Tuesday, January 19, 2016

Hansard: 18 Jan 2016


Cambridgeshire CCG and UnitingCare Partnership


Motion made, and Question proposed, That this House do now adjourn.—(Charlie Elphicke.)

10.1 pm


I thank you, Mr Speaker, for allowing the House the opportunity to consider the extraordinary collapse, after just eight months, of one of the biggest tendering exercises conducted in the NHS: the £800 million contract between UnitingCare Partnership and Cambridgeshire and Peterborough clinical commissioning group.

Two weeks ago, I spent a morning out with an ambulance crew working from the ambulance station on the Addenbrooke’s site outside Cambridge. I saw the NHS at its best: top-quality care, provided quickly; people in pain and discomfort treated with dignity and respect; a swift, seamless transfer into hospital; and fantastic, caring, committed staff, on ambulances and in hospital. It was our NHS at its best and we should be proud of it.

When health leaders in Cambridgeshire said they wanted to create an integrated service for older people that would focus on prevention, it was a worthy aim, albeit polluted by the need for a competitive tender, insisted upon by the Health and Social Care Act 2012. When the contract was finally signed with NHS providers, it should have been the start of a new way to provide care, so what went wrong? That is what I want to quiz the Minister on tonight, because the failure of this contract matters way beyond Cambridgeshire, and it has rightly attracted national attention. A recent editorial from the Health Service Journal said:

“When a five year contract of this size and this importance to some of the most vulnerable people in society fails, it is not enough to shrug and walk away. As NHS England develops capitated, outcomes based contracts for national rollout, it is important to understand and explain what has gone wrong in Cambridgeshire so the lessons can benefit the health service as a whole.”

It is right.

This is a long and complicated story, which some of us have followed closely over many years. You will be glad to hear, Mr Speaker, that I will give an abbreviated account, but I must pay tribute to some of the campaigners who spent many, many months at meetings across Cambridgeshire questioning and challenging: my friend and colleague, Huw Jones; Steve Sweeney, then of the GMB; Jo Rust; Tracey Lambert and Martin Booth from Unison; and many, many others. We always knew something was not right and, sadly, we were proved to be correct.

I believe the story really begins back in 2012, when the future of Cambridgeshire Community Services NHS Trust, which had itself only a few years earlier been separated from the predecessor to the CCG, was thrown into doubt when it failed in its bid to become a foundation trust. Under Government guidance at the time, through the Trust Development Authority, trusts that were not foundation trusts faced being wound up. That was a foolish policy and, as so often, it was later rescinded, but given that this was happening when the infamous 2012 Act was under massive discussion, the real possibility was raised that many care staff would be transferred to private providers. That did not happen, because Cambridgeshire County Council, which many staff had originally worked for, took many of them back, but the consequence was a disintegration of services—the very opposite of what was needed. Integrated teams were ripped apart—an act of vandalism that set care back. The Trust Development Authority, the body overseeing this early-stage debacle, remains the line of accountability for NHS trusts. Those trusts are now being merged, in their regulatory function if not statutorily, into Monitor, which is yet another executive non-departmental public body of the Department of Health.

It is worth noting in passing that, since the time it was denied foundation trust status, causing the disintegration of care, Cambridgeshire Community Services has gone on to be named as the best community trust to work for by the Health Service Journal, and is now doing very well, albeit by working mainly with others outside Cambridgeshire.

Against that backdrop, and because of the 2012 Act taken through Parliament by the then local MP and Secretary of State for Health, the Cambridgeshire and Peterborough clinical commissioning group, in wanting to move to a new model of outcomes-based care, was forced in 2013 to put health services for older people out to tender. The process attracted national attention and was very controversial locally, mainly because of its focus on trying to attract private providers. Shrouded in commercial confidentiality, rumours abounded. Many organisations expressed interest including Virgin Care, Care UK, Circle, Capita and UnitedHealth and more.

Over many months, campaigners and I sat through numerous CCG board meetings and what were described as public consultation meetings where we were assured that all was well and that the many concerns we raised were unjustified. It was announced that the three final bidders for the contract to lead the services were Care for Life, UnitingCare Partnership and Virgin Care. Eventually, in October 2014, it was announced that the five-year outsourcing contract to run older people’s healthcare and adult community care was to be awarded to the UnitingCare Partnership, which was not a private bidder but an NHS consortium of Cambridgeshire and Peterborough NHS Foundation Trust and Cambridge University Hospitals NHS Foundation Trust. The five-year contract was worth £800 million and covered: urgent care for adults aged 65 and over, including in-patient and A&E services; mental health services for people aged 65 and over; adult community services for people aged 18 and over, including district nursing and rehabilitation services; and health services to support the care of people aged 65 and over. It is one of the biggest contracts the NHS has ever tendered.

The partnership started delivering services last April, and regular updates were issued outlining how the new services would work. We now know that, behind the scenes, much wrangling over costs was going on, but that was withheld from public gaze. Then came the bombshell. After eight months, and just one month of the new system operating fully, a joint statement was issued by Cambridgeshire and Peterborough CCG and UnitingCare. Not much detail was given, other than an assurance that services would continue, that patients should be reassured but also that the provider and the commissioner had agreed that

“the current arrangement is no longer financially sustainable.”

The contracts that had been established were to be honoured by the CCG, and patients and carers were promised that services would go on as usual and would not be disrupted. It was also said that it would try to retain the new model of integrated service delivery.

