Algorithms

[psychapp121]

What are people thoughts on algorithms particularly during residency when people are starting to get a grasp of how to approach a patient? Or even as attendings? I saw this Harvard algorithm recently that I hadn’t heard about, is this thing legit? There’s also the Texas algorithm, are there any go to algorithms or is this Harvard one a reasonable one? Thx for the insight.

harvard algorithm: Psychopharm Algorithms

---

Members don't see this ad.

 

[TexasPhysician]

Probably valuable the first 2 weeks of intern year.

 

[psychapp121]

Only the first two weeks? Isn’t this evidence based and should be followed by all clinicians?

 

[downwithDTB]

What are people thoughts on algorithms particularly during residency when people are starting to get a grasp of how to approach a patient? Or even as attendings? I saw this Harvard algorithm recently that I hadn’t heard about, is this thing legit? There’s also the Texas algorithm, are there any go to algorithms or is this Harvard one a reasonable one? Thx for the insight.

harvard algorithm: Psychopharm Algorithms

These particular ones are highly opinion driven and, I think, miss a lot of very reasonable, non-controversial options.

Good psychopharm balances long and short term efficacy, side effects and patient preference and if you limit yourself to the above you're missing a lot and sometimes even making questionable, non-evidence based decisions.

I have the biggest qualms with the schizophrenia algorithm. 1) the minimal RTC data that exists does not support there being any benefit to trying a second equally effective antipsychotic for inadequate response 2) choosing efficacy over side effects should be a collaborative decision (i.e. zyprexa) 3) LAI availability is extremely important 4) I'm not convinced that FGAs shouldn't be offered first line as a choice between side effect profiles given comparative efficacy and wide lai availability.

 

[clausewitz2]

These particular ones are highly opinion driven and, I think, miss a lot of very reasonable, non-controversial options.

Good psychopharm balances long and short term efficacy, side effects and patient preference and if you limit yourself to the above you're missing a lot and sometimes even making questionable, non-evidence based decisions.

I have the biggest qualms with the schizophrenia algorithm. 1) the minimal RTC data that exists does not support there being any benefit to trying a second equally effective antipsychotic for inadequate response 2) choosing efficacy over side effects should be a collaborative decision (i.e. zyprexa) 3) LAI availability is extremely important 4) I'm not convinced that FGAs shouldn't be offered first line as a choice between side effect profiles given comparative efficacy and wide lai availability.

All I would say is that realistically you are switching neuroleptics that are not clozaril to possibly get some effect because you are trying to figure out which particular balance of side effects are acceptable to the patient and is consistent with them taking it. Unreported or denied nonadherence is very high in this group.

 

[psychapp121]

These particular ones are highly opinion driven and, I think, miss a lot of very reasonable, non-controversial options.

Good psychopharm balances long and short term efficacy, side effects and patient preference and if you limit yourself to the above you're missing a lot and sometimes even making questionable, non-evidence based decisions.

I have the biggest qualms with the schizophrenia algorithm. 1) the minimal RTC data that exists does not support there being any benefit to trying a second equally effective antipsychotic for inadequate response 2) choosing efficacy over side effects should be a collaborative decision (i.e. zyprexa) 3) LAI availability is extremely important 4) I'm not convinced that FGAs shouldn't be offered first line as a choice between side effect profiles given comparative efficacy and wide lai availability.

I don’t understand your first point, are you saying that after you try one neuroleptic and it fails you should jump straight to clozapine?

 

[downwithDTB]

I don’t understand your first point, are you saying that after you try one neuroleptic and it fails you should jump straight to clozapine?

There is an argument to be made for that. The only attempt to empirically answer whether switching makes a difference using RTC was optimise which found the answer to be "no." There are several schools of thought about treatment failures, one being that if someone gains too much weight and uses that as an excuse to stop SGA, then you try a FGA or more weight neutral SGA. The other school says that many people keep their antipsychotics when they are working well enough in spite of side effects of, so you should switch to a better antipsychotic (i.e. maybe zyprexa or amisulpride, LAI, otherwise, clozapine). In my reading of the literature both approaches may be completely appropriate and I favor the later; this algorithm ignores a completely reasonable, literature based (but not with great external validity to antipsychotics beyond amisulpride/zyprexa) approach in favor of an experience based approach. Whatever. They also ignore the fair evidence for using abilify augmentation and my person favorite questionably supported combo of luvox and clozapine (can't blame them there).

