As Claims Denials Surge, AI and Data-Driven Insights Equip Clinicians, Hospitals to Fight Back

In recent years, insurance claim denials for essential medications have risen significantly, with a 16% increase from 2018 to 2024, according to Arif Nathoo, MD, CEO and cofounder of Komodo Health. This trend particularly impacts generic drugs used for managing conditions like diabetes and asthma. Denials often arise from cost-containment strategies, such as prior authorizations, which can delay patient access to necessary treatments.1

The challenges posed by these denials reflect broader systemic issues within health care, where practices like step therapy often conflict with clinical guidelines. The American Journal of Managed Care® (AJMC®) spoke with experts about the importance of optimizing data to keep up with these changes and implementing artificial intelligence (AI) to reduce the burden on providers when confronted with increasing denials and prior authorizations while ensuring their patients’ access to care and treatment.

Christopher Gold, DO, an internal medicine physician at Mount Sinai in New York, noted that these denials not only exacerbate patient symptoms but also strain the patient-provider relationship, as providers navigate complex insurance protocols, in an interview with AJMC.

Christopher Gold, DO | Image Credit: © Mount Sinai

“The idea behind the denials and the prior authorization is to lower costs in the short term, but sometimes treatment delays or lapses can potentially increase costs with preventable complications, and emergency and more frequent primary care office visits,” he said.

Insurance claim denials and appeals in Affordable Care Act Marketplace plans are especially relevant given the significant impact these denials have on patients’ access to care. In 2023, 20% of all claims for qualified health plans on HealthCare.gov were denied, with notable variation across insurers and states, according to KFF.2 Nathoo stressed the need to monitor outcomes to understand the effects of these policies.

Harnessing Data to Identify Health Disparities

In an era where health care costs continue to rise and disparities persist, data-driven insights are playing an increasingly vital role in shaping policy, research, and patient advocacy. Komodo Health, a company at the forefront of health care analytics, has spent the past decade mapping the US health care system to better understand gaps in care and unmet patient needs. Nathoo shared how data reveal critical trends—ranging from rare disease diagnosis delays to rising prescription drug denials.

Komodo Health utilizes its “Healthcare Map,” an extensive dataset that tracks health care costs and patient outcomes nationwide, to identify patterns often overlooked by individual physicians. This approach has had a significant impact on rare disease diagnostics, enabling earlier testing and treatment for patients who typically face long diagnostic delays. Additionally, the company has focused on common conditions like obesity, stroke, and heart disease, revealing disparities in access to advanced treatments such as glucagon-like peptide-1 medications.

Arif Nathoo, MD | Image Credit: © Komodo Health

One of Komodo Health’s most striking findings has been the increasing rate of prescription drug denials. According to the company’s recent data, pharmacy claim denials initially declined between 2016 and 2018 but show to be rising based on data from 2018 to 2024.

“Over the last 5 to 6 years, we’ve seen a rising trend in medication denials—not just for administrative reasons, like refill timing, but for more concerning reasons like prior authorizations and medical necessity reviews,” Nathoo said. “That tells us something is happening where payers and pharmacy benefit managers are taking more restrictive actions against patients.”

Commonly Denied Medications: What the Data Show

Among the most frequently denied medications identified in Komodo Health’s analysis are essential drugs such as insulin, albuterol, vitamin D, and apixaban (Eliquis; Bristol-Myers Squibb/Pfizer). Although high-cost branded medications often face barriers to approval, it is particularly concerning that even generic drugs, which are typically more affordable, are also subject to restrictions.

“For medications like inhalers, we see cases where payers may prefer a different formulation, but that alternative may not be therapeutically equivalent for every patient,” Nathoo explained. “These restrictions can push patients out of the system, leaving them without access to the treatments they need.”

Despite national efforts to cap insulin prices and improve affordability,3 insulin denials, in particular, have been a longstanding issue, in which future directions appear uncertain since President Donald J. Trump’s return to office.4 Komodo Health’s data suggest that although some progress has been made in reducing cost barriers, formulary restrictions and prior authorization requirements continue to create obstacles for many patients, providers, and health systems.

“Sharing data across the health care ecosystem is nothing short of a feat of both imagination and strength,” Nathoo said. “But the advantage we have in the US is that, because doctors must code services for reimbursement, we actually get very detailed insights into patient conditions and treatments.”

Komodo Health has spent years working to fill in the gaps in health care data, particularly in underrepresented populations. The company has increased its visibility into Medicaid populations and Medicare Advantage plans, which are often overlooked in traditional research. By shining a light on these areas, Komodo is helping to ensure that policymakers and health care organizations can make more informed decisions about where to allocate resources.

How AI Can Reduce Administrative Burdens in Hospitals

As payers tighten criteria for claim approvals, hospitals across the US are grappling with increasing denial rates, placing additional strain on already thin operating margins. The process of appealing these denials is not only time-consuming but also requires extensive clinical expertise, according to Michael Gao, MD, a physician-turned-health-tech entrepreneur, who has been on the frontlines of tackling this challenge through AI-driven solutions.

