How a cost-control tool meant to review care became, for many patients and clinicians, a test of stamina
Late in 2025, a scan delivered the kind of news that families in cancer treatment spend years dreading and never really prepare to hear. Eliza Cowey, a Pittsburgh teenager with an extremely rare brain cancer called astroblastoma, learned that the disease had spread to a new part of her brain, one surgeons could not reach. By then, she and her family had already endured the long grammar of serious illness: surgeries, scans, chemotherapy, radiation, the suspended calendar of hospital life. What changed with that scan was not only the anatomy of the disease. It was the horizon. Her doctors were no longer speaking in the language of cure. They were speaking in the smaller, harsher language of time.
A blood test suggested that Eliza might respond to a pair of chemotherapy drugs. The goal was not to reverse what had happened. It was to slow it. Her mother, Jody Handley, described the treatment in plain terms: the family was now trying to buy time. But before Eliza could begin the regimen, her doctors needed approval from the family’s insurer, Aetna. One drug was approved. The second was denied. Because the drugs were meant to be used together, the partial approval was, in practice, a refusal.
Eliza’s physician sought the required peer-to-peer review with a doctor hired by the insurer. According to reporting by Pittsburgh public radio station WESA, the doctors never connected. Jody kept calling. The hospital kept working. Eventually, with help from a Philadelphia nonprofit called Persius, an 11-page appeal was drafted. Only then did the answer change. Eliza got the medication.
That detail—the eventual yes—is the center of the story. Not because a reversal makes the system look merciful. It does almost the opposite. A final no is brutal, but at least it is legible. A delayed yes asks a more unsettling question: what, exactly, was being evaluated in the meantime? The clinical case had not fundamentally changed. The patient had not become less sick. What changed was that a family and a set of clinicians had managed to survive the bureaucracy long enough to alter the decision.
No serious person believes that any health-care system can offer unlimited access to everything immediately, with no scrutiny, no limits, and no cost control. Every system rations something: money, time, waiting, drug availability, specialist access, or paperwork. The American argument for prior authorization rests on a rational premise. Someone must review whether a treatment is covered, medically necessary, safe, or cost-effective. The problem is not that scrutiny exists. The problem is the form it has taken in the United States: late, fragmented, repetitive, privately administered, and often inserted only after a doctor and patient have already decided on a plan. What is defended as oversight is experienced, at the bedside, as interruption.
Utilization review has older roots than its current reputation suggests. Hospitals have long examined care for appropriateness and cost. But the broad, routine external review of proposed treatments—the kind Americans now associate with prior authorization—expanded dramatically during the managed-care era of the 1980s, as insurers and employers searched for ways to restrain rapidly rising health costs. What began as a targeted tool gradually widened into a routine feature of medical care.
In theory, the practice is meant to protect both patients and health systems. Managed-care literature still describes prior authorization in terms of patient safety, evidence-based medicine, and appropriate use of finite resources. Some economic research supports at least part of that case. A 2023 analysis from the National Bureau of Economic Research found that prior-authorization restrictions in Medicare Part D reduced use of restricted drugs by roughly 27 percent and lowered drug spending by about $96 per beneficiary per year, while generating about $10 in administrative costs. From the perspective of payers, the policy can work.
But what works economically from above can look very different from below. The clearest public data comes from Medicare Advantage, whose plans must report their prior-authorization activity. In 2024, Medicare Advantage insurers made nearly 53 million prior-authorization determinations. They fully or partially denied about 4.1 million requests. Only 11.5 percent of those denials were appealed. Of the denied requests that were appealed, more than 80 percent were partially or fully overturned.
Those numbers do not prove that every initial denial was incorrect. Some reversals may reflect additional documentation submitted during the appeal process. Still, the pattern is difficult to ignore. A system in which most denials are never challenged, but most challenged denials are later reversed, begins to look less like a clean test of medical necessity than a test of endurance.
This is one reason prior authorization has become so widely disliked. A 2026 survey by KFF found that 69 percent of insured adults describe prior authorization as burdensome, and about a third say it is their single biggest non-cost barrier to obtaining health care. Physicians report similar frustrations. The American Medical Association found in 2024 that medical practices complete an average of 39 prior authorizations per physician each week, spending roughly 13 hours on them. Nearly nine in ten physicians say the process contributes to professional burnout.
What people resent, in other words, is not simply paperwork. It is the feeling that important medical decisions have been placed inside a maze whose chief logic is delay.
Cancer care exposes that logic with unusual clarity because delay is not neutral there. It is not merely a pause between clerical steps. It can mean disease progression, increased pain, or abandoned treatment. In a 2024 survey by the American Society for Radiation Oncology, 92 percent of radiation oncologists said prior authorization causes treatment delays. One-third said it had led patients to abandon treatment altogether. Thirty percent reported cases in which delays contributed to emergency visits, hospitalization, or permanent disability. Seven percent said it had contributed to a patient’s death.
