United Health Group (UHG) Minnesota - Katherine Ward

United Health Group (UHG) Minnesota - Katherine Ward

United Health Group (UHG) is an US managed care company that has significantly improved care to highly dependent older people. The Department of Health has been impressed with their programme and is funding a pilot project to help transfer the model to 10 PCTs in the UK.

The model of care uses intensive end of life planning with patients and relatives to achieve agreement on appropriate care wanted by the patient and their families. Care delivery relies heavily on the use of nurse practitioners and case managers. They assess the need for care, implement care and provide early assessment whenever further worsening of the patients condition occurs. Data on care episodes is regularly collated and fed back to clinicians. As much care as possible is provided at home (predominantly in nursing homes). Admissions to hospital are minimised and the benefit of the planned intervention in hospital is always balanced against the potential loss of function and reduction in quality of life.

Features of the UHG system that are underdeveloped within the NHS but could aid the more appropriate care delivery are:

• Greater flexibility of care providers (care following the patient whenever possible)
• Greater capacity of nursing homes to undertake intermediate care on site (iv infusions, blood transfusions, X-rays) and to retain patients
• Greater use of nurse practitioners and case managers to deliver the bulk of the work in conjunction with physicians

The US programme took 2-3 years to implement as a well functioning system and to show results. User satisfaction is very high, mortality has remained unchanged and hospital admissions dropped by 40%.

This paper describes a visit of 10 UK PCTs (listed in Appendix 1) to UnitedHealth Groups Headquarters in Minnesota, United States in November 2002. It sets out the Evercare programme, a model of care for the frail elderly that has managed to improve quality of care and at the same time reduce inappropriate care. This may offer a useful model for PCTs and Trusts in Bristol and South Gloucestershire.

It also describes phase 1 of the project which has been formalised through a contract between the Department of Health and UnitedHealth Group between November 2002 and January 2003 and phase 2 which would involve a contract between local PCTs and United should the local health community decide to progress this project next year.

The objective of the project is to improve the efficiency and effectiveness of health care of predominantly institutionalised older people in the NHS.

The project followed a visit of DoH advisers to UnitedHealth Group and a subsequent meeting between Lois Quam, the chief executive of Ovations, a subsidiary of UnitedHealth Group and the parent company of Evercare, with Alan Milburn.

UHG visited the UK during September 2002 and met with a number of PCTs, including Bristol, to explore ideas at this stage. They have also met with the BMA, the NHS Confederation and Unison and are planning a meeting with the RCN to discuss their proposals.

In November 2003 eleven senior managers and clinicians from the Bristol and South Gloucestershire community visited Minneapolis, Minnesota, the Headquarters of UHG (list of attendees is attached at Appendix 2). The visit comprised a series of presentations, small group discussions and field visits to hospitals and nursing homes. This report summarises the learning from this visit and sets out the next steps for the project.

UHG made it clear that they are not intending to set up hospitals or manage clinical services in the UK. They are, however, keen to hire and train British people and are looking to develop a portfolio of services to offer PCTs at a set price, including help with commissioning, HR and change management.

Evercare is the name of a company, affiliated to UnitedHealth Group and part of its Ovations Company, which provides services to public sector (federal and state government) and non-profit organisations across the US.

Evercare is also the name given to the programme which the company provides, delivering enhanced medical care and co-ordinating services to 60,000 frail elderly, chronically ill and disabled people in the US.

The programme draws on a set of core principles. These are:

• Treatment for older persons must focus on the whole person with all the interventions meeting the goal of preserving independence, function, comfort and quality of life.
• Primary Care (delivered mainly through nurse practitioners) is the central organising force of ALL medical care for older people.
• Medical care for older persons is delivered in the least invasive manner in the least intensive setting possible.
• Minimise iatrogenesis secondary to polypharmacy
• Data drives decisions

Key elements of the model are:

• Development of mid-level practitioners working in partnership with physicians. These geriatric care extenders are either Nurse Practitioners or Physician Assistants (depending on state licensing regulations in Minnesota they are Nurse Practitioners) and are the only clinical staff directly employed by Evercare. They are substitute geriatricians and work in partnership with 4-6 primary care physicians to manage a caseload of approximately 100 patients in 2-3 nursing homes (on average). Physicians, who are contracted to Evercare have a caseload of up to 500 patients based in several nursing homes. The primary care physician retains the duty of care for the patient.
  
