ABSTRACT
Introduction
High‐quality care transitions are essential for sepsis survivors at risk of readmission and long‐term complications. These complex innovations require coordination across organizations, individuals, and processes. The Consolidated Framework for Implementation Research (CFIR) offers a systematic approach to identifying multi‐level implementation determinants, but translating them into actionable strategies remains challenging. This study examines the representation of implementation determinants across CFIR domains and visualizes their cross‐domain interconnectedness.
Methods
CFIR‐guided, semi‐structured pre‐implementation needs assessment interviews were conducted with stakeholders involved in sepsis care transitions across five hospital‐home healthcare (HHC) dyads. Thirty‐two themes identified through deductive‐inductive coding were mapped to CFIR domains and constructs and visualized using node diagrams.
Results
Sixty‐one interviews with 91 stakeholders identified 32 themes spanning all five CFIR domains, with most themes mapping to multiple domains. Access to Care theme mapped across all five domains, while External Policies and Incentives and Informal Caregiving themes were represented only once. Implementation Process domain was most represented and Innovation least represented. All constructs within the Implementation Process and Individual Characteristics domains were captured. Node diagrams illustrated the depth and interconnected influence of determinants in sepsis care transitions.
Conclusions
Stakeholders identified complexity emerged across technical, structural, and policy domains as both an innovation feature and contextual determinants of sepsis care transitions. CFIR mapping and visualization provided a structured approach to characterizing pre‐implementation complexity in sepsis care transitions. Examining determinant representation and cross‐domain interconnectedness may help identify cross‐cutting, potentially high‐leverage implementation needs and inform strategy selection and tailoring for complex hospital‐to‐home interventions.
Keywords: care transitions, CFIR, complex system mapping, data visualization, home health care, implementation science, qualitative methods, sepsis
1. Introduction
Sepsis is a life‐threatening condition associated with high morbidity, mortality, and healthcare use [1, 2]. Sepsis survivors face elevated risks of rehospitalization [3, 4], persistent physical and cognitive impairments, and psychological distress, creating substantial challenges during care transitions [2, 5].
Home healthcare (HHC) provides skilled, intermittent post‐acute support to monitor recovery, detect deterioration, and facilitate coordination across care settings [6]. Early follow‐up through timely HHC and outpatient care is associated with lower 30‐day readmission rates among sepsis and heart failure patients [7, 8, 9]. As health systems pursue value‐based care, improving care transitions for high‐risk, high‐cost populations like sepsis survivors is a growing priority [10, 11].
Care transition interventions are inherently complex, spanning multiple levels, settings, and stakeholders. While prior studies have examined implementation determinants for transitions among frail older adults [12, 13, 14, 15], sepsis survivors remain underrepresented [16]. The Consolidated Framework for Implementation Research (CFIR) offers a comprehensive structure to examine multilevel contextual determinants [17], which guides strategy development and supports adoption [17, 18]. However, prior CFIR applications have often provided limited insight into how determinants interact across domains/constructs or did not necessarily indicate which should be prioritized or how they should inform implementation strategy selection [19, 20, 21]. Prior work has highlighted the challenge of moving from lists of barriers and facilitators to appropriately matched and contextually relevant implementation strategies [22, 23, 24].
Accordingly, approaches that characterize not only the presence of determinants but also their prominence and relationships across implementation domains may provide additional guidance for prioritization. Thus, this study examines the depth, pervasiveness, and interconnectedness of implementation determinants in a multi‐level sepsis care transitions intervention. By visually mapping the density of theme representation and cross‐domain relationships, this study seeks to illustrate an approach for identifying potentially high‐leverage determinants that may inform strategy selection and tailoring in other complex hospital‐to‐home interventions.
2. Research Interests
This study examines the multilevel and interconnected implementation determinants influencing a sepsis care transitions intervention using CFIR. Specifically, it focuses on characterizing and visualizing the depth, pervasiveness, and cross‐domain relationships among barriers and facilitators across organizational, individual, and policy contexts to generate actionable insights that can inform the design, adaptation, and implementation of complex care transition interventions within learning health systems.
3. Methods
This study draws on qualitative findings from Improving TRansitions ANd outcomeS oF sEpsis suRvivors (I‐TRANSFER), which evaluates a care transitions intervention involving timely skilled HHC and outpatient visits within one week of hospital discharge [25]. Prior to implementation, we conducted a qualitative needs assessment to identify barriers and facilitators relevant to delivering I‐TRANSFER.
The study and qualitative analysis were guided by CFIR, a widely used implementation science determinant framework that supports the systematic translation of evidence into practice. CFIR comprises five domains: (1) Innovation: the “thing” being implemented; (2) Implementation Process: approaches used to carry out the innovation; (3) Individual: roles/characteristics of staff involved in implementation; (4) Inner Setting: organizational environment where implementation occurs; and (5) Outer Setting: external context [17].
Thirty‐two themes derived from 61 semi‐structured interviews with 91 clinicians and staff (e.g., administrators, care coordinators) across five healthcare system–HHC dyads served as the dataset for CIFR mapping. Detailed sampling, analytic procedures, theme definitions, and representative quotes from the qualitative analyses are published elsewhere [26, 27, 28]. Ethical approval was obtained from the University of Pennsylvania (IRB#843958) and VNS Health (IRB #173670‐1). Building on prior I‐TRANSFER work describing pre‐implementation barriers, facilitators, and proposed strategies, this study maps these themes to CFIR domains and constructs and visualizes their interconnectedness.
3.1. Data Analysis
Each theme was reviewed and mapped to the most relevant domains, using the CFIR definitions and codebook [21, 29]. Within each CFIR domain, themes were mapped to the most relevant CFIR constructs; because some themes were broad, they could map to multiple domains or constructs. For example, Augmented Services mapped to Assessing Needs and Tailoring Strategies within the Implementation Process domain because additional services beyond routine care were informed by recipients' assessed needs and adapted accordingly. Similarly, Definition of Sepsis mapped to both the Outer Setting and Inner Setting domains because variation in how sepsis was defined affected implementation both within participating organizations and across the broader care environment. For the Individual domain, Roles and their associated Characteristics were mapped using a Role‐by‐Characteristics matrix to illustrate interaction and highlight differences across roles. Initial mapping was conducted collaboratively by three team members (K.B., P.G., S.C.), and an independent reviewer (S.B.Y.) assessed the results to ensure triangulation and confirm the accuracy. Any disagreements were resolved through iterative discussion until consensus was reached.
To quantify the prominence of CFIR domains, we used raw and normalized counts of theme‐to‐construct mappings to characterize mapping density. Raw counts represented the total number of theme‐to‐construct mappings within each domain, whereas normalized totals represented the average number of themes mapped per construct, accounting for differences in the number of constructs across domains. A higher number of mappings indicated that participants more frequently described barriers or facilitators aligned with that construct, suggesting its perceived salience for implementation delivery. This approach enabled assessment of both absolute and relative representation of CFIR domains. Quantifying prominence in this way provided an objective measure of semantic relationships between CFIR constructs and implementation determinants. To complement this, node diagrams were created using Lucidchart (lucid.co), a web‐based visualization tool, to visually depict relationships between themes and constructs through the density of theme‐construct connections, with construct size reflecting the number of themes and the proximity between CFIR constructs and themes indicating mapping frequency.
4. Results
The 32 themes were mapped to constructs across all five CFIR domains, with most spanning multiple domains. Table 1 outlines I‐TRANSFER's operationalized definitions alongside original CFIR definitions [17] and representative themes. Access to Care was mapped to all five domains, while External Policies and Incentives and Informal Caregiving were each mapped exclusively to a single domain—Outer Setting and Individual, respectively. Table 2 summarizes the frequency of theme‐to‐construct mappings. Raw counts reflect domain prominence, while normalized totals (i.e., average mappings per construct) adjust for the number of constructs per domain. The Implementation Process domain was most represented, with 143 mappings across 9 constructs (average themes per construct = 15.9). The Innovation domain was least represented, with 7 mappings across 8 constructs (average themes per construct = 0.9). Figure 1 provides an overview of how each theme mapped to CFIR. Supporting Information 1 provides detailed theme‐to‐construct mappings by CFIR domain. The included pivot tables (Tables S1 – S5) summarize construct frequency (column totals) and theme prevalence (row totals).
TABLE 1.
