AI-based clinical decision-making systems in palliative medicine: ethical challenges

Effectiveness of initiating advance care planning among older outpatients through intervention by physicians trained in a model discussion video: A randomized controlled trial

AIM

This study aimed to assess the impact of a physician-led intervention on advance care planning (ACP) introduction among older outpatients using a model discussion video.

METHODS

This prospective interventional study included Japanese general practice outpatients aged ≥65 years. All participants received an ACP leaflet. Co-researchers were trained using a model discussion video based on the "Serious Illness Conversation Guide," and provided ACP introduction explanations to the intervention group to ensure intervention content consistency. The control group received the ACP leaflet only, with no standardized explanations. The primary outcome was ACP discussion occurrence. Secondary outcomes included ACP engagement (assessed through the 4-item ACP Engagement Survey), engagement score for advance directives acquisition and score for surrogate decision-maker identification, and anxiety incidence (assessed through Generalized Anxiety Disorder-7).

RESULTS

This study included 48 participants: 25 (52.1%) in the intervention group and 23 (47.9%) in the control group. Regarding primary outcome, the intervention group had significantly more ACP discussions compared with the control group (92.0% vs 26.1%, adjusted odds ratio 50.2, P = 0.0012). Regarding secondary outcomes, the intervention group showed greater readiness (mean scores for the 4-item Engagement Survey [P = 0.0001], engagement score for advance directives acquisition [P = 0.01] and score for surrogate decision-maker determination [P = 0.03]) than the control group, without increasing anxiety.

CONCLUSIONS

This physician-led ACP introduction for older patients in a Japanese general practice setting enhanced ACP readiness and stimulated discussions, without increasing anxiety. The model discussion video for ACP introduction might effectively train physicians in general practice settings. Geriatr Gerontol Int 2024; ••: ••-••.

Prognostication: A fading Hippocratic art?

Over the past 75 years modern medicine has advanced in its ability to diagnose and treat many diseases. However, the medical profession's ability to prognosticate the course and outcome of an illness has not satisfied the needs of many patients. Physicians must not lose the ability, or desire, to consider the whole person in relation to a patient's disease. We need to ask ourselves what person has the disease, not what disease the person has. Here I endeavor to demonstrate why Hippocrates valued prognostication highly, how its importance may have faded from the consciousness of current medical practice, and how modern technology is attempting to reinvent or revise it.

Evaluating the Efficacy of ChatGPT in Navigating the Spanish Medical Residency Entrance Examination (MIR): Promising Horizons for AI in Clinical Medicine

The rapid progress in artificial intelligence, machine learning, and natural language processing has led to increasingly sophisticated large language models (LLMs) for use in healthcare. This study assesses the performance of two LLMs, the GPT-3.5 and GPT-4 models, in passing the MIR medical examination for access to medical specialist training in Spain. Our objectives included gauging the model's overall performance, analyzing discrepancies across different medical specialties, discerning between theoretical and practical questions, estimating error proportions, and assessing the hypothetical severity of errors committed by a physician.

MATERIAL AND METHODS

We studied the 2022 Spanish MIR examination results after excluding those questions requiring image evaluations or having acknowledged errors. The remaining 182 questions were presented to the LLM GPT-4 and GPT-3.5 in Spanish and English. Logistic regression models analyzed the relationships between question length, sequence, and performance. We also analyzed the 23 questions with images, using GPT-4's new image analysis capability.

RESULTS

GPT-4 outperformed GPT-3.5, scoring 86.81% in Spanish (p < 0.001). English translations had a slightly enhanced performance. GPT-4 scored 26.1% of the questions with images in English. The results were worse when the questions were in Spanish, 13.0%, although the differences were not statistically significant (p = 0.250). Among medical specialties, GPT-4 achieved a 100% correct response rate in several areas, and the Pharmacology, Critical Care, and Infectious Diseases specialties showed lower performance. The error analysis revealed that while a 13.2% error rate existed, the gravest categories, such as "error requiring intervention to sustain life" and "error resulting in death", had a 0% rate.

CONCLUSIONS

GPT-4 performs robustly on the Spanish MIR examination, with varying capabilities to discriminate knowledge across specialties. While the model's high success rate is commendable, understanding the error severity is critical, especially when considering AI's potential role in real-world medical practice and its implications for patient safety.

Exploring the Potential of Chatbots in Critical Care Nephrology

The exponential growth of artificial intelligence (AI) has allowed for its integration into multiple sectors, including, notably, healthcare. Chatbots have emerged as a pivotal resource for improving patient outcomes and assisting healthcare practitioners through various AI-based technologies. In critical care, kidney-related conditions play a significant role in determining patient outcomes. This article examines the potential for integrating chatbots into the workflows of critical care nephrology to optimize patient care. We detail their specific applications in critical care nephrology, such as managing acute kidney injury, alert systems, and continuous renal replacement therapy (CRRT); facilitating discussions around palliative care; and bolstering collaboration within a multidisciplinary team. Chatbots have the potential to augment real-time data availability, evaluate renal health, identify potential risk factors, build predictive models, and monitor patient progress. Moreover, they provide a platform for enhancing communication and education for both patients and healthcare providers, paving the way for enriched knowledge and honed professional skills. However, it is vital to recognize the inherent challenges and limitations when using chatbots in this domain. Here, we provide an in-depth exploration of the concerns tied to chatbots' accuracy, dependability, data protection and security, transparency, potential algorithmic biases, and ethical implications in critical care nephrology. While human discernment and intervention are indispensable, especially in complex medical scenarios or intricate situations, the sustained advancements in AI signal that the integration of precision-engineered chatbot algorithms within critical care nephrology has considerable potential to elevate patient care and pivotal outcome metrics in the future.