I ask you, if you had schizophrenia, would you take clozapine first? I would.

 

[splik]

I ask you, if you had schizophrenia, would you take clozapine first? I would.

Hell no, and you would be the exception. When surveyed, most psychiatrists say they would want the more benign antipsychotics first.

If I had a psychotic illness, I would first try to avoid taking any medications and privilege psychotherapeutic intervention, possibly with nutritional interventions. If meds were necessary, I would first want benzos. If I absolutely had to take a neuroleptic, I would probably want to start with something like Abilify. If the big guns were needed, I would go for Zyprexa after the above had proved insufficient.

 

[splik]

What are people thoughts on algorithms particularly during residency when people are starting to get a grasp of how to approach a patient? Or even as attendings? I saw this Harvard algorithm recently that I hadn’t heard about, is this thing legit? There’s also the Texas algorithm, are there any go to algorithms or is this Harvard one a reasonable one? Thx for the insight.

harvard algorithm: Psychopharm Algorithms

The Texas Medication Algorithm Project was corrupt from the beginning (unsurprisingly), with significant drug company funding. It was influential in other states for medicaid patients back in the day too. But it is flawed and not considered best practice these days. The Harvard/Southshore algorithms never caught on. In general guidelines in psychiatry have never been influential in the US for various political reasons. The APA (which produced and is again producing practice guidelines) did not want to upset membership by being too narrow or evidence-based in recommendations or creating a standard of care which would be used as a stick by malpractice attorneys to beat membership with and so the guidelines are pretty vague and useless and the old ones were (unsurprisingly) medication-heavy.

If you are looking for guidelines to inform practice, you need to look overseas. A good place to start is the British Maudsley Prescribing Guidelines. You can also look up the canadian CANMAT guidelines (for bipolar, depression, anxiety), the British Association of Psychopharmacology Guidelines, the british NICE guielines and the world federation of societies for biological psychiatry guidelines. Obviously caveats apply as dosing of some drugs is different, some meds not available overseas and some not available in the US etc.

Psychiatry does not lend itself to algorithms as well as some other specialties (e.g. EM, medicine). That means there is a wide range of different practices including poor care, but it also reflects the nuance of psychiatry and the need for personalized care.

 

[psychapp121]

Hell no, and you would be the exception. When surveyed, most psychiatrists say they would want the more benign antipsychotics first.

If I had a psychotic illness, I would first try to avoid taking any medications and privilege psychotherapeutic intervention, possibly with nutritional interventions. If meds were necessary, I would first want benzos. If I absolutely had to take a neuroleptic, I would probably want to start with something like Abilify. If the big guns were needed, I would go for Zyprexa after the above had proved insufficient.

benzos for psychosis..?

 

[psychapp121]

The Texas Medication Algorithm Project was corrupt from the beginning (unsurprisingly), with significant drug company funding. It was influential in other states for medicaid patients back in the day too. But it is flawed and not considered best practice these days. The Harvard/Southshore algorithms never caught on. In general guidelines in psychiatry have never been influential in the US for various political reasons. The APA (which produced and is again producing practice guidelines) did not want to upset membership by being too narrow or evidence-based in recommendations or creating a standard of care which would be used as a stick by malpractice attorneys to beat membership with and so the guidelines are pretty vague and useless and the old ones were (unsurprisingly) medication-heavy.

If you are looking for guidelines to inform practice, you need to look overseas. A good place to start is the British Maudsley Prescribing Guidelines. You can also look up the canadian CANMAT guidelines (for bipolar, depression, anxiety), the British Association of Psychopharmacology Guidelines, the british NICE guielines and the world federation of societies for biological psychiatry guidelines. Obviously caveats apply as dosing of some drugs is different, some meds not available overseas and some not available in the US etc.

Psychiatry does not lend itself to algorithms as well as some other specialties (e.g. EM, medicine). That means there is a wide range of different practices including poor care, but it also reflects the nuance of psychiatry and the need for personalized care.

very helpful thank you

 

[TexasPhysician]

Only the first two weeks? Isn’t this evidence based and should be followed by all clinicians?

No. There is some evidence for general care in these algorithms, but algorithms don’t work well in the real world. Patients are unique with diverse health issues, past medication trials, and preferences.

If these algorithms were that great, anyone could be a psychiatrist. They aren’t. Memorize them the first 2 weeks of intern year and then start developing a good understanding of the complexities of psychiatry.

