How much of psychiatry is simply following standardized protocols?

[ThatSerb]

I hope this question doesn't seem too naïve. I recently read through a conversation that discussed how much of practicing medicine is actual complex medical decision making, as opposed to more or less adhering to pre-established and standardized protocols. The consensus seemed to be that the actual complex decision making (while really cool) is rare.

How true (if at all) do you guys find this to be with regards to practicing psychiatry?

---

Members don't see this ad.

 

[Psychresy]

Just a resident but I find a lot more room to operate in pyschiatry than I did on the medicine wards for example.

 

[ThatSerb]

Just a resident but I find a lot more room to operate in pyschiatry than I did on the medicine wards for example.

hey thanks for the response. do you think you could maybe expand on that at all? how so?

 

[Stagg737]

Psych is a bit different. We don't have one med that is first line for most problems, we use classes. So different docs will have different preferences for first line med (Zoloft, Lexapro, and to a lesser extent Prozac are the ones I usually hear for depression). Even with that recommendation, I do sometimes start patients on Wellbutrin or Remeron first line if they've got the right constellation of symptoms. PTSD is another area that I've found the medication algorithms to be particularly useless for.

Once you're past second or third line it gets even less algorithmic. What do you do for the moderate to severely depressed patient in therapy who failed 2 SSRIs, an SNRI, and Wellbutrin? Mirtazapine? TCA? Abilify? Adderall? ECT? At a certain point the algorithms fail and it becomes clinical judgment. Even some of the better algorithms out there have dubious/contradicting evidence once they're past the first few steps. Look up the recommendations for insomnia treatment and you'll probably be surprised by what's actually recommended (and not recommended) if you're not already familiar with them.

Another thing to keep in mind is that diagnoses aren't always black and white like in other fields. There's a clear-cut definition for things like HTN and DM. Psych is a little more fluid. Is that bipolar patient with a UDS positive for cocaine truly bipolar or is it drug induced? Or is it BPD? Or PTSD? Or ADHD? I've prescribed Abilify as first line for major depression before just to make sure all the bases were covered because it wasn't clear if the patient had MDD or bipolar/psychotic issues. Have seen that patient a few times since and they're doing great, far better than I would have imagined. Would I have gotten the same result with an SSRI even thought it's technically first-line? Probably not.

Imo, psych is the field least suited to algorithms and cook-book medicine because there are so many nuances to many different aspects (diagnoses, meds, social/psychological/personality factors, etc).

 

[Psychresy]

hey thanks for the response. do you think you could maybe expand on that at all? how so?

For instance I may see someone with depression and comorbid anxiety disorder who has been treated by their PCP without good effect. I have a number of different paths I can take in their treatment. There are more branch points in the treatment algorithms. Whereas on the ward, someone comes in with a hypertensive urgency and we more or less know what we're going to do before we even see the patient.

This of course does not even take into account augmenting with therapy, what kind, how you will deliver it, etc. Things are very NOT black and white in psychiatry which lends itself for one to be more creative when treatment planning (though should still be practicing evidenced based medicine as much as possible).

 

[psychacad]

Medicine is not at all straight forward, so I disagree with the initial premise. Sure there are guidelines, but cases are rarely textbook and there is a lot of thought that can go through in everything from the physical exam, to diagnosis to treatment. It just depends who's doing it, and that's what separates good doctors from the rest.

And yes, pharmacotherapy is less textbook-y than other specialties but not for a good reason. Mostly because we don't have good data to go by, so there are a lot of urban legends (depending on the institution you work in) and people end up making up their theories why this and that works based on their 'experience' or what they read in Stahl..etc

Where it does get more interesting in psychiatry is the multifaceted formulation of what is going on, particularly in regards to character traits, dynamic factors..etc There is this sort of "depth" you don't see in other specialties and it is also very intersubjective and, hence, not really amenable to textbook approach.

 

[dl2dp2]

Heh. I wish people actually followed algorithms.

 

[Alemo]

Heh. I wish people actually followed algorithms.

