Matthew Hotopf: Integrating Mental and Physical Health Care - Research, Training and Services (IMPARTS)

This content relates to the following topics:

Article information

  • Posted:Monday 21 September 2015

Speaking at our 2015 Increasing access to mental health care conference, Matthew Hotopf presents findings from the IMPARTS study, which supports clinical teams to integrate mental and physical health care within acute trusts.


What I’m going to talk about a little bit is the problem of integration between mental and physical health and why this is a challenge and it’s something which was picked up in the Chief Medical Officer’s report on public mental health which came out in 2014 but confusingly has the date 2013 on it. One of the chapters picked up on the difficulties of integration between mental and physical health and dealt with a number of different themes which arise when you’ve got physical health services providing physical health care, mental health services providing mental health care and little integration between the two.  So one of the problems is that commissioning tends to be done in silos between these mental health commissioning and physical health commissioning, there’s difficulties on the ground and throughout the whole system in a sense that informatics doesn’t link up so that it’s very difficult to actually … if you’re in a mental health setting you don’t necessarily know about the physical health needs of your patients and service users you’re seeing and the opposite way round in a physical health setting there’s often a lack of effective joint working, but there are some great examples and liaison psychiatry and health psychology which are around, but there is not a universal model for these sorts of services and there’s a problem about incentives in mental health services to ensure that the physical health agendas are considered.

So a whole range of different issues which were picked up in the report and the question then arises why does this matter and I think, in a way, one of the most stark examples of why it matters is that the life expectancy of people suffering from mental disorders … this is a study we published looking at our population in south east London, the life expectancy of people with significant mental health problems is between 10 and 17 years less than the rest of the general population in the UK which makes them more like the life expectancy of populations like Bangladesh or North Korea. So we have a massive health disparity which is a serious problem in the group of people in secondary health care.

However, it also applied in the community and this was a study we did looking at people in the 1946 birth cohorts, so these are people all born in 1946 followed over the next 65 years and looking at the mortality and what one can see is that at the age of 36 when they did a mental health questionnaire, it was a briefish questionnaire, you see that it predicts when you die. So the people on the blue line are those who have no symptoms, the people with the red line have some symptoms but wouldn’t come anywhere close to seeing a mental health practitioner, the people in the green line are what one might say are clinical cases, they met criteria for having a disorder, they were sufficiently severe and they were dying about twice as much as the people on the blue line.

You notice also the people in the yellow line and if you’re ever asked by an epidemiologist to complete a survey please do so because if you don’t you die younger than you should (laughter) it’s the curse of the epidemiologist, but the problem arises and this is really a problem of multi morbidity. So that looking at our GP databases we start seeing that there’s a huge excess of any physical health condition you care to mention amongst people with mental disorders.  So the red lines are people with severe mental disorders, the blue lines without these and you see a doubling of that effect no matter what the physical health problem is.

So you’ve got a multi morbid population and a population who die young, but from a really hardnosed health economics point of view one of the critical issues I think from the point of view of a general hospital is that the outcomes for the disease you’re actually looking after are worse if you have a mental health problem.

So this is looking at outcomes for rheumatoid arthritis according to whether you have different levels of severity of depression. So the score you’re looking at is the standard measure that rheumatologists use to look at rheumatoid arthritis and what you find is the people who are depressed at the start, which are the top line, and have significant depressive symptoms both start off in their treatment for rheumatoid arthritis as more severe but by the end of two years, very carefully monitored this is data from a clinical trial, they’ve all been very carefully monitored, at the end of two years their endpoint is equivalent to the starting point of the people who didn’t have depression.  So the people who have depression look like they’re people who aren’t responding to treatment.  What does that mean?

Well that means then that after two years of standard conventional care for rheumatoid arthritis you’re very likely to go on to a biologic cost £10,000 a year plus service cost. So you are suddenly seeing a group of people with mental disorders who may very well be unnecessarily being provided with a treatment which is hugely expensive, whereas my argument is that you should be trying to identify and look after the people and get their outcomes to look a bit closer to the people who don’t have depression at the start and you see this again and again in different settings.

