The latest reorganisation of the NHS in England comes at a time of intense pressure on the system. The backlog of unmet need is growing and NHS staffing gaps currently stand at more than 100,000.41 Public satisfaction with the NHS has fallen to a 25-year low.42 Ongoing underinvestment in adult social care means that many people are going without the care they need. These challenges are compounded by a gloomy economic outlook and rising inflation that will eat up a share of planned health and care budgets.38 ICSs face a daunting task when they inherit responsibility for managing NHS services in July 2022.

But the task facing ICSs is not equal. Pressures on services and the health of the population vary widely between ICSs – as do the resources available to address them. ICSs also look very different in their size, complexity, and other characteristics. Variations within the NHS are nothing new. But these differences will shape how ICSs function and their ability to collaborate to improve services. For instance, more complex systems with little history of working together may find it harder to make decisions and agree service changes. Less complex ICSs in more affluent areas may face an easier task.

What does this mean for policy on ICSs? First, national policymakers must acknowledge the wide variation between local systems and be realistic about what different areas can achieve. ICSs have already been set a long list of objectives and national priorities – including improving the responsiveness of urgent and emergency care services, increasing access to primary care, returning the NHS to pre-pandemic levels of productivity, and more.11 NHS England is in the process of updating the NHS’s guiding strategy, the NHS Long Term Plan. This presents an opportunity to define realistic expectations for the performance of ICSs. But it should also acknowledge that progress will differ between systems, given their varied contexts and starting points.

Second, these differences should be reflected in how ICS performance is assessed and reported. The approach to doing this is currently unclear. A mix of policies are under development: a revised system oversight framework from NHS England,5a new approach to assessing ICSs from the Care Quality Commission,43 and an outcomes framework for the ‘place’ level of the system22 have all been promised by national policymakers. Measures to assess previous versions of ICSs were narrowly focused and imbalanced towards hospitals – and used to construct overall ratings for local areas.44

The right approach to measuring local health system performance depends on policy objectives – for example, whether policymakers want to judge ICS performance or support local systems to improve services.45 Any future approach to measuring ICS performance should reflect the broad range of objectives for local systems and avoid overly simplistic comparisons between areas. Publicly reported data should also link ICS performance to the underlying resources and context in each area – for example, to allow comparisons on relevant indicators between ICSs with similar levels of deprivation.

Finally, policymakers must provide targeted support and resources for ICSs with different needs. For example, national NHS bodies are seeking to reduce health inequalities by targeting interventions at people living in the 20% most deprived areas of the country. But the most deprived areas are not evenly distributed between ICSs (Figure 2). Cuts to local government spending have also been greater in more deprived areas.46 As a result, ICSs in some areas will likely require additional resources to help deliver national policy objectives. The same is true within ICSs, where variations in health and use of services exist between ‘places’ and neighbourhoods within them – for example, in the provision of high-quality GP services between richer and poorer neighbourhoods in England.33,47 National NHS bodies must also focus their support for ICSs on the broader range of factors needed for local partnerships to operate effectively – including culture, management, use of data, and more.9