WP 1. Validation of the GEMA model in Spain
Retrospective observational study based on three cohorts: British, Australian and a multicentre cohort of the Spanish National Transplant Organisation (ONT). The ONT cohort includes patients from 26 adult liver transplant centres in Spain. The project is supported by the Spanish Liver Transplant Society (SETH) and the National Federation of Liver Patients and Transplant Patients (FNETH). In ethical terms, the project complies with the principles of the Declaration of Helsinki, guaranteeing patient confidentiality and consent. The study population comprises approximately 16,000 adult liver transplant candidates from the three cohorts, who meet specific inclusion criteria and certain cases are excluded. The GEMA model has emerged from previous research funded by the Consejería de Salud y Familias of the Junta de Andalucía. It has been developed, trained and internally validated in the UK cohort, and subsequently externally validated in the Australian cohort. The model uses variables similar to the original MELD-Na model, with adaptations including the replacement of serum creatinine with the estimated glomerular filtration rate for cirrhosis from the Royal Free Hospital (Development and validation of a mathematical equation to est… : Hepatology (lww.com)). Statistical analyses have shown that the GEMA model improves the discriminative ability of MELD-Na and MELD 3.0, and reduces waiting list mortality, especially among women, suggesting a potential to correct gender disparities in access to liver transplantation.
WP 2. Optimising GEMA with Advanced Artificial Intelligence
Gender disparities in access to liver transplantation within the ONT and RETH cohort will be examined. Indicators such as time on waiting list, probability of transplantation, waiting list mortality and probability of exclusion due to worsening will be compared between men and women. Cox regression will be used to analyse the influence of sex, controlling for the aetiology of liver disease, indication for transplantation and baseline severity parameters. The absolute reduction in the risk of list mortality and the number of deaths avoided with GEMA implementation in Spain will be evaluated. In addition, the use of artificial intelligence to improve the GEMA model (GEMA-IA) will be explored. A model based on artificial neural networks will be built, using evolutionary algorithms for training, especially suitable for unbalanced databases or for extreme analytical values. Different combinations of cohorts will be tested for training and testing, selecting the model that obtains the best results in discriminative ability. Tools such as NNEP and Scikit-learn will be used for neural network model design.
WP 3. National Consolidation of the Prioritisation System in the Liver Transplant List
Organisation and coordination of a national consensus to homogenise the prioritisation system for the transplant waiting list within the Spanish Liver Transplant Society (SETH) and with the collaboration of the National Transplant Organisation (ONT). SETH will coordinate the consensus and will hold several face-to-face meetings of the working team made up of a representative from each of the 26 transplant centres, the two coordinators (Dr. Itxarone Bilbao from Vall D’Hebron Hospital and Dr. Manuel L. Rodríguez Perálvarez from Reina Sofía Hospital in Cordoba), and the ONT representative (Dr. Gloria de la Rosa).
WP 4. Training and Dissemination: Transition to the New GEMA Model
Development of training actions for professionals and patients on the importance of the transition to the new GEMA model and on the agreements reached as a result of the previous action. Training activities for patients and professionals have been planned, including specific patient workshops on transplantation and gender, as well as an online course for professionals to disseminate the advantages of the GEMA model and to instruct on the use of the automatic calculation software tool. Advice and support will be offered to other countries choosing to transition to the GEMA model of liver transplant waiting list prioritisation.
WP 5. Development of App for Transplant Waiting List Management
Design and implementation of an app for the management of the transplant waiting list and free distribution among national centres. A software tool will be designed for the automated calculation of the model and for the management of the waiting list. Its extended version, which will require institutional permission, will allow transplant centres to manage their own waiting list in a secure way. Each centre will be able to include the information of its patients on the waiting list and establish its own specific prioritisation rules. All team members will be able to access the information in real time so that, in the event of a transplant alarm, the software will be very useful for coordinating the different specialists involved. The ownership of this software tool will be transferred to the ONT, which will benefit the National Health Service.