Case Studies
Case studies showcase real-world applications of policies and innovations, offering insights into improving outcomes and efficiency. They enhance problem-solving skills, highlight best practices, and engage through storytelling for education and transparency.
Interest(s): Advanced Practice Professionals, Artificial Intelligence, Donation after Circulatory Death (DCD), Donor Management, Electronic Medical Records (EMR), Emerging Technology, Growth in Transplantation, Hospital Donation Process & Education, Innovation, Marginal / Increased Risk Organs, Organ Allocation (Offer and Acceptance)
Organ(s): Heart, Intestine, Kidney, Liver, Lung, Pancreas, Small Bowel
Patient Group(s): Adult
Share
Increasing Donor Management Communication Efficiencies through Data Capture, Consolidation & Visualization Tools
Lee
Keddie
, CompuMed Inc.
The Challenge
Organ procurement organizations (OPOs) play a vital role in maximizing donor heroes’ gifts and increasing lives saved through organ transplantation. However, variability in data interpretation among organ coordinators, un-secure and fragmented communications between coordinators and key clinical decision-makers, and manual/inconsistent data entry contribute to inefficiencies and missed opportunities for organ utilization.
A central challenge lies in the inconsistency and fragmentation of clinical data—ventilator settings, laboratory values, imaging results and EMR notes. Key clinical decision-makers like AOCs, medical directors or other subject matter expert consultants are frequently overwhelmed by voluminous/insecure communications and unstructured information. Essentially, clinical decision-makers are drowning in data and starving for information.
Compounding the issue is the difficulty in standardizing communication and documentation across diverse systems, especially in time-sensitive situations like rapid lung or DCD cases. Additionally, many hospitals perform three or fewer organ recoveries per year, making full EMR integrations impractical. Instead, coordinators often rely on insecure text messaging or manually relaying subjective impressions, resulting in data variability, delayed decisions and suboptimal organ utilization.
This project was designed to address these pain points by leveraging real-time data capture and visualization tools combined with an AI-driven platform developed by CompuMed in collaboration with ConnectLife and the Center for Donation and Transplant (CDT). The platform’s primary purpose is to capture, consolidate and contextualize donor clinical data in real-time, enabling subject matter experts to make rapid, well-informed decisions based on consistent, verifiable information.
Importantly, the solution had to work within existing workflows and accommodate hospitals without integrated digital systems. To meet this need, the platform was designed with multiple flexible upload methods—screen capture, QR code-enabled mobile phone image uploads, EMR ingestion—and the ability to extract structured insights from unstructured data using AI.
Ultimately, the goal of this initiative is to shift donor data management from reactive to proactive by ensuring that actionable information is captured at the source, intelligently summarized and made available instantly to those who need it earlier in the donor
management process. In doing so, the platform aims to transform donor management from a decentralized, variable process into a streamlined, data-driven workflow that supports better outcomes across the donation and transplant continuum.
The Approach
To address the challenges of fragmented communications, inconsistent coordinator expertise and delayed decision-making, ConnectLife and the Center for Donation and Transplant (CDT) partnered with CompuMed to implement data capture tools and a flexible, AI-driven data platform designed specifically for their workflows. The core principle behind the solution was simple but transformative: capture donor data as close to the source as possible, keep it in clinical context, and use artificial intelligence (AI) to extract and summarize the information in a way that supports real-time decision-making by subject matter experts.
The platform supports four primary methods of data input: (1) direct upload of files (e.g., PDFs, XML exports from EMRs), (2) mobile phone image capture of clinical monitors or ventilators, (3) screen capture of EMR flowsheets or imaging reports, and (4) real-time capture of lab values or monitor data at the bedside. Each piece of data is tagged with a “bucket” or data category to preserve context (e.g., ventilator data, ABGs, CT scans), which helps constrain the AI’s interpretation and improves extraction accuracy. By maintaining clinical context, the AI can reliably differentiate between data types and ensure the correct parsing of numerical values, time stamps and clinical parameters.
Once data is captured and categorized, the platform uses AI tools to extract structured information from unstructured inputs (e.g., parsing values from a screenshot of a ventilator or extracting the impression from a radiology report). These structured data elements are then synthesized into customizable summaries.
Subject matter experts—such as transplant surgeons, pulmonologists, medical directors and AOCs (Administrator on Call)—can access AI-generated summaries alongside the original source material. Importantly, every AI-generated output is traceable, with links back to the original images or documents, supporting transparency and validation. The summaries do not provide clinical recommendations or predictive scoring; instead, they describe the donor’s physiological status in a clear, concise narrative, surfacing modifiable versus non-modifiable factors and changes over time (e.g., improvement in PF ratios after ventilator adjustments).
