Results from the Nuffield Trust’s review of the community based models of care relied on by STPs are a warning against setting expectations too high, says Candace Imison

In 1920 Lord Dawson set out a vision for future medical services that looks eerily like the Five Year Forward View. He argued that hospitals should be supported by a network of supporting community provision and that: “This organisation is needed on the grounds of efficiency and cost, and is necessary alike in the interest of the public and the medical profession”.

A hundred years later, we are still trying to execute that vision.  And while the number of beds in hospitals has fallen, activity within hospitals continues to grow. In the last eight years alone, outpatient activity has grown by 26 per cent and emergency admissions by 14 per cent.

Many believe, like Lord Dawson, that a more community based model of care would be more cost effective. Nationally, £4bn, nearly a fifth of the £22bn savings required by 2020/21 are anticipated to come from “new models of care”. Many of the STPs assume that they will significantly reduce hospital activity, in some cases by up to a third in the next three years, and at the same time deliver savings.

Evidence and expectations

At the Nuffield Trust, we have undertaken a considerable number of evaluations of community based models of care, including the early evaluations of the integrated care pioneers and virtual wards.

While these have often found positive outcomes for patients, they frequently show disappointing results with respect to cost and reductions in hospital activity. So we thought it timely to undertake a more comprehensive review of the initiatives that STPs are relying on, to see how the evidence for demand reduction and cost savings stacks up against expectations. 

We reviewed 27 initiatives in total. They included changes to the elective and emergency care pathways; initiatives to avoid admission and facilitate discharge; and initiatives to better support patients in the community including support for self-care and social prescribing.

The results stand as a warning against setting our expectations too high.

Key success factors seemed to be accurate targeting of the “at risk” population; ensuring staff are well trained and supported; and not attempting to duplicate hospital services

While seven of the 27 initiatives showed relatively good evidence of a potential to reduce hospital activity and deliver savings, the majority had mixed evidence for their capacity to reduce cost. In fact, six could actually increase costs. The seven with the most positive evidence were initiatives to improve GP access to specialist expertise; encourage ambulance or paramedic triage to community; provide additional clinical support to nursing homes; improve community support at the end of life; offer active rehabilitation or remote monitoring for people certain long term conditions and support for self-care.

Key success factors seemed to be accurate targeting of the “at risk” population; ensuring staff are well trained and supported; and not attempting to duplicate hospital services.

We found that services that tried to replicate hospital provision in the community rarely saved money and could cost more. These included consultant and GPwSI clinics in the community and urgent care centres (if not co-located with A&E). The initiatives with mixed evidence include many that are often promoted as cost saving. These include care coordination and case management, virtual wards, shared decision making, hospital at home, rapid response services and bed-based intermediate care services.

Powerful change

A range of factors make savings hard to deliver. By introducing new, more accessible and lower cost alternatives to higher cost services, you may inflate overall demand and reduce referral thresholds. Care coordination programmes are costly, and without very precise targeting, their cost can more than offset the savings from hospital care. Targeting overuse is likely to uncover underuse. And finally, there is a tendency to equate price with cost.

Activity withdrawn from secondary care often leaves the secondary provider with fixed or “stranded” costs that are hard to release, particularly if initiatives do not address the whole referral base, so shifts of activity from secondary care can magnify costs overall.  

Is the real problem that so far we haven’t been radical enough in the scale and scope of change? Maybe the initiatives are too small and underpowered to really change things, particularly in the face of a set of system incentives that are stacked in the hospitals’ favour.

But standing in the way of more powerful change are three things: The first is time. Time is needed to deliver change and to demonstrate impact. As many have argued, the Five Year Forward View should really have been at least a 10 year programme.

The second is people. Over the last fifteen years we have grown the hospital workforce at a much faster rate than the workforce in the community, leaving community services ill equipped to take on what is asked of them.

The third, critically, is money. Transformation requires additional funding to build up capacity in the community first, before we can expect it to be released from hospitals.

The NHS is undertaking the herculean task of changing its modus operandi at the same time as experiencing the leanest years in its history.  Nobody can argue against the principle of better, more appropriate care closer to home. But sadly we cannot assume that this will save money, especially in the short term. Instead, to succeed, we need a relentless focus on what works, and, crucially, to admit when the funding envelope simply isn’t big enough to deliver the transformation needed.

Candace Imison is director of policy at the Nuffield Trust