Artificial intelligence-based clinical decision support for liver transplant evaluation and considerations about fairness: A qualitative study.

BACKGROUND: The use of large-scale data and artificial intelligence (AI) to support complex transplantation decisions is in its infancy. Transplant candidate decision-making, which relies heavily on subjective assessment (ie, high variability), provides a ripe opportunity for AI-based clinical decision support (CDS). However, AI-CDS for transplant applications must consider important concerns regarding fairness (ie, health equity). The objective of this study was to use human-centered design methods to elicit providers' perceptions of AI-CDS for liver transplant listing decisions. METHODS: In this multicenter qualitative study conducted from December 2020 to July 2021, we performed semistructured interviews with 53 multidisciplinary liver transplant providers from 2 transplant centers. We used inductive coding and constant comparison analysis of interview data. RESULTS: Analysis yielded 6 themes important for the design of fair AI-CDS for liver transplant listing decisions: (1) transparency in the creators behind the AI-CDS and their motivations; (2) understanding how the AI-CDS uses data to support recommendations (ie, interpretability); (3) acknowledgment that AI-CDS could mitigate emotions and biases; (4) AI-CDS as a member of the transplant team, not a replacement; (5) identifying patient resource needs; and (6) including the patient's role in the AI-CDS. CONCLUSIONS: Overall, providers interviewed were cautiously optimistic about the potential for AI-CDS to improve clinical and equitable outcomes for patients. These findings can guide multidisciplinary developers in the design and implementation of AI-CDS that deliberately considers health equity.

Multicenter study of racial and ethnic inequities in liver transplantation evaluation: Understanding mechanisms and identifying solutions.

Racial and ethnic disparities persist in access to the liver transplantation (LT) waiting list; however, there is limited knowledge about underlying system-level factors that may be responsible for these disparities. Given the complex nature of LT candidate evaluation, a human factors and systems engineering approach may provide insights. We recruited participants from the LT teams (coordinators, advanced practice providers, physicians, social workers, dieticians, pharmacists, leadership) at two major LT centers. From December 2020 to July 2021, we performed ethnographic observations (participant-patient appointments, committee meetings) and semistructured interviews (N = 54 interviews, 49 observation hours). Based on findings from this multicenter, multimethod qualitative study combined with the Systems Engineering Initiative for Patient Safety 2.0 (a human factors and systems engineering model for health care), we created a conceptual framework describing how transplant work system characteristics and other external factors may improve equity in the LT evaluation process. Participant perceptions about listing disparities described external factors (e.g., structural racism, ambiguous national guidelines, national quality metrics) that permeate the LT evaluation process. Mechanisms identified included minimal transplant team diversity, implicit bias, and interpersonal racism. A lack of resources was a common theme, such as social workers, transportation assistance, non-English-language materials, and time (e.g., more time for education for patients with health literacy concerns). Because of the minimal data collection or center feedback about disparities, participants felt uncomfortable with and unadaptable to unwanted outcomes, which perpetuate disparities. We proposed transplant center-level solutions (i.e., including but not limited to training of staff on health equity) to modifiable barriers in the clinical work system that could help patient navigation, reduce disparities, and improve access to care. Our findings call for an urgent need for transplant centers, national societies, and policy makers to focus efforts on improving equity (tailored, patient-centered resources) using the science of human factors and systems engineering.

Safety and Reactogenicity of 2 Doses of SARS-CoV-2 Vaccination in Solid Organ Transplant Recipients.

BACKGROUND: We studied the safety and reactogenicity SARS-CoV-2 mRNA vaccines in transplant recipients because immunosuppressed patients were excluded from vaccine trials. METHODS: US transplant recipients were recruited into this prospective cohort study through social media; those who completed the full vaccine series between December 9, 2020 and March 1, 2021 were included. We collected demographics, medical history, and safety information within 7 d after doses 1 and 2 (D1, D2). Associations between characteristics and reactions were evaluated using modified Poisson regression. RESULTS: We studied 741 transplant recipients who underwent BNT162b2 (54%) or mRNA-1273 (46%) vaccination. Median (interquartile range) age was 60 (44-69) y, 57% were female, and 10% were non-White. Although local site reactions decreased after D2 (85% D1 versus 78% D2, P < 0.001), systemic reactions increased (49% D1 versus 69% D2, P < 0.001). Younger participants were more likely to develop systemic symptoms after D1 (adjusted incidence rate ratio [aIRR] per 10 y = 0.850.900.94, P < 0.001) and D2 (aIRR per 10 y = 0.910.930.96, P < 0.001). Participants who experienced pain (aIRR = 1.111.662.47, P = 0.01) or redness (aIRR = 1.833.928.41, P < 0.01) were more likely to develop an antibody response to D1 of mRNA vaccines. No anaphylaxis, neurologic diagnoses, or SARS-CoV-2 diagnoses were reported. Infections were minimal (3% after D1, <0.01% after D2). One patient reported incident acute rejection post-D2. CONCLUSIONS: In solid organ transplant recipients undergoing mRNA vaccination, reactogenicity was similar to that reported in the original trials. Severe reactions were rare. These early safety data may help address vaccine hesitancy in transplant recipients.