Productivity: working model

It has been over two years since a prototype of NHS Improvement’s Model Hospital was first released to the acute provider community. It was grown from the seed of an idea in the Carter review of productivity to show what good looks like. But it is slowly becoming an established tool as providers look to address wide-ranging variation in the performance and costs of their services.

The digital information service (see box) now includes more than 7,000 metrics spread across different compartments covering clinical service lines, support services and people. It is subject to some of the same criticisms of benchmarking solutions that have gone before – principally concerns about data quality and comparability.

But there is increasing evidence that it is overcoming this issue – even driving improvements in data quality – and establishing itself as a valuable and increasingly used tool. The national profile given to the programme – and the ministerial backing given to the Carter recommendations – goes some way to explaining it. Some 21 of the Model Hospital’s headline metrics now inform the Care Quality Commission’s use of resources assessment for trusts. This puts metrics such as pay cost per weighted activity unit (WAU), pre-procedure elective bed days, estates cost per square metre and finance cost per WAU right alongside the more established finance metrics that make up the single oversight framework.

There is a concern that this will lead to a narrow focus on this subset of metrics, but in general trusts say the use of resources link has made executive teams sit up and pay attention to the Model Hospital. A number of trusts are now actively using Model Hospital data to inform efficiency and improvement programmes, although they warn that individual indicators don’t always give an absolute indication of performance.

High pay costs per weighted activity unit (WAU) could mean exactly that – a trust is spending more than its peers on staff for the same output of healthcare. Or it could mean that some support services (catering, cleaning and portering for example) have been outsourced through commercial service arrangements with those costs then contributing instead to the non-pay costs per WAU. This would artificially lower a trust’s costs for its own staff, although the overall cost per WAU for a trust would stay the same.

Plymouth progress

Laura Langsford is the model hospital and Getting it right first time (GIRFT) programme manager at University Hospitals Plymouth NHS Trust. She says the tool really comes into its own at clinical service line level and in particular when looking across different care settings – elective, non-elective, day case, outpatients and other (outpatient imaging). The trust has been using the data to track trust-wide improvement programmes, particularly in theatres and outpatients. This builds upon its established service line reporting approach but provides a different dynamic and measure of financial efficiency and effectiveness.

‘I identify anything in the quartile 3 or 4 red position for both elective and day cases by service line and I do the same for outpatients,’ says Ms Langsford. ‘I map their position on theatre utilisation and outpatient clinic utilisation pre-commencement of these projects
and using the latest data. So, we can see how our improvement productivity programmes are starting to have an impact on our cost per WAU metrics.’

And having identified outliers by service line and setting, the trust has something solid to challenge performance. It first investigates any reasons that might explain the apparent high costs – under-recording of activity and higher level of complexity than other providers, for example. However, it would also start to explore the amount of programmed clinical time in job plans compared with the direct clinical care being delivered.

It would also look at efficiency performance – for example, in terms of clinic utilisation and did-not-attend rates within outpatients.

Service lines’ cost per WAU positions are also being factored into decisions around business cases looking to expand services. ‘If a service is more expensive than the national and peer median, Ms Langsford says it is reasonable to expect them to be able to explain their cost position before increasing the size of the workforce.

One area where Model Hospital data has contributed to major service change is ophthalmology day cases where the service was an outlier with its cost per WAU metric.

GIRFT also signposted this through the cost per procedure for cataracts. The trust was higher than the national average cost of £893 and considerably higher than some of the costs of peer organisations following some supporting patient cost benchmarking where some trusts were cheaper by £200 per procedure.

A key difference identified by the ophthalmology clinical lead was the trust’s continued use of a separate anaesthetist. This was despite some other trusts not using this approach for less complex cases and the ‘no anaesthetist’ approach being an established model in the delivery of work in a private setting. This led to a unanimous agreement from all 10 consultants within the trust to deliver one list a month in this way as part of a pilot.

