Unburdening Patients and Clinicians Through Automation and Artificial Intelligence: Informatics Strategies for Reducing Administrative Burden

Abstract

Healthcare delivery systems face mounting administrative complexity that contributes to clinician burnout, medical errors, and reduced access to care for patients. This editorial explores how automation and artificial intelligence (AI) can address key operational inefficiencies—specifically in prior authorization, quality metric reporting, and clinical documentation—by leveraging informatics-driven solutions. We examine the current landscape, quantify the impact of administrative burden, and propose informatics strategies to realign healthcare delivery around patient-centered, efficient care.

Editorial

Patients, clinicians, and administrators face a crisis of mounting administrative burdens that drive physician burnout, associated medical errors [1, 2] and access challenges all while reducing workforce productivity and job satisfaction [3]. While the causes are multifactorial, automation and the use of artificial intelligence (AI) driven informatics tools can transform healthcare by unburdening patients and clinicians from administrative constraints. Broader headwinds in the physician labor force make improving efficiency a policy imperative, with the physician shortage estimated in 2027 to reach 124,180 physicians [4]. Simultaneously, clinical workloads are overwhelming, with a study estimating that completion of all recommended tasks results in a 26.7 h primary care physician workday [5]. Thus, improving administrative efficiency to unburden physicians and expand the workforce is a national policy imperative.

AI and automation can drive efficiency in a more decentralized, personalized healthcare ecosystem through operational process improvement, including prior authorization, quality measurement, and documentation and billing, thus reducing patient and physician administrative burden. This editorial reviews each of these current operational challenges, clinical labor impacts, and opportunities for technological innovation to surmount these barriers.

Operational Challenges in Prior Authorization

Prior authorization is a prime example of an operationally complex, high-friction process ripe for improvement with AI. Workflow frequently relies on manual, redundant information submissions and human-driven reviews for completeness and for first stage clinical review against basic criteria. According to the American Medical Association 2024 prior authorization Physician survey, 94% of Physicians report delays patient care, 19% report resulting hospitalizations and 78% note that patients often abandon treatment due to delays [6]. Costs are significant, with research demonstrating that collectively payers spend $6 billion administering drug utilization management annually while physicians spend $26.7 billion in time in navigation response [7].

Data from analysis of the Medicare Advantage marketplace shows most the media denial rate is 17%, with 57% being overturned on appeal [8]. In some cases, this may be due to inadequate, inaccurate, or incomplete information submitted to health plans via manual processes prone to error. While the insurance industry has announced a commitment to prior authorization reforms, automation and AI will be critical to achieving these goals through decreasing friction for electronic data submission at the point of care and automating approval, not denial [9]. With the Centers for Medicare & Medicaid Services expanding prior authorization in Fee For Service Medicare through the WISeR model [10], efforts to improve and automate prior authorization using informatics tools will take on even greater clinical importance.

Rather than replacing fax machines with the equivalent digital workflows, informatics solutions, including AI-driven chart review can transform the prior authorization process through leveraging large language models (LLMs) to comb through patient charts and extract comprehensive clinical information for review. Empowering patients and providers at the point of care, this would reposition payers from after-the-fact impediments to clinical service to driving point-of-care discussions around clinically efficient and cost-effective clinical practice, fulfilling a modern physician ethos [11]. AI could also help with flagging complex cases for human review, with process improvements decreasing unnecessary friction, improving both the patient and physician experience [12].

Decreasing Quality Metric Reporting Burdens

Clinician administrative frustrations escalate significantly with quality metric reporting. In accordance with Donabedian’s 1966 principle that “one cannot improve what is not measured,” policy efforts have pushed health systems and health plans measure individual and group clinician performance. Consequently, the quality measurement industry has continued to group most recently through the 2015 implementation of Merit-based Incentive Payment System (MIPS), with the Centers for Medicare & Medicaid Services (CMS) maintaining a library of over 1,200 metrics, of these over 500 are active [13]. Compliance costs for MIPS quality reporting along are estimated at an annual cost of $40,069 per physician or $15.4 billion [14].

