Prior Authorization Blues

Psychiatric TimesVol 40, Issue 3

"The rationale for these policies is on a need-to-know basis, and we clinicians evidently do not need to know."




In an email blast to the medical community in September 2022, Gerald E. Harmon, MD, immediate past president of the American Medical Association (AMA), shared a personal account of his difficulties in obtaining insurance approval for medications for a critically ill relative.

He characterized the prior authorization (PA) system as being “out of control” and informed the medical community that the AMA is taking vigorous action to try to address its excesses. The AMA has posted educational material online documenting the scope of the PA problem, the harm it is causing to patient care, and proposals for mitigating the problem.1

An Origin Story

What is the origin of PAs, what are they supposed to be accomplishing, and are they meeting their goals? It turns out to be quite difficult to answer these questions. From my perspective, PAs appeared to develop spontaneously through the actions of many insurance companies, clearly with the initial intent of saving money by limiting the use of expensive procedures and medications when there are cheaper alternatives.

However, if we examine our experience as practicing physicians, it is clear that PAs have evolved and are now being used for purposes other than saving money.

Because there appears to be no official guidance on what PAs are supposed to accomplish and which methods insurance companies are allowed to use, we will have to proceed based on what we practicing physicians encounter daily.

Examining the Rationale

Preventing unnecessary procedures. Overuse of expensive procedures and surgery is a legitimate concern, particularly considering that many physicians’ income is determined on a fee-for-service basis. An externally imposed process of documenting the justification for procedures may help reduce their excessive use. Transcranial magnetic stimulation, ketamine treatment, and electroconvulsive therapy probably fall into this category, but otherwise it is not of much relevance to psychiatric practice.

Prescribing expensive brand-name drugs when generic choices are available. This issue is the main concern about PAs for our profession. On first impression, one would think this would be quite straightforward. When we prescribe an expensive medication, we are generally asked to document the generic drugs that have been tried and “failed”—although the term failed is not defined.

If a blood pressure medicine does not reduce a patient’s blood pressure to the acceptable range, we can all agree it has failed and that switching to another or adding a second medication may be necessary.

As psychopharmacologists, we see plenty of situations in which a medication has little or no benefit and/or unacceptable adverse effects, in which case we can be comfortable reporting the drug as failed. More commonly, the patient has experienced substantial benefit from the medication, and there may be some tolerable adverse effects, but we are seeking another option to obtain a superior result. Yet, in my experience, we are required to assert that all previous medications have failed.

As are usually submitted via online forms, and a few insurers keep the process simple. After we note the required number of failed medications, they usually approve the request. Most forms, however, ask us to specify the dose, dates of treatment, and reason for discontinuation for each failed medication. I usually skip that section and scroll to the bottom where, with luck, there is a blank field for “comments.”

I dictate a 1-paragraph summary of the case listing all previous medication trials that I know of and stating that all of them failed “due to insufficient efficacy and/or unacceptable adverse effects.” This summary usually does the trick, and the PA is approved—except when it is not.

Even if I did fill out the detailed questions about each medication, what then? Is somebody in an office 100 miles away going to read what I wrote, stroke his beard, and say, “Well, that venlafaxine trial 8 years ago was only for a week. That is not long enough to determine whether it failed. Denied!”?

And if I explain that my patient refused a trial of a particular medication because, for instance, his cousin’s wife took that drug and had a bad reaction, how does that affect the decision? Does the lack of that trial justify denying the patient appropriate treatment with another drug?

The simplistic choice of whether a medication can be judged to have failed based on undefined criteria has no relevance to clinical decision-making. The PA process is set up to look like due diligence, making a sober clinical decision based on information, but it is a sham. There is an art to working with individuals; as clinicians, we need flexibility to find the optimal regimen for each patient.

The consequence of a denial is often that the patient must accept an inferior result even if a better one is available. Occasionally, a letter will piously inform us that the denial does not prevent the patient from using the medication if desired because they are free to pay cash—but the cash price is often absurd and unaffordable for most patients. Even when the request is approved, the process causes delays that can cause discomfort or even harm to patients.

We are forced to go through a time-consuming and essentially meaningless ritual in our attempts to provide our patients with optimal care, and the decision is ultimately made in an arbitrary fashion. But there are other aspects of this process that make it even more senseless and onerous.

More Folly Associated With PAs

Standards and unpredictable changes. Each insurance company has its own list of medications and its own criteria, all of which may change unpredictably. There is often no discernible rationale for specific policies.