The system also includes data collection progress indicators and completion checklists to assist coordinators in collecting the required data efficiently and ensuring nothing is missed before expert review. In doing so, the platform creates a closed feedback loop that promotes accuracy, minimizes variability and streamlines communication throughout the donor management process.
The Results
Since the launch of the AI-assisted data capture and visualization platform, ConnectLife and CDT have completed 35 donor cases with immediate and measurable impact across several operational and clinical domains. The most notable benefit has been the reduction in variability and subjectivity in donor assessment among organ coordinators. By capturing data directly from the source—ventilators, imaging, lab values, and EMR flowsheets—and tagging it appropriately, the system enabled accurate, consistent extraction of relevant clinical information, regardless of the coordinator’s background or level of experience.
In practice, the platform allowed AI to generate structured summaries tailored to specific stakeholder needs. For example, Connect Life and CDT’s consulting pulmonologist, who previously communicated with coordinators through fragmented, un-secure text exchanges, now receives a single AI-generated donor summary. These summaries not only consolidate the relevant imaging and lab data but also include contextual explanations and links to source files. This transformation enabled more efficient reviews and faster, more informed feedback. As a result, decisions that once required multiple back-and-forth conversations could now be made with a single data review.
The system also facilitated the real-time identification of modifiable risk factors. In one highlighted case, a 48-year-old female donor initially presented as a poor candidate for lung donation. However, the AI summary visualized a clear path forward by identifying that the majority of limiting factors—such as oxygenation and ventilator settings—were modifiable. A dedicated coordinator was assigned to focus solely on lung management. With targeted interventions, the PF ratio improved from 129 to 347, and the lungs were ultimately transplanted after reaching out to aggressive centers on a second pass through the match run.
In another case, a 60-year-old male DCD donor—the oldest DCD lung donor in the program’s history—was similarly successful through the use of the platform. Real-time tracking of oxygenation improvement and ventilation changes gave the team confidence to continue pursuing transplant, despite an initially poor assessment.
Operationally, the platform also proved valuable in reducing training time for new coordinators. CDT reported a 30% reduction in lung management training requirements, as the focus shifted from rigorous clinical training to learning how to capture data cleanly for expert review. Furthermore, the system allowed both remote and on-site staff to collaborate effectively, enabling real-time updates, targeted decision-making and confident donor advocacy even under tight logistical timelines.
Insights & Lessons Learned
The implementation of this AI-enabled donor data platform represents a paradigm shift in how OPOs can manage clinical data, communicate with clinical decision makers and support both donor advocacy and operational efficiency. Historically, donor management
has depended heavily on the individual experience and clinical background of on-site coordinators, leading to variability in how data is interpreted, summarized, and communicated to stakeholders. This model introduces a more scalable, reliable, and future-ready alternative by offloading data interpretation to a structured, transparent and verifiable AI system that captures information at the point of origin and delivers actionable insights in near real time.
Crucially, this innovation does not remove clinical judgment from the process—it enhances it. Coordinators are still responsible for understanding the context of the donor case and making key decisions about when and what to capture. However, they are no longer burdened with parsing ventilator settings or summarizing lab trends on their own. Instead, they become high-efficiency data gatherers, equipped with tools that allow even new or non-clinical staff to contribute meaningfully. This has enormous implications for OPOs facing high turnover, limited clinical hiring pools and increasing pressure to shorten training timelines while expanding coverage.
By improving the precision and speed of information flow, the platform also increases trust. Subject matter experts no longer need to navigate dozens of PDF files or rely on secondhand reports from coordinators. Instead, they receive focused, consistently structured summaries—customized to their needs—along with direct access to the source data for validation. This dual structure of interpretability and transparency significantly increases confidence in donor evaluations and, as shown in several highlighted cases, can directly influence the decision to pursue marginal or complex donors.
Moreover, the platform’s architecture—with its API integration capabilities and modular tagging system—makes it adaptable across a wide range of environments, including hospitals without full EMR integrations. This is especially relevant given that over 50% of donor hospitals recover fewer than three donors per year, making full digital integrations infeasible. The ability to use secure mobile capture, QR-coded uploads, and screen recordings ensures that the data pipeline remains consistent and compliant, regardless of hospital resources.
In conclusion, this platform enables a new standard in donor management: one that is data-rich but not overwhelming, centralized yet adaptable, and powered by AI but grounded in clinical collaboration. It holds the potential to reduce discard rates, increase transplantability and ultimately save more lives through better, faster decision-making.