How do psychiatrists feel about the SSRI vs Placebo issue?

[NRP]

I'm a 3rd year interested in psych and I'll admit the whole unknown realm of anti-depressants is a bit of a turn off for me. How can something so generally prescribed as SSRIs be something that has a success rate on par with placebos for most cases of depression? Why are they still being used? What do psychiatrists feel about one of their main drugs only being as good as a placebo?

For context, I myself was on SSRIs but I have wondered if I got better due to time or due to the actual drugs so I'm not speaking out of nowhere.

---

Members don't see this ad.

 

[MrChance2]

SSRIs are more effective than placebo. ~30% more effective. Better than that if u add therapy.

 

[HarryMTieboutMD]

There is an emerging body of literature on the placebo response and its a very interesting area academically. Here's a recent paper Patient Expectancy as a Mediator of Placebo Effects in Antidepressant Clinical Trials. - PubMed - NCBI

That said, there are many issues with RCTs wherein you find placebo effects- a controlled setting with lots of attention, support, use of outdate rating scales (sad that in 2017 we are still using decades old paper scales, however valid they may be when we have the capabilities to do it in real time with EMA, etc) which in part prompted efficacy trials like STAR*D (though these have their own issues). Maurizio Fava has also patented a trial design that attempts to obviate these effects

However, in practical day to day practice, most of us care about our patients getting better by whatever means (besides marijuana ), so I don't find it a hindrance

 

[NRP]

SSRIs are more effective than placebo. ~30% more effective. Better than that if u add therapy.

Do you know the source for that claim? I've seen that the one Kirscher study of SSRIs vs placebos has an effect size of .3 if that's what you're referring to?

 

[Stagg737]

I'm a 3rd year interested in psych and I'll admit the whole unknown realm of anti-depressants is a bit of a turn off for me. How can something so generally prescribed as SSRIs be something that has a success rate on par with placebos for most cases of depression? Why are they still being used? What do psychiatrists feel about one of their main drugs only being as good as a placebo?

For context, I myself was on SSRIs but I have wondered if I got better due to time or due to the actual drugs so I'm not speaking out of nowhere.

Care to provide sources? All the meta-analyses I've seen have shown that SSRIs are superior to placebos. Don't look at individual studies, look at the meta analyses, they tell a different story.

Even with the debate, this isn't the only area of medicine where treatment modalities are come into question: narcotics for lower back pain, kayexelate for hyperkalemia, etc. Plus just because one SSRI isn't effective doesn't mean another one won't be.

 

[TexasPhysician]

Both clinically and research demonstrate SSRI's to be effective. The FDA also supports them as statistically effective and safe enough to gain approval.

It is so easy for antidepressants to beat placebo that dozens of antidepressants have 2+ effective studies each.

 

[nitemagi]

The Kirschner stuff is a misuse of statistics. By pooling data you signal:noise ratio gets washed out. There is a definite signal (effectiveness) -- but it's probably for a subset of patients. Such as if they subdivided research into patients based on SERT subtypes (L/L, L/S, S/S), I'm willing to bet we'd see the effect size grow significantly in the L/L. When we lump those all together the effect seems to get diminished through averaging those that responded with those that didn't.

So the real question, IMO, is not if SSRIs work, but for whom?

 

[NRP]

Care to provide sources? All the meta-analyses I've seen have shown that SSRIs are superior to placebos. Don't look at individual studies, look at the meta analyses, they tell a different story.

Even with the debate, this isn't the only area of medicine where treatment modalities are come into question: narcotics for lower back pain, kayexelate for hyperkalemia, etc. Plus just because one SSRI isn't effective doesn't mean another one won't be.

These are the ones that have caught my attention:
Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration
Antidepressants and the Placebo Effect

This layman site SSRIs: Much More Than You Wanted To Know helped to breakdown SSRI vs Placebo arguments. I'm fine with other areas of medicine having multiple drugs that have to be used like you pointed out, but I just wish psychiatry had a more stable footing for one of their most used medications.

I'll be the first to admit that I'm not well versed in psychiatric medication efficacy. I always assumed that all the drugs we were taught worked in some shape or form with some having specific indications over the others (I've done my psych rotation btw). It wasn't until a classmate brought up the question on if SSRIs even work vs natural time healing wounds that I wasn't able to bring up a primary source of information up.