This is a cohort of people with coronary heart disease and this is a single very short questionnaire at the start of a three year follow up and you’re looking at the costs to the system and you can see the costs for people who have mental disorders are higher and it’s the costs related to coronary heart disease are higher. So there’s a cost to the system of not looking after the mental health problem potentially.

So I work in south east London and a while back we formed an academic health science centre which consisted of two big acute physical health providers, Guy’s & St Thomas’ and King’s College Hospital Trusts and the South London and Maudsley, a mental health provider, and King’s College London and one of the things we wanted to do was try and bolster this mental physical health integration argument and the outcome … one of the outcomes from this was developing the IMPARTS program.

So what we did first off was to find out what was out there. We had these complex organisations, what is actually being provided?  If you’re in a physical health setting how much mental health is actually going in to it?  The answer was quite a lot but there wasn’t a clear service model, there were some pockets of excellence with very in-depth good quality liaison between mental health and physical health, there are other areas of physical health where you have absolutely no input from mental health at all.

We also looked at the training needs for staff in these acute Trusts. What did they think they need?  The answer was they really were very keen to learn more about how to look after people with mental health problems and how to offer care which is appropriate.

Then we asked also about what would physical health services want and we developed a toolkit to help them, which is what I’m going to talk about mainly.

There were one or two other things we did about trying to bring people together and having seminar series and we run a course called the IMPARTS course which is designed for people who don’t have any particular mental health training to be able to offer mental health. So some basic mental health skills in terms of looking after people with physical health problems.

So what does the package look like? Well it’s got five main components, informatics at the front end.  So one of the core things we’re trying to do is actually to measure things.  So we’re trying to routinely collect patient reported outcomes as people come into the services.  We devised care pathways often looking at what’s already available and trying just to use that rather than reinvent the wheel.  We have a training scheme for services, so we have our clinical psychologist going out to teams, finding out what they need and then offering training.  So she’s been training orthopaedic surgeons, training rheumatologists in some very simple basic mental health questions we’ll talk about a bit later.

We’ve also devised a package of self-help measures which deal with common problems which people with these different conditions experience. The point is that if you just give people a generic leaflet about depression but they’re actually presenting to the hospital with rheumatoid arthritis the leaflet will end up in the bin, they’re not interested because it’s not salient to them.  So you need to bring together learning from the disorder and the mental health problem into the same place.  So these have been adopted widely by a number of national charities.  Then we’ve got new treatments which we’re trying to develop off the back of this platform.

So when we started we had a social science researcher who was shadowing us who was looking at how you implement things in academic health science centres and at the end of this, having shadowed us, she fed back and she said it was brilliant because what we had was what was called a boundary object and in anthropological studies they’ve got this idea that if two systems are meeting you sometimes need something which focuses attention. The boundary object was an iPad and this suddenly changed the way in which the physical health providers looked at us because they saw rather than mental health being something which was a bit frightening or a bit wishy washy, they had this bit of kit which was exciting to them.  What we did with the bit of kit was say, “Okay, let’s use this to collect outcomes which matter and those outcomes are the things which you define.  So if you’re in rheumatology you want your patient reported outcomes which you’d normally want to have information on, but we’re going to make sure we always ask about depression, anxiety and one or to health behaviours like smoking, let’s make sure we always do that.”

So this is what the questionnaire looks like, people come into the waiting room, they complete this whilst they’re waiting to see the doctor, this is a rheumatology questionnaire called the HAQ and then this is a of couple of depression questionnaires from the PHQ2 which if you answer ‘more than half the days’ on either one of those it becomes the PHQ9 so you have more questions to fill in. That then gets uploaded in real time into the patient’s records.  So the physician then has some information about you, how you’re doing and you see just below the top there’s the result probable major depression, suggestions: referral to liaison psychiatry and so forth.

So this is an idea of trying to integrate some of that information which is collected and then you devise a referral pathway which makes use of what you’ve got already. So the aim of this was not to try and suddenly say, “Well actually we needed a vast army of mental health practitioners coming into these acute hospitals,” we knew we wouldn’t get that but we thought we could actually work with what was already there and for each of these services we effectively set the criteria according to what facilities are available but we never run a service, we never do screening in a setting where you can’t actually provide for the needs you identify.