Ethical Implications of Chatbot Utilization in Nephrology

This comprehensive critical review critically examines the ethical implications associated with integrating chatbots into nephrology, aiming to identify concerns, propose policies, and offer potential solutions. Acknowledging the transformative potential of chatbots in healthcare, responsible implementation guided by ethical considerations is of the utmost importance. The review underscores the significance of establishing robust guidelines for data collection, storage, and sharing to safeguard privacy and ensure data security. Future research should prioritize defining appropriate levels of data access, exploring anonymization techniques, and implementing encryption methods. Transparent data usage practices and obtaining informed consent are fundamental ethical considerations. Effective security measures, including encryption technologies and secure data transmission protocols, are indispensable for maintaining the confidentiality and integrity of patient data. To address potential biases and discrimination, the review suggests regular algorithm reviews, diversity strategies, and ongoing monitoring. Enhancing the clarity of chatbot capabilities, developing user-friendly interfaces, and establishing explicit consent procedures are essential for informed consent. Striking a balance between automation and human intervention is vital to preserve the doctor-patient relationship. Cultural sensitivity and multilingual support should be considered through chatbot training. To ensure ethical chatbot utilization in nephrology, it is imperative to prioritize the development of comprehensive ethical frameworks encompassing data handling, security, bias mitigation, informed consent, and collaboration. Continuous research and innovation in this field are crucial for maximizing the potential of chatbot technology and ultimately improving patient outcomes.

Recent advances in artificial intelligence applications for supportive and palliative care in cancer patients

PURPOSE OF REVIEW

Artificial intelligence (AI) is a transformative technology that has the potential to improve and augment the clinical workflow in supportive and palliative care (SPC). The objective of this study was to provide an overview of the recent studies applying AI to SPC in cancer patients.

RECENT FINDINGS

Between 2020 and 2022, 29 relevant studies were identified and categorized into two applications: predictive modeling and text screening. Predictive modeling uses machine learning and/or deep learning algorithms to make predictions regarding clinical outcomes. Most studies focused on predicting short-term mortality risk or survival within 6 months, while others used models to predict complications in patients receiving treatment and forecast the need for SPC services. Text screening typically uses natural language processing (NLP) to identify specific keywords, phrases, or documents from patient notes. Various applications of NLP were found, including the classification of symptom severity, identifying patients without documentation related to advance care planning, and monitoring online support group chat data.

SUMMARY

This literature review indicates that AI tools can be used to support SPC clinicians in decision-making and reduce manual workload, leading to potentially improved care and outcomes for cancer patients. Emerging data from prospective studies supports the clinical benefit of these tools; however, more rigorous clinical validation is required before AI is routinely adopted in the SPC clinical workflow.

Doing palliative care research on hematologic cancer patients: A realist synthesis of literature and experts’ opinion on what works, for whom and in what circumstances

Background

Research in PC (Palliative Care) is frequently challenging for patient’s frailty, study design, professional misconceptions, and so on. Little is known about specificity in PC research on Hematologic cancer patients, who have distinct characteristics that might influence the enrollment process.

Aims

What works, how and for whom, in increasing enrollment in studies in PC on patients with hematologic malignancies?

Methods

Realist review: a qualitative review whose goal is to identify and explain the interaction between Contexts, Mechanisms, and Outcomes (CMOs). The theory was informed by a narrative, theory-based literature research, including an initialsystematic research, and the addition of papers suggested by experts of the field. We also used 7 interviews with experts in PC about patients with hematologic malignancies research and our own experience from a PC pilot study on patients with hematologic malignancies to refine the initial theory.

Results

In our initial theory we hypothesize that:

 - Access to palliative care could be beneficial to hematologic patients, even in early stages

 - Hematologists tend to under-use palliative care services in general, due to unpredictable disease trajectories and cultural barriers.

 - These factors may negatively impact the patients’ enrollment in PC research

We included secondary literature as narrative reviews, if they presented interesting propositions useful for our theoretical construction. 23 papers met our inclusion criteria.

We also searched for relevant CMOs impacting referral in palliative care, and we selected a list of CMOs that could be relevant also in hematology. We accordingly theorized a group of interventions that could increase the enrollment in PC research and presented them using “social exchange theory” (SET) as a theoretical framework.

Prominent researchers in PC in hematologic malignancies were interviewed on their opinion on our results, and additional CMOs.

Conclusions

Before conducting research in PC on patients with hematologic malignancies, it’s probably advisable to assess:

 - The perception of the different actors (physicians, nurses, other professionals involved), in particular the hematologists, in terms of pros and cons of referral to PC and enrollment in PC trials

 - The existing relationship between PC and the Hematology department

Accordingly, it’s possible to tailor different interventions on the various actors and choose a model of trial to increase the perception of benefits from PC and, consequently, enrollment.

Negotiating the Beginning of Care: A Grounded Theory Study of Health Services for Amyotrophic Lateral Sclerosis

A range of professional figures are needed to preserve the quality of life of people with amyotrophic lateral sclerosis. This study aimed to explore the beginning of the care process as negotiated by people with amyotrophic lateral sclerosis, their caregivers, and healthcare professionals. We designed the study according to the constructivist Grounded Theory method, collecting data through open-ended, semi-structured interviews, employing theoretical sampling and constant comparison, and performing conceptual coding as data analysis. By naming the core category "off-beat interfacing", we were able to show how the demands of the professionals concerned did not correspond to the ability of people with ALS and their proxies to process information, deal with requests, and be at ease in making decisions at the beginning of the shared care pathway. Three categories were generated: (i) navigating different paths, (ii) offering and experiencing a standard, non-personalized pathway, and (iii) anticipating decisions. The network of services must be organized according to guidelines, but must also contemplate a patient-family-centered approach that permits more personalized assistance.