The Fund supports networks of state health policy decision makers to help identify, inspire, and inform policy leaders.
The Milbank Memorial Fund supports two state leadership programs for legislative and executive branch state government officials committed to improving population health.
The Fund identifies and shares policy ideas and analysis to advance state health leadership, strong primary care, and sustainable health care costs.
Keep up with news and updates from the Milbank Memorial Fund. And read the latest blogs from our thought leaders, including Fund President Christopher F. Koller.
The Fund publishes The Milbank Quarterly, as well as reports, issues briefs, and case studies on topics important to health policy leaders.
The Milbank Memorial Fund is is a foundation that works to improve population health and health equity.
February 18, 2025
Report
Yalda Jabbarpour
Anuradha Jetty
Hoon Byun
Anam Siddiqi
Jeongyoung Park
Publication
Feb 18, 2025
Back to Publications
Primary care clinicians spend more time on electronic health records (EHRs) than clinicians in any other specialty.84 While increased utilization of the EHR can improve clinical, organizational, and societal outcomes,85 these gains come at a significant cost.86 Increased EHR usage can impede face-to-face time with patients, reducing clinical effectiveness and increasing physician burnout.87, 88 Current EHR technology also lags in interoperability between different health settings and facilities.89 Similar to last year, data from the American Board of Family Medicine (ABFM) demonstrate that almost half of family physicians view EHR usability (a measure combining ease of finding information and usefulness of alerts) as poor or fair (Figures 13-1 and 13-2), and over one-quarter of family physicians were not satisfied with their EHR. In fact, a slightly greater percentage of family physicians reported that they were very dissatisfied with their EHR in 2023 compared to 2022 (Figure 13-3).
Figure 13. Almost Half of Family Physicians View EHR Usability as Poor or Fair, and Over One-Quarter Are Dissatisfied Overall with Their EHR (2022—2023)
13-1. EASE OF FINDING INFORMATION
13-2. USEFULNESS OF ALERTS
13-3. ELECTRONIC HEALTH RECORD OVERALL SATISFACTION
Data Sources: Data Sources: American Board of Family Medicine Continuing Certification Questionnaire (CCQ), 2022—2023
Notes: A total 6,345 respondents completed EHR usability questions. A total 12,709 respondents completed EHR satisfaction questions.
Because of the high cost of EHRs and the difficulties of integrating them effectively and efficiently into the practice of primary care, many primary care practices are left with EHRs that are suboptimal in meeting the needs of the practice and their patients.90, 91 The cost of implementing a robust system ranges drastically, from $32,000 to $70,000 per full-time employee, and these expenses can easily reach millions for an individual hospital.92–95 This price range can also vary depending on factors such as on-site versus cloud-based EHR deployment type and the level of implementation assistance,94 as well as the overall quality of the EHR, which can significantly impact interoperability.89
Despite the many hours PCPs spend on the EHR doing inbox management, responding to portal messages, and coordinating care for patients, these efforts do not generate RVUs and this time is generally non-reimbursable.87, 96 Since the COVID-19 pandemic, time spent in the EHR has continued to grow, elevating the need to improve EHR technology at the organization and systems level.86, 97
Research has shown that useful tools such as digital scribes,98 artificial intelligence,99 and other investments into EHRs can greatly reduce time spent on the EHR and improve clinical efficiency. The lack of investment in primary care has made implementing these new critical technologies extremely difficult, primarily due to up-front costs, but also technical, time, legal, organizational, psychological, and social constraints.98, 100
ACTION 2.4: CMS should permanently support COVID-era rule revisions such as payment for telehealth services.
ACTION 4.1: The Office of the National Coordinator (ONC) for Health Information Technology and CMS should develop the next phase of digital health certification standards that support relationship-based, continuous, and person-centered care; simplify the user experience; ensure equitable access and use; and hold vendors accountable.
ACTION 4.2: ONC and CMS should adopt a comprehensive aggregate patient data system that is usable by any certified digital health tool for patients, families, clinicians, and care team members.
