Lung transplantation process

Lung transplantation is a complex, highly coordinated, multistep process involving multiple key constituent groups. After a patient with advanced lung disease is listed for transplantation, donor identification and assessment begin. Donors may be declared after brain death (DBD) or after controlled circulatory death (DCD). Donor suitability is assessed using standard or extended criteria [11,25], and then donors are actively managed to optimize organ function [11]. During this period, donor lungs are vulnerable to injury from ventilator-associated lung injury [11,26,27], aspiration and pulmonary infection [11,26], chest trauma related to the cause of death and/or resuscitation maneuvers [11], and fluid overload [26,28]. These vulnerabilities, combined with conservative acceptance thresholds, contribute to the low national lung utilization rate [29]. Donor lung evaluation requires integration of clinical data while balancing recipient factors and logistical considerations.

Organ procurement organizations (OPOs) coordinate donor evaluation, management, and organ allocation, thereby serving as the critical link between donor hospitals and transplant centers. Variation in donor management practices (e.g., ventilation parameters, diuresis, bronchoscopy, and treatment of infection) can substantially influence whether lungs are ultimately transplanted [17,30–34]. Following donor management, donor lungs are offered to a potential recipient on the waitlist according to established allocation policies that integrate disease severity, comorbidities, and predicted benefit of transplantation [7,11,35]. The potential recipient’s transplant team then decides whether to accept the offer, influenced by center-specific thresholds, provider experiences, and recipient acuity. Once donor lungs are allocated, a specialized procurement team retrieves and transports them to the recipient hospital, where the transplantation surgery is performed, completing the process from listing to implantation.

Transplant providers make several critical decisions along this pathway that directly influence organ utilization. They are often made under time substantial time constraints related to donor stability, organ allocation logistics, and recipient readiness. Given this inherent complexity and variability, adoption of standardized, data-driven tools is essential to promote consistent, evidence-based decision-making in donor evaluation and organ acceptance.

Lung donor (LUNDON) acceptability score

To address the need for more standardized, evidence-based tools for lung donor evaluation, our group previously developed and validated the Lung Donor (LUNDON) Acceptability Score [33], a predictive model that estimates the likelihood of donor lung utilization based on nine clinically relevant variables (Table 1) and correlates with post-transplant recipient outcomes [34]. Using a large national cohort from the Scientific Registry of Transplant Recipients (SRTR), we derived and validated the model in over 80,000 brain-dead donors. The LUNDON score demonstrates excellent discrimination and calibration across diverse donor populations [36,37]. The score is operationalized as a user-friendly, web-based tool (https://sites.wustl.edu/lundon/) that provides real-time case-specific probability estimates and supports more objective and reproducible decision-making in donor evaluation and management.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. LUNDON score component variables.

https://doi.org/10.1371/journal.pone.0342383.t001

Four of the nine component variables in the LUNDON score are potentially modifiable (Table 1), including (1) donor PaO2/FiO2 [P/F] ratio, (2) chest x-ray appearance, (3) creatinine, and (4) bloodstream infection. Prior work by our group has shown that lung-protective ventilation strategies and goal-directed donor management can improve oxygenation, radiographic findings, and renal function, corresponding with higher organ utilization rates [17,38,39,41]. These findings indicate that targeted, evidence-based interventions can improve LUNDON scores and expand the donor lung pool, providing actionable targets for OPOs to enhance donor management practices.

Implementation gap

Despite strong validation across multiple studies, the LUNDON score remains underutilized in routine donor management and transplant decision-making. This reflects a broader and well-recognized phenomenon in medicine, where the translation of evidence-based innovations into routine practice is slow, inconsistent, and often incomplete [16,42,43]. Predictive tools such as the LUNDON score demonstrate excellent discrimination for donor lung utilization and post-transplant recipient outcomes, yet their integration into the time-sensitive, high-stakes clinical environment of transplantation remains limited.

Barriers to implementation of evidence-based practices in lung transplantation, such as the LUNDON score, may occur at multiple levels [44,45]. At the individual-level donor management decisions are frequently shaped by subjective judgment [46–49], institutional norms [46,50–52], and provider experience [46,50] rather than standardized data-driven approaches [11]. At the organizational level, variability in OPO workflows, limited interoperability of data systems, and the absence of embedded decision-support tools constrain adoption [8,53–55]. At the system level, the absence of structured implementation frameworks and performance feedback mechanisms perpetuates variation in donor evaluation and acceptance practices across centers [56,57].

Recent studies reinforce the need for deliberate implementation efforts. Our group and others have demonstrated that while the LUNDON score reliably predicts donor lung recovery and post-transplant recipient outcomes, its effectiveness depends on local calibration, consistent donor management strategies, and organizational readiness for change [34,36,37]. Collectively, these findings underscore that predictive validity alone is insufficient to drive behavioral change in the lung transplantation workflow.