 

[downwithDTB]

Hell no, and you would be the exception. When surveyed, most psychiatrists say they would want the more benign antipsychotics first.

If I had a psychotic illness, I would first try to avoid taking any medications and privilege psychotherapeutic intervention, possibly with nutritional interventions. If meds were necessary, I would first want benzos.

Schizophrenia seems really bad to me. There must be a reason that the Catie folk were more likely-ish to stay on zyprexa. I don't know that I have the financial reserves to have more than one psychotic episode, so I'd happily do pretty much anything to reduce the likelihood of a second. Gotta value paying the bills for my dependants above all else. Have you ever seen someone recover from schizophrenia enough to work any job in a timely fashion on psychotherapy alone who didn't have near unlimited financial resources?

 

[bashir]

This is a really interesting thought experiment. I think if I were psychotic I would want zyprexa first (alongside psychotherapy). If I start to blow up like a balloon we'll cross that bridge when we come to it but I'll be in a better position to weigh the risks and benefits after psychosis is rapidly well-controlled.

I might seriously consider hiring a personal chef and personal trainer at the same time. Obviously not an option for most people but in a perfect world it would be.

 

[splik]

Have you ever seen someone recover from schizophrenia enough to work any job in a timely fashion on psychotherapy alone who didn't have near unlimited financial resources?

YES. That said, because the system is broken and does not value paying for non-drug interventions for people experience emotional crises, you do have to be of means in this country to access psychotherapeutic services for psychosis as the primary/sole intervention. This is not true everywhere. There is a rich literature on psychological treatments for psychosis dating back to people like frieda fromm-reichmann, harry stack sullivan, carl rogers, RD Laing, Jay Haley and others. Not all patients can benefit from more intensive psychotherapy, and most therapists also cannot tolerate being confronted with madness. "Schizophrenia" is not a unitary phenomena, and I would say that medications and not therapy are going to be essential in some cases. But there are also many cases where medications (including clozapine) are totally ineffective.

Many people are surprised to learn that Aaron Beck wrote his first paper on cognitive therapy for schizophrenia in 1952. He wrote a piece in Psych Services last month about how his work started with psychosis and now, almost 50 years later he has returned to this field.

 

[Stagg737]

Only the first two weeks? Isn’t this evidence based and should be followed by all clinicians?

Cookbook medicine is generally garbage outside of a few areas. This is especially true for psychiatry. If you've got a patient with a simple diagnosis and minimal comorbities whose psychotropic naive, then sure, you can start with an algorithm. Things rarely fit into nice little boxes though, so algorithms should only really be used as skeletons/jumping off points, imo.

I don’t understand your first point, are you saying that after you try one neuroleptic and it fails you should jump straight to clozapine?

There are actually some studies which have suggested Clozapine should be a first line medication for psychosis. Look up CUtLASS 2 study and Finn 11 trial. Both suggest that Clozapine is both more efficacious and safer overall than every other antipsychotic it was tested against (Olanzapine, Risperdal, Seroquel, Haldol, and others).

benzos for psychosis..?

I could see this for manic psychosis with the goal being to get the patient to sleep with hopes that this would effectively treat the psychosis. I wouldn't solely rely on a benzo to treat mania, nor would I personally use it as a primary med in pure psychosis, but I can see where it could be utilized (other than agitation).

 

[Stagg737]

This is a really interesting thought experiment. I think if I were psychotic I would want zyprexa first (alongside psychotherapy). If I start to blow up like a balloon we'll cross that bridge when we come to it but I'll be in a better position to weigh the risks and benefits after psychosis is rapidly well-controlled.

I might seriously consider hiring a personal chef and personal trainer at the same time. Obviously not an option for most people but in a perfect world it would be.

It really is. I was about to say that I guess I'm another exception who would go straight to Clozapine (especially if it were really severe psychosis), but on second thought I think I would start with Zyprexa first as well and then go to Clozapine as a back-up if that didn't work.

YES. That said, because the system is broken and does not value paying for non-drug interventions for people experience emotional crises, you do have to be of means in this country to access psychotherapeutic services for psychosis as the primary/sole intervention. This is not true everywhere. There is a rich literature on psychological treatments for psychosis dating back to people like frieda fromm-reichmann, harry stack sullivan, carl rogers, RD Laing, Jay Haley and others. Not all patients can benefit from more intensive psychotherapy, and most therapists also cannot tolerate being confronted with madness. "Schizophrenia" is not a unitary phenomena, and I would say that medications and not therapy are going to be essential in some cases. But there are also many cases where medications (including clozapine) are totally ineffective.

