What is the impact of a gradual reduction of antipsychotics?

Antipsychotic medicines reduce the risk of psychotic episodes – and they are recommended for long-term use by people with schizophrenia or recurrent psychosis. But the medicines can have such unpleasant side effects that people prescribed these drugs may want to reduce or to stop them.

In this podcast, Helen Saul, Editor in Chief of NIHR Evidence, speaks with Joanna Moncrieff, Professor of Critical and Social Psychiatry at University College London; and Rachel Upthegrove, Professor of Psychiatry and Youth Mental Health at the University of Birmingham. They discuss an NIHR trial in which people were supported to gradually reduce their antipsychotic treatment

Read a full transcript of the episode here.

The views and opinions expressed in this podcast are those of the host and guests and do not necessarily reflect those of the NIHR or the Department of Health and Social Care.

Transcript

[MUSIC]

0:08

Helen Saul: Hello, and welcome. In this podcast from the NIHR, we will be looking at psychosis and the potential benefits and harms of reducing the dose of antipsychotic medicines. 

During a psychotic episode, people lose some contact with reality. They might see, hear or believe things that other people don’t. Their thoughts, ideas and speech may become fast and confusing; these symptoms can cause severe distress, and change people’s behaviour. 

Antipsychotic medicines reduce the risk of psychotic episodes – and they are recommended for long-term use by people with schizophrenia or recurrent psychosis. But the medicines can have such unpleasant side effects that people prescribed these drugs may want to reduce or to stop them.

Today we'll be discussing an NIHR trial in which people were supported to gradually reduce their antipsychotic treatment. My name is Helen Saul, and I'm the editor in chief of the NIHR Evidence website. I have 2 guests with me today. Welcome first to author Joanna Moncrieff.

1:20

Joanna Moncrieff: Thank you for having me. I'm a professor of critical and social psychiatry at the at University College, London, and I was the chief investigator for the what was called the RADAR program, research into antipsychotic discontinuation and reduction which was funded by the NIHR. As I said.

01:39

Helen Saul: Thank you. And Rachel Upthegrove.

01:40

Rachel Upthegrove: Hi! Thanks for inviting me so. Yes, my name is Rachel Upthegrove. I'm a professor of psychiatry and youth mental health at the University of Birmingham, and a consultant psychiatrist in early intervention and psychosis services, where I've worked for over 20 years. 

01:56

Helen Saul: Thank you. Rachel, could you outline for us, the impact of recurrent psychosis on someone’s life? 

2:04

Rachel Upthegrove: Psychosis in itself can be very impactful for people's function in terms of ability to maintain roles, jobs, relationships. And because the onset of the illness is

between late late teens and early twenties, for the majority of people, this is a critical period for developing those social roles as in, you know, jobs, relationships, place in society. 

2:28

Helen Saul: And you’d see antipsychotic drugs as an important part of treatment?

2:33

Rachel Upthegrove. Coming from the background, I do from from early intervention, our aim has always been to reduce the impact of psychosis and reduce what we call the duration of untreated psychosis by intervening early after someone has made or had a first episode of psychosis.

So antipsychotic medication, we know, is effective in 75 to 80% of people for treating an acute psychotic episode, and in early intervention. That's part of psychological, social, peer support and other interventions, but as a very key. As a very key part. 

3:06

Helen Saul: What about the side effects of these drugs Rachel?

3:10

Rachel Upthegrove: All medication or everything, umm Every treatment would have potential side effects. So psychological interventions, social interventions. But also medication has potential for side effects. In terms of antipsychotics, these can be broadly grouped into things that might affect what we call extrapyramidal side effects, so, muscle stiffness, tremor, rigidity. The more sort of second generation antipsychotics might have a little bit more emphasis on other things like weight gain and sedation and and also can affect sort of hormone levels so prolactin so they can stop periods and and cause and cause other prolactin-related side effects. So they do have a range of side effects that that people want to to avoid. By getting dosing regimes right.

And and and all medication has side effects that that as a as a prescriber you talk to your patient about, and would be honest with them about potential for side effects as well as potential for benefits.

4:14

Joanna Moncrieff: I just think it's important to emphasise they are, they are really significant.

4:18

Rachel: Yeah. And this is a and this is treating a significant illness. So I think that yes.

4:21

Joanna Moncrieff: of course.

4:24

Helen Saul: Joanna, you led a trial called RADAR exploring the impact of gradually reducing antipsychotic treatment. Before we discuss that work, perhaps you could tell us what previous research has shown?

