems participate in ACT. They formed a collaborative to reduce readmissions by 10% by adopting five best practices: • Upon admission, implement a risk assessment tool with a focus on heart failure to identify patients who are at high risk of readmission from social factors • Use the teach-back method during the hospital stay from admission to discharge during key clinical interventions • Provide real-time handover communications • Address timely physician follow-up (appointment to occur within 5–7 days of discharge) • Follow-up with the patient or primary caregiver (or emergency contact) within 48–72 hours of discharge through telephone or home visit The five practices draw from a variety of sources, including Project Better Outcomes for Older adults through Safe Transitions (BOOST) and the Institute for Healthcare Improvement’s STAAR initiative. The ACT rests on two core principles: collaboration and transparency. The Collaborative is regional because patients cross routinely from one community within the Cincinnati region to another. It is not bound by hospital structures; even within a competitive environment, hospitals have to share data and communicate with one another to adhere to the five practices. According to the medical director of the Collaborative, implementation and evaluation have been challenging. The Collaborative is not receiving data in real time, which creates delays in the implementation timeline. This lag in data submission also has hampered the Collaborative’s ability to track dollars saved and number of readmissions reduced; however, self-reported data from hospitals participating in the ACT indicate a downward trend in readmissions. The medical director added, “But regardless of whether we meet our goal, the journey and the process has been so helpful and has improved care for patients in our communities.”

Applying the Care Transitions Framework to the Case Study Below, we apply the flowchart and framework to the case study, selecting a few constructs as examples for each step.

Step 1—Define the Intervention Characteristics • What is the intervention designed to achieve? The goal of this intervention is to reduce heart failure readmissions. In a broader sense, the intervention is building a collaborative and working to coordinate care across disparate organizations. • What are the features of the intervention?

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

48

Journal for Healthcare Quality

The intervention for this case has five key elements, all adapted from established care transitions programs but rebundled for this collaborative: • • • • •

Implement a heart failure readmission risk assessment tool Use the teach-back method during key clinical interventions Provide real-time handover communications Address timely physician follow-up Follow-up with the patient or primary caregiver after discharge

Relevant constructs may include feasibility (e.g., whether all these elements can realistically be carried out at all hospitals, including issues of cost-effectiveness), complexity (e.g., difficulty of implementing five disparate elements of the intervention, which will require involvement of a number of providers, including training), and the workflows and task/process standardization that will be needed to incorporate tasks such as teach-back into daily care (Table 1/Appendix Table 1). • Who is the intended target group? The ultimate beneficiary of the changes in practices is the patient, who is thus a primary stakeholder. However, much of the intervention is focused on changes in processes and workflows within and between hospitals, so other targeted entities include providers and staff, units, and hospitals (Table 1/Appendix Table 1).

Step 2—Define the External Context • What components of the environment will impact the implementation? Key constructs for external context may include external networks or existing relationships with outpatient providers who will need to see the patient in a timely way to achieve timely physician follow-up (element #4), and the external pressure and policy incentives to reduce heart failure readmissions (Table 1/Appendix Table 2).

Step 3—Define the Organizations Involved and Their Characteristics • Which organizations are directly involved in the intervention? For this case, the 19 hospitals and health systems are included, as well as community and national organizations: the Health Improvement Collaborative of Greater Cincinnati, Greater Cincinnati Health Council (hospital organization), and the Robert Wood Johnson Foundation. • Which components of structure and process within and between these organizations will affect the implementation? Key structural characteristics may include the size and organizational resources of the various 19 hospitals, which could influence their capacity, internal support, and ability to be flexible enough to make the multiple changes needed. The implementation climate could affect how willing the individual organizations are to change care processes to improve care transitions. Other important constructs include individual organizational accountability for reducing heart failure readmissions in the larger collaborative, relative priority within the organizations to dedicate to the elements of the intervention as compared with other priorities, and leadership engagement to support the organizations and staff in implementing the intervention.

Step 4—Define the Characteristics and Roles of the Providers • What are the characteristics of individuals who are engaged in the provision of care or treatment?