Ready or Not, Here Comes AI
By Elise Blaseg
Primary care doctors are often overwhelmed by clerical burdens, such as note-taking during visits, that detract from patient care.i The integration of artificial intelligence (AI) into the world of medicine holds promise for freeing up some of their time. Clinicians are using AI to assist with everything from electronic health record updates to predictive analytics like identifying hospital readmission risks.ii
Still, primary care practices will need to be thoughtful about how they safely and effectively introduce these tools. Experts raise concerns such as the potential for overreliance on the technology for clinical decision making and the increased risk of exposing patient information to cyberattacks and data breaches.iii In a landscape deeply imprinted with health disparities, AI algorithms can also perpetuate established biases and further marginalize vulnerable populations.iv Another concern is that the technical and financial burden of AI implementation will worsen disparities by making the tool inaccessible to smaller practices or less-resourced facilities like federally qualified health centers.
St. Mark’s Family Medicine Residency is a training program and small nonprofit independent family medicine clinic in Salt Lake City, Utah, that introduced an AI scribe service into their practice in April 2024. According to St. Mark’s resident John James, MD, “AI is really the first tool that I’ve seen that has actually delivered on [decreasing administrative burden] and . . . sped up my process without compromising on the quality of care that I deliver.” The clinic is financing the residents’ use of the scribe tool with GME funding, while the faculty’s use of the tool is funded by the clinic, cumulatively costing less than 0.1% of the total operating budget. Providers who wish to use generative AI tools for tasks like patient education materials or email templates must pay for these services themselves.
The University of Rochester Medical Center, a large health system in Rochester, New York, has expanded its AI use from integrated scribe services to the inception of their own generative AI triage tool for online patient messages. Michael Hasselberg, PhD, chief digital health officer for URMC, reported that over the eight months they have been testing the triage system, it has shown great potential in reducing the administrative burden of clinic staff and nurses. Additionally, URMC has started to fully automate quality reporting and registry processes. While some of the budget for the AI efforts come from the medical center’s IT and operations departments, URMC is utilizing other university resources to create cost-effective, customized technology. “We put [our] data scientists next to the primary care docs and [other] clinicians to understand what their problems are, and then we develop the tech solutions in-house.”
Both health care facilities have seen a marked improvement in clinicians’ quality of work life with the new tools. At URMC, the initial response has been overwhelmingly positive, with users reporting decreased cognitive load, reduced burnout, and improved clinician well-being. Likewise, Dr. James of St. Mark’s noted, “My job is more satisfying to me because it . . . frees my brain up for the more high-level critical thinking and creative processes that actually make medicine successful and rewarding.”
Of course, clinics, practices, and health systems will need to invest in policies that help ensure the responsible and ethical use of the technology, while developers will need to continue to refine the technology itself to eliminate bias and make it more secure. And at a clinic level, education of clinicians and transparency with patients is critical. According to Hasselberg, while AI will never replace the clinician, “doctors and nurses who use AI will replace doctors and nurses who don’t use AI.”
i. American Medical Association. Physician sentiments around the use of AI in healthcare: Motivations, opportunities, risks, and use cases. AMA Augmented Intelligence Research. Published 2023.
ii. Dankwa-Mullan I. Health equity and ethical considerations in using artificial intelligence in public health and medicine. Prev Chronic Disease 2024;21:240245. https://doi.org/10.5888/pcd21.240245.
iii. Liyanage H, Liaw ST, Jonnagaddala J, et al. Artificial intelligence in primary health care: perceptions, issues, and challenges. Yearb Med Inform. 2019;28(1):41-46. doi:10.1055/s-0039-1677901
iv. Nazer LH, Zatarah R, Waldrip S, et al. Bias in artificial intelligence algorithms and recommendations for mitigation. PLOS Digit Health. 2023;2(6):e0000278. doi:10.1371/journal.pdig.0000278.