  The doctor and the nurse practitioner have a collaborative relationship with flexible roles that are complimentary but not interchangeable. The two professionals have a joint practice agreement, negotiated at a meeting between them and written down. Relationships clearly differ depending on individuals but the physician is expected to respond directly to the nurse when contacted. Contact between the professionals usually decreases (e.g. monthly meetings) as they work together longer and have less need to discuss each individual issue. The role of the Nurse Practitioner is not protocol driven but rather driven by the relationship between them and the physician, although guidelines are agreed between them.

Nurse Practitioners either have a 2-year registered nurse training (generalist) plus a graduate education/continuing education (specialist) or a 4-year degree, which takes them to a masters level. During the graduate education stage, they spend a day a week in a primary care physicians practice. On completion of the graduate programme, which includes modules on physical assessment, they take a national exam to certify them as a nurse practitioner. For the first 6 months of working for Evercare they are mentored. The relationship between them and the physician is described as like that of a doctor in training. They have full prescribing rights, they can undertake full physical, functional and mental health assessments, they order diagnostic tests, admit to hospital and manage chronic disease in the community. They also undertake training of nursing home staff. Nurse Practitioners need revalidation every 5 years, demonstrating a minimum of 150 hours teaching and significant clinical practice.

The US version of the wider primary care team that link to this partnership also includes nurses from the nursing home, social workers, therapists, pharmacists and other physician assistants (for example in orthopaedics). Nursing Homes employ people with a wide range of nursing skills.

Nurse Practitioners can also refer to specialists such as Tissue Viability nurses providing that they are accredited by UHG. They refer to hospital physicians but do not have input into patient care in hospital or discharge planning.

• Comprehensive assessment and end of life planning.
  Initial assessment can take up to 10 hours to complete and is described as high touch. This identifies risk of mortality and hospitalisation and interventions are then targeted on this basis.

A great deal of emphasis is placed on developing Advanced Directives. This involves work with individuals in the programme or with families and/or advocates. High levels of dementia within its constituent population requires Evercare to work closely with families and with local nursing home staff to advocate for patients. The US legal system requires clinical staff to consult with families if the patient is not cognitively functional. Advanced Directives are a legal requirement on admission to hospital and therefore more commonplace than in the UK.

An emphasis is placed on the pros and cons of interventions such as tube feeding or IV antibiotics and indeed on hospitalisation itself. The dangers of iatrogenesis including loss of independence on admission, hospital acquired infection, bedsores, loss of ambulation and confusion are emphasised and balanced against the expected benefit of clinical intervention. Expected progression of chronic disease is also explicitly discussed with patient and families and they are encouraged to think through how they would like their care at each stage. The emphasis is on maintaining quality of life through sustaining functional independence. Every intervention is reviewed on the basis of the patients preferences and the balance of the benefit that it will provide versus the burden it will inflict. Patients are encouraged to state what elements of their functional independence they most value and their care plan is a dynamic rather than static document. 85% of patients do not want to be admitted to hospital.

Worksheets and scripts to help nurse practitioners to undertake end of life planning are made available by Evercare.

• Earliest Intervention
  The comprehensive assessment described above identifies risk factors and Evercare principles dictate that subtle changes in condition are notified so that diagnosis and treatment can be initiated early in the disease process. For example, an older person not eating breakfast or not seeming themselves would be enough for the nursing home to refer to the nurse practitioner.
  
  Checklists are provided at the nursing station and for families: Change in Condition Alert forms to ensure that the nurse practitioner is informed when something has changed.

• Less intensive setting and least invasive intervention.
  Evercare works on the basis that hospital admission for frail older people should be avoided where possible. They are conscious of the susceptibility of this client group to iatrogenesis and of the likelihood of the American fee for service system to result in unnecessary acute care services. Hospitalisation would only be required for a new illness, new injury or exacerbation of a chronic condition if an operating theatre, intensive care or high level technology is necessary and even then, the required number of bed days would be calculated prior to admission and arrangements made for discharge. Average length of stay for Evercare patients is 5 days.

Minimising medications to the useful and necessary is another element to this maxim.

Evercare works predominantly with people in Nursing Homes and sheltered accommodation. Nursing Homes in the US seem to have a more medical environment than those in the UK. Significantly interventions such as transfusions, IV medication and orthopaedic injections were available in the homes (and in the sheltered housing projects which were adjacent to them).