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

49

For this intervention, provider roles may be particularly important, as new roles (especially discharge follow-up) need to be developed within each organization. Collective efficacy, belief that the intervention can be achieved, is needed at the individual and organizational as well as the collaborative level.

Step 5—Define the Characteristics and Roles of Patients and Caregivers • What are the characteristics and roles of patients and caregivers that will affect their ability to engage in the intervention or to benefit from it? The importance of patient goals, needs, preferences, and resources is reflected in the first element of the intervention—risk assessment for readmission, including social factors—and these factors may affect the implementation of the intervention and readmission outcomes. Knowledge and beliefs and skills and competencies are important for the effectiveness of the teach-back method and phone follow-up, and social factors such as access to transportation and a telephone could affect physician follow-up.

Step 6—Define the Processes Required to Achieve Desired Level of Use • What are the implementation processes applied to achieve individual- and organizational-level use of the intervention? For this case, planning is important, with the elements of the intervention chosen from existing programs and rebundled for this collaborative. Engaging patients and caregivers and providers is also critical, with important information exchange in teach-back and postdischarge phone calls for patients/caregivers, and effective handoff communication with providers. Measurement capability and data availability are particularly important for care transitions interventions; in this case, outcomes could not be evaluated due to issues with accessing data from the various organizations involved. This step does not cover how completely an intervention was used; this concept is covered under the Measures of Implementation domain.

Step 7—Define Measures of Implementation • What are the attributes of the implementation process that should be measured to determine how it was carried out and can be sustained? In care transitions interventions, specific elements may not be implemented as planned or may require adjustment during implementation or after initial evaluation. Evaluating what was actually implemented and the measures of implementation is critical to understanding the intervention and outcomes. Potentially useful aspects to be measured reflect many of the constructs described above, including acceptability of the intervention among stakeholders and the fidelity to the established protocol and design within each organization. Reach within the organization would examine the providers involved in care of the patients, while reach within the population would examine patients and caregivers. Better measurement of the implementation process might help to identify barriers and solutions effective in some organizations that could be helpful to others.

Step 8—Define the Outcomes • What are the specific measurable outcomes that will result from the intervention? In this case, the focus was on readmissions as an outcome of healthcare utilization, but many other possible outcomes could have been relevant, including those focused on patients and caregivers, such as achieving patientand caregiver-centered outcomes, the patient/caregiver experience of care, cost effects/impact, and unintended consequences (e.g., the burden of the substantial investment required for follow-up postdischarge).

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

50

Journal for Healthcare Quality

Figure 2. Care transitions framework.

Characteristics and Roles of Providers includes attributes of the individuals (as carriers of cultural, organizational, professional, and individual mindsets, norms, interests, and affiliations) who are engaged in the provision of care or treatment. They may or may not be directly involved in the intervention and/or implementation process. Roles of various providers are key to care transitions interventions. Characteristics and Roles of Patients and Caregivers includes attributes (individual mindsets, norms, interests, and affiliations) of the individuals and caregivers who are the recipients of care or treatment in the given intervention setting. This domain is separate from providers to emphasize the importance of patient-centeredness and of caregivers in care transitions interventions. Process of Implementation is defined as processes (including planning, engaging, and reflecting) to achieve individual- and organizational-level use of the intervention as designed. The construct Physical space includes the presence of organizations, as the physical presence of providers/facilitators from other organizations may be key to building and sustaining collaborations. Engaging relationships between organizations, external context addresses the significance of the external context and community organizations in care transitions interventions. Transition roles addresses the importance of provider roles, including integrators, who

are responsible for building relationships/ collaborations between organizations; their role is central to many sponsored care transitions programs, such as BOOST. Measures of Implementation are the qualities of the implementation and descriptions of how the intervention components are actually implemented, within and between organizations, and changes over time (Proctor et al., 2011). Measurement should involve not just the number and type of interactions with patients and caregivers or between providers but the content and quality of those interactions. These are also referred to as implementation outcomes or intermediate outcomes. The key construct of evolvability reflects the importance of describing adaptations of intervention components and their implementation. The construct reach addresses both reach within the population and reach within the organization. Outcomes are the results of the implementation and the overall targets of the intervention. These include patient-centered outcomes, which include caregivers and emphasize achievement of goals and care consistent with preferences.