Let us briefly review some of the damage. We still do not really know how much the procurement process cost, but on the public side it was certainly millions, and probably at least as much again for private providers working up failed bids—doubtless to be recouped from somewhere else in the NHS later.

Let me turn now to the impact on staff. Back when Cambridgeshire Community Services failed in its bid for foundation status, a transition steering group was established to oversee the future of thousands of its staff. Teams were ripped apart and, with the new contract, more than 2,000 staff were transferred to Cambridgeshire and Peterborough Foundation Trust and Cambridgeshire County Council.

That was a massive task for the Cambridgeshire Community Services Trust, distracting it from other work. There was huge uncertainty and stress for staff over the future of their jobs. Throughout the entire process, across the NHS in Cambridgeshire, senior managers and local health service leaders were spending large amounts of time on all of this. Was it really time well spent, when last year we saw so many major hospitals repeatedly in crisis?

I have said nothing so far about the strategic projects team. What was its role? Many would ask, “Who are they?” To many who follow these things, the STP is, in effect, the pro-privatisation arm of NHS England, and it played a key role throughout this process. Its website tells us that the team specialises in competitive procurement, the re-design of patient pathways via an integrated care model, change management, service reconfiguration and integration, trust development and culture change.

The STP is part of NHS England. We are told by Lord Prior of Brampton, the Minister responsible for NHS productivity, that in its investigation into the collapse of the contract, NHS England will examine the strategic projects team’s role, and will also consider how similar contracts will be managed and assured in future. So it is NHS England that will investigate its own strategic projects team—a hopeless conflict of interest. That is not good enough. We need a genuinely independent and transparent review.

People are right to ask questions about the strategic projects team. Its list of interventions reads like a roll-call of recent NHS disasters: not just this project, but the private hospital saga at Hinchingbrooke in Cambridgeshire, and the failed tendering process for the George Eliot hospital in Warwickshire, among others. Its website leaves one in no doubt about its leading role in the Cambridgeshire older people’s tendering process. It says:

“SPT delivered an open procurement process on behalf of the CCG”.

It delivered—no room for doubt there. On 8 October 2014, when UnitingCare was announced as the preferred bidder, the SPT was again trumpeting its key role. On 12 November, when it was announced that UnitingCare would operate as a limited liability company, the SPT was there again, and it is worth quoting from the press release still on its website to get a sense of just how central it was:

“Andrew Macpherson, Managing Director of the Strategic Projects Team that managed the procurement on behalf of the CCG said: ‘The Strategic Projects Team are once again proud to have supported courageous leadership in the NHS.’”

The SPT may call it courageous; others might describe it rather differently.

Let us be clear—it is the SPT, very much part of NHS England, that has been calling the shots. On the decision to set up UnitingCare as a limited liability company, it was approved by Monitor, the strategic projects team and NHS England at the time, yet all knew that that meant that there would be no room for flexibility, and no room for losses in years 1 and 2, when the model explicitly expected extra cost at the beginning, in expectation of savings later. Looked at from the outside, it is hard to see how that could ever have worked, so why did Monitor, the special projects team and NHS England give the go-ahead? Did none of them spot the potential VAT problems introduced by a limited liability partnership?

Having given a brief outline, let me come to the further questions that I hope the Minister will be able to help us with. First, on the flurry of investigations being announced, although it is right that individual organisations will want to look at their role, there is a danger not only of duplication but of exactly the kind of fragmentation that has caused such problems already. Given the conflict of interest within NHS England that I have already described, would it not be better to have a genuinely independent review carried out by the National Audit Office—a review in which we could all have confidence?

The Minister should surely be able to tell us about the role played by his Department and by Ministers, at two key moments in particular. When it was clear in October/November 2014 that there was insufficient information on costs to agree a final contract, why was the process not delayed until that had been sorted out? Did Ministers know? What exactly was the rush to achieve implementation for April 2015, coincidentally perhaps just weeks ahead of the general election? And what role did Ministers play in the final decision to end the contract in December 2015? There were clearly detailed discussions going on with NHS England and Monitor about how much was needed to keep the contract running. The figure seems to have been about £10 million, a lot of money, but given that killing the contract may well have cost more, it was certainly worth considering. What was the ministerial involvement at that point? Were Ministers consulted? Who made the decision to let the contract collapse?

Looking forward, which is what matters most, patients have been assured that services will be maintained. That may well be true in the short term, but what next? Will the outcomes model be pursued, just with UnitingCare taken out of the equation? Does the CCG have the capacity, and if it does why did we go through that ludicrous tendering process? This has been a sorry saga. It seems that everyone agrees that our NHS and our care services need to be integrated, but years of fragmentation make it extremely hard to achieve.

This was a well intentioned attempt to deal with the perverse incentives that shackle our health and care services, and we need to find out what went wrong. We have dedicated, hard-working staff who want to provide the best care possible to our citizens. We need to find a way of making it possible for them to do that. In my view that means an end to contractualised market models, and a move to a genuinely integrated public system, an NHS solution based not on competition but on collaboration; an NHS solution that patients desperately need, and that staff, I am sure, would cheer.

10.15 pm


I congratulate the hon. Member for Cambridge (Daniel Zeichner) on securing the debate and thank the Members who are present, including my hon. and learned Friend the Member for North East Hertfordshire (Sir Oliver Heald)—I know that he has an interest in the matter—for attending. I also pay tribute to all those working on the frontline in the NHS in East Anglia, particularly at this time of year, when pressures are at their greatest.