A 2025 survey by CancerCare found similar patterns among patients. Eighty-five percent reported facing prior authorization during their cancer treatment, and 40 percent said their most recent authorization delayed care. Yet most of those requests were ultimately approved anyway.
Eliza’s case is therefore both extraordinary and not extraordinary at all. It is extraordinary because she is a child whose doctors were no longer pursuing cure. It is not extraordinary because the broader oncology pattern already exists.
What makes prior authorization especially maddening is that it is often described as though it were faceless, a kind of weather system moving through medicine. In reality it has many authors. Legislators set regulatory boundaries. Employers choose benefit structures. Insurers design coverage rules. Pharmacy benefit managers shape drug formularies. Vendors build the digital portals through which requests must travel. Reviewers apply clinical criteria. Medical directors oversee adverse determinations.
The modern achievement of the system is that responsibility is distributed so widely that no single participant must feel entirely accountable for the burden the whole structure creates. The bureaucracy acquires the moral force of nobody in particular.
That distributed authorship was visible in Eliza’s case. According to WESA’s reporting, the process involved a treating oncologist, a third-party reviewer, an insurer, hospital staff, and finally a nonprofit organization that stepped in to help draft the appeal. A retired oncologist interviewed for the story noted that peer-to-peer calls often do not even connect doctors in the same specialty. Even when real people populate the process, they do not necessarily share the same expertise or accountability to the patient whose care is being delayed.
Legally, patients do have rights within this system, though those rights can be difficult to navigate during a medical crisis. Federal rules require insurers to provide written explanations for denied claims and to offer internal appeals. Patients can request relevant documents and may seek independent external review after exhausting internal appeals. In urgent cases, those reviews are supposed to move quickly.
Yet the effectiveness of these protections often depends on factors patients rarely understand while facing illness—such as what type of insurance plan they actually have. Some employer health plans operate primarily under federal law and are largely exempt from state insurance regulation. The remedies available to patients in disputes with those plans are narrower than many people expect.
This structure helps explain why prior authorization can feel insulated from ordinary accountability. The experience of a patient confronting a frightening medical decision is translated, in legal terms, into a benefits dispute.
The rise of advocacy groups like Persius reflects that reality. Such organizations help patients challenge insurance denials and navigate appeals processes that have grown increasingly complex. Persius has also built a public database of insurance-denial data obtained through public-records requests in Pennsylvania. In another era, one might have expected this kind of transparency to come primarily from regulators. Now it also comes from civil society.
Other countries ration care differently. Britain’s National Health Service often makes coverage decisions at a national level through centralized evaluation bodies, and Australia’s pharmaceutical system allows certain restricted medicines to be prescribed using streamlined authorization codes. These systems still limit access to some treatments. Scarcity does not disappear. But more of the decision-making occurs upstream, before the moment of treatment.
The United States has begun experimenting with reforms. Federal regulators have pushed for greater electronic standardization of prior-authorization processes, and some states have adopted “gold-card” policies that exempt physicians whose requests are consistently approved. Major insurers have also pledged to reduce administrative friction by expanding real-time digital responses.
These changes may help, but they remain incremental because the incentives supporting prior authorization are deeply embedded. Employers want lower premiums. Insurers want tools to control utilization. Administrative vendors are already integrated into the system’s infrastructure. The architecture persists not because it satisfies everyone involved but because it continues to serve powerful institutional interests.
That is why prior authorization occupies such a peculiar place in American medicine. It is widely disliked by patients and physicians alike, yet it endures. The practice has become one of the system’s primary ways of translating the abstract problem of limited resources into a series of individualized ordeals. Instead of appearing openly as a national rule or waiting list, rationing arrives as paperwork and delay.
Eliza’s story should not be asked to prove more than it can. It does not demonstrate that every insurer behaves unlawfully or that every prior authorization is unjustified. What it reveals is how quickly a technical review process can become morally dissonant.
An 11-page appeal is not exactly a price. It is a measure of what the system now demands before it yields. Persius’s document summarized both the oncologists’ clinical reasoning and the contractual language of the insurance plan. In effect, a family already confronting a life-threatening illness had to assemble something close to a legal brief in order to obtain not rescue, but time.
Eliza did eventually receive the drugs. Her mother fought. Her doctors fought. Advocates helped. The answer changed. That victory matters because the time mattered.
But the victory is not the same thing as vindication. An eventual approval does not demonstrate that the system functioned well. It shows only that this family managed to persist long enough to outlast the delay.
And so the modern ritual of American health care sometimes requires something quietly absurd. Before patients can receive treatments their doctors have already judged necessary, they must demonstrate not only medical need but administrative endurance. Eliza Cowey’s story is painful because of the illness at its center. It is instructive because the system added another obstacle to it—and called that obstacle review.
Reference
Koscinski, K. (2026, March 3). Prior authorization delayed Pittsburgh teen’s cancer treatment. This org helped her fight back. 90.5 WESA.
[This article does not constitute legal or medical advice]