Feature / The Virtue of Patients

01 October 2007

Login to access this content

Like many NHS finance managers, accountants at Salford Royal NHS Foundation Trust thought they knew a thing or two about costing. Virtually word perfect in the Department of Health costing manual and years of experience filing reference cost returns – what else was there to know? But a trip to a Rolls-Royce factory in Derby to see how the world-famous engineering firm costed the production of an aero engine made the trust realise that a completely different approach was needed.

Tony Whitfield, finance director at Salford (above), reflects that it was something of a ‘Road to Damascus’ experience for the three accountants he sent to the Rolls-Royce plant. They were taken aback by the level of precision, not just in the engineering process but also in the approach to costing. And they were impressed with the desire of the production people to look at every opportunity to reduce costs in an industry in which lives depend on the quality and reliability of the finished product. 

‘They were costing engines because they had to sell them competitively,’ says Mr Whitfield. ‘If they got it wrong, they might win an order that actually lost them money or turn down an order that they could have made money on.’

He argues that the national tariff means the same dynamic now exists in the NHS. Before recent reforms, costing had little direct relevance in the local health service and even less impact on how services were delivered.

Mr Whitfield adds that the tariff also means ‘benchmarking is no longer an academic exercise’, no longer easily dismissed as failing to recognise unique circumstances or because of suspicions over data accuracy. He says the NHS needs to go beyond the crude averages at the heart of reference costs and start ‘costing for business’ rather than simply costing for the Department of Health returns.

At Salford this has meant an early plunge into the world of patient level costing (PLC) – with six months looking at costing principles and systems and a further four months to design and implement the first cut of PLC. It has been a timely move. During the same period, foundation trust regulator Monitor, predominantly under the leadership of its policy director Robert Harris, has been making the case for the need for organisations to understand their profitability and has started to get vocal about the benefits (and in some cases requirements) for foundation trusts to report at the service line level. It is clear to Salford that PLC can provide the means of identifying the cause and effect of resource utilisation and what is actually driving service line performance.

Stephen Kennedy, Mr Whitfield’s deputy and one of the accountants to have visited Rolls-Royce, admits that Salford is navigating without a map. ‘There is no textbook definition of what constitutes PLC,’ he says. The NHS is currently looking to develop some standards, but trusts need to look at costing their own organisations using their own skill and judgement, remembering that in Salford’s case 40% of income remains outside national tariffs.

The core of PLC is attaching costs to patients – that is, using commercial activity-based costing techniques to produce patient costs that are built up from the component parts of a patient’s stay in the hospital.

Once the patient cost is established, the data can be aggregated in whatever way is useful – for instance, at service line level, healthcare resource group level, or procedure code level (or, for that matter, any currency that is given the final go-ahead by the Department following its review of classification systems).

With reference cost information, looking for trends, say, in theatre costs was a fairly meaningless exercise as the simple apportionment approach made it virtually impossible to draw reliable conclusions. But by building costs up from the lowest possible level – the patient – there is the potential for a much richer conversation with the true influencers of cost – clinical staff.

If the traditional costing manual has been thrown away, where should trusts start? Dedicated PLC systems are not essential, although they can simplify the collection of huge amounts of data from disparate sources. But this is not rocket science – using patient identification numbers to distinguish different episodes and assign costs on the basis of actual consumption of resources (minutes in theatre, for instance, or minutes – not nights – on ward). A key issue for any PLC system will be how the information is presented and how easily it can be analysed – a major factor in Salford’s final decision to use the Bellis-Jones Hill PLC Service Line Management System.

There has been some hesitancy on behalf of trusts to choose a patient costing system for fear that an alternative package could eventually become the de facto national system. But Mr Whitfield warns colleagues that they are wasting precious time. Eight systems were highlighted by the Department of Health (Healthcare Finance, June 2007, page 16) as fit for purpose, with choice depending on local circumstances. He says trusts need to be clear about their requirements before entering the market, and be very careful in understanding how their preferred system will meet their current and future system needs. But he believes that trusts should not be put off by costs. At typically tens of thousands of pounds in upfront outlay, he says the software is probably the smallest investment that organisations will have to make in delivering PLC. 

The real costs lie in the resources needed for implementation. Trusts need to take patient costing seriously, which means freeing up senior managers to lead both the technical development within finance and to engage with clinicians about what they can expect from the project. Whatever the costs, Mr Whitfield says the potential return on this investment may be measured in millions. 

He insists the real imperative is for trusts to get going. ‘You can’t afford to wait for a preferred system to be established,’ he says. Accurate patient level costs will not materialise over night. Salford has embarked on a two-year project in which the quality of the information produced will get better the greater the level of clinical engagement. But PLC will deliver significant benefits throughout that whole journey. He adds that however long trusts stall, there is never likely to be a system that can be implemented off the shelf. There will always be significant local development and refinement needed. 

Mr Kennedy says the first step towards PLC is identifying all the feeder systems – theatres, radiology, pharmacy, pathology, wards – and then rating them in terms of those needing most urgent attention to deliver accurate data at the right level of detail. For Salford, improving nursing time and junior doctor rostering feeder systems are current priorities. He stresses that poor or missing feeder systems are no excuse for inactivity. And he says that going through the motions and getting clinicians accustomed to looking at the underpinning data driving overall performance is important. 