With quality reporting requirements generating operational burdens redirecting efforts away from patient care, researchers have found that MIPS scores often do not align with outcomes [15]. In other settings, quality improvement programs have paradoxical effects, such as an absolute increase in 30-day mortality post-discharge in heart failure [16].

Informatics tools can simplify quality metric and performance reporting, with relative and absolute reporting at the individual, organizational, and regional level by integrating structured and unstructured data across electronic systems. Automation of quality reporting data collection could also facilitate more efficient and effective review of quality metric performance, promoting the growth of a “quality metric lifecycle” wherein old metrics are retired, reiterated, or replaced as appropriate [17].

Streamlining Documentation and Billing

Administrative burden manifests most visibly in daily clinical practice, where primary care physicians now spend nearly 6 hours daily interacting with electronic health records (EHRs), with clerical tasks accounting for nearly half of this time [18]. EHR systems compound this problem through transferring competing documentation demands to physicians at the point of care including the manual entry of diagnosis coding, billing, and electronic clinical documentation. Documentation strains are real, with trauma surgeons spending 1,760.5 hours annually on documentation alone [19], copy and paste style documentation promulgating through the electronic health record [20], and manual billing processes yielding error rates for standard CPT coding in anesthesia have reached as high as 38% [21]. According to CMS, in calendar year 2024 improper payments in the Traditional Medicare program reached 7.66% in 2024 translating to $31.70 billion in improper payments, with an estimated 59.8% due to insufficient documentation [22].

Emerging AI tools addressing the daily burden of documentation and chart review [23] represents another frontier where AI will have a real-world operational impact. Ambient AI scribes offer potential relief [24, 25] by automating significant portions of this workflow, [26] obviating the need for typed or voice dictated clinical notes, transforming documentation into an abbreviated review and editing task. Although these tools do not lead to increased revenue or patient volumes, they do increase clinician workplace satisfaction. Intelligent coding assistants pulling information from prior notes, labs, and imaging can promote automation of diagnosis coding, supporting both better clinical communication through improved diagnostic precision and accuracy, while simultaneously supporting both diagnosis related group-based inpatient billing, time-based evaluation and management outpatient services, and Medicare Advantage diagnosis coding [27].

Conclusion

Administrative centralization has shifted healthcare away personalized care to standardization. To reverse this trend and realign healthcare delivery around clinicians and patients, collectively health systems, technology companies, and policymakers should focus the deployment of automation and AI to reduce administrative burdens and instead strengthen the patient-physician relationship. By focusing innovation at the point of care, we can realign the technology with medicine’s core values with informatics tools supporting automation to help restore the joy and meaning of the practice of medicine.

Author Contributions

C.K. wrote the initial draft. A.D., J.E., J.S., and B.M participated in critical revision and editing of the manuscript. J.E. and B.M. provided supervision. All authors reviewed the manuscript.

Funding

None.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

Caleb Keng serving in an unpaid position as the Product Refinement and Community Engagement Lead at Dendritic Health AI, a technology startup focused on developing AI medical education study tools for students. Dr. DiGiorgio reports unrelated grant support from DePut Synthes, Florida Essential Healthcare Partnerships, and the Charles Koch Foundation. Mr. Spear reports receiving unrelated grant support from Stand Together Trust. Dr. Miller reports receiving unrelated grant support from Stand Together Trust, the Ohio State University Drug Enforcement and Policy Center and service as a Commissioner on the Medicare Payment Advisory Commission.

Disclosures

Caleb Keng serving in an unpaid position as the Product Refinement and Community Engagement Lead at Dendritic Health AI, a technology startup focused on developing AI medical education study tools for students. Dr. DiGiorgio reports unrelated grant support from DePut Synthes, Florida Essential Healthcare Partnerships, and the Charles Koch Foundation. Mr. Spear reports receiving unrelated grant support from Stand Together Trust. Dr. Miller reports receiving unrelated grant support from Stand Together Trust, the Ohio State University Drug Enforcement and Policy Center and service as a Commissioner on the Medicare Payment Advisory Commission.