• Data drives decisions

Evercare focuses on population management. They identify from their population a sub population of vulnerable people at high risk of hospitalisation. Benchmark data is established and variations identified e.g. clinical indicators such as prevalence of pressure sores, incidence of falls or numbers of depressed patients on anti-depressants. Annual targets are set and clinical action plans identified to reach targets at national and site level. Information is analysed fortnightly and urgent interventions set up to impact on negative trends and ensure compliance with the monthly measure. Targets are set using historical data and predictive modelling.

Avoidable hospitalisation is categorised and analysed for example by days of the week, timing, referrer and diagnosis. This then results in training for staff, changes in clinical planning, and links to national initiatives and further detailed monitoring.

The programme is prepaid based on historical trends so Evercare have an incentive to keep their population healthy through prevention programmes, to intervene early and to prevent unnecessary high cost hospitalisation. They also creates incentives for meaningful interventions, for example paying doctors for an evening visit or for participating in an end of life planning meeting with families. Nursing Homes have reduced incentives to admit patients to hospital unnecessarily through profit sharing (the system previously encouraged admissions as nursing homes were paid extra on the return of the patient). Hospitals are paid a flat rate per admission helping to ensure patients are returned home as quickly as possible. Nurse Practitioners have lap tops or palm top computers for patient information linked to a central database.

The programme has been running since 1987. It has been independently evaluated by Dr Robert Kane of the University of Minnesota and has been shown to:

• Reduce hospitalisation of its enrollees by 50% without impacting on mortality rates
• Reducing prescription drugs whilst maintaining health (and reducing side effects)
• Have a 97% satisfaction rating to families and high satisfaction levels to physicians
• Have saved 7% of the federal governments annual Medicare budget

The Ever care Team emphasised that this programme had been developed over time and that it had taken 2-3 years to see the impact on hospital admissions. They described the barriers that they had overcome to implement the scheme and these included:

• financial incentives in the system for nursing homes to admit people to hospital
• shortage of carers in nursing homes
• limited numbers of nurse practitioners
• sceptical doctors

They worked through each of these barriers in turn and described their strategies for dealing with the difficulties. These included:

• listening to rather than talking to the key stakeholders
• focusing on the early adopters rather than the sceptics
• reworking the financial incentives
• finding champions within nursing home staff
• proactive recruitment of nurse practitioners and significant support to those new in post

A visit of Evercare staff to Bristol in January 2003 will explore whether this model can be adapted to be useful for the local health community.

During the presentations and visits, there were a number of other key ideas and concepts which the group picked up. These include:

UnitedHealth Group have collaborated with the BMJ to ensure that "Clinical Evidence" is distributed to each of the 500,000 physicians contracted to the group. Importantly recommendations on evidence-based practice are used alongside systematic feedback to individual clinicians.

Diagnoses, prescriptions, interventions, referrals and policies and protocols from professional bodies are incorporated into the feedback to individual clinicians. Such data form the basis for evidence based clinical profiles and identify outliers amongst participating physicians. Feedback and care process improvement is offered for care which is contrary to the evidence base or not contributing to an improved outcome. The aim is to decrease variation in clinical practice and to promote evidence based medicine.

United are working on a website design which will allow clinicians to be awarded continuing medical education points for interacting with the data online.

They also have myuhc website for patients where patients can access Clinical Evidence and learn about the evidence base of their care.

Use of volunteers seems more common than in the UK. They were visible in day centres, nursing homes and the hospital that the group visited employed 500. Further information is required to establish what roles volunteers play.

The value of good data and IT was stressed throughout the visit. United spent million on IT during 2001. The greater use of IT is partly explained by the need to manage the billing process in a health care system paying to a large extent by item of service and thus collecting a great deal of information on the care provided to individual patients.

Some examples of ways in which data was being used include:

• Linking diagnoses e.g. cancer and depression, coupling this with the evidence that 60% of people with depression will not comply with medication and offering additional intervention through telephone counselling to maximise compliance with medication.

• The myuhc website allows patients to estimate the cost of their care and select a hospital on the basis of location, mortality rate, volume of patients for a particular condition, cost and waiting time. They are currently also implementing a physician choice module.

• Automated patient history: it was thought that better and more accurate information was gained when the patient completes a proforma on the computer. Interestingly volumes of paper records were described as too much information.

In spite of the reliance that UHG now had on data, there was an acknowledgement that their own systems were paper based in the early days and an understanding that the NHS did not have such sophisticated systems at this stage.

