So, you hate prior authorizations?
Why prior authorization exists and what to do about it
Prior authorization (PA) in health care has become one of the most contentious issues in medical practice today. This process, whereby health care providers (and, their patients) must obtain approval from insurance companies before providing certain treatments or medications, has sparked intense debate among practitioners, patients, and policymakers. While intended to control costs and ensure appropriate care, many argue it has created an unsustainable burden on the health care system and harm to patients.
Oh, and by the way, there were 46 million prior authorization requests in just Medicare Advantage alone in 2022. Its a big issue.
While it may be difficult to believe, prior authorization does serve an important purpose—it helps control cost for employers, taxpayers, and to some extent patients who are most often both employees and taxpayers. This cost savings is supposed to result from reductions in unnecessary, duplicative, wasteful, low-value, and fraudulent care. Does the patient really need a $120,000 spinal fusion surgery before trying physical therapy and other options? Does the patient really need another MRI after they had one last week? Does the patient need the brand name drug when the generic is the same thing?
Now you might look at these questions and say, “well I am not sure the answer to that, but I suspect the treating physician probably knows best, and they should be making that call not the insurance company.” You are probably right.
However, if I might argue the counterpoint for the sake of a balanced article, I would say that physicians sometimes1 choose procedures where they make more money even if another one might do just fine, and hospitals might want another MRI because they can charge $1000 for it. These organizations and individuals have their own bills to pay and so revenue must be managed accordingly.
Importantly, we know that certain factors like the profitability of certain procedures and payment models can alter the behavior of organizations and health care providers. Said differently, the rates that are paid to providers for certain procedures affect both the availability and selection of those procedures. It is a well-known concept that fee-for-service (FFS) payment models, the ones the dominate the American health system, lead to a higher volume of care provided than may be necessary.
So, the point is that sometimes prior authorizations are helpful at preventing utilization of services and products that may not be as cost-effective as another one, or that are entirely unnecessary. In these cases the patient and provider may call foul and be angry, but it may be for the better.
However, what happens when prior authorizations go wrong? What happens when the patient has tried all these other treatments and they still face barriers? What happens when the insurance company tries to play doctor2 despite obvious profit motives? What happens when the prior authorization process itself creates undue delay that harms patients who need rapid treatment? And, what can we do about it?
Understanding Prior Authorization and Medical Policy
Prior authorization requires health care providers to obtain pre-approval from insurance companies before proceeding with specific treatments, procedures, or medications. This process is intrinsically linked to medical policy (you can read one here)—detailed guidelines that outline when particular interventions are considered medically necessary. Insurance companies develop these policies based on clinical evidence, cost-effectiveness data, and their interpretation of medical standards. Many of these documents are published and publicly-available, while some are not.
The typical prior authorization process involves multiple steps: submitting detailed clinical documentation, waiting for review, responding to additional information requests, and potentially appealing denials. This process can take days or even weeks, during which patient care may be delayed. The friction that this process creates generates cost and feelings of discontentment among health care providers as well as general disgust from patients who expect their health plans will fund necessary treatments without delay.
It is important to note that many prior authorization requests are processed and approved and the procedures go on without a hitch. Still, based on the stories shared by patients on social media and in the news, we know many more do not get approved or are unnecessarily and detrimentally delayed. We also hear about the very egregious cases where great harm came to individuals who were denied services or who got caught up in lengthy prior authorization (PA) battles.
Thus, there are two major grievances with the concept of PA:
The administrative burdens
The de facto practice of medicine by insurance companies
So, how do we solve this? The problems to be solved can be distilled into the following questions:
How do we ensure the services that are paid for are necessary, beneficial, and based on the highest standard of clinical evidence and not unnecessary or wasteful? We do need checks and balances on these issues.
How do we build processes for PA that are based on the highest standard of clinical evidence and that allow for flexibility and professional judgement in unique circumstances? How can we create a system that removes insurance companies from the medical policy-making processes due to their conflict of interest (and, candidly, their blatant lack of expertise in conducting rigorous reviews of clinical evidence)?
