Article Text

Download PDFPDF

Factors in implementing integrated care… a duty to collaborate?
  1. Katja Behrendt1 and
  2. Parashar Ramanuj2
  1. 1Strategy and Planning Department, Royal Free London NHS Foundation Trust, London, UK
  2. 2Imperial College Health Partners, London, UK
  1. Correspondence to Dr Katja Behrendt; behrendt.katja{at}

Statistics from


System-level integration of health and care provision is currently a major goal in health systems around the world. Developed health systems are seeking to achieve this goal in strikingly similar ways. In the USA, accountable care organisations (ACOs) are a recent form of healthcare delivery mechanism in which the care for a defined patient population is incentivised to be provided in its totality by a number of different providers who are held accountable to meet certain outcomes.1 In England, the National Health Service (NHS) Long-Term Plan (LTP) was published in January 2019.2 The plan builds on the policy platform laid out in the NHS Five-Year Forward View, which articulated the need to integrate care to meet the needs of a changing population.

The LTP outlines a central policy goal to provide better coordinated care and use of resources through the form of integrated care systems (ICSs), underpinned by the management of whole population health. The move towards a more interconnected NHS is to be underpinned by a ‘duty to collaborate’ on providers and commissioners. However, is a duty to collaborate really possible? And can an externally mandated directive actually foster the type of joined-up working needed for collaboration to be successful and effective?

While a whole host of different factors influence effective collaboration within a system, in this paper, we consider three tensions we believe to be most important to balance in order to implement integrated care as laid out within the NHS LTP. These tensions span three major drivers of integrated care (table 1). The first is the direction of incentivisation can be generalised as top-down versus bottom-up directives.3 The second is the motivation to effect change, namely external versus intrinsic motivators.1 4 Lastly, we value may be derived from integrated care, and how this balances against the values of the stakeholders delivering the transformative change.5 6 While our tensions are anchored in the context of England’s healthcare system, we believe that balancing these priorities can provide learning for other systems.

Table 1

The concepts we explore in this paper have multiple definitions

Implementing system change top-down versus bottom-up

Evidence from large-scale implementation of integrated care suggests that there are two main approaches for implementation, a traditional top-down planning approach and a more bottom-up approach that enables local leadership and variation.7 8 These are not mutually exclusive but need to be complementary for effective implementation. Top-down, an enabling policy environment supports the planning, standardisation and resource allocation for change to be achieved at scale and pace. Bottom-up interventions have been highlighted as crucial to engage clinicians and local leadership to support at the frontline of care delivery.3 7

Historically, top-down directives have adopted a command-and-control approach assuming linear change processes rather than systemic thinking. The latter pays deliberate attention to the relationships between building blocks of a system, anticipating synergies and feedback mechanisms that is more appropriate for complex adaptive systems like healthcare than an action and reaction approach.9

It has been suggested that the NHS in the UK, due to its close relationship to politics, is particularly prone to top-down policy change,7 with less focus on equipping staff to lead change and improvement efforts locally. As a result, change is beholden to short-term political cycles and therefore rarely given the time needed to be embedded. This is particularly pertinent to the central policy goal of the LTP for most NHS care to be delivered by ICSs by April 2021.2 This is at odds with the evidence that successful partnership working, requires foundational relationship building often developed organically over many years if it is to bridge organisational cultures between diverse organisations.10

Importantly, experience suggests that change needs to be driven from within organisations in order to be sustainable.11 It has been consistently shown that for high performing health systems strong and long-term leadership is often a prerequisite for success, enabling organisations and systems to work together on a change journey.12 A major challenge to achieving the goal of collaboration within the NHS is how to build stable system leadership while the average length of tenure of an NHS chief executive is only 3 years.12 There is a crucial interplay between centralised policy and localised implementation efforts. The best approach may be one in which local clinical leadership and collaboration is enabled and nurtured from the bottom-up and learning events are used to spread best practice, working synergistically with top-down incentive to measure and spread impact. We would argue that in the NHS a rebalancing is needed in which there is an increased emphasis on bottom-up incentives to balance the over-reliance on top-down directives. A duty to collaborate as laid out in the LTP would not achieve this as a top-down directive.

Internal versus external incentives

Incentives are a way to promote behavioural change and can be divided into intrinsic or external motivators.4 Again these are not mutually exclusive, we argue that both are needed for successful implementation of integrated care but an over-reliance on external motivation can erode internal motivators for change delivery.4 Much of the debate on how to incentivise integrated care has focused on the use of financial incentives to promote collaboration; for example, payment reform that includes a whole episode of care across different providers can promote more joined-up care delivery.13 Yet funding is not the only lever that policy-makers should consider. Phipps-Taylor and Shortell, in their landmark study of ACOs observed that clinicians are motivated by a wide range of drivers of change. External incentives are just one aspect of these motivators; but internal motivators like mastery, autonomy and social purpose are just as important.1 These internal incentives are more difficult to observe or measure and therefore change. Hence organisations rarely use the full complement of possible incentives to promote behavioural change. Yet as Shortell pointed out, financial incentives did not always lead to sustainable change because clinicians did not always perceive them to be closely linked to outcomes that actually mattered to their patients or themselves. Shortell’s work advanced the notion that external incentives can erode intrinsic motivators like autonomy, power to lead change and social relatedness. Shortell suggested there was a tension between promoting internal motivators (like autonomy) versus restricting them to limit unwarranted variation. This is another layer of complexity that needs to be considered by policy-makers.

