Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 42599269

A Sentiment-Based Comparison of AI- and Physician-Generated Empathic Statements in Palliative Care.

Aidin Gharavi·Tyler Luonuansuu·Kathleen Rogers·April Christensen
View original source ↗Export metadata ↓

Abstract

CONTEXT: Empathic communication promotes trust in patient-provider relationships. As healthcare integrates artificial intelligence (AI) into patient communication, we have yet to understand how these models' communication compares to that of physicians. OBJECTIVES: Our primary objectives were to examine patient preferences for AI-generated vs. palliative care physician-generated empathic statements addressing fear and anxiety around cancer treatment, and to analyze associations between linguistic features and patient preferences. METHODS: We conducted a secondary analysis of the PALL-AI trial, a randomized controlled survey comparing cancer patients' preferences of AI- to physician-generated empathic statements. Physicians and AI were provided the same prompt with a maximum sentence length. Patient preferences for each statement were measured in blinded surveys. We analyzed sentiment of the statements using the Valence Aware Dictionary and Sentiment Reasoner (VADER) and the National Research Council Canada (NRC) Emotion Lexicon. We evaluated associations between sentiment scores and patient preferences using Spearman's correlation coefficients. RESULTS: A total of 105 patients completed blinded surveys, preferring the AI-generated statement 72.4% of the time. VADER sentiment analysis showed all three AI statements displayed positive sentiment, while all three physician statements displayed negative sentiment. Controlling for statement length, AI statements used twice as many positive words as human statements. However, they contained a similar number of negative words. Of the eight NRC emotions, "trust" and "joy" demonstrated the strongest correlations with patient preference. CONCLUSION: Patients preferred AI-generated statements around cancer care over those from palliative care physicians when standardized for prompt and statement length. Analysis shows AI-generated statements contain more positive language which may be the factor driving patient preference toward AI.

Explore related subjects

HumansPalliative CareEmpathyFemalePhysician-Patient RelationsMaleArtificial IntelligenceMiddle AgedPatient PreferenceNeoplasmsAgedPhysiciansCommunicationAdultSurveys and QuestionnairesArtificial intelligencecancer carecommunicationrandomized controlled trialsentiment analysisJournal ArticleComparative StudyRandomized Controlled Trial

Keep this discovery

BibTeXRIS

Aidin Gharavi, Tyler Luonuansuu, Kathleen Rogers, April Christensen. 2026-08-14. A Sentiment-Based Comparison of AI- and Physician-Generated Empathic Statements in Palliative Care.. https://doi.org/10.1016/j.jpainsymman.2026.05.011

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

Adult subjectAged subjectArtificial Intelligence subjectComparative Study subjectFemale subjectHumans subjectJournal Article subjectMale subjectMiddle Aged subjectNeoplasms subjectRandomized controlled trials study

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related citations

PubMed2026

Continuous Intravenous Lidocaine for Refractory Cancer Pain in Palliative Care: A Multicenter Feasibility Study.

ObjectivesTo assess the feasibility and tolerability of continuous low-dose intravenous lidocaine infusion in patients with opioid-refractory cancer pain receiving palliative care, and to explore its potential impact on pain outcomes in real-world clinical conditions.MethodsWe conducted a multicenter, randomized, double-blind, placebo-controlled feasibility study in palliative care units to evaluate continuous intravenous lidocaine infusion in patients with opioid-refractory cancer pain. Patients were randomized to receive lidocaine (5&#x2005;mg/kg/day, increased to 8&#x2005;mg/kg/day if pain reduction was <30% after 24&#x2005;h) or placebo for 48&#x2005;h. Pain intensity was assessed using the Numeric Pain Intensity Scale, with a clinically meaningful response defined as a&#x2009;&#x2265;30% reduction from baseline at 40&#x2005;min. Secondary outcomes included pain evolution over time, neuropathic pain, symptom burden, and tolerability.ResultsThirty-five patients were included in the final analysis (18 lidocaine, 17 placebo). No significant difference was observed between lidocaine and placebo for the primary endpoint or for secondary pain outcomes. Reductions in pain intensity were observed in both groups. In the lidocaine group, 61% of patients required dose escalation to 8&#x2005;mg/kg/day. Continuous intravenous lidocaine infusion was generally well tolerated, with mostly mild adverse events and no unexpected toxicity.ConclusionIn this multicenter feasibility study, continuous low-dose intravenous lidocaine did not demonstrate a clinically meaningful analgesic benefit over placebo. As the planned sample size was not reached, the study was underpowered. These findings highlight the challenges of randomized trials in palliative care and may inform future feasibility-oriented designs.