4:37

Joanna Moncrieff: So there have been lots of studies that have compared continuing with antipsychotic medication with coming off it and being given a placebo instead or just coming off it altogether. And those studies generally show that people who come off the medication are more likely to have a relapse and end up in hospital.

But one of the reasons for doing the RADAR study was that those studies have almost all taken people off antipsychotics very quickly and we know that there are withdrawal effects, including agitation and sleep problems, there's also possibly a link between withdrawal and psychosis itself, so that it may be that withdrawal increases the risk of having a psychotic relapse that’s specifically related to the withdrawal process rather than the underlying condition.

5:30

Helen Saul: So what approach did the RADAR study take, Joanna? 

5:34

Joanna Moncrieff: The trial was aimed at people who had recurrent episodes of psychosis or a diagnosis of schizophrenia. And what we did was identify them. We excluded people who were thought to have very, very severe risks of of harming themselves or harming other people. If they were, if they came off medication, but otherwise our inclusion criteria were fairly fairly broad.

We recruited people from secondary care, mental health services and once people consented to to join the study, then they had a baseline assessment, and then they were randomized either to continue on their current dose of medication, more or less. People could make adjustments for side effects, obviously, but they were encouraged to stay stay on their current dose if they could. Or they were randomised to have a gradual reduction of their medication, supported and overseen by their treating clinician, and people came into the the trial on all sorts of different medication regimes, different sorts of antipsychotics, and different combinations of antipsychotics in some cases. So what we did was, we drew up a schedule for each individual for a gradual reduction of their medication that their clinicians could follow.

But we also emphasized that the reduction was meant to be flexible, and you know, people weren't, weren't to be forced to continue to reduce if they, you know, were having a life crisis, or if the reduction was going badly, and their symptoms were coming back. So it was meant to be flexible. And then people were followed up at 6 months, a year, and the main follow up was 2 years after randomization.

7:29

Helen Saul: And what was happening at 2 years? What were your main findings? 

7:33

Joanna Moncrieff: What we found was that people who had been randomised to reduction were more likely to relapse than people who were randomised to maintenance treatment, we defined relapse quite stringently. Our main relapse outcome was readmission to hospital with with a psychotic episode.

So so that that was the result in terms of relapse, which was disappointing, because we hoped that we could reduce the rate of relapse. I don't think we ever thought that there would be no relapses, but we hoped that by doing this gradual reduction process, we could at least have lower relapses than than were seen in other maintenance studies, but in fact, the relapse rate was around about the same as as in most maintenance studies, the readmission rate, the hospitalisation rate, anyway.

8:23

Helen Saul: But your main outcome was social functioning. Could you explain what that means please?

8:28

Joanna Moncrieff: social functioning really means, how, how are people able to get on with their day to day lives. Are people able to go to work? Are people able to look after themselves, attend to their personal care? Do their shopping and cooking? Are people able to get on with other people, both their close family members, but also members of the public, are people able to take part in in leisure activities as well as well as work activities.

So it's that's that's what social functioning means. 

9:01

Helen Saul: And was there an improvement in social functioning with a reduction in antipsychotics?

9:05

Joanna Moncrieff

We had hypothesis ed that maybe there might be an improvement in social functioning at our 2 year outcome, although I think we always thought that was a little bit optimistic. The Dutch study that we'd modeled the study on had only found the improvement in social functioning at their 7 year follow up, in fact. But anyway, we, we looked at social functioning at 2 years. There was no improvement in the group who'd been randomised to reduction, but there was no detrimental effect either. There was no difference between people who had been randomised to reduction, and people who had been randomised to maintenance treatment in terms of their social functioning or in terms of their symptom levels at follow-up or their quality of life. Or their medication side effects, even though they'd reduced their medication quite a bit. So that was a surprising finding. Or their satisfaction levels, or any of the other outcomes that we looked at.

10:01

Helen Saul: So what was your response to your own findings Joanna? You said the increase in relapse was disappointing.  

10:08

Joanna Moncrieff: It was disappointing. As I say. I hoped that we could at least have reduced relapse rates a little bit, and I think it's you know, I think it's interesting and important to know the sort of gradual reduction we did, anyway, which was over about a year to 18 months, doesn't significantly reduce relapse rates. It's possible that doing reduction over a longer period might might be better.