How do we ensure the PA review processes are quick, flexible, and updated as new information and technologies come out? Part of the issue here is that technological and clinical innovation occurs faster than payment systems and policies can keep up.
Quick Historical Context
In all of my articles, I like to provide some historical context because it helps us understand that these practices are not set in stone—they were once created by humans and can be changed by humans once again. Prior authorization emerged in the 1980s as health care costs began rising dramatically. This was a period of time where both public and private health insurance was maturing and medical innovation was increasing at an exponential rate. This then led to efforts to curb the rate of growth in medical expenditures through changes in payment models, like those used by Medicare to pay hospitals (see the Inpatient Prospective Payment System). Initially implemented for high-cost procedures and medications, it was seen as a tool to prevent unnecessary treatments and control spending. Over time, its use expanded significantly, particularly with the rise of expensive specialty medications, advanced imaging technologies, and costly surgical procedures.
Part of the story of prior authorization is that of incredible success in medical and health innovation—unfortunately, more procedures and medical capabilities results in more opportunities to spend money which results in a need to ensure adequate checks and balances.
The Economic Case for Prior Authorization
Proponents of prior authorization argue that it serves several crucial functions in health care economics (and, obviously it pads insurance company bottom lines):
Cost Control: By preventing unnecessary or inappropriate treatments, prior authorization helps contain health care costs. Insurance companies cite significant savings through avoided unnecessary procedures and medications. (The trick is building a system that catches truly unnecessary costs, but allows necessary ones quickly and efficiently.)
Quality Assurance: The process theoretically ensures that patients receive evidence-based care aligned with current medical standards, potentially preventing harmful or ineffective treatments. The system is supposed to be built on clinical evidence, but it is clear that insurers’ incentive to minimize payouts can cloud their medical policies. Insurers are also not supposed to be practicing medicine. It is the role of the individual clinician and their professional society to determine professional best practices.
Resource Allocation: Prior authorization helps direct limited healthcare resources toward interventions with proven clinical benefit, theoretically optimizing the health care system's efficiency.
The Case Against Prior Authorization
Despite its intended cost-control benefits, prior authorization faces well-placed and mounting criticism from health care providers and patients:
Administrative Burden: Health care providers spend countless hours managing prior authorization requests, diverting time and resources from patient care. Studies suggest that practices spend an average of 14 hours per physician per week on prior authorization activities. That is 14 hours not dedicated to patient care services.
Clinical Impact: Delays in treatment approval can lead to adverse patient outcomes. A 2024 American Medical Association survey found that 24% of physicians report that prior authorization led to a serious adverse event for a patient including death, hospitalization, and severe impairment. Further, 90% of physicians indicate that prior authorizations have had a negative impact on patient health outcomes and 78% of physicians indicated that patients have abandoned necessary care due to the prior authorization process.
Questionable Authority: Many health care providers argue that insurance companies, driven by financial interests, should not have the power to override clinical decisions made by qualified medical professionals who directly assess patients. Clinicians in daily clinical practice likely have a much better idea about what will work for a particular patient than a random physician at an insurance company who reviews a prior authorization request.
Cost Shifting: While insurers save money through prior authorization, the administrative costs of the system are largely borne by health care providers, effectively shifting expenses within the health care system rather than reducing overall costs. At insurers, utilization management departments cost money. In order for them to generate a return for the insurer they must deny more care dollars than they cost in terms of salary, operations, and technology. This sets an additional profit motive and a minimum expected prior authorization denial amount.
Real World Impact
Cancer Treatment Delays: A 2023 American Cancer Society survey found that ~90% of cancer patients experienced delays in treatment due to prior authorization. For example, a stage 3 breast cancer patient in Ohio waited three weeks for chemotherapy approval, during which time her cancer progressed to stage 4.