CFIR construct descriptions and operationalization in the I‐TRANSFER context.
| CFIR domains and constructs | Original definition a | Operationalization or example of themes represented in I‐TRANSFER |
|---|---|---|
| Innovation domain | The “thing” being implemented. | I‐TRANSFER protocol: first HHC nurse visit within 2 days and the second nurse visit within a week, and additional outpatient visit within a week following sepsis discharge. |
| Innovation Adaptability | The innovation can be adjusted or customized to suit specific local conditions or requirements. | Due to staffing challenges, sites are providing outpatient visits using telehealth and study team has updated (re‐adapted) their criteria for outpatient visits by expanding to incorporate telehealth visits to count as one of initial outpatient follow‐up appointment after hospital discharge. (Telemedicine; Outpatient Provider) |
| Innovation Trialability | Trialing the innovation on a limited scale with the option to reverse it if needed. | One site adopted a Discharge Planning Tool that uses an algorithm to forecast the follow‐up appointment 10 days after admission and presents available slots with affiliated outpatient providers. This approach allowed care teams to begin coordinating follow‐up visits early in the hospital stay, starting with a small‐scale trial of the process (Care Coordination) |
| Innovation Complexity | Complexity of the innovation, involving multiple steps or interrelated components. | Participants reported that delivering innovation is complex given that the innovation consists of multiple complex steps (from sepsis identification to scheduling HHC and outpatient visit) for care coordination and involves multiple stakeholders from different health settings to achieve three components of the innovation within a week timeframe. (Care Coordination) |
| Implementation Process domain | Approaches and actions taken to carry out the innovation. | Strategies used to meet I‐TRANSFER objectives and deliver timely post‐discharge services. |
| Teaming | Work collaboratively and purposefully, coordinating efforts on shared, interrelated tasks to support the implementation of the innovation. | Participants discussed that close communication and information sharing during discharge planning and referrals were critical for effective care coordination. Given that the innovation spans multiple settings and care team members, collaboration across healthcare settings was essential to ensure timely referrals and follow‐up appointments. (Care Coordination; Scheduling) |
| Assessing Needs |
Gather insights on priorities, preferences, and needs from either innovation deliverers or recipients. Includes innovation deliverers and innovation recipients |
Innovation deliverers: Home health clinicians reported that it was often difficult to immediately recognize that a patient was a sepsis survivor—they had to dig into the charts to uncover this information. (Information Transfer; Use of the Problem List; Structured Documentation; Unstructured Documentation) Innovation recipients: Participants from acute care settings were surprised to learn that resolving sepsis on the problem list could influence how the next level of care perceives and manages the patient. They also emphasized the importance of understanding patient needs and the challenges patients may face in scheduling and attending early follow‐up visits—such as lacking caregivers or support to get to appointments. (Patient Education) |
| Assessing Context | Gather and evaluate information to understand what helps or hinders the implementation and delivery of the innovation. | Sepsis is often resolved by the time patients near discharge, causing the diagnosis to be moved to the medical history section. This makes it difficult for clinicians in post‐acute care settings to recognize that the patient had sepsis, reducing the likelihood that the patient will receive the innovation in a timely manner. Additionally, provider and staff behavior around not adding sepsis to the problem list was influenced by the fact that coding and billing are typically based on clinical notes rather than the problem list, leading them to believe it was unnecessary to document sepsis in that section. (Use of the Problem List; Provider or Staff Behavior, Decisions, and Preferences) |
| Planning | Define roles and responsibilities, establish clear steps and milestones, and set goals and success metrics before starting implementation. | To ensure timely outpatient follow‐up visits, one site implemented a new workflow that involved scheduling the outpatient appointment before discharge, acknowledging the complexity and time required for scheduling. Participants also actively discussed strategies to improve sepsis identification using EHR alerts and attaching sepsis educational materials at discharge to raise awareness about the importance of early follow‐up. (Care Coordination; Scheduling; EHR Alert) |
| Tailoring Strategies | Select and tailor implementation strategies to overcome obstacles, build on existing strengths, and align with the specific context. | Telemedicine programs expanded during COVID, which was mentioned as a facilitator for seeing outpatient provider within a week, aligning with the existing workflow for outpatient appointment. (Telemedicine) |
| Engaging |
Engage and motivate innovation deliverers or recipients to take part in implementing and/or adopting the innovation. Includes innovation deliverers and innovation recipients |
Innovation deliverers: Participants emphasized the value of inviting implementation deliverers to brainstorm potential strategies for improving sepsis identification, documentation, and adherence to the three components of the innovation protocol. (Staff Education) Innovation recipients: Participants also highlighted the importance of involving implementation recipients in sepsis education and care coordination to enhance their understanding of the importance of early follow‐up care after sepsis, ultimately supporting greater compliance with the innovation. (Patient Education) |
| Doing | Roll out the innovation through incremental steps, small‐scale trials, or iterative changes to gradually improve its implementation and effectiveness. | Policies, pathways, and workflows within acute care hospitals and HHC agencies influenced when and how the innovation was initiated and delivered, with each setting using its own processes for identifying sepsis patients, documenting the diagnosis, making referrals, and scheduling follow‐up visits. (Acute Care Policies, Pathways, and Processes; Home Health Agency Policies, Pathways, and Processes) |
| Reflecting & Evaluating |
Gather and review both numerical and descriptive data to evaluate how well the innovation and/or its implementation are working. Includes implementation and innovation |
Implementation & Innovation: Each site led data collection, analysis, and reporting on implementation progress—such as the number of sepsis survivors discharged with HHC, their 30‐day readmission rates, and the timing of nursing visits—for critical evaluation using either qualitative or quantitative methods. However, participants noted that most performance metrics had to be manually extracted from the EHR. They suggested automating data retrieval processes and developing live dashboards, with support from IT teams, to more efficiently monitor the quality metrics related to the innovation. (Quality Monitoring of Service) |
| Adapting | Revise the innovation and/or organizational environment to better align with existing workflows and ensure smoother integration. | A participant suggested that once sepsis is documented on the problem list, automatically add a chunk of text regarding the importance of timely follow‐up care and sepsis educational materials to reduce burden on clinicians to remember attaching educational material and better documentation of sepsis without adding multiple steps for providers. (Use of the Problem List; Acute Care Policies, Pathways, and Processes) |
| Individual domain‐Roles subdomain b | Staff roles in implementation. | Roles of individuals involved in or affected by the intervention. |
| High‐level Leaders | Individual in senior positions of authority and have significant influence over decisions and actions. | Those in senior leadership positions such as Chief Medical or Nursing Officers and Directors overseeing Clinical Operations, Care Management, or Quality. They empower their staff to pursue the eight objectives and implement the I‐TRANSFER protocol. |
| Mid‐Level Leaders | Individuals positioned between senior executives and frontline staff. | Individuals responsible for managing areas such as clinical nursing, critical care, and care coordination, who help operationalize implementation activities within their respective departments. |
| Opinion Leaders | Individuals who may not hold formal roles but significantly shape others' attitudes and actions through their informal influence or social capital. | Individuals with their strong passion and influence in improving care and outcomes for sepsis patients. Their impact is not defined by a formal title but by their commitment and respect within the organization. |
| Implementation Facilitators | Individuals who offer expert knowledge and provide guidance, coaching, or assistance to help carry out the implementation. | Individuals our study team consulted to support and optimize the implementation process. They include IT personnel who can modify EHR systems, quality improvement staff, home care experts, report writers, and coders. |
| Implementation Leads | Individuals responsible for guiding and managing the implementation of the innovation such as a champion of implementation sites. | Individuals who motivate staff, support problem‐solving, and coordinate implementation efforts, such as site champions. Although site champions were identified during the pre‐implementation phase, this construct was not mapped directly to the themes as a standalone role because their responsibilities overlapped with those of other primary roles, such as High‐Level Leaders, and they often did not have dedicated time allocated for this role. |
| Implementation Team Members | Individuals who work alongside implementation leads to support the implementation process, ideally involving those who deliver or receive the innovation. | Individuals who are either directly or indirectly involved in delivering the protocol or improving processes to support protocol adherence at any point in the implementation. |
| Other Implementation Support | Individuals assisting the implementation leads and/or team members in carrying out the implementation of the innovation. | Individuals, such as project managers, who help setting up meetings and coordinating logistics of implementing the innovation. I‐TRANSFER team had a project manager, but this role was not mapped to the themes because the needs assessment interviews focused on determinants identified by clinicians and administrative stakeholders at the participating sites, rather than by members of the implementation study team. |
| Innovation Deliverers | Individuals delivering the innovation, either directly or indirectly. | Clinicians and staff such as doctors, nurses, and care coordinators who are directly involved in delivering care for sepsis patients and facilitating their transitions to home healthcare. |
| Innovation Recipients | Individuals receiving the innovation, either directly or indirectly. | Patients diagnosed with sepsis during hospitalization who are transitioning to home healthcare, along with the caregivers who support them through recovery. |
| Individual domain—Characteristics subdomain | Traits of individuals in implementation. | Attributes of individuals involved in or affected by I‐TRANSFER. |
| Need | Individual experience challenges affecting their survival, health, or quality of life, which the implementation or delivery of the innovation aims to improve. | A Mid‐level Leader emphasized that innovation deliverers needed to understand the importance of listing sepsis on the problem list and how post‐discharge workflows could support better case identification. They stressed the need for leadership to acknowledge this and promote staff compliance. Participants also shared insights about the needs of Innovation Recipients—sepsis survivors and their caregivers—including barriers such as reluctance to accept HHC visits or attend follow‐up appointments. When patients did not understand the importance of early follow‐up, they often delayed or declined visits due to discomfort with unfamiliar providers or competing priorities. (Use of the Problem List; Provider or Staff Behavior, Decisions, and Preferences; Patient Education; Patient Behavior, Decisions, and Preferences; Decline or Delay Home Health; No‐show Outpatient) |
| Capability | Individual has social capabilities, understanding, and expertise needed to carry out their role effectively. |
Innovation Deliverers faced difficulty identifying sepsis patients due to nonspecific symptoms and varying sepsis definitions, complicating timely recognition in acute care. Once active treatment ended, the sepsis diagnosis was often marked as resolved and moved to the patient's history, making it harder for downstream providers to recognize them as sepsis survivors and delaying the initiation of the intervention. (Use of the Problem List) Innovation Recipients frequently missed outpatient appointments due to limited transportation options or not having an established primary care provider, making it difficult to schedule visits within 7 days of discharge and reducing attendance rates according to clinicians and staff participants. (Access to Care; Transportation; Outpatient Provider) |