 

[birchswing]

What's the consensus on fish oil vs SSRIs? I thought I saw someone on these forums once say they were equitable for specific conditions (maybe "regular" depression versus severe?) but memory is fuzzy.

 

[Stagg737]

These are the ones that have caught my attention:
Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration
Antidepressants and the Placebo Effect

This layman site SSRIs: Much More Than You Wanted To Know helped to breakdown SSRI vs Placebo arguments. I'm fine with other areas of medicine having multiple drugs that have to be used like you pointed out, but I just wish psychiatry had a more stable footing for one of their most used medications.

I'll be the first to admit that I'm not well versed in psychiatric medication efficacy. I always assumed that all the drugs we were taught worked in some shape or form with some having specific indications over the others (I've done my psych rotation btw). It wasn't until a classmate brought up the question on if SSRIs even work vs natural time healing wounds that I wasn't able to bring up a primary source of information up.

Keep in mind that psych is a more subjective field than many others in terms of diagnosis because there's far more reliance on history than actual quantifiable data. Patients aren't always accurate historians or sometimes don't understand what the terms we use mean. I'm also only a third year, but I've already seen a dozen patients that didn't know what anxiety was. You'd also be shocked at how many times I've had to explain to fully grown adults what the word "concentration" meant. Plus some patients will flat-out lie, or just exaggerate symptoms. So we may be trying to treat something that isn't even there.

I've had every attending I've worked with tell me all of the anti-depressants have the same overall efficacy for treating depression, the only difference when considering what to prescribe is the side effect profile and what other indications the meds have (like duloxetine for various types of pain). Plus there's a difference between someone going through a major depressive episode because of an event/having a more acute episode and having chronic severe depressive symptoms. For many patients natural time healing wounds just won't happen and antidepressants become essential for them to even function.

 

[jadedphysiotherapist]

These are the ones that have caught my attention:
Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration
Antidepressants and the Placebo Effect

This layman site SSRIs: Much More Than You Wanted To Know helped to breakdown SSRI vs Placebo arguments. I'm fine with other areas of medicine having multiple drugs that have to be used like you pointed out, but I just wish psychiatry had a more stable footing for one of their most used medications.

I'll be the first to admit that I'm not well versed in psychiatric medication efficacy. I always assumed that all the drugs we were taught worked in some shape or form with some having specific indications over the others (I've done my psych rotation btw). It wasn't until a classmate brought up the question on if SSRIs even work vs natural time healing wounds that I wasn't able to bring up a primary source of information up.

SSRIs work. Natural time is ridiculous. Psychotherapy works well over a course of time but the SSRI in conjunction with a drug that increases its potency is the most effective for serious depressive states or episodes. Weaning off medications working together and progressing to just psychotherapy is a good way to go until a patient can self manage with coping skills but it can be difficult for the patient. Some will stay on their whole life.

Reference level 1 evidence for it (I.e. Systematic reviews and meta analyses)

Hope this was informative

 

[smalltownpsych]

Antidepressant Drug effects and Depression Severity: A Patient-Level Meta-Analysis
This article, I believe, is one that stirred up a lot of this controversy about placebo effect being equivalent to SSRI medication effects. Basically it points to the medication being more effective with the more severe cases of depression and equivalent to placebo in mild to moderate depression. Thinking about it logically, this shouldn't be too surprising of a finding because of a number of factors that need to be taken into account and more hypothesis testing be done. The way the research is funded on these medications does make it more difficult to take it to that next level of analysis, however.

A couple of quick hypotheses to explain above finding. First is that the more severe the depression, then the more room we have to see an effect and to differentiate that effect from placebo. Also, maybe the more severe depression is caused by biochemical factors more than environmental or situational factors so addressing the biological factors will be more effective. Possibly the more severe depression causes more biochemical changes and thus the patient needs these changes to be treated. Maybe peole who are really depressed have lost all hope and the placebo effect is lessened so then the medication effect looks better in comparison.

The reality is that we need a lot more quality research to be conducted to parse this. In the meantime, the majority of the people being treated with SSRIs are being treated by PCPs for mild to moderate depression and some remit thanks to either med effect, placebo effect, or regression to the mean and the rest end up in our offices after they get much worse. That raises a lot more questions to research, as well. Maybe the next big class action lawsuit against one of these pharma companies could go to fund some of this much needed research. I'm not saying that big pharma is some evil entity, just that they have their own agenda that doesn't mesh well with what our field really needs in regards to these medications.