Implementation science

Implementation science is a field dedicated to promoting the adoption and sustained use of evidence-based practices [58] across diverse healthcare contexts through structured, multi-level frameworks and methods [59]. These frameworks help identify barriers and facilitators, develop targeted strategies, and evaluate mechanisms that drive behavioral and organizational change. Applying implementation science principles in the context of lung donor management [16,60] offers an opportunity to accelerate the integration of predictive models such as the LUNDON score into routine clinical practice.

Expected impact

Grounded in implementation science methods, the proposed study aims to identify determinants (barriers and facilitators) of LUNDON score adoption in OPOs and lung transplant centers.

The integration of the LUNDON score into routine clinical practice has the potential to standardize donor evaluation, reduce unwarranted practice variation, and expand the pool of transplantable lungs. Ultimately, findings from this study will inform a future hybrid effectiveness–implementation trial designed to assess the real-world impact of LUNDON-guided donor management on organ utilization and recipient outcomes.

Study design and setting

This is a pre-implementation study that will focus on U.S. OPOs and lung transplant centers, the key environments where donor evaluation and organ acceptance decisions occur.

Study participants

Participants (Table 2) will include the following key constituent groups that are directly involved in donor evaluation and acceptance decisions: (1) OPO clinical leadership, (2) OPO donor coordinators, (3) transplant pulmonologists, and (4) transplant surgeons.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Overview of target sample populations.

https://doi.org/10.1371/journal.pone.0342383.t002

These multidisciplinary groups represent the primary potential end-users of the LUNDON score and are integral to understanding contextual barriers and facilitators to its adoption. Participants will be identified through the professional networks of the study team and invited with support from the study’s advisory board to ensure broad representation across institutions and geographic regions.

Sample size determination and sampling strategy

Sample sizes for each aim are detailed below. Sampling strategies were informed by our team’s prior unpublished work within the relatively small, close-knit community of OPOs and lung transplant clinicians. Based these experiences, we anticipate response rates exceeding 70%. Sample size estimates account for this expected level of participation.

To evaluate and minimize potential non-response bias, an interim analysis will be conducted after 40% of the survey samples have been enrolled for Aims 1–3. Key demographic and professional characteristics of responders and non-responders will be compared. If systematic differences are observed, targeted oversampling will be employed to ensure representation of participants with characteristics similar to non-responders for the remainder of enrollment.

Implementation frameworks guiding the study

This study applies two complementary implementation science frameworks to guide the identification of determinants of LUNDON score usage and the development of implementation toolkits to promote adoption:

1. 1). The Capability, Opportunity, Motivation-Behavior (COM-B) model [16,61] posits that behavior is driven by three essential components: capability (knowledge and skills), opportunity (environmental and social factors), and motivation (automatic and reflective processes). This model provides a structured lens for conceptualizing why specific behaviors related to donor management and organ acceptance occur (or fail to occur) in real-world practice.
2. 2). The Theoretical Domains Framework (TDF) [62] consolidates multiple behavior change theories into 14 domains to help systematically identify barriers (factors that impede implementation) and facilitators (factors that promote implementation). The framework supports the selection of tailored strategies to address the identified determinants of behavior.

Together, these frameworks will be used to: (a) identify barriers and facilitators of LUNDON score adoption in clinical practice and map the corresponding behavioral domains [Aim 1]; (b) characterize the behavioral and contextual factors influencing OPO-level donor management decisions related to modifiable LUNDON score variables [Aim 2]; and (c) assess how knowledge of a donor’s LUNDON score and the corresponding estimate of recipient survival influences organ acceptance decisions at the lung transplant center level [Aim 3].

Aim-Specific Protocols and Data Collection

Aims 2 and 3: Assess the impact of the LUNDON score availability on clinical decision making related to donor management [among OPO representatives] and organ offer acceptance [among lung transplant providers]

We will recruit OPO personnel (Aim 2, N = 60) and lung transplant providers (Aim 3, N = 60) to participate in a simulation-based experiment designed to evaluate how access to the LUNDON score influences clinical decision-making. Purposive sampling will ensure representation from low-, medium-, and high-performing OPOs and lung transplant centers. Participants will be recruited via email invitation, with up to three reminders. Informed consent will be implied by initiation of the simulation activity and subsequent survey completion following review of the study information sheet.