| Opportunity | Individual has the time, responsibility, and authority to perform their role effectively. |
Innovation Deliverers, implementation facilitators, and implementation team members highlighted the opportunity to identify sepsis patients and document sepsis in the medical record using available tools and functionalities within the EHR such as EHR alerts. (EHR Alert; Structured Documentation; Unstructured Documentation) Innovation Recipients were often reported to face challenges in seeing an outpatient provider within the first week after discharge, particularly when they did not have an established primary care provider. This was often the case for those who had recently relocated and needed to first identify a PCP before scheduling an appointment within the recommended timeframe. (Access to Care; Outpatient Provider) |
| Motivation | Individual dedicated to carrying out their responsibilities. | High‐ and mid‐level Leaders were motivated to participate in the implementation study, as it aligned with system goals such as early sepsis identification, reducing 30‐day readmissions, and minimizing reimbursement denials. In addition to innovation deliverers, opinion leaders, facilitators, leads, and team members also needed internal motivation to take responsibility for adhering to the protocol and adjusting workflows to improve sepsis identification and documentation in the EHR. (Provider or Staff Behavior, Decisions, and Preferences; Electronic Health Records [EHR]) |
| Inner Setting | The environment where the innovation is carried out (e.g., hospital, school), which may include several inner settings or hierarchical levels. | Environment where the innovation is implemented – acute care hospitals, HHC agencies, and outpatient offices. |
| Structural Characteristics | Structural elements and resources are in place to ensure the inner setting operates effectively and efficiently. This includes physical infrastructure, information technology infrastructure, and work infrastructure |
Physical Infrastructure: The physical location of outpatient provider offices was mapped to the Inner Setting domain, as it influenced access to care for innovation recipients. (Access to Care; Outpatient Provider) Information Technology Infrastructure: This was primarily mapped to areas related to care coordination and communication, such as using EHR systems to identify sepsis early, document it, and monitor quality metrics associated with the innovation. (Electronic Health Records [EHR]; Quality Monitoring of Service) Work Infrastructure: This included policies, pathways, and processes within acute care and HHC agencies related to sepsis identification, documentation, staff/patient education, discharge planning, referrals, and care coordination. These practices, along with the use of EHR systems, varied by health system. (Acute Care Policies, Pathways, and Processes; Home Health Agency Policies, Pathways, and Processes) |
| Relational Connections | Strong formal and informal connections, networks, and collaborative teams exist within and between parts of the inner setting. | Discharge planning and referral decisions, often made during multidisciplinary rounds, relied on established internal relationships among patient healthcare team to coordinate care and schedule timely follow‐up visits. Relational connections with HHC agencies and outpatient offices—especially when part of the same health system with integrated EHRs—facilitated care coordination process and improved scheduling efficiency. (Acute Care Policies, Pathways, and Processes; Care Coordination; Scheduling) |
| Communications | Reliable and effective formal and informal communication practices exist within and between different parts of the inner setting across different departments and roles. | Different modes for communication were used: secure EHR messaging system, in‐basket messages, fax, or phone calls to communicate and coordinate to schedule for HHC visits and follow‐up appointments within hospital and/or between different inner settings (e.g., hospital‐to‐HHC, hospital‐to‐outpatient office). (Information Transfer; Messaging System; Phone‐fax) |
| Culture | Common values, beliefs, and norms are consistently held and practical throughout the inner setting, including human equality‐centeredness, recipient‐centeredness, deliverer‐centeredness, and learning centeredness |
Recipient‐centeredness: Some sites offered telemedicine for individuals with limited access to care. (Access to Care; Telemedicine) Deliverer‐centeredness: Documentation practices for sepsis varied across health systems, reflecting local clinical cultures and norms around how deliverers prioritize and record sepsis in the EHR. (Provider or Staff Behavior, Decisions, and Preferences; Electronic Health Records [EHR]) Learning‐centeredness: Participating sites monitored service quality, using EHR data to generate reports and review their performance in care delivery. (Quality Monitoring of Service; Electronic Health Records [EHR]) |
| Tension for Change | The present circumstances are unsustainable, creating a strong need for change |
Care coordinators expressed that they were already overwhelmed with their existing responsibilities and lacked the capacity or commitment to dedicate additional attention to sepsis survivors in order to ensure timely HHC visits and outpatient appointments within 1 week. (Staffing) Providers reported that since medical billing and coding is based on free text progress notes, providers often did not feel compelled to use the problem list or to change their existing documentation practices. (Provider or Staff Behavior, Decisions, and Preferences; Unstructured Documentation; Use of the Problem List) |
| Compatibility | Alignment with existing workflows, structures, and operational procedures. | Some participants noted that documenting sepsis on the problem list did not align with their usual workflow, as coding and billing practice at their sites are typically based on provider notes rather than entries in the problem list. (Use of the Problem List; Provider or Staff Behavior, Decisions, and Preferences) |
| Relative Priority | A higher priority relative to other ongoing efforts or initiatives. | When a site was enrolled in the Bundled Payments for Care Improvement program, its processes for treating sepsis patients were reinforced, as enrollment tied reimbursement to patient outcomes—creating incentives to reduce readmissions and improve recovery. (Acute Care Policies, Pathways, and Processes) |
| Mission Alignment | Carrying out and sustaining the innovation aligning with the organization's core mission, purpose, and strategic priorities within the inner setting. | One site had a Sepsis Alliance Group, dedicated to improving sepsis care and outcomes of sepsis patients. Members of Sepsis Alliance were enthusiastic about I‐TRANSFER; their awareness of the vulnerability of sepsis survivors strengthened their motivation to support the innovation. (Acute Care Policies, Pathways, and Processes) |
| Available Resources | Accessibility to necessary resources, including funding, space, and materials and equipment, to support the implementation and delivery of the innovation. |
Every setting had an EHR system, but many lacked the functionality needed to support complete documentation or seamless information transfer to the next level of care using in‐basket and secure chat in EHR. (Electronic Health Records [EHR]; Messaging System) One of the participating sites had a Visiting Outpatient Provider Program where doctors and nurse practitioners deliver medical appointments outside of a traditional outpatient office, which was mentioned to support timely follow‐up visits for sepsis survivors. (Augmented Services) |
| Access to Knowledge and Information | Availability to supportive instruction and educational resources to support the implementation and execution of the innovation. | Participants shared that having supportive guidelines that provide a standard definition of sepsis for identification and diagnosis would facilitate more consistent sepsis recognition. Clinicians within and across health systems are using varying definitions, leading to inconsistencies in sepsis identification. Additionally, access to educational materials for both patients and staff was seen as a facilitator. The capability of generating reports from the EHR also helped clinicians understand their current performance and adherence to the care protocol for each site. (Definition of Sepsis; Patient Education; Staff Education) |
| Outer Setting | The broader context surrounding the Inner Setting (e.g., state, nation), which may include multiple levels of influence. | External influences on implementation inner settings, including those from hospitals, HHC, and outpatient offices, that influence implementation. |
| Critical Incidents | Major or unexpected events interfere with implementing or delivering the innovation. | Participants reported that COVID was a critical incident that impacted general access to care. (Access to Care) |
| Local Conditions | External factors related to economic, environmental, political, and technological conditions that may facilitate the implementation and delivery of the innovation. | The physical environment and geographic location of the inner setting played a key role in determining which HHC agency could serve a patient based on their home address or whether another agency would need to be involved. For patients living in rural areas, both transportation to outpatient appointments and the ability of HHC agencies to provide timely visits were affected. These challenges were often tied to broader social determinants of health, such as access to transportation, provider availability, and geographic isolation. (Access to Care; Geographic Location; Transportation; Social Determinants of Health) |
| Partnerships & Connections | Connections between inner setting and outside organizations via referral systems, academic/professional partnerships, etc. | Sites have collaborated with Virtual Health Centers on improve the identification of sepsis patients in inpatient settings. Also, use of software that supports information transfer and communication between external sites were used to facilitate care coordination. (Acute Care Policies, Pathways, and Processes; Information Transfer) |
| Policies & Laws | Laws and regulatory requirements support implementation and/or delivery of the innovation. |
The evolving nature of international guidelines for sepsis definitions emerged as a barrier to identifying sepsis patients, as different definitions have been used over time across settings. Additionally, the lack of standardized processes for information transfer was identified as a major challenge during care transitions, with participants noting the wide variability in how information was communicated between care settings. Participants also noted that some innovation recipients were reluctant to co‐pay for HHC visits due to competing priorities with daily necessities. (External Policies and Incentives; Information Transfer; Financial Insurance) |
| Financing | Accessibility to financial support from outside sources for implementation and/or delivery of the innovation. | Compliance with CMS programs such as the Value‐based Purchasing and Bundled Payments for Care Improvement was considered essential, as reimbursement is tied to adherence to these models and the quality of care provided. Government‐sponsored transportation support for Medicare who need assistance accessing medical care. One participant highlighted a reimbursement‐related barrier tied to state‐specific policies, although nurse practitioner visits met the protocol's timing requirement, they were not documented for billing purposes because physician visits—despite occurring outside the recommended 7‐day window—were reimbursed at a higher rate. This reflects a misalignment between financial incentives and quality care delivery, which may negatively affect implementation fidelity. (External Policies and Incentives; Financial Insurance; Transportation) |
| External Pressure | External forces that push the implementation and/or delivery of the innovation, including societal pressure, market pressure, performance measurement pressure. | Performance Measurement Pressure: The CMS initiatives to reduce early readmission were reported as external financial pressures that incentivized participating sites to reduce 30‐day readmission rates (i.e., performance measure of sites). One health system reported having a Sepsis Workgroup that regularly reviewed sepsis‐related metrics, such as readmission rates, time to antibiotics, and time to blood culture. These reviews encouraged stakeholders to reflect on their performance and identify ways to improve care processes and patient outcomes. (External Policies and Incentives; Acute Care Policies, Pathways, and Processes) |
Note: Constructs that were addressed by themes are shared in this table. The following constructs of respective domains were not mapped for I‐TRANSFER study: For Inner Setting, Incentive Systems construct was not mapped. For Outer Setting, Local Attitudes construct was not mapped. For Innovation domain, Innovation Source, Innovation Evidence Base, Innovation Relative Advantage, Innovation Design, and Innovation Cost constructs were not mapped.
