Theo Georghiou on how many evaluations find very little sign that integrated and community care initiatives produce reductions in costs or emergency admissions to hospital 

The NHS long-term plan, like its predecessors, paints a compelling picture for the future of care for older people: that it will become increasingly more coordinated and responsive, and that as a result, pressure on hospitals will be reduced.

Over the last decade the Nuffield Trust has evaluated more than 40  promising initiatives that have aimed to prevent unnecessary admissions. But in almost every case we have found little evidence of success.

Spot the difference

Our report this week examines Age UK’s Personalised Integrated Care Programme. This innovative scheme gives personalised support to older people who are at risk of a future emergency admission, within a model of care designed to reduce fragmentation. Through this programme, Age UK aims to bring together commissioners, NHS providers, local Age UK partner organisations and others to design and deliver specific models of care for each area.

When we examined an early pilot of the programme in Cornwall as one of a varied group of voluntary sector-led care support schemes the results were encouraging. We found that these kinds of interventions had the potential to help reduce feelings of isolation in older people, and to increase the productivity of health and social care staff.

Our current work looked for evidence that Age UK’s programme helped to reduce admissions to hospital.

We did this by analysing data from eight local Age UK PICP schemes linked anonymously to national hospital records. We compared the hospital activity of clients who were referred to the schemes to that of a very similar “matched control” group. This was a group of older people who hadn’t been referred to Age UK’s service, and who would have been expected to have had very similar patterns of hospital use.

Our results were ultimately disappointing. We found that hospital admissions and costs were actually higher for those referred to the programme than they otherwise might have been, and by substantial amounts.

These higher costs persisted for well over a year after referral to the schemes, and were quite consistent across different client characteristics (for example, across different age groups).

What’s going on?

These findings are not easy to interpret, but a key hypothesis we have is that higher hospital use might have been a result of the programme helping to discover previously unidentified needs that would otherwise have been overlooked.

The older people who were at the lowest apparent risk of future emergency admissions going into the programme ended up with the highest relative increase in costs. This is potentially consistent with the idea that some of these individuals actually had extensive healthcare needs that were not being addressed until the programme had helped to identify them.

We found tentative signs that a group of clients who started to receive Age UK’s service after it had been established for a period of time bucked the overall trend, with no difference in hospital use compared to the matched control group. This is a reminder that new services can take some time to adapt themselves locally, and that the settled effects of interventions may take some time to filter through.

We’ve warned about this before, and Professor Kieran Walshe recently described in HSJ similar dynamics across many different types of healthcare innovation. Yet many people still expect that conclusive evaluations can always be carried out in one or two years.

In addition, a small number of the local PICP schemes also showed no overall difference compared to the matched control group, but it’s not clear why these areas were different from the others.

You can’t always do more for less

Our findings shouldn’t deter innovations like Age UK’s programme. Where there is evidence that such schemes benefit patients, as well as health and care staff and carers, it is right that they are considered as valuable local options for enhancing services.

But these findings – alongside those of our previous evaluations – suggest that we shouldn’t bank on the prospect of significant savings in the shorter term without good evidence suggesting otherwise. We need to stop being over-optimistic about likely financial benefits, especially from reductions in hospital activity. Often, doing more for patients is going to cost more, not less.