Example: For a few years, most of the insurers in my area required failure on venlafaxine before they would approve any other antidepressant, even other generics. But during the same period, 1 company absolutely refused to pay for venlafaxine under any circumstances. The rationale for these policies is on a need-to-know basis, and we clinicians evidently do not need to know.

Unstable insurance coverage. Patients frequently change insurance—their employer switched insurers, they got a new job, they realized they cannot afford their deductible, and so on. Also, there seems to be a constant churning of coverage, with companies buying out one another or switching to a different prescription management company.

Annual PA renewals. Most psychiatric disorders are chronic, yet we must resubmit the PA every year. Consider a patient with schizophrenia on an atypical antipsychotic. If the patient had schizophrenia last year, odds are they will have schizophrenia next year. But each year I need to renew the PA, filling in the same information and answering the same questions. Usually they approve it—except when they do not.

Miscellaneous craziness. Years ago, an insurance company demanded that we submit an adverse reaction report to the US Food and Drug Administration (FDA) about a failed medication before they would consider approving the newly requested drug. This demand disappeared quickly. The policy was so absurd that some authority must have stepped in.

This incident shows that, at least some of the time, these policies are designed by individuals who have no understanding of what is involved in prescribing medications, yet they have the power to intervene in clinical decisions.

Wasted time and money. There have been attempts to quantify the waste of time and money, both for physicians and other staff involved in this activity.2,3 It is hard to know the accuracy of the figures, but an informal calculation can at least give us a sense of the problem’s scope. Let us suppose that each PA event requires 20 minutes of a doctor’s time. So, for every 100 PAs, more than 33 hours are spent on bureaucracy as opposed to patient care.

Dangerous drug-drug interactions. This is an explicit goal presented by some insurance companies.4 These companies portray PAs as an attempt to check that prescribers are hewing to the standard of practice. A few companies have added wordage implying that it saves money by preventing adverse effects.

Here is my favorite example of this nonsense. A patient aged 48 years has been in my care for 15 years. Although I am treating him for anxiety and depression, his chief complaint is multiple chemical sensitivity. In fact, he does not consider himself a psychiatric patient. He takes diazepam 2 mg once or twice a month for moments of anxiety and has been doing so for at least 20 years. Suddenly, I was required to submit a PA request.

This was so odd that I called the insurer rather than sending in a form. Long story short: The prescription was approved as soon as I responded in the affirmative to the question, “In this case, does the benefit of this medication outweigh the risk?”

What is wrong with this picture? First, every medication has risks, and it is part of our job to weigh the risks against the benefits and decide what to prescribe. I would bet any amount of money that in the thousands—or millions—of times this question has been asked during a PA process, no doctor has ever stated that the risk of the requested medicine outweighed the potential benefit. So, the question becomes: How does going through this ritual prevent adverse effects?

The next question is: Why are they focusing on this particular medication? It is true that we are all being bombarded with new warnings about the potential of benzodiazepines for addiction. More recently, there has been alarm due to research purporting to show that benzodiazepine use is a risk factor for dementia.5,6

It appears that the insurance companies think they are supporting the standard of practice with their clumsy oversight. But benzodiazepines are relatively safe and highly useful for a variety of diagnoses,5,6 and it is our job, as the prescribing doctors, to weigh the risks and benefits and make decisions.

We can debate the wisdom of using or not using these medications among ourselves, but insurance companies should not be inserting themselves into this process—imposing their interpretation of the standard of practice through the power to deny authorization. Aside from the inappropriateness of their intervention, their methodology is completely inept and is not contributing anything useful.

Another brief example: A patient aged 73 years with schizophrenia has been prescribed 3 antipsychotics to provide sufficient antipsychotic response while minimizing the resulting extrapyramidal effects. She is on a modest dose of benztropine, which she has been taking since her 20s. Of course, I had to do a PA, in which I was solemnly informed about the anticholinergic effects of benztropine and the potential for impaired memory and other adverse effects. Then I was asked the magic question: “In this case, does the benefit outweigh the risk?”

When I was in medical training, we had haloperidol, chlorpromazine, and their cousins; atypical antipsychotics came a decade later. Almost all patients with chronic psychosis had extrapyramidal symptoms, and almost all were on benztropine or other appropriate medications. We struggled with the anticholinergic effects all the time. I really do not need an insurance company checking on how I practice medicine.

Off-label Prescribing

Off-label prescribing—or the prescription of a pharmaceutical drug for an indication, dosage, age group, or administration route not currently approved by the FDA—is a legal and common practice. The FDA has made the meaning of a formal indication quite clear: There is rigorous evidence that a particular medication is safe to use for a particular diagnosis in a particular population.