Proposed Solutions
So, now, what can we do about this? What have other countries with centralized, government operated health insurance programs done? How can we reduce unnecessary care and ensure we are paying for the most effective care, while reducing administrative burden and conflicts of interest?
The answer is not simple as this is a very complex issue to address.
Technological Innovation
Modern technology offers promising solutions to reduce the administrative burden of prior authorization:
Electronic Prior Authorization (ePA): Implementing standardized electronic systems could streamline the process, reducing processing times and administrative costs. Real-time authorization systems could provide immediate decisions for many routine requests.
Artificial Intelligence: Machine learning algorithms could analyze clinical data and automatically approve straightforward requests, reserving human review for complex cases.
Both of these technologies might speed up the review process, which is positive, but these systems can also be weaponized to save money in new ways and remove critical human empathy from the equation. AI is a promising technology in this area as it can make decisions faster, but it must be trained on good data and taught ethical rules to produce a positive societal benefit and not just new ways to profit for insurers.
Much state and federal policy aimed at curbing these practices focuses on the prevention of delays in responses to increase the speed of the prior authorization processes.
But prior authorization AI or speed regulations do not make the medical policy rules themselves. For now at least, human intelligence is requires to synthetize complex clinical evidence, patient perspectives, and cost-effectiveness data to determine how and when certain services should be prioritized.
Also at the root of this problem is how insurers craft medical coverage policy.
Independent Medical Policy Standards
One way to combat the harms of insurer-created medical coverage policy is to craft an independent body or bodies (lets call these Health and Medical Policy Standards Boards (HMPSBs)) to conduct thorough evidence reviews, write treatment algorithms, and craft medical policies. In the United States, this could be bodies that are mapped to the Department of Health and Human Services regions. This decentralized model would align well with our federal system, would promote independent thinking, and would allow regions to learn from one another (i.e., if Region 1 makes a different policy than Region 2, we can compare outcomes and patient satisfaction).
The HMPSBs would function similarly to the existing United States Preventive Services Task Force (USPSTF), which is tasked with reviewing the clinical and cost-effectiveness evidence for preventive services. The work of this body is not just academic as all insurers are required to cover, with no patient cost-sharing, services rated as an A or a B by the USPSTF. This approach could be adapted to medical policy as a whole.
One of the key benefits of the USPSTF is its independence from conflicts of interest3 and criteria for membership. Membership criteria is as follows directly from the Agency for Health Care Research and Quality:
“Qualified applicants and nominees must, at a minimum, demonstrate knowledge, expertise, and national leadership in the following areas:
The critical evaluation of research published in peer reviewed literature and in the methods of evidence review.
Clinical prevention, health promotion, and primary health care.
Implementation of evidence-based recommendations in clinical practice, including at the clinician-patient level, practice level, and health system level.
Some USPSTF members without primary health care clinical experience may be selected based on their expertise in methodological issues, such as meta-analysis, analytic modeling, or clinical epidemiology. For individuals with clinical expertise in primary health care, additional qualifications in methodology would enhance their candidacy.”
For the HMPSBs, one might want a nice balance between practicing clinicians, clinical researchers, health economists, and health services researchers to minimize the risk of financial conflicts of interest.
This idea is not necessarily unique. The United Kingdom's National Institute for Health and Care Excellence (NICE) is a good analog for how this system might operate in the United States albeit with some unique American decentralization tendencies. NICE provides the following key benefits to the UK:
Standardization: Universal guidelines would eliminate the confusion caused by varying requirements across different insurers in the US.
Scientific Rigor: An independent body could focus solely on clinical evidence without financial considerations, ensuring more objective standards. NICE provides this benefit to the UK’s National Health Services (NHS). And, I might add, that their health system performs much better than the American one despite the constant bad press about the NHS (challenges exist everywhere).
