The persistent problem of integrated care in English NHS hospitals: Is the Mayo model the answer?

Introduction

Despite attempts to provide complete and integrated services in English National Health Service (NHS) hospitals, many patients seem to receive fragmented and episodic care, resulting in multiple hospital attendances, reviews and shuffling between specialities. This cannot be good for patients or for efficiency. We aimed to compare models of care between an organisation reported to provide high-quality integrated clinical care (Mayo Clinic) and the situation in the English NHS, and to draw lessons which could improve patient care as well as patient and staff satisfaction.

Recent reforms to the English NHS have included a strong focus on integrated care (IC) models, which are largely concerned with integration of care across primary, community and secondary care, and between the health and social care systems. While such programmes may improve coordination between the traditional silos of health and social care, encourage multidisciplinary approaches to patient care and offer at least the possibility of cost savings, they do not directly address a pernicious issue within the secondary care sector: the episodic and uncoordinated care of non-emergency patients in NHS hospitals, particularly in the case of complex, multi-morbid patients. In this paper we argue that there may be a strong case to embed (or return to) a culture and practice of bringing multi-specialty expertise around the patient, to offer more holistic and timely diagnosis and intervention, within the setting of NHS hospitals.

In 1997 the Labour Government published a white paper “The new NHS: modern, dependable” which marked a turning point in the organisation of healthcare in the English NHS (Department of Health, 1997). It proposed to abolish the “internal market” in favour of IC, and to invest the supposed savings in targeted service improvements and an expanded focus on population health. Since 1997, government policy has continued to promote better integration of care, for reasons which have been claimed to include cost saving, improved performance and quality, patient safety and satisfaction, a need to respond to changing health and social care technologies and the professional views of clinical staff and service managers (Department of Health, 2008; Ham and Curry, 2011; Gretton and Honeyman, 2016; Monitor, 2014).

The most recent major reorganisations of the English NHS, implemented by the 2010–2015 UK Coalition Government, have carried forward this ambition, with the 2010 health white paper “Equity and excellence: liberating the NHS” arguing for “putting patients in control” and for full integration across health and social care (Department of Health, 2010). In the period between the publication of this white paper and the passing of the Health and Social Care Act, a number of influential health policy organisations published reports which were broadly supportive of this aim (UK Government, 2012; The King’s Fund, 2012, 2013; Goodwin et al., 2012). Integrated, patient-centred care continued to be the focus of a number of studies and reports published up to, and following, the launch of NHS England’s Five Year Forward View (5YFW) in 2014, adding to the sense that IC was the only game in town (Royal College of General Practitioners, 2014; NHS England, 2014).

The 5YFW argued that these changes to NHS structures and practices are broadly necessary to achieve the triple aim of improving the patient experience and population health, and reducing the per capita cost of healthcare. To this end, the 5YFW proposed clear commitments to breaking down barriers in how care is provided, and developing new, integrated models of care and new relationships with patients and communities. These commitments are now embodied in initiatives such as the new care models (and associated “Vanguard” sites), the sustainability and transformation partnerships (STPs) and the emerging accountable care systems (ACSs) and accountable care organisations (ACOs) (NHS England, 2017a, b; Ham et al., 2015; Imison, 2015; The King’s Fund, 2017).

The new care models which are developing throughout the NHS in England have a common theme: each proposes some form of integration which spans traditional silos, between primary, community and secondary care and/or between the healthcare and social care systems. While these models undeniably aim to remove barriers to patient-centred, IC, they still assume that for non-emergency treatment of patients who need specialist care or diagnosis in a hospital environment, the English NHS will continue to work through a hub and spoke model (NHS England, 2017a, b), with general practitioners (GPs) acting as gatekeepers. Patients will continue to be offered an appointment with a hospital consultant in a particular department, which may take place only after weeks or months (NHS Choices, 2017). Following the hospital appointment, patients will generally have to arrange another visit to the GP to review the specialist’s advice, begin treatment and often then to find out what further advice, care or testing may be required. If there is a need to return to the same specialist, to be seen by a different one, or to carry out other tests, patients may need to return for a further appointment, arranged by either the GP or the hospital specialist.