Any SLR/PLC system will only be valuable if it leads to action being taken on the back of the information it provides. It is not an exercise in producing more financial information to sit on finance department shelves or to meet central demands for data. This means ensuring data is analysed and presented in as meaningful a way as possible to the people who dictate how resources are consumed – clinicians. 

Reports from some early implementer sites suggest clinicians are engaged by patient level information, but the idea that they are desperate for such information – a picture sometimes painted – may be stretching matters. Mr Whitfield says the key is to remember that SLR and patient costing is not all about money, but about providing opportunities to improve services, part of which is cutting out waste. The ability to combine quality information and patient experience data will be key to ensuring that clinicians buy in to the idea. 

To support this wider perspective on Salford’s interpretation of SLR, the trust is using the same system to report using a balanced scorecard approach. The trust presents key non-financial information alongside better, and more relevant, financial data. Every department, ward and support area will have its own balanced scorecard presented and updated regularly on the web, with metrics that are relevant to the organisation and the specific area. 

According to head of performance Edd Berry, Salford is advanced in its development of electronic patient records, which contain a rich source of information, such as treatment pathways and clinical outcomes. ‘Placing key metrics selected by clinicians alongside measures of patient and staff experience and detailed patient costs will give staff, for the first time, a truly holistic view of their services,’ he says.

Claire Wilson, financial lead on patient costing at Salford, stresses the importance of taking a rounded view. She says that while there is a major interest in profit and costs, the conversation with clinicians needs to be more than ‘why are you making a loss on these patients?’ Although PLC data will inform regular board reports, the trust does not envisage simply issuing specialties with reports on a monthly basis. ‘Instead, we will update the system on a regular basis and provide training and assistance to specific teams to enable them to analyse the data themselves using a set of facilities and tools,’ she says. 

There is huge enthusiasm at Salford for the project. With a first report that was due to go to the board at the end of September, the project is now being launched on the trust’s service departments. There are already a number of headline issues emerging, which will be investigated further. For instance, one immediate benefit has been the greater precision of cost and activity attribution, resulting in a better understanding of sub-specialty areas such as spinal surgery. 

There is huge interest among foundation and other trusts in PLC. But to date there is little experience of actually producing patient level data. With Salford in such an advanced position, all eyes will be on the North West trust.

Getting Started

Salford’s Claire Wilson provides an insight into the practicalities of setting up the PLC system at Salford and shares the early lessons:

1. Organisational sign-up is vital. The project will fail unless it has high-level support within the organisation. The project steering group at Salford includes most of the executive team and other key senior managers, and the work is given a high priority. The steering group ensures that the work is seen within the bigger picture and that its interaction with other initiatives, such as the quality agenda, is clear.

2. The project needs dedicated resources from both finance and informatics, as well as the support and input of the wider finance team and clinical colleagues.

3. Project planning is vital. With a variety of data sources to pull together, it is crucial that everyone knows the key dates and requirements. Also, when delivering something with such huge potential, it is important to manage expectations. What are the key deliverables for each stage of the project and in what timeframe?

4. You need to spend time early on identifying how the costing model should be set up to deliver the organisation’s information needs. High volumes of patient level information will mean little to anyone without careful thought about the outputs and analysis needed.

5. Feeder systems play a key role. The quality of what you get will vary. Some systems will need to be developed in terms of how they are used. In some circumstances, new systems will be needed.

Using a work-around in the meantime is fine as long as everyone knows how reliable the information is.

6. Data from feeder systems needs to be attached to each ‘patient episode’ using a patient identifier (such as the hospital number), date and ‘what happened’. When episodes of care do not have unique dates, matching will require additional logic to enable activity and costs to be matched to the correct patient episode. For example, in the case of accident and emergency (A&E), where a patient could be admitted on the same day as the A&E attendance, it will also be necessary to differentiate the location from where the request originates – for example, from the A&E department rather than medical ward.

7. Cost weights are central to the accuracy of information produced. Start working early with clinical colleagues to establish accurate cost weights for charging out activities to patients. For example, pathology tests and radiology scans will need a weighting based on staff time and non-pay costs.

8. Ward nursing and medical staff are not part of ‘feeder system’ data but are obviously a significant proportion of expenditure.

There are many ways of understanding how costs are consumed at an individual patient level, which go beyond a crude calculation on length of stay. Consultant job plans, electronic patient records and nursing acuity systems are all rich sources of information.

9. Reconciliation is crucial given the different feeder systems in play.

10. Patient level data cannot normally be viewed or manipulated in Excel because of the number of records, so it is time to brush up on your database skills.

11. Service line management needs a good understanding of all income received at the individual service line and requires non-payment by results income to be identified down to the patient level. The project will need to decide the most appropriate way of reporting certain elements of income. For example, does it make sense to cascade non-elective threshold adjustments down to individual patients or is this better reflected as an adjustment at point of delivery?

12. Levy income/expenditure must be carefully worked through so that the organisation understands the true cost and financial position on research and development and education.

Claire Wilson is corporate reporting and project accountant at Salford Royal NHS Foundation Trust