The friction between healthcare providers and insurance payers has long been a defining challenge of the American medical landscape. At the heart of this struggle lies "prior authorization"—the process by which physicians must obtain approval from health insurance plans before performing a procedure or prescribing a medication. As artificial intelligence (AI) begins to permeate the administrative backend of healthcare, the debate over its role in these high-stakes decisions has reached a boiling point.
At a recent panel discussion at the HLTH conference, industry leaders grappled with a fundamental question: When it comes to determining whether a patient receives life-saving care, where should the line be drawn between algorithmic efficiency and human clinical judgment?
The Consensus: AI as a Tool, Not a Gatekeeper
The overwhelming sentiment among the panelists was clear: while AI possesses the potential to streamline administrative workflows, it must never serve as the final arbiter of medical necessity.
Sean Robbins, Executive Vice President and Chief Operating Officer of the Blue Cross Blue Shield Association (BCBSA), took a firm stance on the autonomy of clinical decision-making. "Artificial intelligence should not deny an authorization for medical necessity," Robbins stated. "A licensed clinician should make the call. Not review the call; they should make the call. Can you use artificial intelligence as a licensed clinician to help you in that? Absolutely, but the person, the human, the licensed clinician, should be the one making the call."
Robbins’ perspective reflects a growing recognition within the payer industry that AI, while capable of processing vast datasets, lacks the nuanced understanding required for individual patient cases. The concern is that delegating such decisions to opaque algorithms could lead to systematic biases, "black box" denials that are difficult for providers to contest, and a degradation of the patient-physician relationship.
A Systemic Failure: The Case for Provider-Led Care
While Robbins emphasized the need for a "licensed clinician" to handle denials, other experts argue that the current structure of that decision-making process is fundamentally flawed. Dr. Warris Bokhari, CEO of Claimable—a firm dedicated to helping patients navigate and appeal insurance denials—argued that the authority to determine necessity should rest with the treating physician, not the insurance company.
"I don’t think it’s an unnamed person at an insurance company who is effectively practicing medicine [who should be making the decision]," Bokhari remarked. "It is a clinical decision that’s being rendered by a human who has never met the patient, never looked at their tests, never met their family, never held their hand, has no context."
Bokhari’s critique highlights a profound disconnect. To the insurance company, a denial may be viewed as a "coverage decision" based on policy guidelines. To the patient and their doctor, it is a barrier to clinical necessity that can have catastrophic health outcomes.
Chronology of Contention: From Administrative Burden to Ethical Crisis
The evolution of prior authorization has been a steady march toward increased complexity. In the early days of managed care, prior authorization was intended as a moderate "guardrail" to prevent over-utilization. However, over the past three decades, it has transformed into a sprawling administrative apparatus.
- The 1990s (The Foundation of Denials): During this era, evidence-based medicine was in its infancy. Payers began implementing standardized protocols to manage costs. Critics point to this era as a period where criteria were established that have remained largely stagnant, even as medical technology has leaped forward.
- The 2010s (The Digital Transition): As healthcare moved to Electronic Health Records (EHRs), prior authorization became digitized. This led to an explosion in volume, as the ease of submitting requests electronically prompted payers to require authorization for an ever-widening array of services.
- 2020–Present (The AI Acceleration): The integration of machine learning and predictive modeling promised to solve the "burden" by automating approvals. However, as these systems were deployed, reports surfaced of automated denials based on outdated clinical data, fueling the current backlash.
Supporting Data and Evidence: The "1994" Problem
One of the most damning arguments presented by Dr. Bokhari involves the obsolescence of evidence. He specifically cited Blue Cross and Blue Shield of Illinois, alleging that the insurer utilized clinical evidence dating back to 1994 to deny liver transplants for cancer patients.
"For anyone who might remember 1994, ER was coming onto television," Bokhari noted, highlighting the absurdity of using three-decade-old data to govern modern surgical outcomes. "We’ve now iterated to The Pitt; medicine has moved on. A lot of things have changed, but they’re still using the same evidence to deny."
This discrepancy between the pace of medical innovation and the rigidity of insurance coverage guidelines is a recurring theme in health policy circles. When insurers rely on static, outdated evidence, they effectively force providers to practice medicine in the past, often to the detriment of patients whose only hope lies in modern, cutting-edge therapies.
Official Responses and Industry Accountability
In the face of these allegations, the industry is struggling to define a path forward that balances fiscal responsibility with patient safety.
Sean Robbins, representing the BCBSA, acknowledged the necessity for improvement. "I don’t want there to be an inappropriate denial in any situation ever," he said. He emphasized that the industry is undergoing a period of self-reflection regarding accountability and transparency. According to Robbins, the goal must be "getting the right answers to the right people quickly."
However, accountability remains an elusive target. Dr. Bokhari argued that there must be tangible consequences for health plans that make erroneous denials—particularly when those denials delay care for terminally ill patients. The pressure is mounting for federal and state regulators to step in, potentially mandating shorter turnaround times for appeals and requiring insurers to publish the clinical evidence supporting their denial algorithms.
A Call for a Structural "Rethink"
Dr. Mandy Cohen, a prominent figure in health policy and a national advisor at Manatt Health, suggested that the industry is missing the forest for the trees. By focusing exclusively on "speeding up" the prior authorization process, insurers and providers are ignoring the root cause of the friction.
"As opposed to saying, ‘How do we speed up the process?’ it’s more, ‘How do we rethink this whole process?’" Cohen argued. She pointed out that the entire model is tethered to outdated payment incentives. "The reason insurance companies exist and have these guardrails is because providers do ask for low-value care. Let’s be honest. That is happening."
Cohen’s perspective provides a vital counter-narrative: prior authorization, in its current form, is a symptom of a healthcare system that struggles with the alignment of incentives. If providers were incentivized solely by patient outcomes rather than the volume of services rendered, the need for stringent, invasive prior authorization might diminish.
Implications for the Future of Healthcare
The debate at HLTH signals a pivot point for health tech. As AI becomes more sophisticated, the temptation to use it for cost-cutting will grow. However, the panelists’ warning is clear: Efficiency at the expense of human judgment is a dangerous trade-off.
Key Implications Include:
- Regulatory Scrutiny: We are likely to see increased legislative efforts to regulate the use of AI in medical coverage decisions. This may include "human-in-the-loop" requirements that are legally enforceable.
- Transparency Requirements: Insurers may soon be forced to disclose the clinical evidence—and the age of that evidence—used by their algorithms to evaluate medical necessity.
- Shifting Incentives: As Dr. Cohen suggested, the industry may move toward value-based care models that reduce the need for administrative "gatekeeping" by aligning the financial interests of payers and providers.
- The Rise of Advocacy Tech: Companies like Claimable, which leverage technology to challenge the status quo, are likely to see increased demand as patients and providers seek tools to fight back against automated denials.
The future of healthcare delivery depends on how we resolve this tension. Technology should empower the clinician to spend more time with the patient, not force the clinician to spend hours fighting a black-box algorithm. As we integrate AI into the fabric of medicine, the goal must remain consistent: technology should be a bridge to better care, not a wall standing between the patient and the treatment they need.
