1 00:00:00,900 --> 00:00:08,520 Welcome to the Oxford University Psychiatry podcast series brought to you today by me, Daniel Moore and I have the pleasure of having Dr Matthew Brew, 2 00:00:08,520 --> 00:00:14,940 who's a senior clinical research fellow at the Department of Psychiatry here at Oxford and also a consultant psychiatrist. 3 00:00:14,940 --> 00:00:23,880 Good afternoon, Matthew. Thanks for inviting me. You have a really interesting research portfolio looking at, well, 4 00:00:23,880 --> 00:00:31,020 really how to predict potential onset of psychosis in people who are developing it perhaps early in their life. 5 00:00:31,020 --> 00:00:41,130 Could you tell us a little bit about that? So this was work that I became involved with back at EU Psychiatry in around 2001. 6 00:00:41,130 --> 00:00:48,150 We began developing the first episode of Intervention Services around the same time, the LEO service, early onset service. 7 00:00:48,150 --> 00:00:54,870 And I guess even back then, we were aware, even with very well-designed intervention services, 8 00:00:54,870 --> 00:01:00,880 people were still presenting quite late with psychosis, with a long duration of untreated psychosis. 9 00:01:00,880 --> 00:01:07,080 The Mental Health Act was being used fairly often and patients were having lost jobs. 10 00:01:07,080 --> 00:01:14,730 Relationships fall out of college or school. So we got very interested quite early on in what was called the high risk paradigm. 11 00:01:14,730 --> 00:01:19,460 That's the work of colleagues in Melbourne, such as Pat McGorry and Alison Young. 12 00:01:19,460 --> 00:01:25,140 And I guess put in a bit of context, there were sort of two ways to think about the high risk for psychosis at that time. 13 00:01:25,140 --> 00:01:32,040 There's a genetic high risk research programme that was led by Chief Johnson and colleagues in Edinburgh, 14 00:01:32,040 --> 00:01:37,170 which followed a cohort of people with firstly relatives with schizophrenia. 15 00:01:37,170 --> 00:01:44,540 And there was the work, Melbourne, which looked at another high risk group identified by symptoms rather than by genetic risk. 16 00:01:44,540 --> 00:01:52,650 Okay, so you've got a group of people who somehow a high risk of developing psychosis. 17 00:01:52,650 --> 00:01:55,200 How do you tell who's who's in the high risk group? 18 00:01:55,200 --> 00:02:00,940 You actually do genetic tests on them and then look at their symptoms or so what do the symptoms look like? 19 00:02:00,940 --> 00:02:04,980 Okay, so the way we did it. So I guess that is genetic. 20 00:02:04,980 --> 00:02:11,700 High risk groups are put to one side and the clinical high risk group is based upon symptoms of psychopathology. 21 00:02:11,700 --> 00:02:17,490 And there are a few different ways of doing that. As I mentioned, we were very much influenced by the Melbourne group. 22 00:02:17,490 --> 00:02:21,480 So we used a measure developed by Alison Young called the CALM's, 23 00:02:21,480 --> 00:02:26,990 which is the comprehensive assessment of the risk mental state, and asked develop that measure. 24 00:02:26,990 --> 00:02:33,960 What I recall initially by retrospective accounts of those with who had developed psychosis, 25 00:02:33,960 --> 00:02:39,780 so asking them to give them give her accounts of the earliest experiences they had. 26 00:02:39,780 --> 00:02:45,180 And in addition, she began reviewing the literature on the programme of phase of psychosis. 27 00:02:45,180 --> 00:02:53,940 Interestingly, that that work is area was triggered by concerns about a DSM three category of prodromal psychosis, and it's non-specific qualities. 28 00:02:53,940 --> 00:03:02,010 So Pat and Alison tried to sort of empirically examine the concept again, afresh, really, and that's what led to the detriment of the cars. 29 00:03:02,010 --> 00:03:03,430 There are other measures are very similar. 30 00:03:03,430 --> 00:03:10,200 So colleagues in Yale had a measure called the Six Insults, which also looks at psychotic symptoms in the early stage, 31 00:03:10,200 --> 00:03:19,680 and colleagues in Germany to develop measures looking at subtle symptoms of psychosis, which include more cognitive and negative features. 32 00:03:19,680 --> 00:03:27,120 But most of UK, Australian and North American psychiatry tends to focus on on using the council. 33 00:03:27,120 --> 00:03:33,060 The Sipson stops and tends to focus on what we would call attenuating positive symptoms. 