Michael Gao, MD | Image Credit: © SmarterDx

As the cofounder of SmarterDx, Gao and his team have developed technology that streamlines the appeals process, helping hospitals secure the reimbursements they are entitled to while reducing administrative burden.

“Your average hospital has a margin of about 1.7% to 2.1%,” Gao explained in an interview with AJMC. “If you think about it in personal finance terms, it’s like living paycheck to paycheck. And on top of that, hospitals have to fight to make sure they are paid correctly.”

For physicians, the frustration is even more pronounced. “When I was a hospitalist at Weill Cornell, my job was to take care of patients, not to argue with insurers,” Gao said. “But when a claim was denied, I’d find myself on the phone debating medical necessity with someone whose full-time job was denials.” This process between highly trained clinicians focused on patient care and payer representatives who specialize in claim reviews can leave hospitals struggling to recover lost revenue.

Denials often contain errors, and overturning them requires an exhaustive review of patient records. The burden of manually combing through thousands of data points in a patient’s medical chart can be overwhelming. SmarterDx aims to bridge this gap by using AI to automate and enhance the claims review process.

“The problem is that a payer might take just a few minutes to deny a claim, but it takes a provider an hour or more to build a clinical argument for appeal,” Gao explained. “What our AI does is automatically pull in relevant clinical data to support a diagnosis or justify a level of service, significantly cutting down on time spent searching through records.”

By integrating AI into the appeals process, SmarterDx allows hospital staff to focus on refining arguments rather than sifting through documentation.

“Instead of taking 2 hours to build an appeal from scratch, it takes 10 minutes to edit a well-structured response,” Gao said.

Beyond saving time, AI-driven tools like SmarterDx help hospitals address claim denials more comprehensively. Many hospitals prioritize high-dollar claims, leaving smaller claims unaddressed due to staffing constraints.

“With automation, hospitals can work through their entire backlog of denials, not just the highest-value ones,” Gao noted. “And even in high-dollar cases, AI allows appeal specialists to focus on crafting stronger arguments rather than spending time pulling data.”

This efficiency gain also helps level the playing field between payers and providers.

“Insurers have been using AI to streamline their processes for years,” Gao pointed out. “Hospitals, especially smaller ones, often lack the resources to develop similar tools in-house. That’s why vendors like us exist: to provide scalable AI solutions that give hospitals the same technological advantages.”

Driving Change Through Data and AI

Addressing these issues is largely hindered by the fragmented nature of the US health care system, which is split across a vast network of private and government payers, as well as tens of thousands of health care institutions.

The ability to analyze data at scale, as done by Komodo Health, has already led to tangible improvements in patient care and policy. By identifying disparities in access to medications and treatments, this approach is helping life sciences companies, insurers, and health care providers refine their strategies and expand access to life-saving therapies.

“We believe that democratizing health care data is key to driving meaningful change,” Nathoo said. “The more we can illuminate these gaps, the more we can empower stakeholders to take action—whether it’s insurers rethinking coverage policies, providers adjusting care models, or policymakers advocating for change.”

While SmarterDx currently focuses on clinical denials, Gao sees potential for AI to transform other aspects of hospital revenue cycle management.

“At the end of the day," he notes, "the revenue cycle is about generating a receipt for the care provided and ensuring hospitals get paid for it."

With $350 billion of the nation’s health care spending tied up in administrative costs, he sees AI as having the potential to significantly reduce inefficiencies and free up resources for direct patient care.

“This isn’t just about automation,” Gao emphasized. “It’s about augmentation—helping providers do their jobs better while improving financial sustainability for hospitals.”

As health care costs continue to rise and the burden of chronic disease grows, data-driven insights will be essential in shaping the future of patient care. The role of technology in health care, demonstrated by Komodo Health and SmarterDx, serves as a powerful example of how data and analytics can be harnessed to create a more equitable and efficient health care system for patients seeking care and the professionals working to deliver it.

References
1. Steinzor P, Grossi G. How insurance claim denials harm patients’ health, finances. AJMC®. February 25, 2025. Accessed March 6, 2025. https://www.ajmc.com/view/how-insurance-claim-denials-harm-patients-health-finances

2. Lo J, Long M, Wallace R, et al. Claims denials and appeals in ACA Marketplace plans in 2023. KFF. January 27, 2025. Accessed February 24, 2025. https://www.kff.org/private-insurance/issue-brief/claims-denials-and-appeals-in-aca-marketplace-plans-in-2023/

3. Santoro C. Lower drug prices announced under Medicare negotiation program. AJMC®. August 15, 2024. Accessed March 6, 2025. https://www.ajmc.com/view/lower-drug-prices-announced-under-medicare-negotiation-program

4. Steinzor P. Trump reverses some Biden drug pricing initiatives, potentially impacting Medicare costs. AJMC®. January 22, 2025. Accessed March 6, 2025. https://www.ajmc.com/view/trump-reverses-some-biden-drug-pricing-initiatives-potentially-impacting-medicare-costs