Discussion Forum for 5/10/18

I was surprised to find that this was the first meta-analysis of its kind, looking at the efficacy of antispychotics in patients with schizophrenia (including schizoaffective, schizophreniform, and delusional disorder).

Sarah, thanks for pointing out that fact that few studies measured social functioning and quality of life (versus decreasing positive symptoms) since that is what patients are most concerned about. This is a good reminder for us to look at the illness from the patient’s perspective. As we know, few patients I see with schizophrenia are fixated on the hallucinations, for example, but many more are concerned about wanting to do well in their job or maintaining their relationships.

Something that journal club is constantly reminding me of is to pay attention to the population sampled to understand generalizability. These subjects were rarely first time antipsychotic responders, but are more representative of the chronically mentally ill population. It’s another reminder of the slow and often unsatisfying results we get when working with this chronically ill population. We often see some improvement with the use of antipsychotic medication, but not “good” improvement, which can be frustrating to both the patient and the prescriber. 

It was interesting to me that the results showed antipsychotics to help with “any” response but less so with a “good” response in short periods of time. This reminds me that most patients must endure a plethora of side effects for some response, but must wait much longer for a “good” response if it comes at all. It’s also a good reminder to me in terms of managing my expectations. I recently converted a patient from oral risperidone to Invega Sustenna, thinking that poor adherence was the cause of her minimal response to treatment. But even after months of knowing that the patient received her medication, she is still so symptomatic. She is representative of the results of this meta-analysis.

(I wasn’t there for the Journal Club discussion re: the China studies being excluded, but… wow!)

And maybe someone who was there for the journal club can speak to the study aim if you all happened to discuss it?

Reference:

Leucht, S., Leucht, C., Huhn, M., Chaimani, A., Mavridis, D. Helfer, B., . . . & Davis, J. M. (2017). Sixty years of placebo-controlled antipsychotic drug trials in acute schizophrenia: Systematic review, Bayesian meta-analysis, and meta-regression of efficacy predictors. American Journal of Psychiatry, 174: 10.

Sarah,

Thanks for kicking off this discussion! And Laura, thank you for stepping up for leading our session--you did a fantastic job!

I really appreciated all of your thoughtful insights, Sarah. I too believe that social integration is an important outcome that deserves more research attention. I also was surprised that only 10 out of the 167 trials focused on this topic.

During our conversation, we discussed how this study had pretty broad inclusion criteria for subjects ("Adults with acute exacerbations of schizophre- via or related disorders (following the Cochrane Schizophrenic Group, we accepted all diagnostic criteria and we also included schizoaffective, schizophreniform, and delusional disorder, because these do not require generally different treatment") and that this may have artificially screwed results. Does it really make sense to talk about antipsychotic drug efficacy when we're discussing so many different conditions? In my opinion, the results of this study would have been more convincing had they focused on a narrower range of conditions (i.e. just schizophrenia).

What does everyone else think about these inclusion criteria? 

Just going to throw in another big THANK YOU to Laura for leading the group discussion this week.  It is a tough task to begin with, made even more daunting by the sheer weight of this article!

I was equally shocked that this analysis was the first of it's kind - just echoes that this field is still right in the middle of figuring out how to treat DSM disorders.  As far as why these studies focused on positive symptoms and not so much on quality of life/ social functioning (in my opinion), I would guess that it has to do with the focus on acute presentation.  I also wish that there were a more objective way to assess Quality of Life....though I guess objective measures wouldn't reflect the client's internal reality, but I digress...  I wonder about the fact that they chose to do a meta-analysis of acute treatment rather than maintenance, as well as including such a broad range of diagnoses, but I wonder if it is more indicative of the availability of the research?  Anyone have any thoughts on this or read the article differently than I did?

Sarah - you also bring up a good point about pharma creating such large sample sizes that their results become smaller, as well as creating such a sheer volume of studies as to find a publishable result.  I have found myself struggling to figure out how I want reps and the pharma companies to play a role in my life.  On one hand, I worry about the influence they can exert over a provider's choice of medication, possibly to the detriment of the patient, but then I also know that their ability to research and provide us with the most up to date research about outcomes on their med could potentially very positively impact practice.  I guess just using what we are learning here in terms of how to critically appraise an article is going to be our best tool - and remembering sample sizes and funding sources....  Also keeping in mind that they can provide us with educational tools, like the ones that Chelsea used in our last class!!

Looking forward to reading everyone else's thoughts on this!  Thanks for your input on the article, even though you weren't able to share in the moment!

Great discussion so far. And yes, special thanks to Laura for facilitating. I had poor service in Maui so I couldn't hear some of the discussion, but I think I got enough to benefit from it. This study was very interesting to read through as I haven't read something like this before with a meta-regression. I was also surprised with their findings from pharma company studies. And you are right, Sara, They have the funds and resources to have a publishable result from whatever study they aim to conduct, which is questionable sometimes because in the end, they aim to benefit for themselves. However, Evelyn also makes a good point with getting the latest data from these pharm companies. We just need to thread lightly along those lines and make sure that we are as unbiased as we can be with the focus on what is best for our patients. 