***Edit to add***. After I posted, I saw that you were asking for psychiatrists' opinions so hope you don't mind that I chimed in with a psychologist's opinion.

 

[singasongofjoy]

When looking at the studies one thing it's important to keep in mind is the severity of symptoms in the population they were treating. If you've got relatively mild depression, maybe the 20% effect you get from a placebo effect or supplements like St. John's Wart and a little bit of therapy support + behavioral activation is plenty. If you've got more severe symptoms it's going to take more than that because even if your symptoms improve a bit, you're still in a pretty bad place with impaired daily functioning. So +1 to the person above who said to look at the meta-analyses; looking at individual studies is not always comparing apples to apples exactly- and also +1 to the real question being who / which patient population benefits from the SSRIs the most, not if.

Edit: smalltownpsych posted like 10 seconds before me and made the point I was trying to make much more eloquently. So +1 to the above

 

[clausewitz2]

These are the ones that have caught my attention:
Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration
Antidepressants and the Placebo Effect

This layman site SSRIs: Much More Than You Wanted To Know helped to breakdown SSRI vs Placebo arguments. I'm fine with other areas of medicine having multiple drugs that have to be used like you pointed out, but I just wish psychiatry had a more stable footing for one of their most used medications.

I'll be the first to admit that I'm not well versed in psychiatric medication efficacy. I always assumed that all the drugs we were taught worked in some shape or form with some having specific indications over the others (I've done my psych rotation btw). It wasn't until a classmate brought up the question on if SSRIs even work vs natural time healing wounds that I wasn't able to bring up a primary source of information up.

Minor aside, the blog you link to is written by a psych resident. Aimed at laypersons, definitely written by someone in the biz.

 

[Merovinge]

I realize you are speaking to depression, but SSRI's work pretty fantastically for many types of anxiety, OCD and can help in some subsets of folks with PTSD. Their ubiquitous use is related to their side-effect profile being so incredibly tolerable, limited lethality in OD (a huge issue when prescribing to people who are likely to OD), and large number of conditions they treat.

 

[deleted343839]

What's the consensus on fish oil vs SSRIs? I thought I saw someone on these forums once say they were equitable for specific conditions (maybe "regular" depression versus severe?) but memory is fuzzy.

The problem with fish oil is that the taste shows up later and this makes blinding difficult. Unblinding --> positive expectancies --> placebo response.

You can quibble with Kirsch's methods, and should, but a key take-home point is that the main predictor of placebo response is the use of a PRO as a primary endpoint. That affects fields well beyond psychiatry, as someone else mentioned.

It is so easy for antidepressants to beat placebo that dozens of antidepressants have 2+ effective studies each.

There are a lot of negative trials buried in the proverbial file drawer, even for "proven" therapies (same is true of behavior therapy trials). So the more interesting question to me is how many trials are needed to get a "win" that you can take to FDA. I'd be curious about the ratio for new drugs versus repurposed/expanded indications. It's harder nowadays to bury negative trials so we may have a better sense of this in the years to come.

If you can count on 30-40% of people meeting your endpoint regardless of group assignment, then you need large sample sizes to show statistical separation in the active arm(s), and that makes for an expensive trial. Perhaps the risk-benefit analysis still favors the expense of drug development, but the stakes are high and the risk is real. Easy, it's not.

 

[tr]

These are the ones that have caught my attention:
Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration
Antidepressants and the Placebo Effect

This layman site SSRIs: Much More Than You Wanted To Know helped to breakdown SSRI vs Placebo arguments.

None of those sites say that SSRIs are no more effective than placebo. Kirsch himself agrees with the generally accepted/frequently quoted effect size of 0.3 which is also discussed on the layman's website you linked.

The fallacious statement that 'SSRIs are no more effective than placebo' is based on an idea Kirsch floated that this effect is actually explained by people in trials figuring out whether they are on the drug or the placebo based on the side effects, and that if you could conduct a truly blinded trial *then* you would find no effect over placebo. That claim remains unsubstantiated at present.

 

[NRP]

None of those sites say that SSRIs are no more effective than placebo. Kirsch himself agrees with the generally accepted/frequently quoted effect size of 0.3 which is also discussed on the layman's website you linked.

