One of the most distinctive characteristics of NHS trusts' work with US company Kaiser Permanente is also one of the least commented upon. Kaiser Health Plan is an insurance company in an exclusive partnership with the Permanente Medical Group, run as a profit share company for the participating partner doctors.

It is the health plan which manages the capital infrastructure, with the medical group concentrating on how best to treat their covered population for an annual sum.

In this context, any use of a hospital bed becomes a cost to the service. Doctors who want to concentrate their spend on diagnosis and intervention only use a bed where this is a necessary part of treatment. The cost of the bed will come from the limited specialty budget and is perceived to be at the expense of pharmaceutical or surgical intervention.

Despite the Department of Health's interest in the Kaiser model, they have presided over the introduction of a payments system which incentivises the opposite behaviour.

Payment by results is perhaps one of the finest misnomers of recent years. It is of course nothing to do with results (outcomes), but entirely about activity (process). Increases in hospital activity bring increased income, making the bed not a cost centre but a source of income generation. This is a fundamentally different model from Kaiser, and strange in the context of a national strategy which claims to be about supporting 'out of hospital' care.

One of the much trumpeted claims of payment by results was that it would enable money to follow the patient. This was most clearly to be the case with emergency medical admissions, the intention being a financial incentive for primary care trusts to manage down demand.

The potential for this to act as a driver was immediately ameliorated so that only half of costs/savings could be realised in each case. Successful demand management may require significant investment in community alternatives by PCTs and seems to be limiting admissions.

But there is little sign of a re-patriation of the money happening at the same pace as re-patriation of activity. Triage systems and prior approval may take huge swathes of activity out of traditional settings, but the traditional settings remain.

The activity may shift, but if that leaves a hospital with capital charges on an under-utilised asset and an expensive workforce, there is every incentive to find new patients to fill the space. We not only need to find a way to shift the work, but to reduce the capacity and overheads at the same time.

Sweden has a state-mandated healthcare system, but taxes are raised locally on 'districts' similar in scale to the reconfigured PCTs. Eighty per cent of the role of the county councils is in commissioning and providing of healthcare. Like the NHS, the traditional model was of monopoly supply through local public provision.

Over the last ten years however, there has been a significant development in the healthcare market, with councils now typically having around 30 per cent of their services commissioned from independent providers.

In the Swedish town of Simrishamn, the local hospital buildings are owned by the council but five years ago, the service was tendered and won by a medical co-operative. That body leased the hospital building from the council and provided a range of elective surgery and beds for non-life threatening medical admissions, in addition to a range of primary medical services.

There was a good relationship between council and provider, but a concern over the life of the contract, that the elective surgery could be provided more efficiently and safely in other dedicated centres. At the contract break point, the council re-tendered with a different specification and awarded it to a provider that was more focused on long-term condition management.

This provider, Carema, sought to make better use of the integration of the primary medical clinics with the hospital infrastructure. They were very aware that every square metre of hospital leased from the council was a cost. Their tender required only half of the hospital space previously used, excluding the elective surgical space, concentrating their expenditure on staff and pharmaceuticals rather than beds and corridors.

This separation of service and asset offers real benefits not only in changing the incentives, but in the flexibility of the service. Rather than being tied into making full use of an existing defined estate, the provider is focused on a service, and having an environment which will support best clinical processes.

Hospitals have very specific and expensive design requirements to safely manage the acute and vulnerable patient, but the majority of emergency work does not require this capital investment. The separation encourages the development of service models able to operate out of a variety of existing buildings with minimal investment, and encourages a culture of leasing, which builds in flexibility.

As assets become a cost to the system, we also become much more conscious of their potential value. A model in which assets were managed as 'trusts' in their own right, separate from service delivery, would ensure that at least one group of people would be focused on ensuring that they added value and didn't just become a cost to the system. In Sweden and Kaiser they sit with the commissioner, but other interesting models may include public-private partnerships such as local improvement finance trusts.

This begs a serious question about our current foundation trust model. Should the asset travel into the new organisation, only part of the NHS at arm's length? Or should the services move and the buildings stay in a different arrangement; one which would genuinely incentivise 'care closer to home'?

Sophia Christie is chief executive of Birmingham East and North PCT.