34 00:03:33,060 --> 00:03:40,830 So mild symptoms that look a bit like hallucinations, delusions, things you focus on. 35 00:03:40,830 --> 00:03:46,980 So, yeah, thanks. What I'm interested in from reading your research, Matthew, 36 00:03:46,980 --> 00:03:55,500 is that you want to have a whole lot of people who have got these sorts of attenuated positive symptoms a little bit like a hallucination, 37 00:03:55,500 --> 00:04:05,190 a little bit like a delusion, not quite either of those yet experiencing some unusual sort of features in their life. 38 00:04:05,190 --> 00:04:07,560 Perhaps that's not going too well. 39 00:04:07,560 --> 00:04:18,090 But it seems to me that it's very difficult to predict from quite large groups sometimes who is going to go on to develop schizophrenia. 40 00:04:18,090 --> 00:04:27,510 And a lot of your work is about how to predict from that large group who is going to develop perhaps a long term severe psychotic illness. 41 00:04:27,510 --> 00:04:32,730 So what have we found out, I guess, is a few things to say about that. 42 00:04:32,730 --> 00:04:42,260 I think you've learnt quite a lot. I think the first thing to say is that the original papers from Yemen, McGorry, 43 00:04:42,260 --> 00:04:50,180 suggested that people who met these criteria, about 40 percent of people develop psychosis within six to 12 months. 44 00:04:50,180 --> 00:04:55,010 That's probably turned out to be not quite so high as that as years have gone by. 45 00:04:55,010 --> 00:05:01,520 And what we call the transition data has sort of become less strong. 46 00:05:01,520 --> 00:05:10,850 But saying that a recent analysis by our polling, I think 2012 still was saying that 22 to 35 percent people will transition within two years. 47 00:05:10,850 --> 00:05:16,940 So just the measure, as is, is still not so bad at predicting precisely the psychosis within within a couple of years. 48 00:05:16,940 --> 00:05:21,710 One in three people will develop psychotic illness. But yes, you're right. 49 00:05:21,710 --> 00:05:26,510 I mean, the reasons why some of the issues I catch myself are very interested in improving 50 00:05:26,510 --> 00:05:31,670 their predictive powers and I guess also what's known as a false positive. 51 00:05:31,670 --> 00:05:37,190 So misidentifying those as being at risk when that whether or not I can see the big 52 00:05:37,190 --> 00:05:43,280 area of psychiatry that this research meshes with is the continuum of psychosis. 53 00:05:43,280 --> 00:05:48,740 So this is the idea that people in the normal population who aren't impaired, aren't seeking, 54 00:05:48,740 --> 00:05:59,960 also may have experiences such as ideas of reference, the occasional auditory experience to voice hearing, and they aren't cases or death cases. 55 00:05:59,960 --> 00:06:05,050 So you're quite right. And your concern that those you might see in an at risk clinic could be. 56 00:06:05,050 --> 00:06:08,210 Remember the general population in their enhanced risk at all. 57 00:06:08,210 --> 00:06:13,790 So we're very much aware of trying to improve the predictive powers of these questionnaires and measures, 58 00:06:13,790 --> 00:06:22,910 given this wider epidemiological research in those people who are non help seeking, non distressed and have psychotic type experiences. 59 00:06:22,910 --> 00:06:29,660 So you've looked at some really quite subtle psychopathological changes in these individuals 60 00:06:29,660 --> 00:06:36,080 to see if they can help predict over and above these tests you might transition. 61 00:06:36,080 --> 00:06:40,340 And what have you found out about that? Well, I guess variety of things. 62 00:06:40,340 --> 00:06:47,170 I mean, we were my kind of main role was I led on some serious evaluation delusion formations, 63 00:06:47,170 --> 00:06:52,940 and therefore I work in other colleagues such as Oliver has spoken about recently, led on top of that. 64 00:06:52,940 --> 00:07:01,280 So we're very interested in. So neuropsychological measures and fMRI correlates predicting psychosis. 65 00:07:01,280 --> 00:07:11,780 I guess it's still at the stage where if we don't know how these variables interact to help, you have a clinical tool to be used for defence. 66 00:07:11,780 --> 00:07:15,500 I guess that's a fair thing to say, but I guess there's some things we do now. 67 00:07:15,500 --> 00:07:19,940 So I guess certainly demonstrated quite convincingly that's trying to take me, 68 00:07:19,940 --> 00:07:27,770 for example, is already elevated in those who might be at risk of psychosis. And the more elevated as people enter the first episode, 69 00:07:27,770 --> 00:07:34,830 spatial working memory and so verbal working memory seems to be quite a key feature replicated in several research groups Ireland, 70 00:07:34,830 --> 00:07:41,090 London, Germany, that seem to predict onset psychosis. How do you test of working memory in that study? 