The fallacious statement that 'SSRIs are no more effective than placebo' is based on an idea Kirsch floated that this effect is actually explained by people in trials figuring out whether they are on the drug or the placebo based on the side effects, and that if you could conduct a truly blinded trial *then* you would find no effect over placebo. That claim remains unsubstantiated at present.

Selective serotonin reuptake inhibitors versus placebo in patients with major depressive disorder. A systematic review with meta-analysis and Trial Sequential Analysis | BMC Psychiatry | Full Text

This recent study is concluding that although SSRIs are statistically significant
To placebos their adverse profile and not clinically significant findings do not lead to them being favorable.



"SSRIs versus placebo seem to have statistically significant effects on depressive symptoms, but the clinical significance of these effects seems questionable and all trials were at high risk of bias. Furthermore, SSRIs versus placebo significantly increase the risk of both serious and non-serious adverse events. Our results show that the harmful effects of SSRIs versus placebo for major depressive disorder seem to outweigh any potentially small beneficial effects."

Thanks to everyone who has responded so far. It's odd to find that there are larger mountains of evidence pointing towards SSRIs being minimally effective to both depression and anxiety (I believe I've found some source for it being significantly for SAD though). It's a little bit disheartening to find this type of research especially when I have to decide what I'm applying into a couple months from now.

 

[Ceke2002]

Antidepressant Drug effects and Depression Severity: A Patient-Level Meta-Analysis
This article, I believe, is one that stirred up a lot of this controversy about placebo effect being equivalent to SSRI medication effects. Basically it points to the medication being more effective with the more severe cases of depression and equivalent to placebo in mild to moderate depression. Thinking about it logically, this shouldn't be too surprising of a finding because of a number of factors that need to be taken into account and more hypothesis testing be done. The way the research is funded on these medications does make it more difficult to take it to that next level of analysis, however.

A couple of quick hypotheses to explain above finding. First is that the more severe the depression, then the more room we have to see an effect and to differentiate that effect from placebo. Also, maybe the more severe depression is caused by biochemical factors more than environmental or situational factors so addressing the biological factors will be more effective. Possibly the more severe depression causes more biochemical changes and thus the patient needs these changes to be treated. Maybe peole who are really depressed have lost all hope and the placebo effect is lessened so then the medication effect looks better in comparison.

There were some interesting responses on some of the patient support forums at the time when this came out. I'd say around 25-30% suddenly panicked and wondered why they'd been prescribed an ineffective medication, and had to be talked down from just going off their medication without actually speaking to their prescribing physician first; and the rest were pretty much along the lines of 'Well, duh! Seriously, you needed a study to tell you this?'.

 

[TexasPhysician]

It's odd to find that there are larger mountains of evidence pointing towards SSRIs being minimally effective to both depression and anxiety (I believe I've found some source for it being significantly for SAD though). It's a little bit disheartening to find this type of research especially when I have to decide what I'm applying into a couple months from now.

If you are looking for a field with highly effective treatments across all conditions, there is a rough road ahead.

On the positive side, I see patients recovering from depression frequently and successfully managing many other conditions to remission/cure weekly. I also do therapy improving family relations. It's a great field, but no field is perfect, including psychiatry.

 

[NRP]

If you are looking for a field with highly effective treatments across all conditions, there is a rough road ahead.

On the positive side, I see patients recovering from depression frequently and successfully managing many other conditions to remission/cure weekly. I also do therapy improving family relations. It's a great field, but no field is perfect, including psychiatry.

Are your patients w/ depression using SSRIs, therapy, or both when you see them get cured? Based on your past post I'm guessing that you are a supporter of SSRIs, but I'm just wondering on how you feel when you see multiple articles mentioning their lack of clinical significance. Does that make you change your regimen or do you continue based on your own personal experiences?

I'd feel a lot "better" if one of the mainstays of depression wasn't being "debunked" so much. I definitely see the real clinical significance of antipsychotics or benzos and their effects so I haven't lost all hope in psychiatry, but for something like SSRIs it seems like such a smoke and mirrors trick. "Oh, it won't work until 4 weeks from now, but after that you'll really see results!" How much of that is just the time healing thing? I know that a lot of people don't like hearing about time solving issues and that for some it may never resolve, but for a large amount of people going to their PCP for depression could their issues resolve eventually? I really don't want to be some professional snake oil salesman and be using a drug that doesn't actually do anything for the "average" depressed patient. I think SSRIs worked for me personally, but hell it might be the placebo talking (although for me that's fine).

