Artificial Intelligence (AI) is emerging as a double-edged sword in primary care, offering groundbreaking opportunities while presenting unique challenges. As a family physician deeply engaged with AI and medical informatics, I aim to dissect these dimensions, emphasizing how our collective voice is pivotal in harnessing AI's potential for primary care.
The Promise
Reducing Administrative Burden: AI can automate routine tasks such as appointment scheduling, patient follow-ups, and electronic health record (EHR) management, enhancing efficiency. A study by Blease et al. (2020) highlights AI's role in reducing clerical workload and improving job satisfaction among healthcare providers [1]. In our work, we have seen how AI can assist physicians and dramatically reduce the burden and time of documentation and chart review. We also see the promises of AI assistants to help tame the EHR inbox. These are products that are already available in the market, and we expect many more to enter.
Expanding Capacity and Capabilities: AI-driven tools like diagnostic algorithms and predictive analytics can augment clinical decision-making. Research indicates that AI can assist in early disease detection, risk stratification, and personalized treatment plans, thereby improving patient outcomes [2].
The Perils
When I think about the challenges related to AI, I put them in two large buckets: intrinsic - challenges inherent to machine learning and AI and extrinsic - forces external to AI that may direct AI to uses that harm primary care.
Trustworthiness: The biggest intrinsic challenges I believe are around trustworthiness. I am sure you have heard about AI bias. There are many types of potential bias that must be mitigated in AI. In this post we will not dive into this very deep topic. The reason being that I believe the market forces are aligned to drive solutions to the intrinsic challenges like bias, hallucination, explainability, and more largely trustiworthiness. We must though continue to hold AI developers to ethical principles so these market forces remain and continue to grow.
Profit-driven Design: I think the more concerning future issues with AI will be focused around the extrinsic challenges of deploying AI in health care. Current healthcare structures often prioritize profit over patient care, potentially misaligning AI's application away from primary care needs. Additionally, the lack of supportive infrastructure for AI integration in primary care can hinder its effective utilization. Currently we are seeing large tech and enterprises driving the development and deployment of AI, which are not two of the most staunch advocates for primary care. Mainly due to the financial incentives driving AI adoption in health care areas outside of primary care or in attempts to disintermediate primary care.
Phyx Primary Care believes the collective voice of physicians is crucial in guiding AI development and implementation. By advocating for ethical AI use principles, emphasizing patient-centered care, and participating in AI research and policy-making, physicians can ensure that AI serves as a tool for enhancing, not replacing, the human elements of primary care. We are working with primary care physicians to establish a collective voice that can be heard to drive AI development and support adoption of AI solutions that truly work for primary care.
AI in primary care is a frontier filled with promise and peril. As we stand at this crossroads, it is imperative that we, as primary care physicians, actively engage in shaping AI's role in healthcare. By doing so, we can ensure that AI augments our capabilities and upholds the values at the heart of patient centered primary care.
Comments