So that was disappointing. I think it's also important to emphasize, though, that our study, like other studies, found that not everyone relapses. In fact, it's a minority of people who reduce, or even who stop their medication altogether, who relapse. So yes, overall, there's an increased risk of relapse. But for each individual it doesn't mean that each individual will relapse if they stop their medication. 

So the point of the RADAR was always to provide information for people to make informed decisions about the the risks and benefits of treatment. And so the RADAR study, along with, along with other studies of antipsychotic reduction does show that there's an increased risk, but not an inevitable risk of having a relapse.

I should also mention that we at the same time, as publishing our the main paper detailing the quantitative results of the study, we published our qualitative study. In our qualitative study, we interviewed 23 people who'd been involved in the antipsychotic reduction programme in detail, and we made a point of selecting people who'd had various outcomes, not just people who'd done well with reduction, but people who'd done badly with the reduction, people who'd relapsed.and people who hadn't been able to even make a reduction. 

And in that study what was really interesting was that that many of the people in the reduction group felt had felt a sense of empowerment through the process of reduction, even if it had not worked out well for them, and had come out in some cases thinking, “Okay, well, I've had a go now. I you know. Now I know I've just got to stay on the medication”, in some cases feeling “Okay, maybe it didn't work out this time, but maybe if I did it more slowly or carefully in in the future it might be possible to work for it to work out.”

So I think that was that was also an an important finding for me that there was something about doing the trial and offering people the possibility of having a supported reduction with a clinician that that that benefited people.

12:41

Helen Saul: Thank you. Rachel. How - do these findings chime with your clinical experience?

12:48

Rachel Upthegrove: Well, I think the first thing to say is, Joanna sums it up very nicely, and that that the study fits in with what we already already know is that in a previous trial that if you reduce antipsychotic medication, there is a significantly increased risk of relapse. And you know, previous trials, trial trials have done that. So whenever we're talking with patients in a collaborative way about treatment, options and treatment decisions, it is weighing up that that evidence and balance of risk.

So the risk of relapse for for somebody might be very different than the risk of relapse for somebody else. Given, you know, past experience of psychosis and everyone's individual risk history. Versus, the side effects of medication. So this is what what we do day to day. This is what this is, what prescribers, psychiatrists, nurse prescribers, allied health specialist prescribers should be doing as part of good care anyway, you know, having that having that individual conversation about the risks of you reducing or stopping medication will be. And you know, particularly in in my practice. This is this is something that we would do. 

So the trial, you know, reinforces that evidence. That that reducing antipsychotic medication in contains with it a risk of relapse in and of itself. Certainly, one of the main, you know, ways that we can prevent relapse is continuing antipsychotic medication. (14.22)

I think that one of the things that that we also need to be really really cognisant of is there's also very clear evidence that for a certain percentage of people every relapse carries with it a risk of treatment non-response. So 15 or 20% of people will have one episode of psychosis and never have another episode. You know, 30% of people, 25% or 30%, people might develop treatment resistance - difficult to treat symptoms from the first episode, and everybody else is in the middle.

But this group in the middle. Every episode, a percentage of people don't recover fully. So there's risk in relapse. And it's just really important that in trials, and in clinical practice, we are very, very clear with those risks of relapse. So it's the risk of social functioning, maintaining work, maintaining relationships, but also the risk of, if we need to re-instigate treatment after a period off treatment that it may

And and so we just need to, you know, to take the evidence from the RADAR trial in this context, as as reinforcing the risks of of relapse. And certainly within 2 years not improving social functioning.

15:45

Joanna Moncrieff: So so I agree agree with Rachel that you know, relapse can have profound personal and social consequences. Some patients desperately want to avoid having a relapse and have had really bad experiences in the past, people who've had recurrent episodes. 

For some people it is less significant. So I think you know that that is an individual thing. There isn't evidence that relapse in itself causes a bad outcome. The majority of studies actually show that people who've had a relapse overall get back to the the level of functioning and and level of symptoms that they were at before. There's nothing about having the relapse that makes this sort of condition worse, other than obviously the social and psychological consequences of it. umm

And and actually, we've been looking at data from the radar study on this as well. And people who've had a relapse don't do any worse in terms of their social functioning and symptom scores at the end of follow up.

16:44

Helen Saul: Clearly there is ongoing debate about the specific harms of relapse. Rachel – what would you see as a research priority?