As the hon. Gentleman has described, the contract between Cambridgeshire and Peterborough CCG and UnitingCare Partnership has very recently been terminated. I need to say right away that NHS England has launched an investigation into the circumstances surrounding the contract. Its terms of reference are to establish, from a commissioner perspective, the key facts and root causes behind the collapse of the contract in order to draw out any recommendations and lessons to be learned. I understand that the CCG is also undertaking a review, as is right and proper.

We should let the NHS complete that process. I hope that nothing I say today can be taken as an assumption that Ministers have in any way prejudged the outcome of that process. Clearly there are different views about what has happened, and I want to wait for the reports of the reviews before deciding what, if anything, needs to be done, either by the NHS or by the Government. Once the reports are published, Ministers will be briefed on their conclusions. I am happy to invite the hon. Gentleman to that meeting, although I cannot say today exactly when it will take place. I know that he is in regular contact with his local NHS, and I encourage him to keep that up.

The core scope of services in the contract with UnitingCare was acute unplanned hospital care for older people—those 65 and over—older people’s mental health services, older people and adult community services and a range of supporting voluntary sector services. The underlying principle was to create an integrated care pathway between all these services. The UnitingCare service model was designed by local clinicians during the procurement process and had a high degree of local health and social care support. Its detail and assumptions were subsequently ratified by two independent auditors. It was designed to: join up services around the patient and reduce service fragmentation; to focus on better outcomes for patients and carers, rather than activity levels; to invest in out-of-hospital services in order to better address the needs of a rapidly ageing and growing population; and to deliver £170 million of savings to the local health economy by 2020 by reducing inappropriate emergency admissions to hospital and inappropriate A&E attendances.

UnitingCare began introducing those new services with an investment of £5.4 million over the first six months of the financial year. They included a number of important local improvements, such as: care based around neighbourhoods, with 17 neighbourhood teams working closely with GPs; access to specialist services, with neighbourhood teams and the support of four integrated care teams to offer more specialist care; a 24/7 helpline, called OneCall; urgent care and support, with joint emergency teams to assess and treat people most at risk of admission to hospital; health and wellbeing, with voluntary organisations working together; a single view of the patient record, called OneView, providing professionals with a summary of all information about a person’s health; and a health analytics service to target interventions at those most at risk of admission.

To achieve those improvements, a contract was needed between the provider and the CCG. The main components of the contract were: a new framework for improving outcomes; a new contracting approach to align incentives in a better way; a five-year contract term; and a new lead provider, UnitingCare. It was therefore a high-value contract; it had a total value of around £800 million. Having taken legal advice, the CCG went to open procurement, using a standard three-stage process—pre-qualification, an invitation to submit outline solutions, and an invitation to submit final solutions. The CCG prospectus set out the CCG budget and the evaluation criteria. It was a contract entered into in good faith. This included submitting bids within the CCG budget. The CCG budget incorporated forecast population growth, an acuity factor, and QIPP—quality, innovation, productivity and prevention—savings for each year.

In 2014, there was in some quarters, as the hon. Gentleman said, concern that the process was “stealth privatisation”. Clearly no one, on any objective criteria, would agree that that was the case; it was merely, as he said, a service reconfiguration placed with a not-for-profit company set up by local health providers. The boards of Cambridge University Hospitals NHS Foundation Trust and Cambridgeshire and Peterborough NHS Foundation Trust held the firm belief that only by introducing radical change led by the NHS would the local health economy under the CCG become viable for patients, staff and the respective trusts across the region. For that reason, they decided to submit a joint bid and, following commercial and legal advice, opted to create a limited liability partnership to fulfil the role of prime vendor, as required by the CCG.

The CUHFT and CPFT consortium was appointed as preferred bidder at the end of September 2014. In October, it formed UnitingCare LLP to hold the contract. The strategic projects team was appointed as procurement adviser to the CCG through a competitive process and its role was to manage the procurement process. The strategic projects team is a specialist unit hosted by the Arden and Greater East Midlands commissioning support unit, which has substantial experience in managing complex procurements. The CCG also appointed legal advisers, Wragge Lawrence Graham, and financial advisers, Deloitte, to support the procurement process.

Much information about the costs of the current services, staffing details and timescales could not be provided by the CCG to UnitingCare until it was at preferred bidder stage. As a result, UnitingCare’s bid was heavily caveated and based on assumptions. To illustrate this point, at the time of preferred bidder award status, there were 71 outstanding clarification questions from the procurement process. The contract signed between the CCG and UnitingCare also included several protection clauses to be utilised in the event of the financial distress of either party. Subsequent to contract signature, additional clauses were agreed that allowed for the rapid exit of the contract in the event of the financial destabilisation of either party. With these protections in place, trust boards, the CCG and Monitor allowed the contract to be signed in November 2014 and for the necessary mobilisation activities to facilitate service commencement on 1 April 2015.

There were clear improvements in patient care. For example, in November 2015 emergency admissions for over-65s reduced by just short of 8% compared with the previous year and by 9% when taking into account population growth; admissions of more than two days’ duration for people over the age of 65 reduced by 14%; and A&E attendance reduced by 3.2% when taking into account population growth. However, in December the contract was terminated by mutual agreement—


As my hon. Friend says, there were advantages to this project and it produced good outcomes. If it is a good concept, will the Department of Health support the services that so need to be provided?


My hon. and learned Friend makes an excellent point. The service is currently being continued, albeit by the CCG rather than through the company that was created for the purpose. As she says, the reforms that were put in place were the right reforms. Indeed, they were led by local clinicians and designed with that in mind.