Within this episodic model, patients with complex multiple conditions may continue to find themselves spending long periods waiting to see the next clinician in the chain, often unsure which physician ultimately has oversight of and responsibility for their care. Waiting for specialist advice, and uncertainty concerning the coordination of care, may be sources of anxiety and stress for patients—particularly when appointments or procedures are postponed or cancelled due to other pressures within the health system (NHS England, 2016). Furthermore, the NHS patient population contains an increasing number of such multi-morbid individuals (Department of Health, n.d.), who often need to be assessed by a range of different hospital specialists, sometimes because a first appointment reveals a previously undiagnosed, untreated or under-treated condition, because the patient simply requires expert advice and treatment from a different specialist or sub-specialist, or because of lowered thresholds for diagnosis (Starfield, 2011; Rowland and Paddison, 2013). In essence, there is a growing mismatch between the increasing complexity and acuity of many patients requiring non-emergency care in a hospital setting, and a system still largely designed to treat them in a linear, episodic manner.

For these patients, it could be very beneficial to organise care differently within the hospital environment, such that they would have rapid access to a range of specialist opinion and diagnostic procedures which could address multiple care needs in a short space of time. In 2016, a team of researchers from Durham University’s School of Medicine, Pharmacy and Health set out to find examples of hospital organisations which organise diagnosis and care in a more holistic way, using multidisciplinary teams of clinician specialists, particularly for patients with complex conditions and/or multi-morbidities. An early, rapid scan of the literature showed that one of the best-known examples of such an organisation is Mayo Clinic in the USA, which reports itself as being organised around a highly patient-centric care model, based on multi-specialty practice and which makes extensive use of a “coordinating physician” to manage the patient’s care during elective, outpatient or ambulatory episodes of care (Mayo Clinic, 2014). Other hospital organisations in the USA have also adopted some elements of hospital-based, integrated clinical care, based on the use of the “hospitalist” doctor model which emerged in the late 1990s (Wachter, 1999). However, Mayo Clinic is reported to have operated this type of care model for many decades and could therefore be regarded as an exemplar organisation.

The same scan of the literature confirmed that Mayo Clinic is frequently cited as one of the highest performing hospital organisations in the USA (Kodner and Spreeuwenberg, 2002; NHS England, 2015). For that reason, coupled with the distinctive approach to the organisation and management of patient care within the hospitals operated by Mayo Clinic, the research team considered that there would be merit in pursuing a two-stage action research project, to investigate whether there are aspects of the Mayo model which could be adopted within English NHS hospitals in order to address the continuing challenge of episodic and fragmented patient care, as outlined earlier in this section. To this end, the team proposed to carry out the following phases of work:

1. Conduct a rapid review of the literature into the Mayo Clinic’s model of care, with particular focus on evidence concerning any links between the “Mayo Model” and patient outcomes, staff and patient experience, hospital performance data and finances.

2. Carry out qualitative research within an English Foundation Trust hospital to investigate the potential barriers to implementing a care model similar to that used by Mayo Clinic.

The overall aim of this study is to determine whether there are aspects of the distinctive Mayo Clinic model of care which could be translated into hospital settings within the English NHS, in order to overcome the fragmented and episodic nature of non-emergency patient care.

This paper reports on the first phase of the study. The findings from Phase 2 of the research, including a fact-finding tour of Mayo Clinic in May 2016, are to be separately reported in a forthcoming article.

The research study protocol was granted ethical approval by the Ethics Committee of Durham University’s School of Medicine, Pharmacy and Health in April 2015.