71 00:07:41,090 --> 00:07:45,110 So what I did was a simple an and back task. 72 00:07:45,110 --> 00:07:54,380 So I'm not one of my task looking at people's ability to to relate a stimuli to one that occurred, a certain number of stimuli before. 73 00:07:54,380 --> 00:07:59,480 So what you're saying is people are unable to or less good at remembering what's being said 74 00:07:59,480 --> 00:08:04,580 to them before having been interrupted with other other things that have been said to them. 75 00:08:04,580 --> 00:08:13,610 And and that that's sort of perhaps that disability or that lack of ability or reduced abilities is potentially predictive of of transition. 76 00:08:13,610 --> 00:08:20,850 I think so. So I think certainly that kind of feature in combination with the clinical signs which would make the slightly higher elevated risk, 77 00:08:20,850 --> 00:08:29,380 are the things that are important anxiety and depression. So this group, although they don't have a of psychosis, definitely they're at risk. 78 00:08:29,380 --> 00:08:37,700 For one, they will have very likely to have a comorbid DSM diagnosis for anxiety disorder. 79 00:08:37,700 --> 00:08:40,430 And again, there's some evidence to suggest I mean, all sorts of reasons, 80 00:08:40,430 --> 00:08:45,260 probably because of psychosis, whereas more biological models of these things can increase. 81 00:08:45,260 --> 00:08:51,110 Risk treatment also can can can be helpful. Other things you've a work about won't matter. 82 00:08:51,110 --> 00:09:01,370 Changes, productivity changes, structural MRI changes, as well as more lifestyle things as a sleep disturbance, substance misuse. 83 00:09:01,370 --> 00:09:11,060 So there's a kind of wide variety of risk factors that are potentially intervene in in this kind of group of people who are high risk. 84 00:09:11,060 --> 00:09:16,310 And I guess the challenge clinically is trying to use the least dangerous interventions. 85 00:09:16,310 --> 00:09:29,270 Yes, that was my next question, because, OK, so you've got this group of people who are at an increased risk of transitioning into psychosis. 86 00:09:29,270 --> 00:09:35,980 How do you feel ethically about potentially giving somebody perhaps a low dose of an antipsychotic in that situation? 87 00:09:35,980 --> 00:09:40,780 To potentially reduce the risk because you might be giving somebody an antipsychotic 88 00:09:40,780 --> 00:09:45,610 for no reason at a time in their life where they might be affected by that so. 89 00:09:45,610 --> 00:09:49,400 Well, how do you cope with that? That's a really important question. 90 00:09:49,400 --> 00:09:52,480 I think when we do debated quite a lot in the early days. 91 00:09:52,480 --> 00:10:00,550 So I think empirically, the economy supports at present that it a treatment for at risk to abort the onset of psychosis. 92 00:10:00,550 --> 00:10:06,700 And in practise, some of these patients will have an additional haptics, I think, by far the minority. 93 00:10:06,700 --> 00:10:11,240 I think what I recall when I worked in our and our service in South London, it may have been, you know, 94 00:10:11,240 --> 00:10:18,970 fewer than 10 percent are the reasons why you might consider it would be because the the patient may have very distressing psychotic experiences, 95 00:10:18,970 --> 00:10:28,730 may be very high risk of self-harm or violence, for example. So, yes, I would generally one would be to avoid any psychotic medicine in this group. 96 00:10:28,730 --> 00:10:31,760 You could try out alternative treatments. 97 00:10:31,760 --> 00:10:40,270 So I think it's not you might want to consider the use of a view of depression or anxiety disorders as an SSRI or hypnotic, 98 00:10:40,270 --> 00:10:45,220 for example, for sleep problems. But I guess the mainstay would be a psychosocial intervention. 99 00:10:45,220 --> 00:10:51,290 So psychological treatment and, you know, family work and case management could be very, 100 00:10:51,290 --> 00:10:58,330 very powerful because what you don't want to do is to give a treatment to inappropriate somebody who has no risk whatsoever. 