16:53

Rachel Upthegrove: The real question that that remains in my day to day practice, and other people working with with young people with first episode psychosis that we're seeing when they've had one episode. How do I know? How do I know if I'm going to be in the 20% or you know. And is there any way of improving our individual prediction for what we need to do for this patient? And that wasn’t what the RADAR study was designed to do. But I think to move ahead, we need to move from big trials of groups of people who perhaps already have a significant risk of relapse to understanding for for me, with all of my risk factors well, my experience. Is it a sensible thing? Is there any evidence at an individual patient level that actually dose reduction or dose discontinuation is a sensible thing to think about?

17:45

Helen Saul: Would you like to come in on this, Joanna?

17:49

Joanna Moncrieff: People have been trying to do this for decades and decades to identify people who might have a relapse, and people who might not, whether that’s during medication reduction or just in general and it's proved very, very difficult to identify. And again, we've been looking at this and not really found any any strong predictors, and there are almost no studies that replicate any relapse predictors. So I think we need to accept that we can’t predict who’s going to have a relapse or not, and we should just assume that everyone has the same chances and go on that.

18:30

Rachel Upthegrove: I would I would just come come back on that little bit because I agree with Joanna that we've we've not been able to, you know. Imp, you know, develop a tool that is clinically useful in terms of relapse prediction.

On the other hand, we just haven't had the access to the, to, to the size of the data we need to do accurate individual level prediction. So I think that's still a challenge for our field in terms of using the harmonized outcomes, having sort of a collaborative approach. So we can, we can have access to that data. And I wouldn't say that that's ever going to be something we should give up on trying to do until we've until we've given it as good as go as there have been another branches of medicine, or health, with the same volume and scale of data.

19:49

Helen Saul: Thank you Rachel. Joanna, where do you see the next steps in research?

18:03

Joanna Moncrieff: I'd like to see research that helps support patients to make informed decisions really about their treatment and be more involved in treatment decisions, particularly people who've had, who’ve had multiple episodes because I think first episode services are are quite good at in engaging patients in these sort of discussions. But the trouble is often patients who've had a number of episodes just get left, and everyone assumes they, you know, they should just stay on medication forever. And

and I think some of those people have really quite a poor quality of life, because of all the side effects that we've been talking about.

20:31

Helen Saul: Thank you. And Rachel, what would your takehome message be?

20:33

Rachel Upthegrove: I certainly agree with the last point in terms of making sure and advocating for best care for people with severe mental illness throughout their lives and throughout their, you know, the duration of their illness and and it comes back to from what this trial and other trials has shown is that because of that risk of relapse, you know, when we're reducing doses and and discontinuing doses of medication, this really does need to be an engaged individual discussion with a trusted prescriber who can help people weigh up their risk of relapse and work with them to find the optimum dose of medication. That that will help for symptoms and reduce the symptom side effect profile. 

And that's just good medicine. Right? That's what we should be doing with our patient and saying these medications work. They have side effects. What is the best thing that we can do with you? Because if we don't have that that conversation, patients don't take the advice that we're giving them, and they don't take the medication. They they will stop on their own and and stop suddenly and and stop, because that that interaction with your prescriber is, you know, is very important, and and that individual choice based on the evidence of risk. It is something that we should, we should be doing.

I think a a take home message is is that that relapse contains risk, and that varies by individual to individual. For some people that risk includes serious self-harm, suicidality, or risk of, you know, impact on their social functioning. And so for every individual patient, we need to have an informed conversation with people before they reduce medication or stop medication. But amongst everything else, work with patients. In a trusted relationship around their prescribing choices.

20:53

Helen Saul: Thank you Rachel. And a final word from you Joanna?

20:55

Joanna Moncrieff: I would go along with a lot of that. I do agree. There are some people who you know the the risks of the risks that may they may run if they have a relapse, are very high, but there are many people who where that's not such a consideration. who, I think, really deserve to be given the opportunity to try and reduce their medication, if that's what they want to, as long as they are informed of the risks, and have all the information, both from RADAR and from other evidence about the pros and cons of that process, and as long as they, you know, have have clinical support during it.

21:30

Helen Saul: Thank you Joanna Moncrieff and Rachel Upthegrove for joining me today and sharing your views.

This is an episode of the NIHR podcast, I'm Helen Saul, and thank you for listening. If you have thoughts or comments on this or any other episode, please contact us at evidence@nihr.ac.uk and do visit our website, which is evidence.nihr.ac.uk.

[MUSIC OUT].