 

[hamstergang]

I'd feel a lot "better" if one of the mainstays of depression wasn't being "debunked" so much. I definitely see the real clinical significance of antipsychotics or benzos and their effects so I haven't lost all hope in psychiatry, but for something like SSRIs it seems like such a smoke and mirrors trick.

I'm not sure what you're looking for at this point. Multiple people have responded already explaining why the literature still supports the efficacy of SSRIs for depression. Do you think that you, as a medical student, have a better grasp of the body of literature than the attendings and residents (and psychologists) in this thread?

Lots of people like to attack psychiatry, claiming that we don't help anyone or we make people worse. I don't see these people much in person and it doesn't affect my day-to-day work. If it bothers you that much than don't apply to psych. You can see here many people happy in the field of psychiatry willing to defend it and explain it against the criticisms you raise. You could be like these people and happy in psych too. Or maybe you decide you can't, that's up to you. But the question of SSRI efficacy has been sufficiently addressed already, so I'm not sure what you're expecting to hear by challenging it again.

 

[sloop]

I just watched this earlier today and it was interesting and explained pretty well:

 

[TexasPhysician]

Are your patients w/ depression using SSRIs, therapy, or both when you see them get cured? Based on your past post I'm guessing that you are a supporter of SSRIs, but I'm just wondering on how you feel when you see multiple articles mentioning their lack of clinical significance. Does that make you change your regimen or do you continue based on your own personal experiences?

I let the patient determine what type(s) of treatment after education and discussing risks/benefits.

One of the most valuable lessons learned in my medical education is the ability to analyze and critique primary research. I can design a study that can reach any conclusion, but that doesn't mean my research methods reached an accurate result.

No treatment is perfect. My daughter had to go through 3 antibiotics to defeat 1 ear infection. When I was a teenager, I could bathe in Accutane and still get acne.

An effective psychiatrist doesn't just reach a DSM5 diagnosis and choose the same antidepressant for everyone or one at random. The same goes for therapy. I believe that there is an art to matching antidepressants/therapy to the patients list of symptoms, comorbid conditions, age, typical responses to other medications, etc.

How many studies have evaluated response rates in psychiatry after customized treatment plans agreed upon by a group of excellent clinical psychiatrists? I'd love to construct a good study comparing a team's customized treatment vs me pulling treatment strategies out of a hat. I just haven't found anyone willing to put up the $$$ yet.

 

[NRP]

I'm not sure what you're looking for at this point. Multiple people have responded already explaining why the literature still supports the efficacy of SSRIs for depression. Do you think that you, as a medical student, have a better grasp of the body of literature than the attendings and residents (and psychologists) in this thread?

Lots of people like to attack psychiatry, claiming that we don't help anyone or we make people worse. I don't see these people much in person and it doesn't affect my day-to-day work. If it bothers you that much than don't apply to psych. You can see here many people happy in the field of psychiatry willing to defend it and explain it against the criticisms you raise. You could be like these people and happy in psych too. Or maybe you decide you can't, that's up to you. But the question of SSRI efficacy has been sufficiently addressed already, so I'm not sure what you're expecting to hear by challenging it again.

Actually nobody here provided any sources or data refuting it. Could you link the literature that you're referring to that supports the efficacy of SSRIs from depression. I'd be happy in psych but I'd rather not be ignorant and I don't understand why that is an issue. I'm expecting to hear a professional in the field provide something concrete data as to why we use SSRIs at all on a large scale. Anecdotal evidence doesn't really go far and that's what's mostly coming through right now.

EDIT: There were multiple posts before I posted this. Thanks for those posts and their information! Honestly, I never even considered the idea that a placebo itself is psychotherapy. That definitely shows my inexperience in all this, but that in itself is pretty comforting to hear.

 

[TexasPhysician]

Actually nobody here provided any sources or data refuting it. Could you link the literature that you're referring to that supports the efficacy of SSRIs from depression. I'm expecting to hear a professional in the field provide something concrete data as to why we use SSRIs at all on a large scale.

Get your hands on the original Treatment for Adolescents with Depression Study in JAMA 2004. Let us know your thoughts.

