Mapping study on AI-based technologies in palliative care - a scoping study

Abstract

Background: The aging population and rising prevalence of chronic illnesses emphasize the importance of palliative care (PC), which focuses on enhancing patients' quality of life (QoL) while supporting their families and caregivers. PC integrates multidisciplinary interventions to alleviate the physical, psychological, social, and spiritual suffering of individuals facing serious or terminal illnesses. Concurrently, Artificial Intelligence (AI) advancements have been transforming the healthcare sector, particularly through Clinical Decision Support Systems (CDSS). Leveraged by advanced algorithms and machine learning (ML), these tools analyze large volumes of data to support diagnostics, personalized treatments, and early interventions. In PC, AI has demonstrated potential to enhance early diagnosis, identify support needs, and personalize end-of-life care. ML algorithms help predict symptoms and complications, enabling timely and effective interventions. However, challenges remain, including data privacy concerns, integration into clinical workflows, and ethical implications of AI in sensitive care contexts.

Methods: We conducted a scoping review to map and analyze AI applications on PC. Articles published until May 2024 were identified in two electronic databases. From 542 records, 57 studies met the inclusion criteria. The review explored trends, benefits, and limitations of AI applications, highlighting tools for diagnostic and prognostic support, symptom tracking, shared decision-making, and communication with patients and families.

Results: The findings highlight how digital technologies and AI are revolutionizing communication, care coordination, and symptom control in PC, unlocking remote care options. The review identified key advancements in symptom management, communication, decision support, telemedicine and education areas, while addressing barriers like ethical, legal, and accessibility concerns.

Conclusions: By compiling evidence on AI use in PC, we aimed to empower professionals, researchers, and policymakers to promote more effective, ethical, and person-centered strategies. Ultimately, we provide insights for developing new technologies and establishing protocols that support the safe, equitable, and person-centered implementation of AI in palliative care, and highlight the need to prioritize early identification of patient needs, promote integration between hospital and community care, and establish protocols.

Keywords: Artificial Intelligence; Digital Health Technology; Palliative Care; Patient-Centered Care; Scoping Review.

Fig. 1

Flowchart of the literature selection process for the included studies. A total of 542 records were identified in two literature databases. After duplicates removal, 443 articles were screened, of which 129 full-texts were assessed. The final selection of included articles consisted of 57 studies

Fig. 2

Summary of targeted technology applications in palliative care. Schematic overview of the main areas of intervention for technology in palliative care, along with the percentage representation of each area in the included articles. "Telemedicine" emerged as the most frequently addressed domain, featuring in approximately 50% of the studies, followed by "communication," cited in 40% of the records. "Symptoms control" was explored in 33% of the articles, while "advising & decision-making" appeared in 29%. The least represented area was "education/training," reported in 22% of the included studies