101 00:10:58,330 --> 00:11:03,470 Ethically, is really interesting is that the clinics are hoping to develop with this thing 102 00:11:03,470 --> 00:11:08,920 a the department is sort of embedded ethical piece of work with the service 103 00:11:08,920 --> 00:11:18,820 in the trust to get it kind of slightly difficult interaction directly where 104 00:11:18,820 --> 00:11:24,190 what you try and do therapeutically is to normalise the abnormal experience. 105 00:11:24,190 --> 00:11:27,750 So quite often when you see a patient, who would you think might be high risk? 106 00:11:27,750 --> 00:11:34,180 You trying to do is to alleviate the anxiety because the symptoms are to normalise them. 107 00:11:34,180 --> 00:11:36,790 It's quite, you know, sort of saying what people do hear voices now and again, 108 00:11:36,790 --> 00:11:43,120 you may feel feel like a bit of paranoia and look at the emotional charge around to try to lessen that. 109 00:11:43,120 --> 00:11:46,510 But at the same time, it's given that normalising message also getting the message. 110 00:11:46,510 --> 00:11:54,610 But can I still see you every month for monitoring? So that kind of difficult conversation is hard. 111 00:11:54,610 --> 00:12:02,770 So there's a bit of controversy. You may be aware about DSM five and incorporating the attenuated psychosis within that. 112 00:12:02,770 --> 00:12:10,190 The illness seems quite interested in young people's experience of early psychosis services and how the at risk category is used. 113 00:12:10,190 --> 00:12:15,670 So we're hoping to embed some quite detailed qualitative research into the clinical services in the future, 114 00:12:15,670 --> 00:12:22,540 because that very reason how young people might feel about these diagnoses and terms and acknowledge probability. 115 00:12:22,540 --> 00:12:32,560 And there's discussions within clinical practise. So when we think about early intervention today in the UK, 116 00:12:32,560 --> 00:12:42,610 do you think we've made strides forward in knowing who to who to treat and how to treat them? 117 00:12:42,610 --> 00:12:53,940 Do you think progress has been made? I think so progress is is bumpy and sometimes goes backwards to get forward. 118 00:12:53,940 --> 00:12:59,160 That makes sense. I guess what I what I'm aware of is how naive we were. 119 00:12:59,160 --> 00:13:04,140 I look back to when we set up protocol services 14 years ago, 120 00:13:04,140 --> 00:13:12,780 we had this very simplistic view that having this adverse mental state impact on psychosis and we think much more than that. 121 00:13:12,780 --> 00:13:19,170 So I guess one of the things that has come out of this research is the adverse mental state described 122 00:13:19,170 --> 00:13:26,220 by you and colleagues might be more a marker of general risk for psychiatric illness across the board. 123 00:13:26,220 --> 00:13:33,870 So I think where people are getting interested in is is more general diagnostic view of adolescent young, young adult health. 124 00:13:33,870 --> 00:13:38,440 And do these markers predict problems down the line? Americans may end up being psychotic illness. 125 00:13:38,440 --> 00:13:43,170 They may end up being personality difficulties, mood disorder, et cetera. 126 00:13:43,170 --> 00:13:46,230 So I guess you got a bit more nuanced about that. 127 00:13:46,230 --> 00:13:57,210 And that's why in parallel with people being less siloed about diagnostic types and clarity, he generally has been it's been great. 128 00:13:57,210 --> 00:14:03,600 I think it's been a sort of trailblazer for high community care can be delivered an optimistic way. 129 00:14:03,600 --> 00:14:11,950 Really focussing on on outcomes are important to young people such as, you know, jobs, education, relationships. 130 00:14:11,950 --> 00:14:18,420 I think I'd be very good at that kind of ideological approach. Psychiatry, and I'm one of them is Ed Bradley. 131 00:14:18,420 --> 00:14:24,310 I'm very compelled by the argument that he is just doing psychiatry well and it's nothing special. 132 00:14:24,310 --> 00:14:31,320 What I do could do it. Any kind of patient group with any to any diagnostic group, it remains very cost effective way of delivering care. 133 00:14:31,320 --> 00:14:40,320 I guess what we have seen the last several years has been a kind of pulling back into the service provision of eEye because of economic cuts. 134 00:14:40,320 --> 00:14:48,730 But I certainly feel in this, trust me, more nationally that's beginning to be reversed slightly and go optimistic again, which is good. 135 00:14:48,730 --> 00:14:53,283 Matthew, it's been great to speak to you today. Thank you for your time and a pleasure. I enjoyed your time with.