Prior authorization friction is pushing patients out of care before appeals are filed

A new national patient survey suggests prior authorization is doing more than delaying treatment. For many patients, it is ending treatment altogether.

Among respondents who experienced a delay, denial, or medication restriction tied to prior authorization, 62% said they gave up on care. In the full sample, 48% reported the same outcome. The pattern is especially concerning because nearly half of patients who actually appealed eventually won, indicating that a substantial share of abandoned cases may have been winnable had the process continued.

The survey covered 500 U.S. adults in November 2026 and was commissioned by Penguin Ai, a health care artificial intelligence company that sells prior authorization tools to payers and providers. Every respondent had needed at least one prior authorization for themselves or a family member in the past two years. Among the 273 respondents who reported a delay, denial, or medication restriction, the abandonment rate reached 62%. The release does not describe recruitment or survey administration, so the findings are best interpreted as a snapshot of patient experience rather than definitive national prevalence data.

The clinical consequence is plain. When a patient walks out of the office with a treatment plan and never gets past the insurer’s gate, disease does not pause for administrative review.

Appeals often succeed, but most patients never try

Appeals were not futile in this survey. Forty-four percent of respondents appealed. Among those who did, 47% had the decision overturned, 24% saw it upheld, 20% were still waiting, and 8% gave up during the appeal itself.

That mix matters because it shows a process with real yield sitting beside a much larger pool of patients who never enter it. Patients were not uniformly refusing to challenge denials; many simply stopped before the appeal stage, often after days of uncertainty, paperwork, and repeated phone calls. In practice, the barrier is not only the denial itself. It is the cumulative burden of trying to reverse it.

Physician behavior mirrors that pattern. The AMA’s 2025 Prior Authorization Physician Survey found that 32% of physicians said prior authorization requests are often or always denied. When asked why they do not appeal, 59% cited past experience suggesting the appeal will not succeed, 52% cited insufficient staff time, and 49% said patient care cannot wait. The persistence problem is shared on both sides of the exam room.

Delays are turning into out-of-pocket costs, skipped visits, and worse health

The downstream effects reported by patients were broad and clinically familiar. Across the survey population, 36% paid out of pocket, 32% delayed care, 21% skipped care entirely, 17% turned to an emergency department or urgent care center, and 13% said their health worsened. Only 20% reported none of those outcomes.

Those numbers align with what physicians report seeing after the fact. In the AMA survey, 26% of physicians said prior authorization had led to a serious adverse event for a patient, including hospitalization, permanent impairment, or death. Twenty percent reported hospitalization tied to prior authorization, while 95% said the process delays access to necessary care. Seventy-nine percent said patients had abandoned recommended treatment because of prior authorization barriers.

The mechanism is straightforward. Delay favors progression. For chronic disease, that can mean symptoms worsen while medication initiation stalls. For acute problems, a missed treatment window can shift care toward emergency services or hospitalization. Prior authorization is often presented as a utilization control tool, yet both the patient survey and physician survey point toward a familiar downstream pattern: higher friction, later care, and more expensive rescue services.

Waiting four to seven days can be enough to break the chain

Time is the most visible pressure point in this process. In the survey, 62% waited four days or longer for a decision, 14% received a same-day answer, and the most common wait time was four to seven days at 29%.

Those delays were not just inconvenient. Seventy percent of respondents spent at least an hour on calls and paperwork, 16% spent more than five hours, and 49% rated their stress as 4 or 5 out of 5. Extended administrative effort can become a clinical barrier in its own right, especially for patients balancing work, caregiving, transportation, and symptom burden.

A large part of the frustration came from silence. Thirty-eight percent of patients received little or no communication while waiting for approval, and 47% said a lack of transparency left them anxious and uncertain. Dean Erhardt of D2 Solutions has described the confusion this creates, noting that patients often do not know whom to call when a request is pending. Administrative errors also appear to be a major driver of delay, which means a meaningful share of these problems may be preventable with cleaner submissions and clearer status updates.

AI may reduce missing paperwork, but patients distrust it without human oversight

The survey also explored patient reaction to artificial intelligence in prior authorization, a topic that increasingly shapes payer workflows. Only 8% said they were told AI or automated software was involved in the decision. Another 32% suspected it, and 32% had no idea.

Patient sentiment was cautious at best. Sixty-nine percent said they would be more upset if they learned AI had made or recommended a denial, and 54% said they would switch insurers if AI were used to review claims. Yet support was not absent. Thirty-eight percent would support AI if it made approvals faster and more accurate, 49% would support it only with human oversight, and 13% said no.

Physicians are wary as well. In the AMA survey, 60% said they are concerned that insurers’ use of AI-assisted decision-making will increase denial rates. In theory, AI can reduce avoidable denials by catching missing documentation before submission. Brian Covino, chief medical officer of Cohere Health, said incomplete submissions drive many denials and described a drop in missing-information issues when prompts are used to improve documentation. Fawad Butt, co-founder and CEO of Penguin Ai, argued that complete clinical information at the first submission could keep more patients from abandoning the process.

The clinical issue is not whether technology can speed a workflow. It is whether faster processing preserves human review where medical judgment is needed.

What practices can change now

Most structural reform remains outside the practice’s control, but the survey points to practical steps that can reduce patient drop-off.

First, submission quality matters. If a large share of denials reflects missing or incorrect paperwork rather than genuine disagreement over medical necessity, the best defense is a complete first submission. Administrative precision can prevent the kind of delay that leads patients to stop pursuing care.

Second, communication needs to be more active. Patients who know a request is pending, understand the usual timeline, and have a clear contact point are less likely to disengage. A brief follow-up call on day three or day five may sound modest, but for a patient waiting on a medication or procedure, it may be enough to keep the process moving.

Third, appeals deserve a more selective but less defeatist approach. A default assumption that appeals will fail is not consistent with the patient data here, where nearly half of appealed cases were overturned. Practices may not have capacity to appeal everything, but high-stakes treatment decisions warrant a closer look. Patients also need to know they can appeal on their own if the practice cannot take it on.

Prior authorization remains more than a billing nuisance. It is a filter that can reshape care trajectories, push patients into higher-acuity settings, and leave clinicians unaware that a recommended plan never reached completion.


Reporting by Dr. A.R. Athira, MBBS for Medicine World Foundation, with additional data provided by Medical Economics.