This pilot is now live, but it is estimated that the change of approach will reduce the unit cost per procedure by £110 (12% of total procedure cost). This will save £66,000 in a full year at pilot volumes.

If scaled up to 50% of cases, the saving would be £155,000 and the trust believes that eventually 60% or more of activity could be undertaken this way – with no impact on productivity, outcomes or safety.

‘This sounds transformational, says Ms Langsford. ‘But it is actually what is done in the private sector, and clinicians are familiar with the practice. This will neatly reduce the cost per WAU for day cases within ophthalmology and our benchmarking will improve.’

She adds that the next step is to ‘work the outputs’ of this through with the service and identify the releasing benefits for the trust as a whole. This could mean redeploying the clinical workforce to make most effective use of resources, which may also translate into offsetting waiting list initiative payments.

Model Hospital metrics are increasingly helping trusts to reinforce business cases for transformation. For example, they helped Guy’s and St Thomas’ NHS Foundation Trust, Lewisham and Greenwich NHS Trust and Dartford and Gravesham NHS Trust make the case for their SmartTogether shared procurement service.

David Lawson, chief procurement officer at Guy’s and St Thomas’, which hosts the service, said the Model Hospital and Carter metrics had crystallised the potential for improvement. Lewisham did not have any catalogues in place before the services came together and was paying 4.5% above average prices according to the purchase price index benchmarking tool within the Model Hospital. And there were recruitment challenges across the trusts’ procurement teams.

But he thinks the Model Hospital has done more than make the case for the shared service, which has reduced the procurement overhead by 10% and already seen Lewisham’s variance from average price fall to 2.9%. ‘The Model Hospital forces accountability on performance,’ he says. ‘For us it was helpful as a catalyst to recognise our own performance [across the different trusts].’ He says it has also helped to raise the profile of procurement in the trusts. ‘With previous benchmarking tools, procurement wasn’t really visible,’ he says.

He adds that the increased transparency works both ways. It creates a ‘healthy pressure’ on procurement departments to improve, but it also enables them to demonstrate any improvement that they do make. Some of the Model Hospital metrics now feature as part of a balanced scorecard that helps all three organisations to monitor performance.

Maidstone milestones

Maidstone and Tunbridge Wells NHS Trust is another trust that is starting to embed the Model Hospital in its way of working. Patrick McGinley, head of costing and service line reporting, believes he has noticed a difference between presenting Model Hospital data and

earlier attempts to encourage teams to focus on improvement on the back of local cost data.

‘There is a real power in the fact that everyone can see the data,’ Mr McGinley says. People either want to improve performance or improve the data being recorded – both of which are good results.

More than 120 people across the trust now have access to the Model Hospital, including an increasing number of consultants. At the moment, getting teams to use the data to challenge performance involves Mr McGinley and a 12-strong project management office. But the trust hopes teams will slowly start to use the data themselves.

There have been some early wins. Following the Carter review of productivity, the trust was told it had a £44m saving opportunity if it brought high-cost areas in line with the national averages – and cardiology was one of 10 services making up the bulk of this figure. As part of a deep dive into the service, Model Hospital data has helped inform a change in cardiology outpatients.

With medical staff costs in the upper quartile and productivity below the median, cardiologists pointed out that part of the problem was the escalated admission of non-cardiology patients to angio wards.

A new model – already in operation on the Tunbridge Wells site and due to be rolled out – now sees nurse specialists leading outpatient clinics for steady state follow-up patients. This frees up consultants to spend more time on wards and support the emergency department, reducing the level of inappropriate admissions.

The change hasn’t reduced direct costs, but it has avoided the need to recruit further cardiologists, which had been proposed in a business case to cope with increased demand.

Model Hospital and GIRFT data has also been important in making the case for frailty and ambulatory care units – both of which are also aimed at getting patients the right treatment and avoiding inappropriate admissions.

‘It is too soon to see the impact in the Model Hospital – as data needs to be refreshed – but we used the Model Hospital to see the cost of operating an inefficient model,’ says Mr McGinley.

Model hospital refresh