Method

We employed a rapid review method to examine the literature on the Mayo Clinic model and other models of hospital-based ICC. Rapid reviews have become a common means of obtaining rigorous but relatively quick insights into the state of knowledge and evidence relating to a particular field of enquiry (Khangura et al., 2012). They are popular with policy makers and public-sector institutions, as they can be more responsive to the timeframes encountered when translating from policy development to implementation (National Collaborating Centre for Methods and Tools, 2012). This streamlined approach to synthesising evidence has been particularly welcomed by decision makers in healthcare settings, owing to the rapidly evolving landscape of health and healthcare systems (Khangura et al., 2012). Systematic reviews may be the “gold standard” under some circumstances, but they generally require significant resources over a 6- to 24-month timeframe, and they often focus on a tightly defined clinical question. Neither of these conditions applied to our study.

Our rapid review was carried out in late 2016. The principal aim was to locate peer-reviewed articles, books, “grey” literature, reports and any other material concerning the implementation and use of hospital-based, ICC models in general, and the Mayo Clinic’s model of care in particular. In addition to the rapid review, we also consulted some literature reviews and syntheses concerning the general concept of IC in the English NHS and other health systems, to discover whether these included discussions of hospital-based ICC, but also to establish a context for the more generic use of IC in health system research.

For the rapid review, we used the following search terms:

• “integrated clinical care” (in title only);

• “Hospital”+(“integrated clinical care” OR “hospitalist”) (in abstracts);

• “NHS”+(“integrated clinical care” OR “hospitalist”) (in abstracts); and

• (“Mayo Clinic” OR “Mayo Model”)+(“integrated care” OR “integrated clinical care”; OR “coordinated care” OR “multidisciplinary care” OR “hospitalist”) (in abstracts).

The search terms were chosen to reflect our particular interest in the Mayo Clinic model of care and ICC in a hospital setting, but with some additional terms included to reflect some of the most commonly used descriptors which are close in meaning to “integrated clinical care”. The term “hospitalist” was included because it is commonly used in the USA to describe a class of in-hospital physician who is charged with medical management of patients while they are present within the hospital environment (Wachter, 1999).

The rapid review was carried out by two members of the research team, and the search terms were used to interrogate the following sources:

• Academic databases: Medline, Web of Science and Google Scholar (the latter for background material on IC in the English NHS).

• Websites of authoritative, UK-based organisations known to have conducted research into IC, including the King’s Fund, the Health Foundation and the Nuffield Trust.

• “Grey” literature sources.

The literature search was updated in early 2017 to capture any peer reviewed or other relevant literature that was published during Phase 2 of the project. The researchers also consulted other members of the research team, who have in-depth, expert knowledge of the field, to sense check the results of the searches and to ensure that there were no omissions.

Results

Table I summarises the results of the literature search in relation to peer-reviewed journal articles potentially relevant to hospital-based ICC. Table II presents information on publications specifically concerned with the Mayo model of care.

Table I shows that a search on “integrated clinical care”, whether associated with Mayo Clinic or not, produces very few results of any relevance.

There were 95 results from Medline and 108 from Web of Science, when searching on (integrated + clinical + care). However, on reading the abstracts associated with these results, the articles were found to focus on integrated clinical pathways across primary, secondary, community and healthcare, rather than on multi-specialty, IC solely within a hospital setting.

There were only six results from (Mayo + Clinic) + (integrated + care), and these were concerned with change strategies, the application of team-based care to ACOs, creating a quality framework in a clinical environment, managing relations between physicians and hospital organisations in the USA and use of centralised patient records in a team-based clinical environment. The use of (Mayo + model) as search terms yielded only three further results, which concerned a single specialty within Mayo Clinic, care coordination across inpatient and outpatient settings and the patient experience of community-based healthcare.

There was also scant evidence concerning the use of hospitalists in the Mayo Clinic, with just one article concerning the work of hospitalists with teams of orthopaedic surgeons and another which cites the work of hospitalists with Mayo Clinic geriatricians.

When the search terms were widened to include Hospital + (integrated + clinical + care), although a large number of results were returned, these could be broadly grouped into articles concerning the integration of professional groups into hospital organisations, integration between hospital, primary and community care organisations, “product lines” within hospitals, overcoming professional silos, integration of financial and IT systems, hub/spoke hospital models, integration of specific specialties (e.g. ICU and palliative care) or managed care programmes in and out of hospital.