In Royston we have the Royston NHS and social care hub, which will include beds as well as other services. Does my hon. Friend agree that there is no reason at this stage for people to become anxious that the difficulty with the contract will lead to any change in the quality of services that are planned for the future?


That is right. As both questions have highlighted, the change in the care pathway is being pursued by the CCG and there is no reason for patients—the users of the system—to fear any dramatic change to the service. The remaining issue is the residual issue of how the contract came to be put in place. The dispute between the parties is about their different conceptions of the financial and contractual situation. I do not want to prejudge the investigations, but the service reforms will continue.

The final decision to terminate was taken after extensive discussions between the CCG, UnitingCare, Cambridge University Hospitals NHS Foundation Trust, Cambridge and Peterborough NHS Foundation Trust, NHS England and Monitor. Prior to escalation to NHS England and Monitor, the CCG, CUH and CPFT worked hard to try to reach a resolution locally.


Could the Minister enlighten me on the role played by Ministers in that final decision? Did they know it was happening? Who ultimately terminated the contract?


As I will come on to say, due process was followed in the correct way. One of the reasons for listing all these acronyms is so that the hon. Gentleman can be reassured that the right bodies carried out their due diligence. I do not believe that there was any reason for Ministers to be concerned at any point until the dispute between the parties became clear. Indeed, the reforms had been generated locally by clinicians and an accountable CCG led by clinicians. As the questions I have been asked have illustrated, the reforms were and remain very sensible. This is a better care pathway, with improved outcomes.

The issue is contractual and relates to a dispute between the parties about liabilities in the contract. As I have said, I do not want to prejudge the ongoing investigations, the point of which is to work out what should have been done differently. I can absolutely reassure the hon. Gentleman and the House that we are hungry to learn any lessons from that commissioning experience. We need novel commissioning. We need commissioners around the country to look into different ways of commissioning the reforms to our integration of health and care, and lessons need to be learned when it goes wrong. I emphasise that this was a contract between the parties. As I have said, the Department is looking forward to the reviews and wants to hear the lessons that others can learn.

The CCG has now taken over all relevant contracts with providers that were previously held by UnitingCare, to ensure that there is no service disruption to patients and carers. In addition, the CCG and CPFT, which employ the majority of the affected staff, have worked closely together to ensure that frontline staff are clear that, while the contractual model has now changed, the service model remains in place.

Of course, I agree with hon. Members that it is a matter of extreme concern that the new arrangements lasted barely six months. That is not ideal. We need to work out how the parties got it wrong and what mistakes were made. There are questions for the reviews to address. For instance, there is the question of why, given full procurement and assurance of the process, the result fell so far short in practice, along with other associated questions.

To describe modern commissioning as back-door privatisation is wilfully to misrepresent what is going on. These are clinician-led improvements to the care pathways, and I do not believe that most service users would consider it privatisation. We are talking about two public sector organisations coming together to form a company for the purposes of jointly commissioning care pathway innovation put together by clinicians in the local CCG. If Labour considers that privatisation, it has a serious problem, because most people would consider it enlightened commissioning for modern care pathways. This is a contract issue. The parties to the contract did not get it right, and we are keen to understand why and what can be done to make sure it does not happen again. I want those answers as much as the hon. Gentleman, and I repeat my invitation, to him and other hon. Members with an interest, to meet in due course to learn the lessons and make sure that the benefits of commissioning for integration go ahead without the contractual errors that have bedevilled this project.

Question put and agreed to.

10.30 pm

House adjourned.


Monday, January 18, 2016

Management of first-episode psychosis


Elizabeth Spencer, Max Birchwood, Dermot McGovern
Over recent years early intervention in psychosis has fired the imagination of clinicians and researchers, following the publication of several studies linking a long duration of psychosis prior to receiving treatment with a poor treatment outcome: a relationship that has been cogently argued to be independent of illness factors (e.g. Scully et al, 1997). Most influential was Wyatt's (1991) review of 22 studies, in which relatively similar groups of patients suffering from schizophrenia were, or were not, given antipsychotic medication early in the course of their illness. This pointed towards an improved long-term course in schizophrenia with early treatment.
At the same time, evidence was emerging that the early phase following the onset of a first psychotic illness could be conceived of as a critical period, influencing the long-term course of the illness. During this period individuals and their families may also develop adverse psychological reactions to psychosis and the circumstances of its management. It was therefore proposed that timely and effective intervention at this stage might alter the subsequent course of the illness (Birchwood et al, 1998). The early intervention concept has even been extended, with some success, in an attempt to identify people experiencing the prodrome of a first episode of psychosis (Yung et al, 1998).
These developments resulted in the establishment of specialised services for the management of first-episode psychosis, notably in Australia and Scandinavia. In the UK, the National Service Framework for Mental Health stresses the necessity of prompt assessment of young people at the first sign of a psychotic illness, in light of the “growing evidence that early assessment and treatment can reduce levels of morbidity” (Department of Health, 1999: p. 44). The Government's plan for the National Health Service (NHS) specifically targets the funding of 50 early intervention services by 2004.
The primary aims in the management of first-episode psychosis are detailed in Box 1. In this paper, we will describe best-practice principles for the management of first-episode psychosis, illustrating them with reference to their application in our service, the Birmingham Early Intervention Service. This is the UK's first dedicated service for young people with a first episode of psychosis, which has recently been granted NHS Beacon Service status.
Box 1.