Many people are surprised to learn that Aaron Beck wrote his first paper on cognitive therapy for schizophrenia in 1952. He wrote a piece in Psych Services last month about how his work started with psychosis and now, almost 50 years later he has returned to this field.

So from what I've been taught psychotherapy, especially insight-oriented therapies, are generally counter-productive to treating the truly psychotic patients. Where's a good place to start in terms of therapy for psychosis? At what point does it become appropriate and what modalities are actually effective?

 

[Mass Effect]

this algorithm ignores a completely reasonable, literature based (but not with great external validity to antipsychotics beyond amisulpride/zyprexa) approach in favor of an experience based approach

An experience-based approach is the way to practice, in my book. Obviously, you want to build a foundation on the evidence and not veer completely out of left field, but docs who rely solely on the evidence with no flexibility for experience tend to be the ones with very ill patients who never get better. An NP can follow the evidence (and some do, believe it or not). The point of being a doctor is to use your vast knowledge of the human body (and in this case, neuroscience/neurobiology) to build upon the evidence.

I ask you, if you had schizophrenia, would you take clozapine first? I would.

No way in hell. I'm with @splik on this one. I would want clozapine to be a last resort.

 

[splik]

So from what I've been taught psychotherapy, especially insight-oriented therapies, are generally counter-productive to treating the truly psychotic patients. Where's a good place to start in terms of therapy for psychosis? At what point does it become appropriate and what modalities are actually effective?

The idea that expressive therapies could worsen psychosis comes from the Chestnut Lodge Studies. Even those studies showed that dynamically oriented supportive psychotherapy could be helpful. The studies were unfortunately misinterpreted and weaponized by the biological psychiatrists (psychiatry was much more polarized in those days) and led to a generation of patients not being offered any psychological treatment at all. And the literature on CBT for psychosis and those ultrahigh risk shows that the supportive therapy intervention is pretty good (even though it was meant to be a control/placebo) and in some circumstances can outperform CBT. There is a strong literature looking at family work for schizophrenia, based on the data that families with high expressed emotion have higher relapse rates. Recently, there has been an interested in mindfulness based interventions, ACT, and compassion-focused therapy for psychosis.

Psychotherapy is an operator dependent procedure. Few therapists have the stomach or skill to do deep psychological work with patients experiencing extreme breaks from reality. It's really hard. But possibly one of the few areas where psychiatrists could advocate to be uniquely placed to provide psychotherapy. Unfortunately most psychiatrists are uninterested or incapable or providing such treatment.

In my experience, intelligent college students and grad students with family pathology and difficulty navigating the transition to adulthood are examples of the kind of patients who do well with a more psychotherapeutically-oriented approach (or at least having this incorporated w/ meds).

 

[bashir]

The idea that expressive therapies could worsen psychosis comes from the Chestnut Lodge Studies. Even those studies showed that dynamically oriented supportive psychotherapy could be helpful. The studies were unfortunately misinterpreted and weaponized by the biological psychiatrists (psychiatry was much more polarized in those days) and led to a generation of patients not being offered any psychological treatment at all.

Thank you for this historical background, Splik. I never understood how psychological treatment came to be so underutilized for psychosis. I can see how this would have happened now. It makes me really mad and there's no excuse for our field not doing a better job of advocating for all effective treatment modalities to be made widely available for our sickest patients.

I consider myself to be doing psychotherapy with almost all of my patients with psychosis, and when I am not quite sure exactly what I'm doing I take great solace in the evidence for supportive therapy and even "befriending."

 

[clausewitz2]

The idea that expressive therapies could worsen psychosis comes from the Chestnut Lodge Studies. Even those studies showed that dynamically oriented supportive psychotherapy could be helpful. The studies were unfortunately misinterpreted and weaponized by the biological psychiatrists (psychiatry was much more polarized in those days) and led to a generation of patients not being offered any psychological treatment at all. And the literature on CBT for psychosis and those ultrahigh risk shows that the supportive therapy intervention is pretty good (even though it was meant to be a control/placebo) and in some circumstances can outperform CBT. There is a strong literature looking at family work for schizophrenia, based on the data that families with high expressed emotion have higher relapse rates. Recently, there has been an interested in mindfulness based interventions, ACT, and compassion-focused therapy for psychosis.