This is the price to pay for psychiatry's freedom from cookbook medicine that generates burn out in so many other fields. It's easy to play fast and loose with the diagnostic categories–what's more, they can be difficult to understand and recognize (I say this as a resident). The course of illness and medication trials/treatment consequences of someone who is misdiagnosed may not be so dissimilar to the natural course of one who was correctly diagnosed from the outset. Such is the state of things now...

 

[birchswing]

This is the price to pay for psychiatry's freedom from cookbook medicine that generates burn out in so many other fields. It's easy to play fast and loose with the diagnostic categories–what's more, they can be difficult to understand and recognize (I say this as a resident). The course of illness and medication trials/treatment consequences of someone who is misdiagnosed may not be so dissimilar to the natural course of one who was correctly diagnosed from the outset. Such is the state of things now...

I agree with all that, and that's all in the framework of well informed, conscientious providers.

But it also seems given the current treatment options available in psychiatry, it's hard to directly kill a patient no matter what you choose, at least not quickly, which may lead to the permissibility and invisibility of some bad choices. Which is why, for example, not giving diabetes causing drugs for sleep or benzos first line for anxiety would be good parts of an algorithm to follow. And maybe deviating from that should require some more informed consent than it currently does.

 

[TexasPhysician]

Algorithm medicine is basic cases. A midlevel generally practices at this level. They learn the algorithms and follow them.

Most physicians will make medical decisions beyond algorithms on an hourly basis. I’d wager that whoever was quoted originally meant that complex medical decisions are events where you sit with multiple physician peers debating the diagnosis and treatment options. Truly complex cases in any field are more rare outside of large academic facilities (Mayo, Harvard).

 

[dl2dp2]

Algorithm medicine is basic cases. A midlevel generally practices at this level. They learn the algorithms and follow them.

What I find a lot is midlevels can't provide streamlined medication management. Their med lists are extremely redundant, as they just write whatever they think is right at the time, which then results in lamictal, klonopin and adderall, Seroquel, Buspar, etc. for everyone. This works for a while as patients become dependent, but eventually causes all kinds of unpleasant side effects.

People come to me and I give 2 diagnoses and remove meds that are not evidence-based for these diagnoses. This is what I mean by I wish people actually followed algorithms. This is also much easier done in a combined med/therapy context. I think while this sounds easy on paper it might actually be nontrivial in real life, because you do have to appreciate the evidence in a way that makes you confident enough to not be driven by the patient's symptomatic agenda.

PMDs are very good with knowing their limits. Typically PMD's problem is underdosing, but they rarely write meds they don't feel comfortable with or lots of benzos. I get referrals from PMDs with 25mg fluoxetine for 4 weeks and no effect. This is easier to correct and train. Midlevels try to practice independently and don't know when to refer.

 

[Stagg737]

Medicine is not at all straight forward, so I disagree with the initial premise. Sure there are guidelines, but cases are rarely textbook and there is a lot of thought that can go through in everything from the physical exam, to diagnosis to treatment. It just depends who's doing it, and that's what separates good doctors from the rest.

And yes, pharmacotherapy is less textbook-y than other specialties but not for a good reason. Mostly because we don't have good data to go by, so there are a lot of urban legends (depending on the institution you work in) and people end up making up their theories why this and that works based on their 'experience' or what they read in Stahl..etc

Where it does get more interesting in psychiatry is the multifaceted formulation of what is going on, particularly in regards to character traits, dynamic factors..etc There is this sort of "depth" you don't see in other specialties and it is also very intersubjective and, hence, not really amenable to textbook approach.

I agree with all of this and the bolded is part of what drew me to psychiatry. There's still so much research to be done and in terms of understanding of the field and disease processes, we're way behind many other areas of medicine. I find that potential very exciting, but it dioes unfortunately lead to the use of "anecdata" far more frequently than it should (and I admit I'm guilty of this as well). I also agree that social factors play a much larger role in our field than most others, which adds depth to the diagnoses and treatment plan which is not as relevant as many other fields.