Aims in the management of first-episode psychosis

To reduce the time between onset of psychotic symptoms and effective treatment
To accelerate remission through effective biological and psychosocial interventions
To reduce the individual's adverse reactions to the experience of psychosis and to maximise social and work functioning
To prevent relapse and treatment resistance

The Birmingham Early Intervention Service

This service takes clients aged between 16 and 35 years, who are experiencing their first episode of psychosis and are also in their first year of treatment by mental health services. It is one of a number of functionalised teams within Northern Birmingham Mental Health Trust and it provides clients with comprehensive and streamed mental health care for a 3-year period, the initial ‘critical phase’ of psychosis. It serves an inner-city catchment area of high psychiatric morbidity with a multi-ethnic population (about 28% African-Caribbean and 30% Asian). Research in similar populations has shown that African-Caribbean patients in particular may be poorly engaged with traditional mental health services (McGovern & Cope, 1991).
At the core of the service is an assertive outreach team consisting mostly of community psychiatric nurses (CPNs). They act as keyworkers, with support from multi-disciplinary health professionals and ex-service users. The service, in conjunction with a youth housing agency, also operates a community-based respite unit. Within this context we deliver intervention programmes, including protocol-based pharmacotherapy, family work, social recovery, work and training schemes and cognitive therapy.

Best-practice management of first-episode psychosis

The principles detailed in Box 2 are drawn from the literature on early intervention in psychosis. These are further elaborated in a set of clinical and service guidelines for use in the West Midlands, prepared by the Initiative to Reduce the Impact of Schizophrenia (IRIS) team. An audit tool to assess the adherence of services to these guidelines has also been prepared (both are available from M.B. upon request).
Box 2.

Principles for best-practice management of first-episode psychosis

A strategy for early detection and assessment of frank psychosis
A specific focus on therapeutic engagement
A comprehensive assessment
An embracing of diagnostic uncertainty
Treatment in the least restrictive setting using low-dose medication

Strategy for early detection and assessment of frank psychosis

The pathway to care in first-episode psychosis can be long and traumatic, and most people have active psychotic symptoms for 1–2 years before treatment (e.g. Larsen et al, 1996). Long duration of untreated psychosis is associated with ineffective and demoralising help-seeking and a variety of traumatic events, including high rates of involuntary hospital admission (Johnstone et al, 1986).
A strategy identifying pathways from the onset of psychosis to successful engagement with specialist treatment is vital: this may reveal circuitous and undesirable routes and opportunities for earlier intervention.
In the UK, most cases of first-episode psychosis pass through primary care, so the interface with secondary care is of particular interest. However, British data on the treatment of first-episode psychosis in general practice are sparce. Studies of schizophrenia in primary care describe problems with reviewing psychiatric status and difficulties of communication with secondary services (Lang et al, 1997). Overseas data suggest that general practitioners (GPs) have a low index of suspicion of the presence of psychosis and little confidence in making the diagnosis and that they see mental health services as slow to respond to referrals (Lincoln & McGorry, 1995).
A number of strategies have been used to strengthen the links between primary and secondary care for people with serious mental illness. One such strategy has been the introduction of standards for and audit of communication between the two services and the development of shared care. These could become part of contractual arrangements (Bindman et al, 1997). Another has been to site a liaison CPN within primary care, taking care to ensure that the appropriate patient group is prioritised (Lang et al, 1997). The Early Psychosis Prevention and Intervention Centre (EPPIC) programme in Melbourne, Australia, has published on the internet guidelines for GPs on the early detection and management of psychosis (http://home.vicnet.net.au/~eppic/).
The Birmingham Early Intervention Service is developing a shared-care programme with local GPs that involves GP training, increased communication regarding client progress and treatment and shared care of selected clients after a period of intensive management by the early intervention service. The service also maintains good relationships with the local police and has conducted some training with them in the recognition of psychosis. Regular audits of pathways to care are conducted.