Discussion This framework for implementation research for complex interventions and hospital to ambulatory care transition

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

interventions focuses on organizations rather than settings (including communitybased organizations), is conceptualized around patient-centeredness and roles of those involved, and addresses accountability, communication, coordination, and customization of the intervention. Key constructs relevant to the complexity of the intervention include relationships between settings and organizations, measurement capability and data availability across organizations, and adaptability of the intervention. Finally, because these interventions often include elements bundled and adapted for local context, evaluating how implementation occurs and the impact of the different elements are critical issues. In care transitions interventions, context strongly affects details of development and implementation (Brock et al., 2013). Therefore, instead of prescribing a set of domains or constructs for evaluation, the framework provides a comprehensive set of constructs from which implementation researchers can select based on context and research goals. The flowchart (Figure 3) presents step-by-step guidance on how to use the framework to choose where to focus. In many interventions, the discharging setting will have minimal control or relationship with the receiving setting or the external context. Ideally, interventions would try to build links between settings, but when this is not possible, focusing evaluation on the primary setting may be most appropriate. Implementation researchers should also select constructs closely tied to intervention outcomes of interest such as utilization (e.g., readmissions, completion of followup tests) and patient satisfaction with the discharge process. Researchers could conduct initial explorations with providers, administrators, and implementers to determine which evaluation constructs are likely to be most relevant. Selection of constructs and their specifications may also be refined based on emerging data or changes in the intervention process and context, and constructs may be adjusted for specific research needs. The process of developing the framework raised several issues that would benefit from further exploration in im-

51

plementation research. Care transitions interventions have packages (e.g., RED, BOOST) with established steps and protocols, which are often recombined to develop intervention packages tailored to local context and preferences. Methodological research is needed to improve evaluations of how different elements of a bundle are implemented, which are more associated with specific outcomes, and the relative importance of the components in effectiveness (Prvu Bettger et al., 2012). The appropriate role of the patientcentered/population health perspective also requires further development. Intervention design is usually centered around organizations, based on issues such as local policies, structure, payment, funding and research opportunities, and organizations that lead this work; institutional outcomes such as readmissions and cost, rather than patient-centered outcomes such as quality of life, are usually the primary goal. However, the primary purpose of a care transitions intervention is improving patient outcomes, and if the intervention cannot be personalized to heterogeneous patient and caregiver needs such as the high percentage of patients with multiple chronic conditions, it will likely be less effective (Naylor, 2006). The strengths of the comprehensive methods used to develop this framework include building on the established CFIR and similar processes for other complex system interventions, reviewing the literature, and obtaining iterative feedback from experts. Limitations include the focus of the literature reviews on hospital to ambulatory transitions, which could have missed some relevant resources, and restricting expert panel composition to a size amenable to discussion, which may have not included some viewpoints. The qualitative nature of the expert input enabled detailed descriptive evaluation but did not use quantitative surveys or consensus processes. Finally, although we asked the experts to consider if and how they would apply the framework to their implementation research, further evaluation of this framework requires application in practice. In conclusion, this project developed a framework for implementation research

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

52

Journal for Healthcare Quality

Figure 3. How to use the care transitions framework.

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

Vol. 37 No. 1 January/February 2015

on interventions for care transitions, which researchers can use to define potentially relevant constructs and outcomes for implementation studies. Evaluating these interventions requires focusing on patientand caregiver-centered care and engaging organizations and attention to key constructs such as adaptability, accountability, and measurement capacity and data capability. Further evaluation focusing on these and other key constructs can help refine the framework for improving the value of evaluations of these interventions. On the left side of the figure, five domains of interventions for care transitions are shown in a circle that includes the attributes of the intervention itself and the individuals and organizations or systems participating in the intervention. The outer ring of the circle represents the External Context. An arrow to the right of the circle points to the Measures of Implementation, which influence Outcomes. The flowchart presents step-by-step guidance on how to apply this framework. The flowchart presents a series of questions, each tied to a particular domain in the framework. As these questions are considered, the user should refer to the appropriate domain in the framework table to see which constructs are relevant. For example, Step 1 corresponds to the intervention domain, and as the team considers the various issues related to this domain, they should refer to the framework to choose those constructs relevant to their context. The framework does not mandate which constructs should be selected, given the heterogeneity of settings and clinical conditions encountered in these programs.