Transparency: Clear, publicly available standards would help providers and patients understand and prepare for authorization requirements. Part of the frustration with the current US PA models is that the processes are intentionally obfuscated to make it more difficult and vary across insurance types (a practice might accept 15 different plans they have to manage).
The NICE Model
NICE provides a compelling (and, nice!) example of how independent medical policy-making can work effectively. Established in 1999, NICE evaluates medical treatments and technologies based on both clinical- and cost-effectiveness.
NICE's Evaluation Process
NICE follows a rigorous, multi-step evaluation process for each medical intervention:
Scoping Phase: NICE first determines the exact questions to be addressed, consulting with stakeholders to ensure all relevant aspects are considered. This includes defining the patient population, intervention, comparators, and outcomes to be evaluated.
Evidence Gathering: NICE's evaluation committees systematically review all available clinical evidence, including:
Published clinical trials
Real-world evidence
Patient outcome data
Expert testimonies
Economic analyses
Technology Assessment: Independent academic groups conduct detailed assessments, analyzing both clinical and cost effectiveness. This includes:
Systematic literature reviews (like this one, here)
Meta-analyses of clinical trials (like this one, here)
Economic modeling
Quality of life impact assessments
Stakeholder Input: Throughout the process, NICE actively seeks input from:
Healthcare professionals
Patient advocacy groups
Industry representatives
NHS representatives
Public health experts
Draft Guidance Development: Committees develop preliminary recommendations based on:
Clinical effectiveness data
Cost-effectiveness analysis using the £20,000-30,000 per QALY threshold (here describes what a QALY is if you want to know)
Equity considerations
Implementation feasibility
Public Consultation: Draft guidance is published for public comment, allowing all stakeholders to provide feedback before final recommendations are made.
Final Guidance: After considering all feedback, NICE issues final guidance that is regularly reviewed and updated as new evidence emerges.
This is a very systematic process that takes time. It is very important that any new system implemented in the United States take into account the need for flexibility at the treating provider level. Importantly, it is critical to note that these systems rely on clinical and economic research to make determinations. It takes many years after a new procedure or drug is introduced for that evidence to come out in sufficient quantities. It is critical that a new system for making medical policy does not hinder the adoption of new technologies.
Importantly, systems that dictate health insurance coverage policy exist, in some form, within the US already, such as how Medicare Advantage plans must not have coverage policies that are more limiting than traditional Medicare which is controlled by the Centers for Medicare and Medicaid Services. Similarly, state Medicaid agencies are supposed to control coverage and benefits for their contracted managed care plans. Enrollees in both have escalation processes available when their plans make an adverse coverage determination. The problem with these escalation processes to independent medical necessity reviews of individual cases is that they further prolong access to care.
Conclusion
Prior authorization in its current form represents a significant challenge in healthcare delivery. While its goals of cost containment and quality assurance are legitimate, the current system imposes excessive burdens on healthcare providers and can harm patient care. Technology solutions and independent policy-making bodies offer promising paths forward.
The implementation of a NICE-like system in combination with advanced technology could maintain the benefits of prior authorization while dramatically reducing its negative impacts. Such reforms would require significant changes to the current healthcare system, but the potential benefits—improved patient care, reduced administrative burden, and more efficient resource utilization—make such changes worth pursuing.
As healthcare continues to evolve, finding the right balance between cost control and efficient care delivery becomes increasingly crucial. The solution likely lies not in abandoning prior authorization entirely, but in transforming it into a more streamlined, standardized, and evidence-based system that serves both patients and providers while maintaining necessary cost controls.
Note: I purposefully say “sometimes” here because I generally believe that most physicians do their best to take cost into account.
Insurance companies do employ physicians and other clinicians as medical directors and advisors to make many of their medical policy decisions and to act as “peer-reviewers” when prior authorizations escalate past the initial phases. So, in a sense, they do have physicians practicing medicine, but ones who are often commenting on specialties and procedures they may never have performed. Many of these Medical Directors have not practiced medicine in some time.
To the extent possible for humans.