Because their position has been widely misunderstood, they have stated explicitly that it does not imply that the drug is unsafe to use in nonapproved conditions or populations: “From the FDA perspective, once the FDA approves a drug, health care providers generally may prescribe the drug for an unapproved use when they judge that it is medically appropriate for their patient.”7

Off-label prescribing is extremely common:

  • About 20% of prescriptions in the United States are for off-label uses.8
  • About 30% of psychiatric prescriptions are off-label.8
  • More than 50% of pediatric office visits involve off-label prescriptions.8 Because investigational studies are not usually completed with children, the FDA cannot approve them formally.

Yet, the PA system for off-label prescriptions is nebulous. One of the companies with whom I participate states that it will approve off-label use for medications if it is recommended in at least 1 of 2 references the company considers valid, presumably as documentation that it meets the standard of care. Yet the FDA clearly states that treating physicians can use medications off-label at their discretion.

My informal observation is that insurance companies generally challenge off-label use of newer, more expensive drugs, but usually do not challenge off-label use of generic drugs. There is a pretense that the challenges are based on safety concerns, but it is clear that the off-label issue is just another excuse to make it difficult to prescribe more expensive medications.

What Can Be Done?

Mark Twain was fond of pointing out, “Everybody talks about the weather, but nobody does anything about it.” We certainly complain about PAs, but we should not view them like the weather—as an inescapable force of nature. When it rains, we can open an umbrella to protect ourselves, but it goes on raining. Similarly, we have extensive discussions about how to make PAs more efficient—keeping us drier, so to speak.

I am sure we would all be grateful if the PA process were standardized and made more efficient, but if the incidence of PAs continues to grow, more efficiency will be cold comfort indeed. An efficient waste of time is still a waste of time. Perhaps we should stop complaining about PAs and find a way to do something about them.

The PA phenomenon was not developed in a thoughtful way by an authoritative agency setting standards and rules that would govern the process. Instead, their development involved hundreds of companies and thousands of individuals devising requirements for numerous procedures and medications, all of which evolved over time. It should be no surprise that the system is chaotic and ineffective.

Therefore, the problem must be addressed through a political and legal process, the nature of which I can only dimly visualize. It must begin with individual doctors and the organizations that represent us speaking up and demanding to be heard. The AMA appears to be taking steps in the right direction.

Whatever structure is created to manage this problem, it must at least:

  1. Define the policies: what information may be required and the basis for decisions. These policies should be developed with extensive involvement of practicing physicians and the organizations that represent them.
  2. Create standardized forms that are efficient and reasonable.
  3. Collect data from insurance companies and other sources to determine objectively if the PAs are accomplishing their stated goals, and if they are causing harm to patients by interfering with medical care.
  4. Assert in a clear policy that insurance companies have no role in policing the standard of care and may not dictate how practicing physicians should be prescribing.

Attempting to reform and rationalize the PA system would be much more doable if we had a single-payer system, so there would be only 1 set of rules. In my opinion, a single-payer system would be horrible—but about a thousand times better than what we have now. Nevertheless, any hope of improving our current, dysfunctional system will require those of us who work in the trenches of clinical medicine to speak up.

Dr Goderez is a psychopharmacologist and integrative medicine practitioner in private practice.


1. AMA is tackling the overused prior authorization process. American Medical Association. Accessed January 31, 2023.

2. Casalino LP, Nicholson S, Gans DN, et al. What does it cost physician practices to interact with health insurance plans? Health Aff (Millwood). 2009;28(4):w533-w543.

3. Morley CP, Badolato DJ, Hickner J, Epling JW. The impact of prior authorization requirements on primary care physicians’ offices: report of two parallel network studiesJ Am Board Fam Med. 2013;26(1):93-95.

4. Reinke T. Prior authorization for safety, not just for economy. Manag Care. 2012;21(12):33-35.

5. Osler M, Jørgensen MB. Associations of benzodiazepines, Z-drugs, and other anxiolytics with subsequent dementia in patients with affective disorders: a nationwide cohort and nested case-control study. Am J Psychiatry. 2020;177(6):497-505.

6. Salzman C. Do benzodiazepines cause Alzheimer’s disease? Am J Psychiatry. 2020;177(6):476-478.

7. Understanding unapproved use of approved drugs “off label.” FDA. February 5, 2018. Accessed January 31, 2023.

8. Syed SA, Dixson BA, Constantino E, Regan J. The law and practice of off-label prescribing and physician promotion. J Am Acad Psychiatry Law. 2022;49(1):53-59.

Recent Videos
Dune Part 2
© 2024 MJH Life Sciences

All rights reserved.