Another thing I thought about with antipsychotic responses was participants genetic response to these medications, their adherence, and number of medication trials they've had. We know that it gets harder and harder to treat psychosis with every psychotic episode. So someone may be experiencing an acute psychotic episode but may still have a poor response to some of these antipsychotics due to their history. They may not have a good response but could still experience side effects. 

In regards to the exclusion of Chinese studies, I agree with most of you on here. I have come across and cited some very impressive Chinese studies, but there is always that questioning of if it is legitimate to use. It's almost become like the story of the man who cried wolf. 

Hi All, 

Thank you for another good discussion and thank you, Laura, for facilitating so well! 

As I read this article I had a few thoughts that I wanted to share but was unable to since I was driving back from clinical during the session. 

First of all, I wanted to highlight some of the strengths and weaknesses  that I noticed. 

One major strength was that that this review looked at "any" response versus "good" response to the antipsychotic medications under study. As the authors explained, far too often individual studies only evaluate "any" response to a medication which is VERY different than a "good" response, especially in light of the serious side effects associated with antipsychotics. Similarly, this review's focus on quality of life and functioning was a strength, as these are incredibly important outcomes that are often overlooked in medication trials, as Sarah and Sareen mentioned in their posts. "Good" vs "any" response and the quality of life outcomes are some of the most clinically important pieces to consider when prescribing medications, so the fact that this review examined them is a real strength. 

 Another strength of this study was that the authors were able to consistently use the PANSS or the BPRS to measure primary outcomes for all but one study within their meta-analysis, adding to the consistency of their study. I also liked that the authors analyzed bias and other factors that may have affected individual study outcomes  study duration, randomization/blinding, and sponsorship by drug companies. 

A third strength that I found was this review's broad inclusion criteria that spanned multiple diagnoses. I discussed this further in a response post to Marlene, but I did feel that this inclusion made the review more clinical relevant to real life settings. 

A weakness that I thought about when reading this study was that the authors did not look at studies of clozapine or IM formulations of antipsychotics. Given how efficacious these can be, I felt that it would have been interesting and clinically useful to see how these matched up to the rest of the oral antipsychotics under review. I also found it strange that the authors included both published and unpublished studies, and worry that this may have led to the inclusion of some research that was not peer reviewed. 

In addition to strengths and weaknesses, I was also intrigued by the authors' discussion of the large 1964 NIMH trial, which showed that antipsychotics were more effective than placebos. Initially, the authors presented this study as lending support to antipsychotics, in contrast to other later "failed" antipsychotic trials. Later, however, the authors explained that half of the participants in the early NIMH study were suffering from their first episode of psychosis and were drug naive, whereas the later "failed studies" and the studies in their meta-analysis were not limited to first onset psychosis. Instead, participants were mostly chronic patients who tend to respond less to antipsychotics. This is a very useful comparison to draw and important for clinical practice. I feel that it makes much clinical sense to differentiate between first episode and chronic psychosis when evaluating treatment efficacy. I appreciated that the authors brought this point to our attention but wish that they had gone deeper into exploring this as a variable in the different studies that they reviewed. 

Overall, like others, I found this review to be interesting, despite the study aim being somewhat confusing. 

Thank you for the posts everyone. Although this post is a little delayed, I agree with the responses.

The author points out that the aim of the study is trying to figure out if antipsychotic drug efficacy may have decreased over the years. I found it interesting that most placebo-controlled studies were conducted by pharmaceutical companies, although I am not completely surprised since pharmaceutical companies have their own motives and agendas to profit.

After reading this article about antipsychotic effectiveness I learned how the rate of efficacy was manipulated. I assumed that antipsychotics are generally pretty effective if taken continuously and help decrease most positive and negative symptoms. As discussed in the article about the 1964 meta-analysis sponsored NIMH revealed that there is a significant difference between antipsychotics and placebos. Taking a deeper look, the author reveals much of the discrepancies  in the study such as increasing placebo response rate and analyzing only predictors of placebo response, which artificially inflated the statistics.

The most significant part of this study that struck out to me was that antipsychotic drugs increased the patients with “any” response (51% vs 30%) to patients that had “good “ response (23% vs 14%), which is different than the 1964 NIMH benchmark study used for antipsychotic efficacy. 51% of patients reported at least a minimal response meaning a 20% significant change in positive or negative symptoms. Also, I think it is important to note that participants receiving placebo were likely to discontinue the study prematurely due to various reasons. I also found it interesting that the antipsychotic effect size for positive symptoms were similar to overall symptoms and that negative symptoms and depression were even smaller. It is important for me to remember as a provider that many symptoms may not be alleviated by antipsychotics and can cause the patient to be deterred from treatment. Having a better understanding of the effectiveness of these drugs gives the provider a better understanding of how patient may be feeling.

This article raises the question of generalizability in studies and encourages me as future provider to take a closer look at research.