 

[smalltownpsych]

I just watched this earlier today and it was interesting and explained pretty well:

Very clear and fair explanation of medications and psychotherapy and the complexity of parsing out all of the factors. I especially liked the emphasis on not treating adjustment reactions with medications. "If you cry in your docs office there is a good chance you'll walk out with an SSRI." I disagree that the patient should necessarily get a referral for psychotherapy either. This is where some clinical judgement and experience of the physician should come into play. Just like not every sprained ankle needs an ortho consult or at least so I would hope.

 

[Salpingo]

It's been awhile, but if you look at number needed to treat (NNT), my recollection is that psychotropics beat out most other medications. I think fluoxetine was 6, then ASA for MI was something like 42.

 

[erg923]

Very clear and fair explanation of medications and psychotherapy and the complexity of parsing out all of the factors. I especially liked the emphasis on not treating adjustment reactions with medications. "If you cry in your docs office there is a good chance you'll walk out with an SSRI." I disagree that the patient should necessarily get a referral for psychotherapy either. This is where some clinical judgement and experience of the physician should come into play. Just like not every sprained ankle needs an ortho consult or at least so I would hope.

I see this too in primary care. It must be taught in medical school that crying is some type of pathological reaction that must be intervened upon.

 

[Mad Jack]

The Kirschner stuff is a misuse of statistics. By pooling data you signal:noise ratio gets washed out. There is a definite signal (effectiveness) -- but it's probably for a subset of patients. Such as if they subdivided research into patients based on SERT subtypes (L/L, L/S, S/S), I'm willing to bet we'd see the effect size grow significantly in the L/L. When we lump those all together the effect seems to get diminished through averaging those that responded with those that didn't.

So the real question, IMO, is not if SSRIs work, but for whom?

The trouble is, at least to me, that they're just passed off as standard of care even though they don't work for the majority of patients. I guess that means we need to sort out the whom better, because I do believe that there are many people who have a robust response to SSRI therapy. But our current system is akin to giving everyone with a URI antibiotics- it just feels like bad medicine.

 

[smalltownpsych]

The trouble is, at least to me, that they're just passed off as standard of care even though they don't work for the majority of patients. I guess that means we need to sort out the whom better, because I do believe that there are many people who have a robust response to SSRI therapy. But our current system is akin to giving everyone with a URI antibiotics- it just feels like bad medicine.

That isn't supposed to happen?

 

[sloop]

The trouble is, at least to me, that they're just passed off as standard of care even though they don't work for the majority of patients. I guess that means we need to sort out the whom better, because I do believe that there are many people who have a robust response to SSRI therapy. But our current system is akin to giving everyone with a URI antibiotics- it just feels like bad medicine.

It feeling like bad medicine and it being bad medicine are two separate things.

Of course being able to more accurately define the population to benefit from an intervention is always desirable, but the truth is that the medical science which defines treatment guidelines for medicine in general is almost always focused on a level remote from the individual. What I mean by this is that, with the limited exceptions of genetic guidance of therapy (which I think is what the future may hold more of), medical science only tells us about aggregate benefits and risks to populations. The hippocratic ideal of "do no harm" (i.e. total nonmaleficence) is a myth. The goal of medicine is to benefit the most people while minimizing harms, but there is always a tradeoff. Sometimes this means unnecessarily medicating people because they are part of a population the science tells us will benefit on aggregate. A lot of people could probably live without adverse effects from mild essential hypertension, but we still treat them despite the side-effects of these medications in hopes of achieving aggregate benefits.

My point is that of course we should aim to exclude from treatment populations that evidence shows receive no benefit, but in the absence of convincing evidence of this and in the face of evidence of a broader aggregate benefit, the obvious decision is to treat. This is not a unique psychiatric problem. It's a function of how the science works.

 

[tr]

Actually nobody here provided any sources or data refuting it. Could you link the literature that you're referring to that supports the efficacy of SSRIs from depression.

Your *own citations* that you linked support that there is a statistically significant effect over placebo.

There is a lot of argument over whether this is clinically significant (which is a clinical judgement call, not something that can be definitively established in a study), whether it's worth the side effects, whether it's useful for patients with mild/moderate vs only with severe depression, etc. The bottom line is that even the most skeptical investigators find that there is a statistically significant benefit over placebo. The interpretation of the clinical meaningfulness of that finding and the contexts in which it applies is what is under debate.

