How AI is Transforming Advanced Care Planning: A Guide for Physicians

By Campion Quinn, MD

Introduction

Advance care planning (ACP) is the process of discussing and documenting a patient’s preferences regarding future medical care, particularly in cases where they may become unable to make decisions for themselves. ACP ensures that patients receive care aligned with their values, improves communication between patients and healthcare providers, and reduces unnecessary medical interventions. It also helps families and clinicians make informed decisions, reducing stress and uncertainty during critical moments.

Despite its importance, ACP remains underutilized. Conversations often occur too late, and many patients' preferences go undocumented. This is where artificial intelligence (AI) is making a significant impact. AI is being leveraged to identify patients who may benefit from ACP, streamline documentation, facilitate conversations, and enhance decision-making, ensuring that ACP becomes a proactive, integrated part of patient care rather than an afterthought.

This essay will explore how AI is used in ACP, its impact on clinical care, administrative efficiency, patient outcomes, and how it reshapes the physician’s role in end-of-life planning.

The Role of AI in Identifying Patients for ACP

One of AI’s most valuable applications in ACP is predictive analytics, which helps clinicians identify patients needing ACP discussions earlier in their care journey. AI algorithms analyze vast amounts of electronic health record (EHR) data, lab results, imaging, and physician notes to predict disease progression, hospitalization risks, and survival probabilities.

For instance, machine learning models have been developed to predict 30-day mortality or the likelihood of ICU admission based on historical patient data. These predictions help physicians determine the optimal time to initiate ACP conversations, ensuring that patients’ preferences are documented before they lose decision-making capacity 1. AI is a “clinical radar,” flagging high-risk patients who might otherwise be overlooked.

AI in Facilitating ACP Discussions and Patient Engagement

AI-powered virtual assistants and chatbots facilitate ACP conversations, making them more accessible and less daunting for patients and families. Many people avoid discussing end-of-life care due to discomfort or uncertainty. AI tools help by structuring these conversations and offering guided discussions tailored to each patient’s condition and preferences.

For example, AI-driven Serious Illness Conversation Guide bots engage patients in interactive discussions about their treatment goals. These tools present various scenarios, such as choosing comfort care and aggressive interventions, helping patients articulate their preferences. 2. By allowing patients to explore these options in a structured, non-threatening way, AI increases engagement and ensures ACP discussions happen earlier in the disease trajectory.

Additionally, sentiment analysis tools analyze patient-clinician conversations to detect emotional cues, ensuring that discussions are compassionate and appropriately timed. This enhances communication strategies and helps clinicians adjust their approach to better support patients. 3.

AI for Automating and Managing ACP Documentation

A significant challenge in ACP is ensuring that preferences are appropriately documented and accessible when needed. AI-driven natural language processing (NLP) tools assist in automating ACP documentation, reducing the burden on clinicians, and minimizing errors.

These NLP systems scan EHR notes, physician dictations, and patient interactions to detect ACP-related content. They then generate structured documentation, ensuring patient preferences are recorded in a standardized, easily retrievable format. 4.

Moreover, AI-enhanced advance directive management platforms such as MyDirectives and Vynca integrate with EHRs, making it easier for healthcare providers to store, access, and update ACP documents. This prevents situations where a patient’s wishes are unknown or overlooked due to misplaced paperwork.

AI-Assisted Decision Support Tools

AI-assisted decision aids guide patients and families through complex medical decisions by providing tailored risk-benefit analyses of various interventions. These AI-driven systems evaluate options such as mechanical ventilation, dialysis, or palliative chemotherapy, offering personalized insights that help patients and their caregivers make informed choices. 4. Additionally, AI-powered ACP tools such as MyDirectives and Vynca integrate seamlessly with EHRs, ensuring that documented preferences are readily accessible when critical care decisions arise. By leveraging AI to enhance decision-making, these tools streamline the ACP process, improving clarity and reducing the burden on clinicians and families. 2.

