Clinical applications of artificial intelligence in symptom management and decision making in oncologic palliative care: a systematic review
Loading...
Files
Date
2025-06-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Medicine Palliative in Practice (Early publication accepted 09.06.2025)
Abstract
Abstract
Introduction: Artificial intelligence (AI) is increasingly being integrated into healthcare,
offering innovative tools to improve symptom management and support clinical decisionmaking
in patients with advanced cancer receiving palliative care (PC). The study aimed to
systematically evaluate recent evidence (2021–2024) on the clinical use of AI-based tools for
symptom management, prognosis prediction, and clinical decision support in adult oncology
patients in PC settings.
Methods: A systematic review was conducted following the PRISMA-P 2015 guidelines.
Databases searched included PubMed, Scopus, Cochrane Library, BVS, Scielo, and
ScienceDirect, using MeSH terms related to AI, cancer, pain, and palliative care. Studies
were included if they involved adult oncology patients using AI tools in PC and reported
outcomes related to symptom control, clinical decisions, or mortality estimation. Two
independent reviewers conducted the selection and methodological quality assessment using
STROBE, PRISMA, and CONSORT guidelines. Only studies rated as medium or high
quality were included.
Results: From an initial pool of 3,018 records, 20 studies were selected. AI applications were
grouped into prognosis and mortality prediction (n = 9), symptom identification and
monitoring (n = 5), clinical decision support (n = 4), and communication tools (n = 2). Models
included neural networks, eXtreme Gradient Boosting (XGBoost), decision trees, natural
language processing (NLP), and chatbots. Most studies demonstrated high accuracy in
retrospective or real-world clinical settings.
Conclusions: AI has shown potential in the early identification of palliative needs, symptom
control, and care planning. Prospective validation and implementation studies are needed to
ensure ethical and safe integration into palliative care.
Description
A systematic review published in Palliative Medicine in Practice analyzes 20 studies on clinical applications of artificial intelligence in symptom management, prognosis prediction, and decision support for patients with advanced cancer receiving palliative care. Tools such as XGBoost, neural networks, NLP, and chatbots are identified as being applied to mortality prediction, symptom control, clinical support, and physician-patient communication. The need for prospective validation and ethical and contextual integration in palliative care is highlighted.
Keywords
cancer, pain, palliative care, artificial intelligence
Citation
Botía Pinzón, S. M., Leal Arenas, F. A., & Molina Arteta, B. M. (2025). Clinical applications of artificial intelligence in symptom management and decision making in oncologic palliative care: A systematic review. Palliative Medicine in Practice. Advance online publication. https://doi.org/10.5603/pmp.105709