Abbreviations: CMS, Centers for Medicare and Medicaid Services; COVID, Coronavirus disease 2019; EHR, electronic health records; HHC, home healthcare; I‐TRANSFER, improving transitions and outcomes of sepsis survivors (i.e., Innovation).
Definitions are rephrased from the CFIR 2.0 definition [17].
For Individual Role domain, definitions of each role rather than a summary of the themes represented were shared.
TABLE 2.
Cross‐domain and construct representation of themes a .
| CFIR domains b (impacted constructs/total available constructs) | |||||
|---|---|---|---|---|---|
|
Themes Subthemes |
Innovation (3/8) | Implementation Process (9/9) | Individual | Inner Setting (10/11) | Outer Setting (6/7) |
| Roles (7/9) * Characteristics (4/4) | |||||
| Access to Care | 3 | 5 | 4 | 2 | 5 |
| Augmented Services | — | 2 | — | 2 | 1 |
| Financial Insurance | — | — | 1 | — | 3 |
| Geographic Location | — | — | 2 | — | 3 |
| Outpatient Provider | 1 | — | 5 | 1 | — |
| Telemedicine | 1 | 5 | — | 4 | — |
| Transportation | — | — | 2 | — | 4 |
| Acute Care Policies, Pathways, and Processes | — | 10 | — | 11 | 3 |
| Care Coordination | 1 | 6 | — | 5 | — |
| Scheduling | — | 7 | — | 5 | — |
| Competing Priorities | — | 4 | 2 | 2 | 1 |
| Social Determinants of Health | — | 2 | 1 | — | 3 |
| Definition of Sepsis | — | — | — | 3 | 2 |
| Electronic Health Records (EHR) | — | 7 | 21 | 11 | — |
| EHR Alert | — | 8 | 9 | 10 | — |
| Structured Documentation | — | 8 | 3 | 6 | — |
| Unstructured Documentation | — | 8 | 5 | 6 | — |
| Use of the Problem List | — | 8 | 4 | 10 | — |
| External Policies and Incentives | — | — | — | — | 3 |
| Home health agency Policies, Pathways, and Processes | — | 6 | — | 9 | 3 |
| Informal Caregiving | — | — | 3 | — | — |
| Information Transfer | — | 8 | 4 | 3 | 2 |
| Massaging System | — | 4 | — | 4 | — |
| Phone‐fax | — | 3 | — | 4 | — |
| Patient Behavior, Decisions, and Preferences | — | 4 | 4 | — | — |
| Decline or Delay Home Health | — | 4 | 4 | — | — |
| No‐show Outpatient | — | 4 | 4 | — | — |
| Patient Education | — | 5 | 7 | 7 | — |
| Provider or Staff Behavior, Decisions, and Preferences | — | 7 | 22 | 12 | — |
| Quality Monitoring of Service | 1 | 7 | — | 8 | — |
| Staff Education | — | 4 | 24 | 6 | — |
| Staffing | — | 7 | — | 8 | — |
| Number of Themes Mapped to Each CFIR Domain | 5 | 25 | 20 | 23 | 12 |
| Total of Constructs Mapped to Each Theme c | 7 | 143 | 131 | 139 | 33 |
| Average Themes per Construct d | 0.9 | 15.9 | 14.5 | 12.6 | 4.7 |
Abbreviation: CFIR, the Consolidated Framework for Implementation Research.
Given the high‐level nature of the themes, they were often mapped to multiple domains and constructs rather than being exclusive to one.
Study definition of CFIR domains: Definition of each CFIR domain in the context of I‐TRANSFER: Inner Setting is defined as an environment where the intervention is implemented, including the acute care hospital, HHC agency, and outpatient offices working together to deliver I‐TRANSFER; Outer Setting includes all external factors outside of the acute care and HHC dyad that influence implementation, such as policies, partnerships, and community‐level conditions; Individual includes roles and characteristics of individuals involved in or affected by the I‐TRANSFER protocol; Innovation refers to the I‐TRANSFER protocol, which outlines a care pattern of two nursing visits (within 2 and 7 days) and an outpatient visit within 1 week after sepsis discharge; Implementation Process encompasses the strategies and actions taken to carry out, monitor, and refine the I‐TRANSFER protocol to meet the eight objectives.
Total count for each domain reflects the cumulative number of times themes were mapped to constructs within that domain. Because each CFIR domain contains a different number of constructs, these raw frequencies reflect useful insights into which domains were most frequently represented in the data.
Average number of themes mapped to a construct (i.e., the total number of theme‐construct mappings divided by the number of constructs within each domain).
FIGURE 1.

Overview of themes‐to‐CFIR relationship. The distance of the themes from CFIR domains represents the frequency with which themes (both themes and subthemes) were mapped to constructs within each CFIR domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
4.1. CFIR Domain 1: Innovation
The innovation is the I‐TRANSFER protocol, comprising two nursing visits and an outpatient appointment within 7 days post‐discharge [25]. Five themes (15.6%) were mapped to this domain. Overall, these five themes were mapped to three of eight Innovation constructs (37.5%), reflecting attributes of the intervention itself.
Innovation Complexity was the most represented construct (n = 3; 42.9%), followed by Innovation Adaptability and Trialability (n = 2 each; 57.1%). While the Innovation domain was less frequently represented overall, its relevance emerged at key intersections between the three protocol components and CFIR constructs. Table 1 provides definitions of each CFIR construct along with representative themes.
The theme Access to Care reflected the structural and logistical demands of the protocol: patients with established outpatient relationships experienced smoother transitions, whereas lack of such connections was perceived as an implementation barrier, contributing to Innovation Complexity. Telemedicine, expanded since COVID‐19, was viewed as a facilitator that enhanced Adaptability by reducing reliance on in‐person visits and increasing flexibility. To address barriers to Access to Care, participants proposed piloting discharge planning tools and telemedicine workflows before broader implementation. Testing and refining these approaches on a limited scale to facilitate process reflected the CFIR construct of Innovation Trialability.
Care Coordination further illustrated Innovation Complexity, given the need to align acute, outpatient, and HHC services across multiple providers, organizations, and EHR systems. Early scheduling of follow‐up appointments prior to discharge was identified as a key facilitator of timely protocol execution.
Quality Monitoring of Service theme highlighted reliance on manual tracking or patient self‐report to monitor outpatient appointment attendance. Some participants noted upcoming collaborations with information technology (IT) teams to develop real‐time tracking tools, supporting Trialability and local workflow integration. Figure 2 illustrates the relationships between Innovation constructs and mapped themes, with each theme linked to one to three constructs.
FIGURE 2.

