Allied treatment algorithms to predict mortality risk –

by time news

Transplant a life-saving procedure for patients with end-stage organ disease. Immunosuppressive therapy and post-transplant care continued to improve survival rates. International case studies tell us that 88% of transplant recipients still alive at one year from transplant e about 70% after five years. Affecting long-term life expectancy are events such as organ failure, infections, metabolic and cardiovascular complications, cancer. It would be possible to predict them, thus allowing to improve the time horizon of patients? Artificial intelligence seems to be able to give an encouraging answer.

I study

In a study published in The Lancet Digital Health, a group of researchers from the University Health Network of Toronto (Canada) developed and validated a deep learning model to provide clinicians with the computational tools to identify the risk of mortality and plan therapeutic interventions effective. The model was developed, feeding it with data from 42,146 transplant recipients from the Scientific Registry of Transplant Recipients (SRTR) in the United States. The possibility of applying it in other contexts was further evaluated by comparison with a dataset of 3,269 transplant recipients of the University Health Network.

Is the model really so innovative?

Liver transplantation is a complex procedure and susceptible to immediate or late complications, related to the surgical procedure, the age of the donor, the age of the recipient or the severity of the disease, as well as immunosuppressive therapies. professor Paolo De Simone, director of the Operative Unit of Hepatic Surgery and Liver Transplantation, Pisa University Hospital. The novelty of the methodology of dynamically evaluating the factors at play, as these can change over the life of the transplant patient (a patient may develop hypertension two years after the transplant), this approach could replace the one currently used in most studies, where the predictors of mortality are employed in a static way, i.e. evaluated at a given time (eg: at the time of listing, at the time of transplantation or at a given time of follow-up) without taking into account their possible change over time.

This new approach can be useful to the transplantologist and why?

The results of the study are more useful than the methodology used to obtain them. Basically, they tell us that to predict the survival of transplant patients, a dynamic assessment must be made, i.e. repeated over time, of all the clinical factors involved. In other words, complications can arise with the passage of time, even if some affect more in the short distance (infections) while others in the long distance (cardiovascular complications). This study tells us that clinical practice requires constant refinement of diagnosis and interventions on individual patients and that transplant recipients require attention for the duration of their existence concludes De Simone.

May 26, 2021 (change May 26, 2021 | 17:19)


You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.