Acknowledgments We would like to acknowledge the support of AHRQ—Stephanie Chang, MD, MPH, Christine Chang, MD, MPH, and Michael Harrison, PhD. We acknowledge the support of Allen Zhang, Christiane Voisin, Jennifer Cook Middleton, PhD, and Ellen Shanahan, MA, as well as Tim Carey, MD, MPH, and Meera Viswanathan, PhD. The Technical Expert Panel included Mary Blumber, MS, RD, Amy Boutwell, MD, MPP, Jane Brock,

53

MD, MSPH, David Dietz, MSW, MHSA, Brian Jack, MD, Margaret Leonard, MS, RN-BC, FNP, Kathleen McCauley, PhD, RN, Sumant Ranji, MD, Juliana Tiongson, MPH, Anne Sales, PhD, RN, and Mark Williams, MD. This work was supported by funding from the Agency for Healthcare Research and Quality (AHRQ), U.S. Department of Health and Human Services (Contract No. HHSA-2902007-10056I). All statements expressed in this work are those of the authors and should not in any way be construed as official opinions or positions of AHRQ or the U.S. Department of Health and Human Services.

References Agency for Healthcare Research and Quality. Project RED (Re-engineered Discharge) training program. Available at: www.ahrq. gov/professionals/systems/hospital/red/ index.html. Accessed May 1, 2013. American Medical Association. 2009. Care transitions performance measurement set. Available at: www.ama-assn.org/ama/pub/ physician-resources/physician-consortiumperformance-improvement/pcpi-measures. page? Accessed May 1, 2013. Brock, J., Mitchell, J., & Irby, K., et al. Association between quality improvement for care transitions in communities and rehospitalizations among Medicare beneficiaries. JAMA 2013;309:381–391. Centers for Medicare and Medicaid Services. 2013. Community-based care transitions program. Available at: http://innovation. cms.gov/initiatives/CCTP. February 8, 2013. Coleman, EA. The care transitions program. Available at: www.caretransitions.org. Accessed May 1, 2013. Coleman, EA., & Boult, CE. Improving the quality of transitional care for persons with complex care needs. J Am Geriatr Soc 2003;51:556–557. Coleman, EA., Parry, C., Chalmers, S., & Min, SJ. The care transitions intervention: results of a randomized controlled trial. Arch Intern Med 2006;166:1822–1828. Damschroder, LJ., Aron, DC., & Keith, RE., et al. Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science. Implementation Sci 2009;4:50. Forster, AJ., Murff, HJ., & Peterson, JF., et al. The incidence and severity of adverse events affecting patients after discharge from the hospital. Ann Intern Med 2003;138:161–167. Geary, CR., & Schumacher, KL. Care transitions: Integrating transition theory and complexity science concepts. ANS Adv Nurs Sci 2012;35: 236–248.