Node diagram of the Innovation domain. The size of the CFIR constructs reflects the number of connections to the themes, while the distance of the themes from the CFIR constructs represents their frequency of mapping to any of the constructs within the Innovation domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
4.2. CFIR Domain 2: Implementation Process
In the context of I‐TRANSFER, this domain captured known or anticipated barriers and facilitators to delivering I‐TRANSFER, particularly those aimed at ensuring timely HHC and outpatient follow‐up after sepsis discharge. A total of 25 themes (78.1%) were mapped to this domain. All nine Implementation Process constructs were represented by the 25 mapped themes, demonstrating the complexity of executing the I‐TRANSFER protocol.
The most frequently mapped constructs were Engaging (n = 25; 17.5%), Assessing Needs (n = 23; 16.1%), Assessing Context (n = 22; 15.4%), and Tailoring Strategies (n = 21; 14.7%). These reflect the iterative, participant‐informed nature of the planning phase, during which strategies were tailored to meet local needs, barriers, and available resources. For example, some sites planned to leverage existing programs (e.g., paramedic or care coordination services targeting high‐risk patients) to deliver I‐TRANSFER components, while others identified Telemedicine as a tool to support timely follow‐up and enhance protocol adherence.
EHR‐related themes were most frequently mapped to the Implementation Process domain (n = 39; 27.3%), aligning with constructs such as Planning, Assessing Needs, Engaging, and Tailoring Strategies. Participants described the EHR as central to delivering I‐TRANSFER, including identifying eligible sepsis patients, initiating HHC referrals, facilitating information transfer, scheduling outpatient visits, and monitoring protocol adherence through dashboards and reports. Shared EHR systems across hospitals, HHC, and outpatient settings were viewed as critical facilitators of real‐time communication, coordination, and ongoing quality monitoring, supported through Teaming, Planning, and Engaging clinicians, IT staff, and leadership in delivering I‐TRANSFER.
Information Transfer (n = 15; 10.5%) was the second most frequently mapped theme and was closely associated with processes related to referrals, communication, and outpatient follow‐up. Effective handoffs across care settings were described as essential for continuity of care. Care Coordination theme (n = 13; 9.1%) was mapped to constructs including Planning and Teaming, highlighting the roles of discharge planners, HHC coordinators, and outpatient schedulers in executing time‐sensitive follow‐up. Early scheduling prior to discharge was emphasized as a key facilitator.
Patient‐related themes such as Patient Behavior, Decisions, and Preferences (n = 12; 8.4%) were mapped to Assessing Needs and Context, Engaging, and Tailoring Strategies. Participants noted that successful implementation ultimately depended on patients' willingness and ability to engage, with missed visits or service refusals posing common barriers. Targeted education and individualized assessment were identified as strategies to enhance adherence. Figure 3 illustrates the distribution of themes within the Implementation Process domain.
FIGURE 3.

Node diagram of the Implementation Process domain. The size of the CFIR constructs reflects the number of connections to the themes, while the distance of the themes from the CFIR constructs represents their frequency of mapping to any of the constructs within the Implementation Process domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
4.3. CFIR Domain 3: Individual: Roles and Characteristics
The Individual domain captured the roles and characteristics of individuals involved in or affected by the intervention. Study participants included only clinicians and staff involved in sepsis care and hospital‐to‐HHC transitions; however, Innovation Recipients (i.e., patients and caregivers) were also represented when clinicians and staff described their needs, behaviors, decisions, and preferences. Examples of key roles included Innovation Deliverers, individuals directly providing care and coordinating transitions; Innovation Recipients, referring to sepsis patients and their caregivers; and Implementation Facilitators, defined as individuals without formal authority but who influenced others' behavior through informal leadership or social capital. Table 1 includes description of roles identified during the needs assessment interviews that could play an influential part in I‐TRANSFER delivery.
Twenty of the 32 themes (62.5%) were mapped to the Individual domain, representing 7 of the 9 Individual Roles constructs (77.8%) and all four Individual Characteristics constructs, highlighting the diverse involvement of individuals across a range of roles and characteristics. With respect to Roles, Innovation Deliverers were represented the most across 14 themes (43.8%), followed by Innovation Recipients across 12 themes (37.5%). Regarding Characteristics, the constructs of Capability and Opportunity were each mapped 37 times (28.2% each), followed by Motivation (n = 31; 23.7%) and Need (n = 26; 19.8%). This suggests that across roles and settings, successful implementation was most often associated with selecting team members who demonstrated sufficient Motivation, equipping them with the necessary skills, tools, and systems (Capability), ensuring that supportive structures were in place (Opportunity), and cultivating a shared understanding of the intervention's purpose (Need).
Table 3 summarizes how themes aligned with individual characteristics across roles. For example, High‐level Leaders were mapped to all four Characteristics, reflecting their broad influence on implementation. Leaders shaped perceptions of Need by prioritizing post‐sepsis follow‐up, reinforced Capability through investments in Staff Education and EHR infrastructure, and created Opportunity by aligning systems (e.g., sepsis alerts, discharge documentation automation workflows) with I‐TRANSFER objectives. Repeated messaging and education also supported Motivation by reinforcing expectations and promoting behavior change of Provider or Staff Behavior, Decisions, and Preferences.
TABLE 3.
Characteristics of each Individual domain Roles and their relation to implementation determinant themes.
| Individual domain | Characteristics subdomain | |||
|---|---|---|---|---|
| Roles subdomain | Need | Capability | Opportunity | Motivation |
| High‐level Leaders |
|
|
|
|
| Mid‐level Leader |
|
|
|
|
| Opinion Leaders |
|
|
|
|
| Implementation Facilitators |
|
|
|
|
| Implementation Leads | — | — | — | — |
| Implementation Team Members |
|
|
|
|
| Other Implementation Support | — | — | — | — |
| Innovation Deliverers |
|
|
|
|
| Innovation Recipients |
|
|
|
|
Note: Transportation, Geographic Location, Outpatient Provider, and Financial Insurance are child themes of Access to Care. Social Determinants of Health is a child theme of Competing Priorities. EHR Alert, Structured Documentation, Unstructured Documentation, and Use of the Problem List are child themes of Electronic Health Records (EHR). No‐show Outpatient and Decline or Delay Home Health are child themes of Patient Behavior, Decisions, and Preferences.
EHR‐related themes and Provider or Staff Behavior, Decisions, and Preferences were among the most frequently mapped within the Individual domain, affecting 7 out of 10 roles. Participants highlighted the EHR's central role in documentation, communication, and action. While structured tools such as alerts facilitated sepsis identification, prompts to add sepsis to the problem list were used inconsistently because of individual and institutional variation, creating barriers to reliable sepsis documentation. Participants noted that progress notes were more consistently maintained as they are closely tied to billing. Differences between structured and unstructured documentation further influenced provider behavior and implementation fidelity. Figure 4 visualizes the connections between Individual Roles and mapped themes.
FIGURE 4.

Node diagram of the Individual: Role domain. The size of the CFIR constructs reflects the number of connections to the themes, while the distance of the themes from the CFIR constructs represents their frequency of mapping to any of the constructs within the Individual domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
4.4. CFIR Domain 4: Inner Setting
The Inner Setting domain was defined as acute care hospitals, HHC agencies, and outpatient offices participating in I‐TRANSFER. Twenty‐three themes (71.9%) were mapped to 10 of the 11 Inner Setting constructs (90.9%), underscoring the broad influence of local context on implementation. These mappings reflect the role of institutional infrastructure, workflows, communication standards, and organizational priorities in shaping the context and determinants of protocol delivery.
Among Inner Setting constructs, Structural Characteristics was most frequently mapped (n = 28; 20.1%), followed by Compatibility and Culture (n = 19 each; 13.7%). Among the subconstructs of Structural Characteristics, themes primarily reflected Information Technology Infrastructure (n = 17; 60.7%), referring to EHR availability and integration; and Work Infrastructure (n = 10; 35.7%), encompassing roles and logistical workflows. Participants emphasized that successful implementation depended on integrating I‐TRANSFER into existing systems without overburdening staff. Alignment with organizational values, routines, and priorities—captured by Compatibility and Culture—was viewed as critical; sites with workflows similar to other care pathways (e.g., heart failure) were perceived as better positioned to adopt I‐TRANSFER.
The themes Acute Care and Home Health Agency Policies, Pathways, and Processes were mapped to Available Resources, Structural Characteristics, Mission Alignment, Culture, and Compatibility, underscoring how institutional priorities shape feasibility. Information Transfer was mapped to Structural Characteristics, Comparability, and Communication, highlighting the importance of EHR interoperability, strong interorganizational communication channels, and organizational Available Resources for effective handoffs.
Themes such as Patient or Staff Education, Staffing, and Quality Monitoring of Service further reflected organizational capacity like Access to Knowledge & Information or Available Resources: sites with dedicated roles (e.g., sepsis coordinators) and analytic infrastructure were better equipped to support timely follow‐up and to extract, monitor, and act on data to streamline the workflow using summary reports. Figure 5 visualizes the complexity of connections between Inner Setting constructs and themes.
FIGURE 5.