Focus on therapeutic engagement

Engagement, or the formation of a therapeutic alliance, is an independent predictor of treatment retention rates and a range of good of symptomatic and functional outcomes in psychosis. However, it is difficult to achieve (Frank & Gunderson, 1990). In particular, young people doubt the usefulness of professional help, and their negative stereotypes of mental illness and fear of mental health services form barriers to self-referral (Lincoln & McGorry, 1995).
In the face of the failure of clients with first-episode psychosis to attend out-patient appointments or to take antipsychotic medication, the temptation to ‘wait and see’ is considerable. However, this approach often ends in a crisis and preventive opportunities are lost. For example, after a first episode of psychosis, clients who cease contact with mental health services are more likely to be readmitted under the Mental Health Act (McGovern & Cope, 1991).
There has been little systematic study of engagement of clients with first-episode psychosis. Opinion leaders in the field (e.g. Birchwood et al, 2000a), however, agree that it is fostered by a search for common ground with the client, an avoidance of a premature confrontation of their explanatory model of illness and the delivery of treatment in as flexible a manner as possible. If a client does not accept that he or she is ill, practical assistance and an initial focus on the resolution of a problem of importance to the individual, but unrelated to the psychotic disorder, may be helpful. Studies of substance misuse suggest that aspects of the service-delivery process, such as home visits, short waiting lists and frequent client contact with a single worker, can also maximise client engagement (e.g. Henggeler et al, 1996).
For the client to develop a positive, trusting relationship with a mental health worker, a single keyworker needs to be allocated as early as possible, regardless of whether the client is perceived as having ‘complex needs’. This worker should preferably be assigned for the entire ‘critical period’. Failure to take prescribed medication or attend appointments should not lead to reduced contact or discharge; in such cases contact should be increased. Staffing levels must take into consideration the fact that frequent contact is often needed to engage such clients.
Furthermore, since most people suffering from first-episode psychosis are young, best-practice services for early intervention must also be ‘youth attractive’. In their approach to the engagement of clients and in the interventions offered they should reflect youth culture and the legitimate aspirations of young people concerning work and autonomy.
The assertive outreach model provides a good structure for workers, and different aspects of the model encourage engagement. Low case-loads allow the time required for the development of a therapeutic relationship and for persistent follow-up of individuals in danger of being lost to services. The ‘team’ model, which demands that all the staff know all the clients, enables the best matching of staff skills to client needs as well as continuity of input when there are problems in particular relationships or breaks owing to holidays and so on. The use of extended hours and community outreach provides the flexibility to see patients at the most appropriate time and place. In our own service, keyworkers have an individual case-load of 15, somewhat higher than recommended in a strict assertive outreach model. All members of the team know all the clients. The service operates 7 days a week, with flexibility of hours if necessary. Workers attempt to engage the client on the basis of the latter's self-perceived needs, building on his or her strengths. Much of the work involves practical help with aspects of everyday life, including benefits and housing. Providing the majority of services within the team gives continuity, which enhances engagement. The team is involved with clients during admissions and the team psychiatrist is responsible for in-patient care. The service also employs former service users, to whom clients may better relate, as support workers. A vocational assistance programme led by an occupational therapist is available if needed. The client does not have to accept an illness model of psychosis for initial engagement, and cognitive-behavioural therapy (CBT) is available to address problems that the client perceives. The service also has a small social budget, which is very important in allowing keyworkers to engage with clients in low-stigma settings such as sporting events. This is also formalised in a weekly social recovery group. There is a staff v. clients football match most Friday afternoons, at which relatives and friends are welcomed as players or spectators.

Comprehensive assessment

The stress-vulnerability model conceptualises the development of psychosis as the action of social or psychological stressors acting on the vulnerable individual. This implies the need for a full assessment of mental state, risk and biological, psychological and social vulnerabilities and stressors.
People with first-episode psychosis often come to care after a prolonged symptomatic period that may have led to unstable living or financial circumstances. Furthermore, comorbidity with problems such as substance misuse, depression, suicidal thinking, social avoidance and phenomena similar to post-traumatic stress disorder (PTSD) are common, and need assessment and treatment both in their own right and because of their potential to act as stressors provoking relapse (Birchwood et al, 1998).
In our service, initial assessment involves a traditional psychiatric history and mental status examination plus relevant physical investigations. However, assessment is not considered complete until information is obtained about the client's sense of stigma, his or her own explanatory model of the illness and view of the future. The individual's access to appropriate social resources, opportunities and social networks is also assessed, as are the current and ‘best ever’ levels of occupational functioning. Most clients will allow us to talk to their family, and the quality of family relationship is noted, as well as the family's attitude to, and understanding of, the client’s illness. These data are summarised as the psychological and social needs and strengths of the client, the strengths and needs of the family and the client's own perception of his or her needs.

Embracing diagnostic uncertainty

The embracing of diagnostic uncertainty is crucial in work with first-episode psychosis, since many of the symptoms found in the prodromal or early stages of psychosis are sometimes found in normal teenagers (McGorry et al, 1995). For true early detection, referring agents must feel free to refer clients for expert assessment on the basis of a suspicion rather than a certainty of psychosis, and a low threshold for reassessment must be set.
The application of a diagnosis too early in the course of a functional psychosis not only has been found to be unreliable, but may also be harmful. The early course of psychosis is often characterised by changing symptoms, and premature diagnosis (e.g. of schizophrenia) tends to engender pessimism in clients, families and staff, which may exert a negative influence on the course of ensuing psychoses (McGorry, 1995).
Provided that organic causes of psychosis are excluded, a symptom-based approach to treatment has been advocated. This might include prescribing, for example, mood stabilisers in the presence of a full or partial affective syndrome. Decisions about diagnosis can then be deferred until some degree of symptom stability has emerged.
Our service, like most similar ones, has adopted the above practice. We have observed that clients generally find the term ‘psychosis’ more tolerable than ‘schizophrenia’ and we focus psychoeducation on risk and vulnerability factors for psychosis in general. A series of booklets have been developed for this purpose (available from M.B. upon request).

Treatment in the least restrictive setting using low-dose medication

McGorry et al (1991) found high levels of symptoms of PTSD following hospitalisation for an acute psychotic episode. It is therefore recognised that exposure of treatment-na¨ve patients to the adverse experiences associated with acute psychiatric wards must be minimised. While some services have tried to reduce the trauma for clients of an in-patient admission for first-episode psychosis by providing special nursing, single rooms and accommodation for family members, others have found treatment at home to be a viable alternative to hospitalisation.
For example, Fitzgerald & Kulkarni (1998) successfully treated 22 of 37 first-episode psychosis patients entirely at home, using home visiting as frequent as twice daily. The success of home-based treatment depended not on the degree of initial psychopathology, but on the degree of family support available to the client. In our service, we have used the 24-hour trust-wide home-treatment team to provide out-of-hours assistance to clients, with good results.
The aim of pharmacotherapy in first-episode psychosis should be to maximise the therapeutic benefit while minimising adverse effects. A number of guidelines for this can be drawn from the literature. These are listed in Box 3 and discussed in detail below.
Box 3.