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

54

Journal for Healthcare Quality

Graham, CL., Ivey, SL., & Neuhauser, L. From hospital to home: assessing the transitional care needs of vulnerable seniors. Gerontologist 2009;49:23–33. Implementation Science. Available at: www. implementationscience.com/about. Accessed February 8, 2013. Institute for Healthcare Improvement. State action on avoidable rehospitalizations (STAAR). Available at: www.ihi.org/engage/ Initiatives/completed/STAAR/Pages/default. aspx. Accessed February 8, 2013. Jencks, S.F., Williams, M.V., & Coleman, E.A. Rehospitalizations among patients in the Medicare fee-for-service program. New Engl J Med 2009;360:1418–1428. McCarthy, D., Cohen, A., & Johnson, M. Case study series synthesis: Innovations in care transitions. gaining ground: care management programs to reduce hospital admissions and readmissions among chronically ill and vulnerable patients. The commonwealth fund. 2013. Available at: http://www.commonwealthfund. org/publications/case-studies/2013/jan/ care-transitions-synthesis. Accessed 1, May 2013. Moore, C., McGinn, T., & Halm, E. Tying up loose ends: discharging patients with unresolved medical issues. Arch Intern Med 2007; 167:1305–1311. National Transitions of Care Coalition. National transitions of care coalition, transitions of care measures. Available at: www.ntocc.org/ Portals/0/PDF/Resources/TransitionsOf Care_Measures.pdf. Accessed May 1, 2013. Naylor, MD. Transitional care: a critical dimension of the home healthcare quality agenda. J Healthc Qual 2006;28:48–54. Naylor, M. Transitional care conceptual model. Available at: www.transitionalcare.info. Accessed May 1, 2013. Naylor, MD., & Sochalski, JA. Scaling up: Bringing the transitional care model into the mainstream. Issue Brief (Commonw Fund) 2010;103:1–12. Proctor, E., Silmere, H., & Raghavan, R., et al. Outcomes for implementation research: conceptual distinctions, measurement challenges, and research agenda. Adm Policy Ment Health 2011;38:65–76. Prvu Bettger, J., Alexander, K.P., & Dolor, R.J., et al. Transitional care after hospitalization for acute stroke or myocardial infarction: a systematic review. Ann Intern Med 2012;157: 407–416. Rennke, S., Nguyen, OK., & Shoeb, MH., et al. Hospital-initiated transitional care interventions as a patient safety strategy: a systematic review. Ann Intern Med 2013;158:433–440. Robert Wood Johnson Foundation. Aligning forces for quality: Spotlight on Cincinnati. Available at: http://forces4quality.org/ spotlight-cincinnati. Accessed May 1, 2013.

Rojas Smith, L., Ashok, M., & Dy, SM., et al. Contextual frameworks for research on the implementation of complex system interventions. Methods research report. (Prepared by the RTI international– university of North Carolina at Chapel Hill evidence-based practice Center under contract No. 290-2007-10056-I.) AHRQ Publication No. 14-EHC014-EF. Rockville, MD: Agency for Healthcare Research and Quality, 2014. Available at: www.effectivehealthcare.ahrq. gov/ehc/products/490/1882/contextualframeworks-complex-interventions-report140318.pdf. Accessed 1, March 2014. Society of Hospital Medicine. Project BOOST. Available at: www.hospitalmedicine.org/ boost. Accessed May 1, 2013.

Authors’ Biographies Sydney M. Dy, MD, MSc, is an Associate Professor of Health Policy and Management, Oncology, and Medicine at Johns Hopkins University and core faculty of the Johns Hopkins Evidence-Based Practice Center and Armstrong Institute for Patient Safety and Quality. Her research focuses on quality of care, including quality measurement, improvement, and synthesis and implementation of evidence-based practices, particularly for patients with serious and advanced illness. Mahima Ashok, PhD, MS, is a Public Health Researcher with the Social and Health Organizational Research and Evaluation program at RTI International. Her research focuses on evidence-based medicine, implementation science, chronic disease, program evaluation, and health information technology. Before joining RTI, Dr. Ashok was a Fellow at the Division of Cancer Prevention and Control at the Centers for Disease Control and Prevention. Roberta C. Wines, MPH, is a Research Associate with the RTI-UNC Evidence-based Practice Center. Her areas of expertise include health services research, policy analysis, and project management. She has contributed to several systematic reviews, metaanalyses, future research needs, and methods projects. Lucia Rojas Smith, DrPH, MPH, is Director of the Social, Health, and Organizational Research and Evaluation Program at RTI International. Her work focuses on evaluating the implementation of organization and system-wide initiatives to transform healthcare and improve patient outcomes. For more information on this article, contact Sydney M. Dy at [email protected]. Supplemental digital content is available for this article. Direct URL citations appearing the printed text and in the HTML and PDF versions of the article at www.jhqonline. com. The authors declare no conflicts of interest.

Copyright 2015 National Association for Healthcare Quality. Unauthorized reproduction of this article is prohibited.

A framework to guide implementation research for care transitions interventions.

Evaluating implementation of complex interventions to improve care transitions and comparison across studies is challenging due to issues such as vari...
547KB Sizes 2 Downloads 7 Views