Node diagram of the Inner Setting domain. The size of the CFIR constructs reflects the number of connections to the themes, while the distance of the themes from the CFIR constructs represents their frequency of mapping to any of the constructs within the Inner Setting domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
4.5. CFIR Domain 5: Outer Setting
The Outer Setting domain captures contextual factors external to the implementing organizations (namely, acute care hospitals, HHC agencies, and outpatient clinics) that may influence implementation. These include structural, economic, political, and social conditions at the local, state, and national levels that shape the broader environment in which care transitions for sepsis patients occur. Twelve themes (37.5%) mapped to 6 of the 7 Outer Setting constructs (85.7%).
Within the Outer Setting domain, Policies & Laws was the most frequently mapped construct (n = 12; 36.4%), followed by Partnerships & Connections (n = 7; 21.2%); whereas Critical Incidents was the least represented (n = 1; 3%). Themes related to Access to Care mapped across multiple constructs, including Policies & Laws, Partnerships & Connections, and Local Conditions, highlighting policy‐ and system‐level barriers to implementing the care transitions intervention. Participants cited HHC service area boundaries, limited public transportation, and Medicaid coverage constraints, including narrower provider networks and lower reimbursement, as barriers to timely follow‐up. Critical Incidents was particularly associated with the COVID‐19 pandemic. Participants described increased patient hesitancy toward in‐home services because of infection concerns related to unfamiliar visitors, alongside greater acceptance of telemedicine, which was viewed as an alternative strategy to improve access to outpatient care.
Structural and socioeconomic conditions—such as limited insurance coverage, financial insecurity, and transportation barriers—were frequently cited as constraints on Access to Care. Competing Priorities was mapped to Financing, Local Conditions, and Policies & Laws, reflecting how Social Determinants of Health (e.g., food insecurity, transportation costs) can affect patients' ability to prioritize follow‐up care. Upstream influences, including reimbursement policies, insurance benefit design (e.g., out‐of‐pocket costs or co‐pays), and public transit infrastructure, were also described as shaping the Local Conditions in which I‐TRANSFER will be delivered.
The theme Acute Care and Home Health Agency Policies, Pathways, and Processes captured how national regulations influenced institutional workflows, including documentation and coding practices tied to reimbursement. These themes were mapped to Policies & Laws and External Pressure. Participants noted that organizational quality initiatives often prioritized conditions such as heart failure over sepsis, reflecting alignment with national incentive programs (e.g., bundled payment initiatives). Partnerships & Connections reflected the role of community resources and interorganizational collaborations, such as transportation programs for Medicare beneficiaries or regional hospital‐HHC networks, in supporting implementation of I‐TRANSFER. The theme Definition of Sepsis was mapped to Policies & Laws and External Pressure, reflecting how variation in diagnostic criteria, reimbursement requirements, and payer policies influenced institutional approaches to sepsis identification and documentation for I‐TRANSFER.
The External Policies and Incentives theme captured how payment and regulatory structures influenced implementation readiness and was mapped to External Pressure, Policies & Laws, and Financing. Participants cited policy alignment—such as CMS requirements to initiate HHC within 2 days of discharge and the Hospital Readmission Reduction Program—as drivers that heightened institutional attention to high‐risk conditions like sepsis and shaped readiness for implementation.
Finally, Information Transfer theme was mapped to Policies & Laws and Partnerships & Connections, underscoring how limited standardized data‐sharing policies contributed to communication gaps during patient transitions from hospital to post‐acute care. Figure 6 illustrates the interconnected relationships among Outer Setting constructs and the themes.
FIGURE 6.

Node diagram of the Outer Setting domain. The size of the CFIR constructs reflects the number of connections to the themes, while the distance of the themes from the CFIR constructs represents their frequency of mapping to any of the constructs within the Outer Setting domain. Node diagram created in Lucid (lucid.co). CFIR, consolidated framework for implementation research.
5. Discussion
This study used CFIR to examine context and implementation determinants shaping a sepsis care transitions intervention through pre‐implementation interviews. By mapping themes to CFIR at both construct and theme levels, the study provides a nuanced view of how determinants interact with program‐specific needs, illustrating their cross‐cutting and relational nature while offering context‐rich insights within a standardized framework. All five CFIR domains were represented across 32 themes, underscoring the multi‐level complexity of care transitions involving diverse stakeholders and interdependent systems. Complexity emerged both as a feature of the innovation and a broader contextual determinant spanning technical, structural, and policy domains. Identifying and addressing these factors during the pre‐implementation is critical for tailoring strategies to local context and creating conditions for successful and sustainable sepsis care transitions [30]. The identified determinants also aligned with all core elements of the national Hospital Sepsis Program, including leadership engagement, designated accountability, multidisciplinary expertise, structures and processes, tracking progress, reporting outcomes, and educating stakeholders to optimize patient care and improve hospital outcomes [31].
5.1. Innovation: Complexity of the Sepsis Care Transition Intervention
Care transition interventions, such as I‐TRANSFER, are inherently complex due to their multicomponent design and the need for cross‐setting coordination [14], which in this case required 31 distinct tasks just to complete referral to HHC [32]. These interventions rely on timely follow‐up, communication, and education—areas where breakdowns can result in missed care and adverse outcomes, especially for sepsis survivors [7, 15, 33]. While such complexity has been noted in prior research, our coordination adds nuance by showing that factors related to access to care and care coordination contributed most to the complexity of sepsis care transitions. Sites shared that they could navigate this complexity by leveraging existing workflows for managing other high‐risk conditions (e.g., heart failure) and adapting them for I‐TRANSFER, consistent with literature emphasizing Adaptability as essential for addressing innovation Complexity [34, 35]. Tailoring the strategies, such as adopting telemedicine, was seen as critical [17, 36, 37], though excessive adaptation risks undermining core components [37]. Thus, balancing local tailoring with fidelity and incorporating continuous monitoring were recommended to manage complexity as a dynamic, navigable feature rather than a barrier.
5.2. Implementation Process: Complexity of Coordination and Information Transfer
Implementation planning required navigating local workflows, resource constraints, and coordination across organizational boundaries. Participants described challenges in aligning referrals, EHR documentation, scheduling, and communication, consistent with prior studies [14, 16]. EHR functionality and information transfer emerged as key areas for improvement. Despite widespread EHR adoption, interoperability gaps and inconsistent documentation practices hindered timely patient identification and follow‐up [27]. While national policies promote data exchange and patient access [38], fragmented systems persist. When effectively implemented, transparency may enhance patient engagement and recovery, though the potential for information overload or misinterpretation highlighted the need for tailored patient education [39].
5.3. Individual Characteristics: Complexity of Roles and Behavior Change
Multiple stakeholders, including leaders, frontline clinicians, and IT staff, shaped implementation success, consistent with literature emphasizing early engagement across roles [40]. Role ambiguity increased coordination challenges, suggesting the need for clear expectations, targeted education, and designated implementation champions. Identifying champions and using snowball sampling may help sustain engagement and surface informal influencers [41, 42]. Individual capability and motivation also influenced adherence and were shaped by leadership support, workflow alignment, and incentives. Patient and caregiver engagement emerged as additional threats to protocol adherence [27, 43], as barriers such as transportation, caregiver availability, and perceived need could limit participation despite adequate planning [28, 44]. These barriers may lead some recipients to be unwilling or unable to engage with the innovation, such as refusing home care visits or missing outpatient appointments.
5.4. Inner Setting: Readiness and Workflow Compatibility
Organizational structures, resources, and cultures shaped how complexity was perceived and addressed across sites. Existing workflows for similar conditions (e.g., heart failure) signaled organizational readiness and facilitated smoother integration [35]. In contrast, sites lacking infrastructure—such as EHR alerts, care coordinators, or analytic capacity—faced greater implementation challenges. Assessing organizational readiness with tools like the Organizational Readiness for Implementing Change scale can help tailor strategies to local conditions [45]. Upon our assessment, post‐acute care participants demonstrated higher readiness for delivering I‐TRANSFER compared with their acute care counterparts. While larger systems may have greater resources, they also face added complexity due to bureaucratic processes for change [46]. Given that organizations varied in access to resources and workflows, cross‐site collaboration was suggested to support shared learning and adaptation, reinforcing the value of iterative knowledge exchange in navigating complexity within local contexts and aligning with the goals of learning health systems [47, 48].
5.5. Outer Setting: Social and Structural Complexity
External policy and social factors shaped organizational policies, patient adherence, and provider implementation efforts. Programs such as the Hospital Readmissions Reduction Program created incentives but did not always align resources with expectations. Transportation barriers, insurance coverage, and unmet social needs frequently impeded patient adherence, particularly among Medicaid beneficiaries [49]. The ICD‐10‐CM code for Sepsis Aftercare (Z51.A) was intended to support continuity but may see limited uptake due to reimbursement misalignment, highlighting the need for policy engagement [50].
EHR documentation practices further illustrate the need for alignment with external policies. While structured fields like problem lists support information transfer, they are often underused because they lack billing relevance. In contrast, progress notes—required for reimbursement—receive more attention, leading to inconsistent communication of key conditions like sepsis that require follow‐up [51, 52]. Addressing this disconnect will require stronger leadership engagement at both institutional‐ and policy‐levels.