What I find a lot is midlevels can't provide streamlined medication management. Their med lists are extremely redundant, as they just write whatever they think is right at the time, which then results in lamictal, klonopin and adderall, Seroquel, Buspar, etc. for everyone. This works for a while as patients become dependent, but eventually causes all kinds of unpleasant side effects.

People come to me and I give 2 diagnoses and remove meds that are not evidence-based for these diagnoses. This is what I mean by I wish people actually followed algorithms. This is also much easier done in a combined med/therapy context. I think while this sounds easy on paper it might actually be nontrivial in real life, because you do have to appreciate the evidence in a way that makes you confident enough to not be driven by the patient's symptomatic agenda.

PMDs are very good with knowing their limits. Typically PMD's problem is underdosing, but they rarely write meds they don't feel comfortable with or lots of benzos. I get referrals from PMDs with 25mg fluoxetine for 4 weeks and no effect. This is easier to correct and train. Midlevels try to practice independently and don't know when to refer.

To clarify my previous statement, I do think that the basic algorithms should be followed. Especially for patients who are naive to psychiatric meds. That being said, even the algorithms don't typically name specific medications like they do in other areas of medicine and we still can choose one of several meds from a specific class based on our clinical judgment. I also think that once you're past step 2 or 3 of common algorithms, things get much more fluid than in other fields. Since many of our patients who come to us are already past that point, it allows us to have more space for clinical judgment as opposed to just following a general guideline.

I do agree with your statement regarding midlevels and even some PCPs regarding inappropriate med choices or doses. Recently I inherited a patient with MDD from an NP who started the patient on 3 antipsychotics (Seroquel, Abilify, and Olanzapine) in the first 2 visits and I wanted to put my fist through the computer. D/c'd Seroquel and Olanzapine, decreased Abilify from 15mg to 10mg, and started Prozac and the patient is suddenly doing much better 2 months later (shocking, I know). I'm also quickly tiring of PCPs who start patients on a benzo shortly after a traumatic event then refer them to me for further treatment. I've had more than one discussion where I gave the patient the option of either tapering the benzo or allowing the PCP to manage it. Unfortunately, they almost always choose the latter.

 

[Alemo]

I agree with all that, and that's all in the framework of well informed, conscientious providers.

But it also seems given the current treatment options available in psychiatry, it's hard to directly kill a patient no matter what you choose, at least not quickly, which may lead to the permissibility and invisibility of some bad choices. Which is why, for example, not giving diabetes causing drugs for sleep or benzos first line for anxiety would be good parts of an algorithm to follow. And maybe deviating from that should require some more informed consent than it currently does.

I agree completely. In almost every other field of medicine (surgical and medical), a misdiagnosis and erroneous treatment has significant consequences. Aside from missing catatonia or bipolar disorder, I can’t really think of many psychiatric conditions that could be fatal if missed/wrongly treated.

 

[sloh]

I agree completely. In almost every other field of medicine (surgical and medical), a misdiagnosis and erroneous treatment has significant consequences. Aside from missing catatonia or bipolar disorder, I can’t really think of many psychiatric conditions that could be fatal if missed/wrongly treated.

Similar to my specialty, PM&R. Outside of more rare radiculopathies; even rarer myelopathies and cauda equina syndrome, because pain is a subjective experience there are a lot of non-evidence based procedures performed and medications prescribed. Harm is never really immediate and/or life/limb threatening

 

[NickNaylor]

Treatment algorithms exist in psychiatry, but I think the foundation of psychiatric illness and the innumerable inputs that contribute to that illness make the practice less susceptible to an algorithmic approach. Psychiatry also suffers from very few head-to-head trials, thus many basic questions about pharmacotherapy in the management of our bread-and-butter illnesses are unanswered by the evidence. Our algorithms tend to be fairly general and are rarely completely prescriptive. And if you find yourself engaging in cookbook psychiatry, I would argue that you likely aren't being as thoughtful in your assessment and management of patients as would be due.