There are three important concepts that are kept in focus when planning and commissioning services:

Utilisation rates and unit cost are calculated across the number of members of the programme at one time to give a costing of Per Member Per Month (PMPM).

A presentation on the overview of actuarial capabilities and processes demonstrated impressive tools for analysing commissioning decisions (capitation versus fee for service) to balance utilisation and cost and for demand forecasting, for example for inpatient utilisation projections.

Predictive modelling is also used to identify and manage high users and those at risk of high use of health services. An artificial intelligence tool (CCG) combines a sophisticated rules engine with a comprehensive medical knowledge base. It uses professional, facility, ancillary and prescribing claims data. All these clinical facts are combined to a measure: Burden of Illness (BoI).

This BoI measure is then taken alongside predicted cost, indirect factors (such as health care seeking behaviour of the individual), trend factors (for example utilisation trend of the individual over time) and creates a relative risk the predicted cost of the individual relative to the average member cost.

The healthcare commissioner can then identify the individuals at high risk of high cost, decide which diseases to focus on and find opportunities to make a difference, e.g. identify missing interventions, monitor compliance, ensuring access to care.

United operate a transplant network where they credential providers and contract with a national network of selected providers, significantly improving clinical outcomes and costs.

A short academic presentation on improving chronic care presented some interesting ideas. The sense was that much of this was still to be implemented in the US which was criticised during the presentation for its emphasis on acute care.

The role of the patient in managing their own care was emphasised. It was suggested that visits to clinicians should be on the basis of come back when you notice that your condition has deteriorated rather than come back in 3 or 6 months.

Clinical glidepaths should be developed which map how the disease would be expected to progress in the absence of care and in the presence of good care. This would provide feedback to staff who would be aware of what they had accomplished and helpful to patients and families, who can be more informed in decision making. Mechanisms should be developed to signal when patients deviate from the expected path which may then allow for appropriate intervention. Patients, family members and home helps could be trained to observe and monitor deviation from the pathway. IT (hand held palm tops) could be used to ensure that the managing clinician receives this local information.

Evercare employed Case Managers to work with frail older people in their own homes. They had a caseload of 100 people. They undertook a full health and social care assessment of the individual and their families. They undertook regular home visits and contacts with the patients. They evaluated this model using a sub population of patients who were expected to be hospitalised in less than 12 months, who were frail with complex conditions and for whom it was felt they had an intervenable condition there was something which could be changed. In this pilot, case managers worked with 35 patients maintaining contact regularly. The scheme broke even on cost with visits to the primary care physician reduced and hospital admissions reduced. This was the new model the physician described where people stayed for up to 400 days on the programme. They were not focused enough on outcomes and discharging appropriately and the physician did acknowledge that for that reason it was only available to a limited population.

Community care Co-ordinators: the assessors who dealt with people in their own homes, had approximately 100 on their caseload, did not have a hands on role but who assessed, managed the patients pathway and co-ordinated discharges from hospital. They assessed every 6 months or sooner if hospital admission instigated. These people were from either a nursing or social work background.

Orthopaedic care in a sheltered housing project was of note. Orthopaedic physicians assistant was a regular visitor to the home and would give orthopaedic injections to residents to avoid them having to visit the hospital. In the case of suspected fracture, an X ray could be taken in the home (results available in 2 hours) and if it didnt require surgery, then the cast technician would visit and apply a cast in situ. If surgery was required, the individual could stay in their own bed over night and be transported directly from there to theatre.

UK visitors were particularly impressed by a prescribing machine which was a cash point for drugs (Instymed machine) the patient entered a card and pin number and received appropriate medication. The machine was linked to the hospital pharmacy for topping up and monitoring.

A mobile X ray machine the size of a (large) vacuum cleaner and used to take chest and skeletal X rays in nursing homes was also of note.

6.1 Phase 1
The Department of Health has signed a contract with UnitedHealth Group to translate their learning in working to improve health services to an at risk population of frail older people to the NHS. This contract involves a diagnostic visit to each of the 10 PCTs involved in the programme. The visit will take approximately 2 days per PCT (5 days in total for the Bristol and South Gloucestershire health community, 13-17 January 2003) and all visits will be completed by mid January 2003. The visit will involve key people from each PCT who will help Evercare to understand care pathways, roles of different practitioners, barriers to change, roles of different agencies and functions of the local hospitals.

The visit will be led by Marcia Smith, former chief executive of Evercare, together with RuthAnn Jacobson, a Geriatric Nursing Practitioner and founder of Evercare, Ann Routier and Mary Paoli, both former Executives of Evercare. Deborah Matson-Beale is the project director. They will need to meet with key managers, professionals and clinicians during the visit.

