Tuesday, September 1, 2026

The Deprescribing Hype...

 


One of the best psychiatrists I know ran a large antipsychotic long-acting injectable (LAI) clinic and the largest clozapine clinic in the state. There was no a side effect he had not seen or addressed related to medications. He also happened to be one of the most well-liked clinicians I have ever known.  Years after he left our clinic his former patients would approach me in the hallways and ask me if I knew how he was doing.

Since I was in acute care, I would occasionally see one of his patients on my inpatient service and call him about that. One day we had a conversation about polypharmacy and all of the guidelines at the time.  Quality assurance guidelines said that monotherapy in any general class of antipsychotics, anxiolytics or mood stabilizers was preferred at the lowest effective dose. We both recalled many patients where that was not true and even slight changes in their medications resulted in significant destabilization.  In some cases the destabilization was so severe it resulted in long term care in a state hospital. 

Both of us had the privilege of practicing in a time when the only atypical antipsychotics were clozapine and later risperidone and olanzapine. But we also saw a lot of people who had been stabilized on first generation antipsychotics that were used for treatment for bipolar disorder, psychotic depression, and schizophrenia.  In those days it was very appealing to consider starting lithium and stopping the antipsychotic, stopping the antipsychotic due to a movement disorder (tardive syndrome or akathisia), stopping the antipsychotic trying to get to antidepressant monotherapy, stopping the antipsychotic to get to antipsychotic monotherapy, stopping the antipsychotic due to more typical side effects, or stopping the antipsychotic just on the general thought that they were bad medicines that should be taken only in desperate situations.

We reviewed a number of clinical scenarios involving polypharmacy.  He told me that some of his patients were taking an embarrassing list of medication.  He used the term because the medication list flew in the face of conventional wisdom at the time about how medications should be prescribed.  He had tried to alter the medications to approach those idealized guidelines but the patients did not do well and he invariably had to go back to the original combination.  Most of these patients had come out of state hospitals where they were observed for months or in some cases years on this medication.  He followed everyone closely, knew their family and social situations, and whether he would admit it or not did supportive psychotherapy. He saw people on a weekly basis during these medication changes.

I shared my same experience with him.  I had been trained by the authors of The Lithium Encyclopedia.  They also ran the Lithium Information Center at the time. My training was that people with bipolar disorder should be treated with lithium. I was seeing people with clear cut bipolar disorder who were stabilized on chlorpromazine – the oldest antipsychotic medication. I knew how to start lithium and gradually taper chlorpromazine so why not do it?  The person I am thinking of has dry mouth as the most significant side effect of chlorpromazine.  When the lithium level was therapeutic for a few weeks, I started the slowly taper what was considered a low dose of chlorpromazine.  When the uneventful taper was over it took just 2 weeks for the patient to experience a manic episode that required hospitalization and restabilization on chlorpromazine.  

And so, the stories went.  Apparently successful transitions with weeks to months off of a particular medication, nothing that would be considered a discontinuation or withdrawal effect, and then relapse to the primary disorder.  My personal experience was meager compared with what I witnessed as an inpatient doc.  I saw the worst possible outcomes of these medication discontinuations suicide attempts, homicides, high speed chases and other confrontations with the police including shoot outs, inadequate self care including stopping life saving medications like insulin, and destabilization of living situations and homelessness. I even had to put up with blowback from people I had stabilized. I was criticized by one of my bosses for an aggressive patient on the inpatient unit who was there because an outpatient provider decided to change the medication, I had just stabilized the patient on because of concerns about weight gain.  

To be clear, there are many reasons to stop medications or change them.  I am an expert at that.  There is also no assurance that a specific dose of medication will continue to work well forever.  Environmental changes like sleep deprivation can trigger relapse in people who are on effective maintenance medications.  All of the usual lifestyle factors need to be addressed. But it is a clear mistake to think that lifestyle factors or psychological interventions alone are enough to keep people stable in those who need them.  Risk stratification is also important.  I have a review coming out that I co-wrote with Ron Pies, MD coming out next month.  One of the  studies we reviewed looked at all of the antidepressant prescriptions in the country.  Only 3% of that population was ever hospitalized.  It is a good idea to include that risk in any risk/benefit decision on either side of the prescribing equation.    

Political approaches to complex medical problems never work.  The idea that a government official with no knowledge of the field can start a massive deprescribing program or proclaim that he knows the cause of autism is hopefully the absurd conclusion that medications rather than mental illness is the main problem.  At a more subtle level, deprescribing is has infiltrated the popular culture as a magical solution.  It is widely promoted but has no more legitimacy than the old term that I have used for decades – discontinuation.  Deprescribing carries with it the implicit notion that medications, especially psychiatric medications are either harmful at best or lethal at worst. None of that is consistent with the literature or the experience of clinical psychiatrists.

As I have written many times before nobody that I ever met inherently wants to take a medication every day.  That includes myself. That decision requires a careful risk/benefit decision. Rational thinking would lead to a decision to keep taking a medication with a favorable risk/benefit/analysis.  That decision is complicated by personal values, cognitive ability, and medication tolerance. It is further complicated by placebo effects, nocebo effects, and habituation to side effects.  As an experience clinical psychiatrist, I can say unequivocally that it is rare to find a person who needs a medication and cannot tolerate all of the medication in that class.  Notice I said rare, because I am sure those people exist.    

Whether you are a psychiatrist or a patient considering the current deprescribing hype either in social media or the professional literature keep this in mind.  These statements are made with a degree of certainty that does not exist.  Carefully consider the circumstances and err on the side of caution. I don’t anticipate that people advocating deprescribing at a global level will be measuring any outcomes. I hope that some of the experts in observational studies will be able to design outcome studies. The harms of medication discontinuation and deprescribing need to be documented before anyone can truly give informed consent.

 

George Dawson, MD, DFAPA