Leveraging AI To Return Clinician Time to Patient-facing Care

During and after the COVID-19 pandemic, burnout among healthcare workers reached crisis levels, with 3 in 10 considering leaving the profession. Now well past the acute phase of the pandemic, the crisis continues. In 2025, 41.9% of physicians reported at least one symptom of burnout, and 58% of nurses reported feeling burned out most days.

Why are so many healthcare workers burning out? Work-related bureaucratic tasks are the primary reason for physicians, according to a Medscape survey. That burden extends across the healthcare system, in which workers spend large chunks of their day managing administrative tasks.

This creates an opportunity to use AI for a very practical purpose: reducing administrative burden in healthcare. The most promising AI applications are often the least flashy, with the greatest immediate value and least immediate risk coming from automating repetitive work so healthcare teams can spend more time with patients.

The burnout loop: When care becomes paperwork

The scale of administrative work in healthcare is staggering. A study of four specialties found that in ambulatory settings, physicians spend 49% of their time on electronic health records (EHR) and desk work, compared to just 33% on direct patient interaction.

Take the medication access process, for example. Prescribing a therapy may take a doctor only a short patient visit, but getting a patient onto treatment can require multiple teams coordinating hours of work, especially for specialty and high-cost therapies. Staff regularly have to manually pull information from EHRs, re-enter it into other platforms, comb through medical records, draft detailed patient histories and justification letters, search for financial assistance programs, verify eligibility, and track filing deadlines.

The work is both repetitive and highly detail-oriented and comes at a high cost. Administrative spending as a whole accounts for 20–25% of US healthcare expenditures, equivalent to roughly $1 trillion annually.

But the indirect costs may be even greater. For healthcare workers, spending day after day navigating repetitive administrative workflows can cause burnout, which drives turnover, which worsens staffing shortages, which creates even greater operational strain across the system. It’s a growing global problem, with an estimated shortage of 18 million healthcare workers worldwide.

Reducing administrative overload with AI

This is where AI presents one of its biggest opportunities in healthcare, not by replacing healthcare workers, but by reducing the operational burden surrounding care delivery.

AI can dramatically improve time-consuming healthcare workflows without taking decision-making away from care teams. In the medication access process, for example, the physician still evaluates the patient, makes the diagnosis, and decides on the appropriate therapy. But once that clinical decision is made, AI and automation can take over the operational work that follows to get patients onto treatment more efficiently.

Instead of staff manually gathering and re-entering information across multiple systems, AI and automation tools can pull relevant data from the EHR, organize lab results and supporting documentation, summarize patient histories, and draft prior authorization materials. What was once significant, repetitive administrative work can be assembled in minutes. The prescribing team simply reviews the information, makes edits when necessary, and submits the request.

Beyond prior authorizations, AI-powered automation can also assist with intake and documentation, automate appointment scheduling, provide routine patient education, and perform other routine tasks. Clinicians themselves already see these as some of the most valuable healthcare AI applications. In a recent study, clinicians rated AI-generated authorization letter drafting as highly helpful, giving it a usefulness rating of 5 out of 5.

Importantly, AI and automation technologies also have the potential to help keep patients engaged in therapy over time. Refill reminders, digital education, and other adherence support workflows can be automated to help keep patients on therapy, with AI escalating matters to the care team when human input is needed.

AI as support, not substitution

At Cinnamon, we believe that AI in healthcare is not about replacing care teams, but reducing the operational burden surrounding care delivery so healthcare professionals can spend more time on the parts of medicine that require human expertise and judgment.

Healthcare is fundamentally relational. Strong provider–patient relationships, trust, and shared decision-making remain central to effective care. Physician empathy, for example, has been shown to improve patient outcomes. The opportunity for AI is not to replace those interactions, but to support them by handling the repetitive administrative work that increasingly pulls clinicians away from patients.

Researchers studying AI in healthcare are increasingly arriving at the same conclusion: AI should support human-centered care. For example, one study found that large language models (LLMs) lack the advanced clinical reasoning to provide differential diagnosis. But used as support infrastructure for things like PA, AI can speed up the repetitive and time-intensive processes in healthcare to help ease the operational strain contributing to the healthcare workforce crisis.

The emerging shift in healthcare infrastructure

At Cinnamon, we believe the most meaningful healthcare AI solutions will not replace clinical expertise, but rather remove the operational friction surrounding care delivery so clinicians can focus more fully on patients.

This is an urgent and necessary change at a time when burnout and administrative overload are driving healthcare workers out of the profession while pulling existing teams away from direct patient care. The healthcare system already faces major workforce shortages, and the Health Resources and Services Administration projects the U.S. could face a shortage of more than 141,000 physicians and nearly 109,000 registered nurses by 2038.

Our goal is to reduce the exhausting administrative work for staff and help patients get onto therapy faster with far less friction. Cinnamon’s technology combines algorithmic logic and pharmacist-trained AI to automate and coordinate workflows like prior authorizations, onboarding, and care coordination so care teams can spend less time navigating systems and more time delivering care.

Healthcare must retain its humanity and employ systems to stop exhausting the humans who provide care. We believe AI’s greatest benefit to healthcare will not be replacing healthcare personnel, but instead helping restore the time and capacity needed for them to care for patients well.

Photo by Nappy on Unsplash