The rapid review of literature which directly addresses the Mayo Clinic model of care (Table II) reveal only two peer-reviewed journal articles (Viggiano et al., 2007; Berry and Beckham, 2014), and these are largely descriptive accounts. Other literature comprises two books (Mayo Clinic, 2014; Berry and Seltman, 2008) and a small number of reports produced by professional associations or healthcare policy and practice foundations such as the Nuffield Trust, the Royal College of General Practitioners, the Commonwealth Fund and the Mayo Foundation (Casalino, 2011; Royal College of General Practitioners, 2014; Curtright et al., 2000).

Discussion

This section covers two areas: the background and context of IC as recently understood in the English NHS, and the Mayo Clinic model of hospital-based ICC.

Conclusions

Within the context of the English NHS, “IC” is a loosely defined term which has become associated with a number of different approaches to bringing about organisational, clinical, cultural and financial reconfiguration and reform, mainly in relation to better communication between, and coordination of, staff working within healthcare and/or between health and social care. For the purposes of our research study, however, we are interested in integrated clinical care within a hospital setting which allows physicians from a range of specialties to focus their attention, in as efficient a way as possible, on the holistic health needs of their patients, supported by non-clinical colleagues and effective technologies.

The forms of integration currently being developed in the English NHS through the new care models (Vanguards), STPs, ACSs and ACOs may (or may not) lead to better patient outcomes, improved patient and staff satisfaction and greater productivity, but they do not appear to address directly our focus of interest: the patient’s journey specifically within the hospital setting, following referral to a hospital specialist. In particular, it is not clear at present if any of the proposed models of IC will overcome the episodic and fragmented pathways within hospitals, experienced by patients with chronic illness or multi-morbidities—frequently elderly people and the most vulnerable patients.

Mayo Clinic has reportedly developed an organisational culture and associated processes which ensure that most patients are attended by a multi-specialty group of clinicians who collectively address the “whole patient”, managed by a coordinating physician. A small number of authors and commentators have identified a range of key enabling factors in this model, as summarised in Table III. These authors and commentators have not provided evidence of original independent empirical academic research into the efficacy of the Mayo model and, on the basis of our rapid literature review, it would be difficult to reach definitive conclusions about which elements of Mayo Clinic’s practice apparently have led to success in terms of the quality of care provided to patients. Nonetheless, the anecdotal evidence overwhelmingly suggests that Mayo Clinic’s clinical model could address our research focus: the patient journey on referral to a hospital specialist and the management of multi-morbid, chronically ill patients within the hospital environment.

Furthermore, it is arguably the case that, in the context of English NHS hospitals, some of these factors are clearly present in most English NHS hospitals, while others are less well represented or currently absent, as indicated in Table III.

If the Mayo Model offers non-emergency, multi-morbid patients a more holistic, multi-specialist and highly coordinated approach to their care, achieving high-quality outcomes and greater patient and staff satisfaction, Table III suggests that a trial of the Mayo Model in the English NHS would require development of two key staff competencies:

1. Specialist hospital doctors with the means and authority to collaborate with colleagues in other specialities to provide advice and guidance in a timely way, in order to provide a highly patient-centred care plan which takes account of the totality of the patient’s needs.

2. Administrators who are solely tasked with obtaining optimal, in-hospital patient flow, including the scheduling of patient appointments, testing and diagnostic procedures and timely, error-free provision of clinical information.

In addition, a Mayo Model for the English NHS would also require significant further investment in information and communication technology, to ensure near real-time access to diagnostic information and to enable the closest possible coordination with primary and community care clinicians, before and following hospital visits.

We therefore argue that it would be worthwhile, assuming that the results of Phase 2 of our research study do not suggest insurmountable barriers, to test this model in an NHS hospital environment, and to investigate whether it could lead to improved outcomes for patients, greater resource efficiencies and, potentially, cost savings overall. The key enabling factors of the Mayo model, as described above, could provide a testable framework against which to evaluate the success, or otherwise, of such a pilot project.