AI’s Impact on Healthcare Efficiency and Patient Outcomes

Beyond improving individual patient care, AI also optimizes ACP-related administrative workflows within healthcare systems. AI-driven solutions reduce clinicians' time manually identifying ACP candidates, documenting preferences, and coordinating care. By automating these processes, AI ensures that ACP becomes a seamless part of routine care rather than an afterthought.

From a patient outcomes perspective, early and well-documented ACP discussions lead to higher satisfaction, reduced hospitalizations, and lower healthcare costs. Patients who engage in ACP are more likely to receive care that aligns with their values and avoid unnecessary aggressive interventions 5. AI helps scale these benefits across entire healthcare systems by increasing the efficiency and accuracy of ACP implementation.

Ethical Considerations and Challenges

Despite its potential, AI in ACP faces several challenges. Algorithmic bias is a significant concern—AI models trained on limited or non-representative datasets may produce inaccurate predictions for specific patient populations, leading to disparities in ACP recommendations 5. To provide equitable care, healthcare organizations must ensure that AI models are trained on diverse, representative data.

Another issue is data privacy and security. Since ACP involves highly sensitive patient information, AI tools must comply with strict confidentiality and security measures to prevent data breaches and misuse. Robust governance policies are necessary to maintain trust in AI-driven ACP solutions.

Additionally, physician acceptance and trust in AI remain challenges. While AI provides valuable insights, clinicians must interpret these suggestions critically and integrate them into shared decision-making with patients. AI should be viewed as an augmentation tool rather than replacing human expertise.

The Future of AI in ACP

As AI continues to evolve, several advancements are expected to enhance ACP further:

1. AI-Powered Voice Recognition: AI-driven voice assistants may soon be able to document ACP discussions in real time, further reducing administrative burden.

2. Multimodal AI Models: Future AI systems will combine EHR data with imaging, genomics, and social determinants of health to create even more personalized ACP recommendations.

3. AI-Generated Nudges: Automated reminders for physicians and patients to revisit ACP periodically will ensure that care preferences evolve alongside the patient’s health status and life circumstances.

By integrating these technologies into routine practice, AI has the potential to make ACP a proactive, patient-centered process that improves both healthcare efficiency and end-of-life care quality.

Conclusion

AI is revolutionizing advanced care planning by improving patient identification, facilitating discussions, automating documentation, and optimizing workflows. By leveraging predictive analytics, NLP, virtual assistants, and sentiment analysis, AI ensures that ACP happens earlier, is well-documented, and aligns with patient values.

While challenges like bias, privacy concerns, and clinician acceptance remain, AI’s ability to enhance ACP is undeniable. Rather than replacing the human element in end-of-life discussions, AI is a powerful tool that supports physicians in delivering compassionate, patient-centered care.

As AI advances, its integration into ACP will become increasingly essential. This will help healthcare systems ensure that every patient’s care aligns with their preferences and dignity.

References

1. Jung K, Kashyap S, Avati A, et al. Predicting need for advanced care planning in a primary care population. J Am Board Fam Med. 2021;34(6):1034-1044. doi:10.3122/jabfm.2021.06.210210.

2. Callahan A, Steinberg E, Fries JA, et al. Machine learning-based prediction of serious illness conversations in hospitalized patients. J Gen Intern Med. 2021;36(12):3731-3737. doi:10.1007/s11606-021-07000-1.

3. Shah NH, Milstein A, Bagley SC. Making machine learning models clinically useful. JAMA. 2019;322(14):1351-1352. doi:10.1001/jama.2019.10306.

4. Li RC, Smith M, Lu J, et al. Using machine learning to predict advance care planning in a primary care population. J Am Geriatr Soc. 2022;70(1):123-131. doi:10.1111/jgs.17490.

5. Pfohl SR, Foryciarz A, Shah NH. A comparison of approaches to improve worst-case predictive model performance over patient subpopulations. Sci Rep. 2021;11(1):18962. doi:10.1038/s41598-021-98491-8.