Commentary

Implementing Trustworthy AI in VA High Reliability Health Care Organizations

Fed Pract. 2024 February;41(2):40 | doi:10.12788/fp.0454

Author(s):

Background: Artificial intelligence (AI) has great potential to improve health care quality, safety, efficiency, and access. However, the widespread adoption of health care AI needs to catch up to other sectors. Challenges, including data limitations, misaligned incentives, and organizational obstacles, have hindered implementation. Strategic demonstrations, partnerships, aligned incentives, and continued investment are needed to enable responsible adoption of AI. High reliability health care organizations offer insights into safely implementing major initiatives through frameworks like the Patient Safety Adoption Framework, which provides practical guidance on leadership, culture, process, measurement, and person-centeredness to successfully adopt safety practices. High reliability health care organizations ensure consistently safe and high quality care through a culture focused on reliability, accountability, and learning from errors and near misses.

Observations: The Veterans Health Administration applied a high reliability health care model to instill safety principles and improve outcomes. As the use of AI becomes more widespread, ensuring its ethical development is crucial to avoiding new risks and harm. The US Department of Veterans Affairs National AI Institute proposed a Trustworthy AI Framework tailored for federal health care with 6 principles: purposeful, effective and safe, secure and private, fair and equitable, transparent and explainable, and accountable and monitored. This aims to manage risks and build trust.

Conclusions: Combining these AI principles with high reliability safety principles can enable successful, trustworthy AI that improves health care quality, safety, efficiency, and access. Overcoming AI adoption barriers will require strategic efforts, partnerships, and investment to implement AI responsibly, safely, and equitably based on the health care context.

PDF Download

References

1. Sahni NR, Carrus B. Artificial intelligence in U.S. health care delivery. N Engl J Med. 2023;389(4):348-358. doi:10.1056/NEJMra2204673

2. Borkowski AA, Jakey CE, Mastorides SM, et al. Applications of ChatGPT and large language models in medicine and health care: benefits and pitfalls. Fed Pract. 2023;40(6):170-173. doi:10.12788/fp.0386

3. Moyal-Smith R, Margo J, Maloney FL, et al. The patient safety adoption framework: a practical framework to bridge the know-do gap. J Patient Saf. 2023;19(4):243-248. doi:10.1097/PTS.0000000000001118

4. Isaacks DB, Anderson TM, Moore SC, Patterson W, Govindan S. High reliability organization principles improve VA workplace burnout: the Truman THRIVE2 model. Am J Med Qual. 2021;36(6):422-428. doi:10.1097/01.JMQ.0000735516.35323.97

5. Sculli GL, Pendley-Louis R, Neily J, et al. A high-reliability organization framework for health care: a multiyear implementation strategy and associated outcomes. J Patient Saf. 2022;18(1):64-70. doi:10.1097/PTS.0000000000000788

6. National Institute of Standards and Technology. AI risk management framework. Accessed January 2, 2024. https://www.nist.gov/itl/ai-risk-management-framework

7. Executive Office of the President, Office of Science and Technology Policy. Blueprint for an AI Bill of Rights. Accessed January 11, 2024. https://www.whitehouse.gov/ostp/ai-bill-of-rights

8. Executive Office of the President. Executive Order 13960: promoting the use of trustworthy artificial intelligence in the federal government. Fed Regist. 2020;89(236):78939-78943.

9. Biden JR. Executive Order on the safe, secure, and trustworthy development and use of artificial intelligence. Published October 30, 2023. Accessed January 11, 2024. https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/

10. US Department of Veterans Affairs. Trustworthy AI. Accessed January 11, 2024. https://department.va.gov/ai/trustworthy/

Artificial intelligence (AI) has lagged in health care but has considerable potential to improve quality, safety, clinician experience, and access to care. It is being tested in areas like billing, hospital operations, and preventing adverse events (eg, sepsis mortality) with some early success. However, there are still many barriers preventing the widespread use of AI, such as data problems, mismatched rewards, and workplace obstacles. Innovative projects, partnerships, better rewards, and more investment could remove barriers. Implemented reliably and safely, AI can add to what clinicians know, help them work faster, cut costs, and, most importantly, improve patient care.¹

AI can potentially bring several clinical benefits, such as reducing the administrative strain on clinicians and granting them more time for direct patient care. It can also improve diagnostic accuracy by analyzing patient data and diagnostic images, providing differential diagnoses, and increasing access to care by providing medical information and essential online services to patients.²

High Reliability Organizations

High reliability health care organizations have considerable experience safely launching new programs. For example, the Patient Safety Adoption Framework gives practical tips for smoothly rolling out safety initiatives (Table 1). Developed with experts and diverse views, this framework has 5 key areas: leadership, culture and context, process, measurement, and person-centeredness. These address adoption problems, guide leaders step-by-step, and focus on leadership buy-in, safety culture, cooperation, and local customization. Checklists and tools make it systematic to go from ideas to action on patient safety.³

Leadership involves establishing organizational commitment behind new safety programs. This visible commitment signals importance and priorities to others. Leaders model desired behaviors and language around safety, allocate resources, remove obstacles, and keep initiatives energized over time through consistent messaging.⁴ Culture and context recognizes that safety culture differs across units and facilities. Local input tailors programs to fit and examines strengths to build on, like psychological safety. Surveys gauge the existing culture and its need for change. Process details how to plan, design, test, implement, and improve new safety practices and provides a phased roadmap from idea to results. Measurement collects data to drive improvement and show impact. Metrics track progress and allow benchmarking. Person-centeredness puts patients first in safety efforts through participation, education, and transparency.

The Veterans Health Administration piloted a comprehensive high reliability hospital (HRH) model. Over 3 years, the Veterans Health Administration focused on leadership, culture, and process improvement at a hospital. After initiating the model, the pilot hospital improved its safety culture, reported more minor safety issues, and reduced deaths and complications better than other hospitals. The high-reliability approach successfully instilled principles and improved culture and outcomes. The HRH model is set to be expanded to 18 more US Department of Veterans Affairs (VA) sites for further evaluation across diverse settings.⁵

Implementing Trustworthy AI in VA High Reliability Health Care Organizations

References

High Reliability Organizations

Pages

Recommended Reading