Recommendations for pharmacotherapy of first-episode psychosis

  1. An antipsychotic-free observation period
  2. A low threshold for the use of atypical antipsychotic medications
  3. The use of low-dose antipsychotics plus benzodiazepines
  4. The aim of remission
  5. Early assessment of treatment resistance
  6. Maintenance of medication for at least 1–2 years in non-affective psychosis (except in some cases with short duration of untreated psychosis)
  1. If possible, have an antipsychotic-free observation period during which the diagnosis of psychosis can be confirmed and organic causes can be excluded. During this time benzodiazepines may be used for tranquillisation, as an alternative to antipsychotic medication.
  2. At present, there is no methodologically sound evidence that novel antipsychotics (excluding clozapine) are more effective than conventional ones in treating the positive and negative symptoms of psychosis in unselected first-episode psychosis populations. However, they have been repeatedly found to cause fewer side-effects (e.g. Emsley, 1999), with resultant improved adherence to treatment. The escalating cost of atypical antipsychotics and their impact on service budgets have led to greater scepticism about their use, especially since trials have not compared their side-effects with those of very low-dose regimes of conventional antipsychotics. If conventional antipsychotics are chosen as a first-line treatment, a low threshold for change to atypical antipsychotics is needed in the face of extrapyramidal side-effects, poor response or non-compliance. Similarly, factors that may increase the likelihood of the early development of tardive dyskinesia (e.g. organic brain disease) should prompt the early use of atypical antipsychotics, which it is hoped will lead to a lower incidence of this side-effect (see Bebbington, 2000, for a review of the area).
  3. The positive symptoms of most patients with first-episode psychosis will respond to 2–3 mg haloperidol equivalents daily, but they may take 2–4 weeks to do so. There appears to be no response advantage in exceeding doses of 6 mg haloperidol daily and doses above this are likely to cause unpleasant side-effects and potential non-compliance. Similarly, 2–4 mg risperidone daily should be adequate in most cases. There may be some advantage in prescribing higher doses of antipsychotics in the treatment of hostility, but benzodiazepines are to be preferred for this purpose owing to their absence of extrapyramidal side-effects (see Remington et al, 1998, for a review of the area).
  4. With sufficiently assertive treatment involving changes in antipsychotics as necessary, symptoms in the vast majority (Lieberman et al, 1993, report 83% in their study) of patients with first-episode psychosis will remit. Therefore, remission of symptoms, not adjustment to them, should be the aim of treatment.
  5. Studies have shown that, even with changes in antipsychotics (excluding clozapine), remission rates of positive and negative symptoms tend to plateau after 3–6 months of treatment following a first episode of psychosis (Edwards et al, 1998). Failure of symptoms to remit after 6 months of treatment with adequate doses of two antipsychotics (preferably including one atypical agent) and psychosocial input as appropriate should trigger prompt action. For example, review for unsuspected organic factors, mood disorder, substance misuse or non-compliance with medication: then consider using CBT and/or clozapine for persistent symptoms.
  6. It is not clear for how long people with first-episode psychosis should continue maintenance antipsychotic medication. Current consensus guidelines suggest ceasing antipsychotics after 1–2 years of remission following a first episode of non-affective psychosis. Future guidelines may recommend a longer period of treatment, in light of Robinson et al's (1999) 5-year follow-up of people with first-episode schizophrenia, which showed that relapse rates rose quickly to 51% over the first 2 years after index admission, then plateaued to a certain extent, reaching 78% by 5 years. Some people with first-episode psychosis stay well with shorter periods of antipsychotic treatment (Crow et al, 1986). Identifying this group is a challenge for future research, but it is unlikely to include those who have had psychotic symptoms for more than 1 year prior to treatment (Crow et al, 1986).

Focus on pychological adjustment and maintenance of social roles

Adverse reactions to the experience of psychosis and its treatment, including depression, PTSD and suicide, are well established. While denial of illness may lead to non-compliance with medication and subsequent relapse, its acceptance can lead to pessimism, loss of self-efficacy and absorption of the pejorative stereotypes of mental illness (Birchwood et al, 1998).
In our service, we use a theoretical framework focusing on individuals' appraisal of their illness in relation to ‘loss’, ‘humiliation’ and ‘entrapment’: this reveals their psychological adjustment to the illness (see Birchwood & Spencer, 1999, for a review of the area). We strive for a blame-free acceptance of illness, together with the encouragement of a sense of mastery over it. We aim to achieve this through client education in cognitive-behavioural strategies of symptom control, the management of early signs of relapse and highlighting and challenging cultural stereotypes of mental illness. Most of this is provided in the context of an ongoing therapeutic relationship with a keyworker and group psychoeducation.
Social roles and goals, particularly work, are highly prized by young people. They provide a source of self-esteem that can affect the psychosis itself (Warner, 1994), and their loss has been linked to depression and suicidal thinking in psychosis (Birchwood et al, 2000b). Since alleviating symptoms is no guarantee of a good social outcome (Birchwood et al, 1998), clients must be given the opportunity to regain and develop life skills in a supportive milieu, particularly one in which a peer group is present. Preliminary results suggest positive effects from the EPPIC recovery programme, which is delivered in such a manner (Albiston et al, 1998).
Our service has developed a structured approach to help clients enter paid employment. The service occupational therapist delivers an 8-week accredited pre-vocational training guidance programme, funded through a local college. A short, intensive work-preparation course has also been developed, which acts as the referral point to a specialised employment service for people with mental health problems. It is run by an agency linked to social services and the National Schizophrenia Fellowship. A less-structured approach is taken to social recovery, with social, cooking and mother and baby groups held weekly at our community base.