Humans↗
PubMed2026

Access to palliative care in rural settings: A mixed-methods systematic review.

BACKGROUND: Rural populations experience persistent inequities in access to palliative care. Existing evidence often describes individual barriers separately, with less attention to how access breaks down across the care pathway or how different service configurations shape access. OBJECTIVES: To synthesise evidence on access to palliative care in rural settings and examine how access barriers, service models, and implementation conditions interact across the care pathway. METHODS: A mixed-methods systematic review using a convergent integrated approach searched nine databases (PubMed, Embase, CINAHL, Web of Science, Scopus, PsycINFO, CNKI, WanFang, SinoMed) from inception to 15 March 2026, supplemented by hand-searching. Eligible studies were primary qualitative, quantitative, and mixed-methods studies on access to palliative care for adults in rural or non-urban settings. Two reviewers independently screened studies, extracted data, and assessed quality using the Mixed Methods Appraisal Tool. Findings were mapped to the Levesque access framework, analysed using the updated Consolidated Framework for Implementation Research, and integrated through mixed-methods synthesis, with additional coding of service models. RESULTS: Thirty-four studies were included, of which 26 were conducted in high-income countries and eight in low- and middle-income countries. Service configurations included specialist or hospice-oriented care, generalist or primary-care-oriented care, mixed specialist-generalist models, home-based and caregiver-centred care, nurse-coordinated services, telehealth-supported care, and community or implementation-oriented approaches. Access broke down cumulatively across four interdependent stages: recognition, entry, reach, and use and continuity, with affordability constraining every stage. Recognition was limited by low awareness, poor service visibility, and delayed identification of need. Entry was shaped by stigma, trust, family expectations, and unclear referral processes. Reach was constrained by distance, transport, workforce shortages, limited specialist capacity, and weak infrastructure. Use and continuity were affected by fragmented coordination, weak transitions, unstable follow-up, and reliance on family caregivers. Access problems varied across service configurations. Evidence on service innovations was methodologically less certain, and the overall evidence base remained concentrated in high-income countries. CONCLUSIONS: Access to palliative care in rural settings is best understood as a pathway and service-configuration problem rather than simply a deficit in service availability. Improving access requires earlier recognition, clearer referral routes, stronger specialist-generalist and nursing links, better support for family caregivers, and greater attention to affordability, continuity, and rural settings with limited resources. REGISTRATION: International Prospective Register of Systematic Reviews: CRD420261340783.

Health Services Accessibility↗
PubMed2026

Aortoesophageal Fistula: Mending the Lethal Connection-A Systematic Review and Meta-Analysis.

BACKGROUND: Aortoesophageal fistula (AEF) is a rare, life-threatening condition with limited high-quality evidence to guide management. METHODS: We conducted a systematic review and meta-analysis of PubMed and Scopus through December 2025, evaluating survival and complication outcomes according to treatment strategy and morphological severity. Patients were categorized into 5 groups: TEVAR alone, staged TEVAR followed by surgery, primary open or hybrid surgery, TEVAR with esophageal stenting, and supportive/palliative care. Outcomes were pooled as proportions with 95% confidence intervals (CIs). RESULTS: A total of 167 reports representing 528 patients were included. Overall 30-day mortality was 31.2% (95% CI 27.0% to 35.7%) and 1-year mortality 41.2% (95% CI 36.2% to 46.4%). Infection occurred in 32.6%, reintervention in 20.2%, and recurrence in 13.2%. Staged TEVAR followed by surgery showed favorable survival (30-day 12.7%, 1-year 26.1%) but high infection (66.7%) and reintervention (43.2%) rates. TEVAR with esophageal stenting had the lowest early mortality (5.0%) but frequent reintervention (54.5%) and 1-year mortality of 33.3%. TEVAR alone demonstrated the lowest 1-year mortality among definitive strategies (25.5%) but notable recurrence (26.9%). Primary open or hybrid surgery carried higher early and late mortality, while supportive/palliative care had the worst outcomes. Morphological severity correlated strongly with mortality, infection, reintervention, and recurrence, with type IV lesions showing particularly poor prognosis. CONCLUSION: Despite contemporary management, AEF carries high early and late mortality. Definitive strategies, particularly staged TEVAR followed by surgery, offer improved survival but increased complications. Morphology-guided, individualized management is recommended.

Adult↗