Psychotherapy is an operator dependent procedure. Few therapists have the stomach or skill to do deep psychological work with patients experiencing extreme breaks from reality. It's really hard. But possibly one of the few areas where psychiatrists could advocate to be uniquely placed to provide psychotherapy. Unfortunately most psychiatrists are uninterested or incapable or providing such treatment.

In my experience, intelligent college students and grad students with family pathology and difficulty navigating the transition to adulthood are examples of the kind of patients who do well with a more psychotherapeutically-oriented approach (or at least having this incorporated w/ meds).

Psychotheraputic interventions probably have the broadest scope and most promise early on, in the first episode space, but they also can certainly be very effective even in more chronic folks. I did a lot of ACT for psychosis work and punnily enough it was critical in helping someone transition off of an ACT team. Still in supportive housing but that is still pretty big. I am not sure any particular psychosis specific therapy for individuals has a robust evidence base v. a supportive approach alone but I think they are helpful as frameworks for motivating the therapists and preventing therapeutic nihilism if nothing else. Family focused and cognitive rehab approaches are well supported, criminally underused, and hard to get paid for in many places.

Re: what neuroleptic first, I am actually going to go out on a limb and say I would want something like molindone first, pretty pure dopaminergic blockade without the weight gain. After that I would accept Zyprexa, but would want to try a couple of things before moving on to clozapine, like amoxapine or minocycline since if two heavily anti-dopaminergic neuroleptics aren't getting the job done dopamine pathway dysfunction is probably not what is going on.

100% on preferring some Ativan and psychotherapy, though, and especially in a UHR/CHR population this is the way to go.

 

[Jules A]

I never thought of this before and at my age it would be dementia or delirium but if I was 25 years younger and presented with psychosis my initial thought would be: time is of the essence. I'd start with rapid titration of Abilify + small dose Klonopin qhs followed by Zyprexa and if not efficacious Clozaril. The benzo would be to take the edge off of the fact that I was psychotic which I would anticipate having me nearly hysterical. And at that point the size of my arse and metabolic implications would have to take a backseat to the fact that I was experiencing perceptual disturbances.

 

[downwithDTB]

An experience-based approach is the way to practice, in my book. Obviously, you want to build a foundation on the evidence and not veer completely out of left field, but docs who rely solely on the evidence with no flexibility for experience tend to be the ones with very ill patients who never get better. An NP can follow the evidence (and some do, believe it or not). The point of being a doctor is to use your vast knowledge of the human body (and in this case, neuroscience/neurobiology) to build upon the evidence.

I don't entirely disagree with you. My point wasn't that the algorithm should say fail one antipsychotic, start clozapine. My point was that 1) these algorithms aren't even always based on evidence and 2) if you stick to them, you're giving up some things that make a lot of sense, work by experience and have some science behind them. I don't think there is a right answer to "when" clozapine (except maybe s/p ICU suicide attempt or after making it to state or NGRI facility, but I don't see those in the algorithm anyways), but I certainly disagree that first or second isn't always wrong. Some folks do very well on early clozapine.

 

[FlowRate]

The problem I have with the Harvard (South Shore) algorithms is that, unless I'm missing something, they don't directly cite each recommended step with any literature.

some not available in the US

Complete aside, but wondering if you have a take on agomelatine? I don't have much knowledge of it other than finding it in Maudsley and then also finding out it's not available in the US.

 

[downwithDTB]

The problem I have with the Harvard (South Shore) algorithms is that, unless I'm missing something, they don't directly cite each recommended step with any literature.

You have to click on the box on the page for their citations. They also publish full articles on them in Harvard rev psych, but they are pretty paywalled...

 

[FlowRate]

You have to click on the box on the page for their citations. They also publish full articles on them in Harvard rev psych, but they are pretty paywalled...

Ah it was the specific one I was looking at (ADHD) that wasn't working.

 

[whopper]

These algorithms can you a foundation but by the time they're published new meds have likely already come out making them outdated. Further any algorithm is not sophisticated enough for a skilled psychiatrist.

Then on the other than you got idiot psychiatrists prescribe pretty much the same antipsychotic to every patient they see. I'd rather someone use a published algorithm if the extent of their own internal algorithm is to give everyone the same meds, but if that's the case this physician is terrible. Unfortunately these terrible psychiatrists aren't a 1% phenomenon but more like at least 1/3 of them.