Focus on the entire family

Since most people with first-episode psychosis are young, many will be living with a family or partner. There is now little doubt that family interventions in schizophrenia are effective in postponing psychotic relapse over periods of up to 2 years. Furthermore, high levels of emotional distress are found among the relatives caring for a person with schizophrenia. Despite their obvious relevance, family interventions in first-episode psychosis remain underresearched (see Birchwood & Spencer, 1999, for a review of the area).
The traditional concepts, such as expressed emotion that form the rationale for family interventions, may be of less relevance in a family experiencing first-episode psychosis than are concepts involving trauma and loss. Indeed, the mitigation of loss and negative experiences of caregiving within such families in the early stages of psychosis may be crucial in order to prevent the development of a critical family atmosphere later.
In our service, we are in the process of training all keyworkers in skills-based family intervention, following Falloon's (1985) model. We also facilitate a monthly family and carer support group, as we found that families appreciate a forum in which to share their feelings of loss, guilt and confusion, drawing on support from similar families. Families seem to benefit more when this is driven by issues they themselves bring to the group, rather than by a structured ‘agenda’.

Prevention of relapse and resistance to treatment

It has been argued that the early course of psychosis is sharply predictive of its longer-term course, and that there may be a toxic effect of cumulative exposure to positive symptoms. Given this, the prevention of relapses and treatment resistance might alter the long-term outcome of psychosis, presumably by limiting the postulated neurotoxic effect of active psychosis (Wyatt, 1991).
The basis of relapse prevention is the modification of stress and vulnerability factors by means of the best-practice interventions discussed above. Clients and their families should be informed about the risk factors, within the context of the stress-vulnerability framework. A shared and documented relapse prevention plan for each individual can then be developed and rehearsed with the client and his or her social network. An active relapse-monitoring procedure can also be instituted.
We have developed an approach to relapse prevention involving collaboration with the client and the family to identify the client's unique early warning signs of psychotic relapse and to prepare and rehearse a response to these. This response draws on cognitive therapy of emerging delusional themes and the use of coping mechanisms, drug interventions and resources within the family and treatment network (Birchwood et al, 2000c).
A system for the early identification of treatment resistance is needed, so that a wait-and-see attitude is not encouraged beyond the appropriate period. This has been discussed above with respect to pharmacotherapy. Similarly, the early application of CBT should be encouraged, since controlled studies have found these to lead to a significant reduction in both the duration of an acute psychotic episode and persisting positive psychotic symptoms (see Birchwood & Spencer, 1999, for a review of the area).
On the basis of a study of time to recovery, the EPPIC programme has identified 3 months after initiation of treatment as the critical point at which the presence of persisting positive symptoms should motivate assertive action. Clients are screened for persisting positive symptoms at this time and, if they are present, they are referred to a specialised subservice (the Treatment Resistance Early Assessment Team; TREAT), which offers reassessment and a variety of interventions including CBT and assertive pharmacotherapy (Edwards et al, 1998).

Conclusion

Services dedicated to young people experiencing psychosis for the first time have not hitherto been afforded a high priority in routine psychiatric care. This is changing as it is recognised that the early phase is a stormy period in which disability and relapse flourish, and one often preceded by a long period without treatment. This, combined with the high risk of suicide, provides an ethical basis for improved practice. Accumulating evidence supporting the critical-period concept holds out the hope of improving early and long-term outcome by intensive early treatment. Early intervention does not simply involve ‘bringing forward’ best practice to this early phase; it requires special care in recognition of the biological, psychological and familial challenges and changes that are active in this period and that have informed these guidelines.

Multiple choice questions

  1. The ‘critical period’ following a first psychotic illness is:
    1. the time during which individuals decide whether or not to stop treatment
    2. inversely related to the duration of untreated psychosis prior to admission
    3. a period during which relapse is common
    4. associated with increased risk of suicide
    5. a period when it is important to offer intervention.
  2. In the pharmacotherapy of first-episode psychosis:
    1. it is often sufficient to treat with benzodiazepines
    2. novel antipsychotics are the first-choice treatment
    3. in most cases, remission of symptoms can be expected within 6 months
    4. symptoms often remit with low doses of an antipsychotic
    5. treatment can usually be safely stopped 6 months after remission of symptoms.
  3. Best-practice management of first-episode psychosis includes:
    1. strategy for the early detection of psychosis
    2. the early establishment of a definite diagnosis
    3. early admission and assessment in hospital
    4. the maintenance of valued social roles
    5. a strategy for relapse prevention.
  4. Psychological adjustment following a first-episode psychotic illness:
    1. is usually uncomplicated if psychotic symptoms resolve completely
    2. is helped by developing a sense of mastery
    3. can be problematic if the individual feels trapped by the illness
    4. is nearly always successful if the individual accepts the illness
    5. is helped by learning how to self-manage relapse.
  5. Therapeutic engagement of young people with a psychotic illness:
    1. improves if staff insist that the young person accepts the diagnosis as soon as possible
    2. is helped if staff search for common ground to understand clients' experiences
    3. can be helped by an assertive outreach approach
    4. requires regular out-patient appointments
    5. is helped by persistent contact with a keyworker.

References

  1.  
  2.  
  3.  
  4.  
  5.  
  6.  
  7.  
  8.  
  9.  
  10.  
  11.  
  12.  
  13.  
  14.  
  15.  
  16.  
  17.  
MCQ answers