5.6. Implications for Implementation Strategy Selection
This study approach illustrates how CFIR‐guided pre‐implementation assessment can move beyond identifying barriers and facilitators to support prioritization of determinants for strategy selection and tailoring. This approach may be particularly useful when pre‐implementation assessments yield numerous CFIR determinants and it is unclear which should be addressed first. Visualizing these relationships through node diagrams can help implementation teams move beyond a descriptive inventory of barriers and facilitators toward a more structured consideration of determinant prominence and interconnectedness. Such information may help identify high‐leverage targets, particularly in complex hospital‐to‐home and care transition interventions that span multiple organizations, roles, and care settings. This approach may complement implementer prioritization and other determinant‐to‐strategy matching methods in future implementation studies. More broadly, this approach illustrates how learning health systems can translate context‐specific implementation data into actionable knowledge to support strategy selection, adaptation, and continuous learning in real‐world settings.
5.7. Strengths and Limitations
Strengths of this study include its multi‐site design, which enhances the transferability of findings, and the use of visual CFIR mappings to illustrate interconnected implementation determinants, offering a structured, pragmatic guide for implementation teams. Limitations include the lack of direct perspectives from patients and their caregivers, which were indirectly inferred from clinicians and staff interviews. Although post‐implementation outcomes were not examined in this study, they are evaluated in a separate analysis. In addition, CFIR constructs are not mutually exclusive; in multi‐level interventions like I‐TRANSFER, themes often span multiple constructs and domains rather than mapping to a single construct. CFIR may also overlook dynamic interactions across macro‐, meso‐, and micro‐levels, highlighting the need for future work to integrate complementary frameworks and evaluate determinant influence during implementation.
6. Conclusions
This study applied the CFIR framework to a multi‐level, cross‐setting sepsis care transitions intervention to identify key determinants across components, processes, roles, and systems. CFIR offered a structured approach to capture these determinants, while node diagrams highlighted their pervasiveness and interconnections to inform context‐specific strategies. Findings provide insights to improve sepsis care transitions and may guide interventions for other high‐risk populations.
Author Contributions
Sang Bin You: conceptualization, formal analysis, methodology, writing – original draft, writing – review and editing. Michael A. Stawnychy: formal analysis, methodology, writing – review and editing. Jiyoun Song: conceptualization, formal analysis, writing – review and editing. Karen B. Hirschman: methodology, writing – review and editing. Melissa O'Connor: formal analysis, writing – review and editing. Elaine Sang: formal analysis, writing – review and editing. Sungho Oh: formal analysis, writing – review and editing. Katherine Pitcher: writing – review and editing. Patrik Garren: project administration, formal analysis, writing – review and editing. Samantha Cueto: formal analysis, writing – review and editing. Kathryn H. Bowles: conceptualization, formal analysis, methodology, supervision, writing – review and editing, funding acquisition.
Funding
The research team of PI Kathryn Bowles Co‐Investigators Karen Hirschman, Michael Stawnychy, Melissa O'Connor, research assistant Sang Bin You, and staff Patrik Garren are supported by the National Institute of Nursing Research (NINR) grant number 2R01NR016014. Elaine Sang, Sungho Oh, Jiyoun Song, and Katherine Pitcher were supported by NINR T32NR009356. In addition, Sang Bin You was supported by the National Institute on Aging (NIA) F99AG095132, Jiyoun Song was supported via the National Heart, Lung, and Blood Institute (NHLBI) K99HL169940, and Elaine Sang was supported by NINR F31NR021242. Melissa O'Connor was supported by a Betty Irene Moore Fellowship for Nurse Leaders and Innovators. These funding sources had no role in the design of this study and will not have any role during its execution, analyses, interpretation of the data, or decision to submit results. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINR, NIA, NHLBI, the Gordon and Betty Moore Foundation, or the National Institute of Health.
Ethics Statement
Ethical approval was granted by the University of Pennsylvania (IRB#843958), VNS Health (IRB#173670‐1), and other participating sites that verified no direct involvement with their patients or staff.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Theme‐to‐construct mapping for the Innovation domain.
Table S2: Theme‐to‐construct mapping for the Implementation Process domain.
Table S3: Theme‐to‐construct mapping for the Individual domain.
Table S4: Theme‐to‐construct mapping for the Inner Setting domain.
Table S5: Theme‐to‐construct mapping for the Outer Setting domain.
Acknowledgments
We gratefully acknowledge the participants for generously sharing their insights to help enhance care and outcomes for individuals affected by sepsis.
Data Availability Statement
The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.
References
- 1. Jones T. K., Fuchs B. D., Small D. S., et al., “Post‐Acute Care Use and Hospital Readmission After Sepsis,” Annals of the American Thoracic Society 12 (2015): 904–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. van der Slikke E. C., Beumeler L. F. E., Holmqvist M., et al., “Understanding Post‐Sepsis Syndrome: How Can Clinicians Help?,” Infection and Drug Resistance 16 (2023): 6493–6511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. You S. B., Song J., Hsu J. Y., and Bowles K. H., “Characteristics and Readmission Risks Following Sepsis Discharges to Home,” Medical Care 63 (2025): 89–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Sun A., Netzer G., Small D. S., et al., “Association Between Index Hospitalization and Hospital Readmission in Sepsis Survivors*,” Critical Care Medicine 44 (2016): 478–487. [DOI] [PubMed] [Google Scholar]
- 5. Duda J., “The Walking Wounded: Sepsis Survivors Grapple With Post‐Sepsis Syndrome,” Lancet. Respiratory Medicine 13, no. 9 (2025): 786–788. [DOI] [PubMed] [Google Scholar]
- 6. CMS , Medicare and Home Health Care (CMS, 2003), https://www.cms.gov/Medicare/Quality‐Initiatives‐Patient‐Assessment‐Instruments/HomeHealthQualityInits/Downloads/HHQIHHBenefits.pdf. [Google Scholar]
- 7. Deb P., Murtaugh C. M., Bowles K. H., et al., “Does Early Follow‐Up Improve the Outcomes of Sepsis Survivors Discharged to Home Health Care?,” Medical Care 57 (2019): 633–640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Song J., Zolnoori M., McDonald M. V., et al., “Factors Associated With Timing of the Start‐Of‐Care Nursing Visits in Home Health Care,” Journal of the American Medical Directors Association 22 (2021): 2358–2365.e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Murtaugh C. M., Deb P., Zhu C., et al., “Reducing Readmissions Among Heart Failure Patients Discharged to Home Health Care: Effectiveness of Early and Intensive Nursing Services and Early Physician Follow‐Up,” Health Services Research 52 (2017): 1445–1472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. CMS , Hospital Readmissions Reduction Program (HRRP) (CMS, 2025), https://www.cms.gov/medicare/payment/prospective‐payment‐systems/acute‐inpatient‐pps/hospital‐readmissions‐reduction‐program‐hrrp. [Google Scholar]
- 11. CMS , Hospital Value‐Based Purchasing (CMS, 2024), accessed July 14, 2025, https://www.cms.gov/medicare/quality/value‐based‐programs/hospital‐purchasing. [Google Scholar]
- 12. McNeil D., Strasser R., Lightfoot N., and Pong R., “A “Simple” Evidence‐Based Intervention to Improve Care Transitions for Frail Patients With Complex Health Conditions: Why Didn't It Work as Expected?,” Healthcare Quarterly 19 (2016): 67–72. [DOI] [PubMed] [Google Scholar]
- 13. Fakha A., de Boer B., Hamers J. P., et al., “Systematic Development of a Set of Implementation Strategies for Transitional Care Innovations in Long‐Term Care,” Implementation Science Communications 4 (2023): 103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Fakha A., Groenvynck L., de Boer B., et al., “A Myriad of Factors Influencing the Implementation of Transitional Care Innovations: A Scoping Review,” Implementation Science: IS 16 (2021): 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Nurjono M., Shrestha P., Ang I. Y. H., et al., “Implementation Fidelity of a Strategy to Integrate Service Delivery: Learnings From a Transitional Care Program for Individuals With Complex Needs in Singapore,” BMC Health Services Research 19 (2019): 1–14, 10.1186/s12913-019-3980-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Collet R., Van Grootel J., Van Der Leeden M., et al., “Facilitators, Barriers, and Guidance to Successful Implementation of Multidisciplinary Transitional Care Interventions: A Qualitative Systematic Review Using the Consolidated Framework for Implementation Research,” International Journal of Nursing Studies Advances 8 (2025): 100269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Damschroder L. J., Reardon C. M., Widerquist M. A. O., and Lowery J., “The Updated Consolidated Framework for Implementation Research Based on User Feedback,” Implementation Science 17 (2022): 75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Fernandez M. E., “Implementation Mapping: Using Intervention Mapping to Develop Implementation Strategies,” Frontiers in Public Health 7 (2019): 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Bourke J. A., Jerram K. A. S., Arora M., Craig A., and Middleton J. W., “Using the Consolidated Framework for Implementation Research to Integrate Innovation Recipients' Perspectives Into the Implementation of a Digital Version of the Spinal Cord Injury Health Maintenance Tool: A Qualitative Analysis,” BMC Health Services Research 24 (2024): 390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lewinski A. A., Crowley M. J., Miller C., et al., “Applied Rapid Qualitative Analysis to Develop a Contextually Appropriate Intervention and Increase the Likelihood of Uptake,” Medical Care 59 (2021): S242–S251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. McCullogh E., Giles A., Macpherson A., et al., “Applying the Consolidated Framework for Implementation Research (CFIR) to Examine Barriers and Facilitators to Built Environment Change in Five Canadian Municipalities: Lessons From Road Safety and Injury Prevention Professionals,” Journal of Transport and Health 27 (2022): 101478. [Google Scholar]