The visit is intended to identify what would be required to adapt the Evercare model to the UK. It is anticipated that the skeleton of the model would be the same across all the PCTs, but that it may be adapted locally to suit individual circumstances. This scoping process will also identify the local opportunities and barriers to implementation, describe how to get the organisation ready for change, identify what resources we have locally and suggest the resources we need to be successful. A final outcome of the visit will be the design of a pilot project for implementation in phase 2.

A personalised PCT report will be developed through an iterative process and will be owned by the DoH but shared with the PCTs. A main document will be prepared for the DoH, Alan Milburn and Tony Blair which summarises the findings across the country.

There needs to be a process during February and March 2003 of discussion locally about moving to the next phase, which, if we agree to progress, will take place from April 2003.

Phase 2 would involve the Bristol and South Gloucestershire PCTs contracting directly with UnitedHealth Group to facilitate the implementation of the model. The DoH representatives on the trip emphasised that there are clauses within the contract that allow PCTs to opt out at different stages in the process. The DoH will provide solicitors to help negotiate phase 2, a template contract with room for adaptation, an evaluation of the process and briefings and presentations to facilitate local ownership if required. Financial resources will be made available to each PCT although PCTs are also expected to contribute resources for local project management. Phase 2 will be last until November 2004. Some PCTs are receiving their Phase 1 visits mid November 2002 and may therefore progress more quickly to Phase 2.

United would provide three teams across the 10 PCTs made up of a Change Agent, a Clinical Expert and a Medical Director. They would identify a subpopulation of older people who are at high risk of hospitalisation. This population may be residential care based or in their own homes rather than in Nursing Homes. Initially they would work with the health community to identify measures of success and a measurable baseline. They would work with PCTs and Trusts on developing an assessment tool and provide training for nurses locally to ensure early intervention. They will work with us on identifying local champions. A central United back up team will be based in this country and would include administrative and financial support.

A learning set of the 10 PCTs, managed by NatPact will be set up throughout the process. The DoH will be responsible for media management.

This report presents the first stage in feeding back from the visit to health and social care organisations across Bristol and South Gloucestershire. Presentations and discussions need to take place in a number of forums across the community. Different organisations need to be clear about sign up to this project and ownership as a health community.

A provisional timetable for the 5-day visit 13-17 January 2003 was negotiated with Evercare in Minnesota but detailed planning needs to take place to ensure the value of the visit is maximised.

Evercare have requested some key information to help their scoping exercise in January. PCTs and Trusts need to work together to ensure the information is high quality and to work through alternative means of extracting data which is not easily available.

A Project Board/Steering Group needs to be set up to oversee the project, with an appropriate stakeholder forum to ensure engagement and ownership of the project. This needs to link to local groups and forums concerned with Intermediate Care, Capacity Planning, Older People and Modernisation.

There are clearly radical differences between US and UK healthcare systems. During the visit Executives of United were explicitly calling for a universal basic health care benefit package: they did acknowledge some of the weaknesses of the US system, for example its fragmentation, its preference for clinically invasive interventions, its high cost and its lack of universal coverage. Evercare were keen to stress that in the absence of one clear political or financial voice, they use the patients point of view to plan their programmes of care.

Parallels were drawn between a programme such as Evercare whose responsibility is to ensure an integrated system that manages the delivery of comprehensive health services for an enrolled population with PCTs with their registered populations and defined budgets. Both are charged with commissioning a network of providers. United are used to working to stringent goals and objectives and reinventing the way that they work to achieve them; they have to respond to multiple constituencies and are subject to increasing public scrutiny. Whilst cognisant of the fact that the UK system is very different, they were keen to see their programme as translatable and open to adapting the model to make it work. Evercare works across 13 different cities in the US with different health systems.

The principles of Evercare do mirror those of the NSF for Older people in assuring standards of care, extending access to services and developing services to promote the independence of older people. This project will also build on the integration of health and social care services for older people in Bristol and South Gloucestershire. The project will also be significant example of modernisation for local PCTs, Trusts and Social Services Departments.

The trip to the US presents a significant opportunity to the PCTs and Trusts in Bristol and South Gloucestershire to learn from another healthcare system, to critically evaluate local healthcare systems and processes and to modernise care for older people. There are also transferable lessons to other care groups and specialties.