Simulation activities will be conducted in SimUNet, a web-based platform developed by the United Network for Organ Sharing (UNOS) [63] that mimics real-time clinical workflows in DonorNet, the official system used by the Organ Procurement and Transplantation Network (OPTN) for management of donor information and organ offers. UNOS developed and maintains DonorNet as the OPTN contractor and created SimUNet for research and experimental purposes. SimUNet has been validated in national studies assessing kidney offer decisions, demonstrating both feasibility and realism for simulating transplant workflows [64,65].

After orientation to SimUNet, participants will view a brief educational video introducing the LUNDON score and its potential applications for lung donation and then complete 16 simulated case scenarios representing a range of donor risk profiles and clinical complexity. The order of scenario presentation and the visibility of the LUNDON score will be randomized to reduce order effects and individual bias. For each case, OPO personnel will provide donor management recommendations, while lung transplant providers will accept or decline organ offers; all participants will rate their decisional confidence and provide open-text justifications.

Quantitative data will be analyzed using generalized estimating equations to model within-subject differences in decision-making between scenarios with and without the LUNDON score. Prespecified subgroup analyses will assess heterogeneity by transplant center volume and by Lung Composite Allocation Score (CAS), a metric used by OPTN to assess transplant center performance based on waitlist, transplant, and post-transplant outcomes [12].

Aim 4: Develop and pilot-test implementation toolkits for LUNDON score adoption.

Based on the findings from Aims 1–3, we will follow a structured process [66] to develop a set of participant-informed implementation toolkits to support real-world use of the LUNDON score. This will proceed in three phases:

• Phase 1: Develop an Implementation Research Logic Model (IRLM), to graphically depict relationships among study elements, aid future study replicability, and evaluate the implementation process [67]. A proposed IRLM for this study is provided in Fig 2. In Aim 4, the IRLM will be revised to incorporate findings from Aims 1–3, using the COM-B model and the TDF to systematically map identified behavioral determinants of LUNDON score adoption. The research team will then link these determinants to corresponding implementation strategies and potential toolkit resources.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Proposed implementation research logic model that will be adapted to incorporate findings from Aims 1-3.

https://doi.org/10.1371/journal.pone.0342383.g002

• Phase 2: Develop tailored toolkits for OPOs and lung transplant centers, comprised of resources to address the site-specific contextual needs for LUNDON score implementation. Guided by COM-B and TDF, toolkit development will draw on findings from Aims 1–3, clinical experience, and the input of multiple key constituent groups. Toolkit components may include decision-support aids, brief user training modules, electronic health record integration templates, and/or communication scripts tailored to clinical workflows.
• Phase 3: Refine toolkits through a brainwriting [68] pre-mortem exercise [69], which is a proactive planning method that uses prospective hindsight to anticipate potential reasons for implementation failure and generate solutions before the intervention is deployed. Focus groups will engage key constituents (n = 48) to assess content, format, acceptability, and usability, and will proactively identify weaknesses, barriers to adoption, and potential points of failure in the toolkit design. Participants will be offered a $125 Visa gift card as compensation.
• Phase 4: Pilot test the toolkits at three OPOs and three lung transplant centers. Pilot testing will include brief user testing followed by structured interviews with decision-making personnel at each site (n = 12) to assess acceptability, feasibility, and suggestions for improvement. Qualitative data will inform iterative refinement of materials. Interview participants will be offered a $125 Visa gift card as compensation.

Safety considerations

This is a minimal-risk study. The primary potential risks is a breach of confidentiality, which will be mitigated through use of secure data platforms and IRB-approved data protection procedures. No biomedical interventions or physical risks are associated with participation.

Data management

All study data will be managed using secure, access-controlled platforms in compliance with institutional policies and relevant federal regulations. Each participant will be assigned a unique study ID, and all personally identifiable information will be stored separately from analytic datasets. Data access will be restricted to approved study team members.

Survey data (Aims 1–3) will be collected and stored in REDCap. Interview recordings will be transcribed verbatim by a secure transcription service and uploaded to NVivo for qualitative coding and thematic analysis.

Routine data quality checks will be conducted weekly to identify missing or inconsistent entries. Periodic data backups will be stored on encrypted institutional servers. Final analytic datasets will be de-identified, with documentation maintained for transparency and reproducibility. Following study completion, data will be made available to qualified researchers upon request and IRB approval.

Ethics approval and consent to participate

This study protocol has been reviewed and approved by the Institutional Review Boards (IRBs) of Washington University [IRB # 202504071] and Duke University [IRB # Pro00118530]. All participants will provide informed consent prior to participation.

Dissemination plan

Study findings will be disseminated through peer-review publications, including this protocol and subsequent results manuscripts. Results will also be presented at national conferences focused on transplantation and implementation science. Upon study completion, finalized implementation toolkits will be made freely available through an open-access study website and shared with the broader transplant community via UNOS and professional society networks to support real-world application.