We then generate data and this just shows you the kind of … I mean the whole point about this is to demonstrate that there’s quite a lot of data and you’re not going to be able to read any of the figures, but we have about 25 services now using the IMPARTS platform with varying success. Some of them are doing it absolutely religiously on every person who comes through the door and they have sequential data as they repeatedly visit, others it’s more haphazard.  But what you start finding is a pattern of problems according to the disorder you’re looking after.

So just taking this measure of depression which is simply symptoms of depression on a questionnaire and reassuringly what you find in some of the places we’ve implemented this are essentially healthy participants, they’re people who are coming to hospital for a check for a problem which is kind of a bit in the past, it’s been parked. So adults who as a child had congenital heart disease but are coming back for an assessment seem to have very low prevalence of mental health problems.  That’s what we’d want to see.  Similarly renal patients, renal transplant patients who have had a transplant and are doing well their levels are very low.  Then you see a group of people that I’ve called moderate prevalence between 5% and 15% who have an ongoing chronic condition and you see an increased prevalence of depression but it’s not massive. Then you see the high prevalence a group of people who I think their conditions are particularly characterised by pain, hidradenitis suppurativa this is a rather specialist clinic but it’s a very nasty skin condition which leaves you lesions all over your body which are painful and unsightly and rheumatology are dealing with a group of people with chronic inflammatory joints.

Interestingly trauma. So you have services looking after people who have fallen off a motorbike, broken their leg in three places, the leg is not healing, there’s a threat of amputation, they might well have a problem from the trauma, they might also have a problem before the trauma.  It might have been that they had a bit of a problem with alcohol which led to the accident.  They also had this difficulty and uncertainty following the trauma.  It’s not surprising that they’ve got quite high rates of mental health problems but the services traditionally have just not considered that at all and the orthopaedic surgeons in the service were incredibly keen to actually get this up and running and make the case and then start having clinical psychology and a bit of liaison psychiatry built into the service in the same way that they have physio.  So the idea is you build in integration rather than seeing this as something separate which involves going over the road to a clinic which says mental health on it.  It’s something you make integrated.

Then right at the end extreme prevalence pain clinic, no surprises there. A highly, highly morbid group of people with very severe difficulties which go on year in year out and very high rates of mental health problems.

The staff training I’ve mentioned already but we start off looking at where they’re at, what do they see as their training needs, what sort of encounters they find difficult, what sort of things do they want to be able to talk to people about. We give them some training which might include things like motivational interviewing, assessing risk, helping people just do a very simple mental health assessment.  Some of our clinical nurse specialists in different settings are getting some basic CBG skills in anxiety or depression management and we review this and come back to the services as we go along.
Then I said self-help materials, I won’t dwell on this, but what we’re pleased about is national charities for these different diseases are starting to adopt some of the materials we’ve put on, and they all go on our website, they’re freely available, you’re welcome to look.

It is also about research and one of the things we want to be able to do is interrogate the database so we have a governance oversight system which is service user led, which means that we will merge the screening data with the electronic patient record data to form an anonymised research database which can then give you some of the sorts of results you’ve seen and also that we’re looking at this as a means of trial recruitment. So you can try and get people into trials.  We’ve got one trial running at the moment which is looking at an eHealth intervention looking at distress in people with renal disease.  So the idea is it forms a platform for developing trials and just to give you a glimpse of some of the sorts of data we start getting out of this.  This was looking at IMPARTS data to see how long it took someone with rheumatoid arthritis to have a so called remission in terms of their rheumatoid arthritis according to whether they had depression or not, and the red line unfortunately are the people who score high on depression and anxiety and the blue line are people who didn’t.  So you see that people, as I started saying at the beginning of the talk, people with mental health problems have a poorer outcome than those without.

Finally we run a course and this is really designed to get someone like a clinical nurse specialist in one of these disease areas a higher level of training, it’s at a Master’s level, it deals with a number of topics which are very problem orientated, so dealing with things like confusion and agitation are very common in general hospital settings, medically unexplained physical symptoms and some of the sort of basic communication skills. One of the most popular parts of the course is actually dealing with conflict, dealing with difficult encounters which often people just lack the skills to be able to ask questions right.  There’s a little bit of promotional material there showing that people doing the course seem to like it and that’s all, thank you very much.


Add your comment