Like other health systems, the NHS has focused on payment structures to support a shift away from activity-based to population-based funding. From 2019/2020 onwards, further reforms will give ICSs greater control over their resources through a process of ‘earned financial autonomy’ based on their financial and operational performance. The LTP also announces an ‘integration index (…) which will measure from patient's, carer's and the public's point of view, the extent to which the local health service and its partners are genuinely providing joined-up, personalised and anticipatory care’.2

However we would argue that there is an overemphasis on external drivers of change that risks stripping away internal incentives to improve care. Insights can be drawn from the Gesundes Kinzigtal (GK) ICS in Germany which operates on a 10-year shared savings contract incorporating a general practitioner (GP) network as well as specialist and community providers.14 Unlike other integrated care mechanisms, Key Performance Indicators (KPIs) on outcomes are not part of GKs shared savings agreement. Instead, the length of the contract incentivises long-term thinking and investment in health promotion and illness prevention activities. GPs are crucial in these activities and meet regularly in quality ‘huddles’ to review data and discuss how improvements across the system can be made. An external evaluation has confirmed that people live longer and there is a comparative reduction in healthcare cost and emergency admissions, as well as increased staff and patient satisfaction.15

The GK example demonstrates that providing a framework for collaboration, allowing time for relationship-building and holding providers accountable for improving outcomes, but not directing internal processes appeals to the autonomy of clinicians and a sense that they are able to direct change for a social purpose. In terms of encouraging a ‘duty to collaborate’, this approach can achieve positive deviation without the need for heavy-handed external regulation or tight scrutiny.

Value-based healthcare versus values-based healthcare

The concept of value-based healthcare has been galvanised by Professor Michael Porter from Harvard Business School. In a departure from volume-based care, Porter defines value in healthcare as patient outcomes over cost.5 Observational and anecdotal evidence suggests some success in using this concept of value in improving patient outcomes and/or saving resources. However, these successes have not yet been replicated consistently or through interventional studies, and there has been increasing criticism about this definition of value being too narrowly focused on economic value. We would argue that for effective implementation of integrated care, a broader, values-based approach is needed.

Groenewoud argues that an overly strict economic definition of value in healthcare could lead to several perverse incentives6:

  1. A tendency to neglect patients’ personal values when they diverge from what is regarded as a desirable outcome by the system.

  2. Focus on formal healthcare vs informal care, discounting the intrinsic value of care and caregiving.

  3. A tendency to replace trust in professionals with accountability. An example of this ‘targets and performance management’ approach could be the integration measure the NHS is developing to mandate a ‘duty to collaborate’.

  4. An emphasis on choice which, although important, can undermine solidarity, which is important in universal healthcare systems.

Further work to translate the Porter concept of value to universal health systems has been conducted in the UK by Sir Muir Grey and others who expound the Triple Value Healthcare model16 and the Value-Based healthcare definition by the European Commission.17 These extend the concept of value beyond the technical dimension (outcomes over cost) to include personal and allocative dimensions, incorporating the needs and expectations of the patient and society, respectively, and not just of the system.

This broader concept is based on values that are important to integrated care: beliefs or desirable goals that healthcare professionals regard as essential to achieve the best outcomes for their patients.18 The NHS LTP has begun the process of balancing value and values by starting to develop a ‘leadership code’ that is expected to enshrine cultural values and behaviours in a professional registration scheme for senior NHS leaders. It aims to foster and embed cultures of compassion, inclusion and collaboration across the NHS. Specific actions include programmes and interventions to ensure a more diverse leadership cadre, a focus on increasing staff understanding of improvement knowledge and skills, and new pledges to better support senior leaders (including improving the approach to assurance and performance management).2

We very much welcome this plan, but remain concerned about a perceived reliance on institutional over systemic leadership. Sally Lewis’ work in Wales seems to us to strike the right balance between value and values-based healthcare—she emphasises the need for outcomes metrics to be meaningful for patients and clinicians alike in order to be worthwhile.19 While it is true that Porter’s value-based healthcare delivery model also incorporates the patient’s perspective, as outcomes measured are co-developed with the healthcare user, we would argue that it is the balance between societal, systemic and individual values from which true value emerges.18


A duty to collaborate… but is it possible?

Across the world, health systems are seeking to deliver integrated care. In this paper, we have looked at three tensions that are important to balance in order to implement integrated care. We argue that two of these tensions are mutually exclusive: top-down versus bottom-up approaches (directionality) and internal versus external incentives (motivators). These require a rebalancing so that one side of the dichotomous tug of war is not overly emphasised. In the case of directionality, an overemphasis on top-down approaches inhibits sustainability of change and scaling and spreading of transformative change.

In the case of motivation for change, an overemphasis on external incentives actually erodes internal motivators. While it is important to remove financial disincentives more importance needs to be given to non-monetary incentives to generate buy-in from the front-line staff responsible for delivering change. In the case of the final tension we have considered: the focus of healthcare delivery, value and values share a bidirectional (as opposed to a dichotomous) relationship, in which the values held by important stakeholders define the value achieved by a system. Therefore, not fully considering all the stakeholders in a system—those with ‘skin in the game’—leads to the delivery of value for the few and not for the system as a whole.

We conclude that collaboration is a value in and of itself, but that a ‘duty to collaborate’ is a top-down, external motivator that emphasises a need to generating value, rather than considering the values that matter. Hence collaboration cannot be mandated—there needs to be a focus on responsibility and trust, not just accountability. Rather than trying to quantify collaboration, policy-makers and regulators would do better to give time and enable the system to be internally motivated and live to its values. And by the system, we mean all those that derive benefit from it—staff, patients and society alike.



  • Twitter @beherendt_katja

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests PR is on the editorial board of this Journal. KB has collaborated on an NHS-funded project with Optimedis in a former role in 2017/2018 (their project ‘Gesundes Kinzigtal’ is mentioned in article).

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.