 

[tr]

None of those sites say that SSRIs are no more effective than placebo. Kirsch himself agrees with the generally accepted/frequently quoted effect size of 0.3 which is also discussed on the layman's website you linked.

The fallacious statement that 'SSRIs are no more effective than placebo' is based on an idea Kirsch floated that this effect is actually explained by people in trials figuring out whether they are on the drug or the placebo based on the side effects, and that if you could conduct a truly blinded trial *then* you would find no effect over placebo. That claim remains unsubstantiated at present.

Reviving this thread because a new publication came out that explicitly tested the Kirsch hypothesis (namely, that the apparent clinical effect of antidepressants is actually due to people figuring out whether they are on active drug based on side effects).
The answer is No. Presence and severity of side effects are not related to clinical benefit (well not at all for Celexa and only minimally so for Paxil). Effect sizes were 0.48 for Celexa and 0.33 for Paxil.

Hieronymus et al., Molecular Psychiatry doi: 10.1038/mp.2017.147
Molecular Psychiatry - Efficacy of selective serotonin reuptake inhibitors in the absence of side effects: a mega-analysis of citalopram and paroxetine in adult depression

 

[Mad Jack]

Your *own citations* that you linked support that there is a statistically significant effect over placebo.

There is a lot of argument over whether this is clinically significant (which is a clinical judgement call, not something that can be definitively established in a study), whether it's worth the side effects, whether it's useful for patients with mild/moderate vs only with severe depression, etc. The bottom line is that even the most skeptical investigators find that there is a statistically significant benefit over placebo. The interpretation of the clinical meaningfulness of that finding and the contexts in which it applies is what is under debate.

Paroxetine versus other anti-depressive agents for depression. - PubMed - NCBI

I mean, this gives you a few solid NNT numbers, but basically your NNT for antidepressants is 8 or higher, which to me shows that 7 out of 8 people, at best, shouldn't be on antidepressants if they're not responding, and we should be working on better ways to sort out who should and shouldn't remain on them. But hey, that would take a lot of research into optimizing who ends up on or off meds that no one is ever going to fund, so the status is likely to remain quo indefinitely.

 

[Mad Jack]

It should also be noted that the NNH of meds such as paroxetine is pretty damn low in regard to certain indicators, such as the NNH of between 2 and 3, depending on the study, for sexual dysfunction- at the higher harm rate, you're basically giving 4 people sexual dysfunction for every 1 person you successfully treat. And that's just one side effect...

 

[nitemagi]

Kirsch is a placebo researcher. He views everything through that lens. He has basically taken the stand that hypnosis is also largely placebo, which is true to a small extent (belief is part of the puzzle), but misses the larger amount of data showing the rest of the picture.

 

[cbrons]

Care to provide sources? All the meta-analyses I've seen have shown that SSRIs are superior to placebos. Don't look at individual studies, look at the meta analyses, they tell a different story.

Even with the debate, this isn't the only area of medicine where treatment modalities are come into question: narcotics for lower back pain, kayexelate for hyperkalemia, etc. Plus just because one SSRI isn't effective doesn't mean another one won't be.

Meta analyses are incredibly worthless compared to good RCTs. With meta analyses if you dont know the quality of the individual studies you wont have any idea whether the MA is good or bad. Vast majority are incredibly bad to quite bad.

 

[thoffen]

1. SSRIs are better than placebo no question
2. Not treating major depression is dangerous and debilitating. Even if the efficacy of your intervention is placebo effect and therapeutic relationship, the effect is critical. We start it because it is better than placebo. If something works, we don't know if it would have worked similarly if it's placebo. We do know that doing nothing wouldn't likely have worked and that treatment working for any reason is life changing or life saving.

The bigger hazard is prescribing SSRIs when they are not indicated.

 

[Mad Jack]

1. SSRIs are better than placebo no question
2. Not treating major depression is dangerous and debilitating. Even if the efficacy of your intervention is placebo effect and therapeutic relationship, the effect is critical. We start it because it is better than placebo. If something works, we don't know if it would have worked similarly if it's placebo. We do know that doing nothing wouldn't likely have worked and that treatment working for any reason is life changing or life saving.

The bigger hazard is prescribing SSRIs when they are not indicated.