- 22. Powell B. J., Beidas R. S., Lewis C. C., et al., “Methods to Improve the Selection and Tailoring of Implementation Strategies,” Journal of Behavioral Health Services & Research 44 (2017): 177–194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Waltz T. J., Powell B. J., Fernández M. E., Abadie B., and Damschroder L. J., “Choosing Implementation Strategies to Address Contextual Barriers: Diversity in Recommendations and Future Directions,” Implementation Science 14 (2019): 42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Balis L. E. and Houghtaling B., “Matching Barriers and Facilitators to Implementation Strategies: Recommendations for Community Settings,” Implementation Science Communications 4 (2023): 144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. O'Connor M., Kennedy E. E., Hirschman K. B., et al., “Improving Transitions and Outcomes of Sepsis Survivors (I‐TRANSFER): A Type 1 Hybrid Protocol,” BMC Palliative Care 21 (2022): 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Bowles K. H., Stawnychy M. A., O'Connor M., et al., “Context and Determinants for Implementing a Sepsis Survivor Care Transition Intervention Reported From Five Health Systems and Home Health Agencies,” Frontiers in Medicine 12 (2025): 1632083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. You S. B., Hirschman K. B., Stawnychy M. A., et al., “Qualitative Study of the Context of Health Information Technology in Sepsis Care Transitions: Facilitators, Barriers, and Strategies for Improvement,” Journal of the American Medical Directors Association 26 (2025): 105606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Sang E., Hirschman K. B., Stawnychy M. A., et al., “Patient‐Level Barriers, Related Facilitators, and Proposed Strategies for Timely Home Health Care and Outpatient Appointments for Sepsis Survivors: Perspectives From Healthcare System and Home Health Care Informants,” Patient Education and Counseling 138 (2025): 109207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Melder A., Robinson T., Mcloughlin I., et al., “Integrating the Complexity of Healthcare Improvement With Implementation Science: A Longitudinal Qualitative Case Study,” BMC Health Services Research 22 (2022): 1–12, 10.1186/s12913-022-07505-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Graham N. D., Graham I. D., Vanderspank‐Wright B., Nadalin‐Penno L., Fergusson D. A., and Squires J. E., “Planning for Implementation Success: Insights From Conducting an Implementation Needs Assessment,” JBI Evidence Implementation 23 (2025): 90–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. CDC , Hospital Sepsis Program Core Elements (Sepsis, 2024), https://www.cdc.gov/sepsis/hcp/core‐elements/index.html. [Google Scholar]
- 32. Oh S., Sang E., Stawnychy M. A., et al., “Application of a Human Factors and Systems Engineering Approach to Explore Care Transitions of Sepsis Survivors From Hospital to Home Health Care,” Human Factors: The Journal of the Human Factors and Ergonomics Society 66, no. 11 (2024): 2468–2484, 10.1177/001872082312223999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Hirschman K. B., Shaid E., Bixby M. B., et al., “Transitional Care in the Patient‐Centered Medical Home: Lessons in Adaptation,” Journal for Healthcare Quality 39 (2017): 67–77. [DOI] [PubMed] [Google Scholar]
- 34. Wiltsey Stirman S., Baumann A. A., and Miller C. J., “The FRAME: An Expanded Framework for Reporting Adaptations and Modifications to Evidence‐Based Interventions,” Implementation Science 14 (2019): 1–10, 10.1186/s13012-019-0898-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Chambers D. A., Glasgow R. E., and Stange K. C., “The Dynamic Sustainability Framework: Addressing the Paradox of Sustainment Amid Ongoing Change,” Implementation Science 8 (2013): 1–11, 10.1186/1748-5908-8-117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Ezeamii V. C., Okobi O. E., Wambai‐Sani H., et al., “Revolutionizing Healthcare: How Telemedicine Is Improving Patient Outcomes and Expanding Access to Care,” Cureus (2024): 1–9, 10.7759/cureus.63881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Lyng H. B., Macrae C., Guise V., et al., “Balancing Adaptation and Innovation for Resilience in Healthcare – A Metasynthesis of Narratives,” BMC Health Services Research 21 (2021): 1–13, 10.1186/s12913-021-06592-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. CMS , “CMS Interoperability and Patient Access Final Rule,” 2025, https://www.cms.gov/priorities/burden‐reduction/overview/interoperability/policies‐and‐regulations/cms‐interoperability‐and‐patient‐access‐final‐rule‐cms‐9115‐f.
- 39. D'Costa S. N., Kuhn I. L., and Fritz Z., “A Systematic Review of Patient Access to Medical Records in the Acute Setting: Practicalities, Perspectives and Ethical Consequences,” BMC Medical Ethics 21 (2020): 1–19, 10.1186/s12910-020-0459-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Potthoff S., Finch T., Bührmann L., et al., “Towards an Implementation‐STakeholder Engagement Model (I‐STEM) for Improving Health and Social Care Services,” Health Expectations 26 (2023): 1997–2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Santos W. J., Graham I. D., Lalonde M., et al., “The Effectiveness of Champions in Implementing Innovations in Health Care: A Systematic Review,” Implementation Science Communications 3 (2022): 1–48, 10.1186/s43058-022-00315-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Naderifar M., Goli H., and Ghaljaie F., “Snowball Sampling: A Purposeful Method of Sampling in Qualitative Research,” Strides in Development of Medical Education 14 (2017): 1–4, 10.5812/sdme.67670. [DOI] [Google Scholar]
- 43. Aarons G. A., Hurlburt M., and Horwitz S. M., “Advancing a Conceptual Model of Evidence‐Based Practice Implementation in Public Service Sectors,” Administration and Policy in Mental Health 38 (2011): 4–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Fuji K. T., Abbott A. A., and Norris J. F., “Exploring Care Transitions From Patient, Caregiver, and Health‐Care Provider Perspectives,” Clinical Nursing Research 22 (2013): 258–274. [DOI] [PubMed] [Google Scholar]
- 45. Shea C. M., Jacobs S. R., Esserman D. A., Bruce K., and Weiner B. J., “Organizational Readiness for Implementing Change: A Psychometric Assessment of a New Measure,” Implementation Science 9 (2014): 7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Sang E., Quinn R., Stawnychy M. A., et al., “Organizational Readiness for Change Towards Implementing a Sepsis Survivor Hospital to Home Transition‐In‐Care Protocol,” Frontiers in Health Services 4 (2024): 1–9, 10.3389/frhs.2024.1436375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Powell B. J., Waltz T. J., Chinman M. J., et al., “A Refined Compilation of Implementation Strategies: Results From the Expert Recommendations for Implementing Change (ERIC) Project,” Implementation Science 10 (2015): 1–14, 10.1186/s13012-015-0209-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Friedman C. P., “What Is Unique About Learning Health Systems?,” Learning Health Systems 6 (2022): e10328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Kepper M., Walsh‐Bailey C., Owens‐Jasey C., Gunn R., and Gold R., “Integrating Social Needs Into Health Care: An Implementation Science Perspective,” Annual Review of Public Health 46 (2025): 151–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Crable E. L., Lengnick‐Hall R., Stadnick N. A., Moullin J. C., and Aarons G. A., “Where Is “Policy” in Dissemination and Implementation Science? Recommendations to Advance Theories, Models, and Frameworks: EPIS as a Case Example,” Implementation Science 17 (2022): 80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Hodge C. M. and Narus S. P., “Electronic Problem Lists: A Thematic Analysis of a Systematic Literature Review to Identify Aspects Critical to Success,” Journal of the American Medical Informatics Association 25 (2018): 603–613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Wright A., McCoy A. B., Hickman T.‐T. T., et al., “Problem List Completeness in Electronic Health Records: A Multi‐Site Study and Assessment of Success Factors,” International Journal of Medical Informatics 84 (2015): 784–790. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Theme‐to‐construct mapping for the Innovation domain.
Table S2: Theme‐to‐construct mapping for the Implementation Process domain.
Table S3: Theme‐to‐construct mapping for the Individual domain.
Table S4: Theme‐to‐construct mapping for the Inner Setting domain.
Table S5: Theme‐to‐construct mapping for the Outer Setting domain.
Data Availability Statement
The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.