Cognitive therapy vs. medications for depression: Treatment outcomes and neural mechanisms

But if therapy is as effective as medication and has none of the side effects, why not just utilize therapy as a primary intervention and add medication if therapy fails? Particularly in light of only 1 out of 8 people seeing benefit from SSRIs, but well over half having substantial side effects?

 

[OldPsychDoc]

Cognitive therapy vs. medications for depression: Treatment outcomes and neural mechanisms

But if therapy is as effective as medication and has none of the side effects, why not just utilize therapy as a primary intervention and add medication if therapy fails? Particularly in light of only 1 out of 8 people seeing benefit from SSRIs, but well over half having substantial side effects?

Because it takes time and effort...and this is America.

 

[hamstergang]

Particularly in light of only 1 out of 8 people seeing benefit from SSRIs, but well over half having substantial side effects?

That's not what NNT means. Over half the patients see benefit from SSRIs, but many see benefit from placebo as well.

Work as a psychiatrist for a while and see how many patients you can get to engage in therapy. For mild to moderate (sometimes) depression, I do recommend therapy before meds.

 

[thoffen]

Cognitive therapy vs. medications for depression: Treatment outcomes and neural mechanisms

But if therapy is as effective as medication and has none of the side effects, why not just utilize therapy as a primary intervention and add medication if therapy fails? Particularly in light of only 1 out of 8 people seeing benefit from SSRIs, but well over half having substantial side effects?

Availability of study quality therapy, cost of said therapy, transportation to said therapy, time away from work/school/family for said therapy, agreeableness to said therapy, etc.

But I agree therapy is underutilized and underrecommended particularly as a first line option.

 

[Mad Jack]

That's not what NNT means. Over half the patients see benefit from SSRIs, but many see benefit from placebo as well.

Work as a psychiatrist for a while and see how many patients you can get to engage in therapy. For mild to moderate (sometimes) depression, I do recommend therapy before meds.

For patients that aren't willing to engage in therapy, I'm all for SSRIs. Just seems like things have swung pretty heavily toward the pharmaceutical end of the pendulum.

Availability of study quality therapy, cost of said therapy, transportation to said therapy, time away from work/school/family for said therapy, agreeableness to said therapy, etc.

But I agree therapy is underutilized and underrecommended particularly as a first line option.

Totally agree. Out of 10 patients that should be getting therapy, there's probably 3 with the means and 1 willing to commit the time for quality therapy, and then actually finding someone to provide that can be a challenge unto itself. Pretty much answers my question lol. SSRIs aren't the best answer for most people, but they're the best answer that people are able to afford, put up with, maintain compliance, and have access to.

 

[thoffen]

Totally agree. Out of 10 patients that should be getting therapy, there's probably 3 with the means and 1 willing to commit the time for quality therapy, and then actually finding someone to provide that can be a challenge unto itself. Pretty much answers my question lol. SSRIs aren't the best answer for most people, but they're the best answer that people are able to afford, put up with, maintain compliance, and have access to.

SSRI risk: material cost (e.g. $4/mo) + access (Dr.'s apts, insurance, etc.) + tolerance (side effects)

psychotherapy risk: material cost (e.g. $100/wk) + access (more frequent apts, insurance, etc.) + tolerance (agreeableness, fit with therapist, tolerance of distress kicked up from psychotherapy)

For the average comer, benefit is roughly equivalent. How are those tolerance factors for therapy different in your risk assessment than SSRIs? Neither is clearly a superior treatment, but I would argue in the real world SSRI is likely to be better tolerated.

Nonetheless, it is disturbing when it isn't offered as an option for treatment from the get-go. It is more disturbing when SSRIs aren't clearly indicated. And it is more disturbing when psychotherapy isn't presented for conditions to which it is the gold standard treatment. I feel sick and angry when I see inpatients who have had classic PTSD or OCD for years whom have never been told that therapy could treat their problem.

 

[smalltownpsych]

As a therapist, I recommend that the patient chooses after we provide the best available information. Nothing worse than a therapy referral that the patient came because their "prescriber" (I hate that word) told them they had to. I won't refer for a medication evaluation until after we discuss the pro's and cons' and patient decides that it is something they want to do. For some patients we might even discuss this over several sessions before they decide. On the other hand, I do tend to refer the more severe cases immediately. The analogy the other way would be the patient is pretty clear from the start that they have a significant amount of interpersonal stressors or trauma that they just can't stop talking about.