Costing: patient progress

02 March 2020 Steve Brown

Login to access this content

CostingThe Costing Transformation Programme reached a significant milestone with February’s publication of the national cost collection, according to NHS England and NHS Improvement. For the first time, the cost schedule for acute providers (covering the financial year 2018/19) has been built solely from patient-level costs for admitted patient care, outpatients and accident and emergency.

There remain major challenges ahead as mental health, ambulance and community services face extremely demanding deadlines if they are to join the patient-level cost fold.

And there are outstanding issues with the acute sector too, with continuing concerns about the complexity of the costing process, the time taken to publish the data and the balance for costing practitioners between time spent compiling cost data and actually using it to inform improvement work.

But first, that milestone. Jack Hardman, head of costing at NHS England and NHS Improvement, says the move to patient-level costing is a journey. It has taken five years to get here – former provider regulator Monitor announced plans for a new approach to costing using patient-level information and costing systems (PLICS) at the end of 2014.

Start of the process

While the two national bodies are keen to show what has been achieved, the publication of acute costs based on a patient-level submission really marks year one in the use of patient-level cost data, not the end of the process.

If you look just at the national publication, it may be difficult to spot what has changed. The newly named national cost collection looks very like the reference cost publications from previous years. There is still a schedule of costs, setting out national average costs for a range of currencies: including healthcare resource groups (acute activity), outpatient attendances, mental health clusters, attendances and contacts. There is also a national cost collection index to replace the previous reference cost index – providing a comparison of costs at the aggregate level for each provider (100 = average costs, 90 means costs 10% below average, 110 means costs 10% above average).

However, the process to arrive at these average costs is very different for the acute sector. Instead of a top-down approach allocating costs to individual HRGs, and trusts then submitting their average cost for each group, with the new approach providers submit costs for individual patients. The averaging calculation is then conducted centrally.

The crucial difference is that the average costs are now underpinned by huge detail. Previously, a higher than average cost for an HRG might leave a trust scratching its head as to the cause. Now it should be able to pinpoint the reasons. Are the higher than average costs due to greater complexity or patient age within the relevant HRG? Are they distorted by small numbers of high-cost outliers? Or are the costs being driven by theatre, ward care or pathology costs, for example?

Changes in approach
While the publication may look familiar, there are small differences. The schedule of reference costs no longer separately identifies the costs of excess bed days beyond HRG-specific trim points, for example. After all, the patient-specific costs submitted by providers will inherently reflect their length of stay. Data relating to an activity count below eight has also been suppressed, in line with NHS Digital’s data disclosure protocol.

At a more general level, however, is the patient-level derived version of the average costs more robust than in previous years? Mr Hardman believes this is likely to be the case. The past five years have seen a new detailed costing methodology developed and rolled out across trusts, which should mean providers are now costing in a more consistent way.

In reality, Mr Hardman says, the improvement will have been gradual. The new standards have been out for a number of years and trusts have been moving towards a bottom-up approach to costing over that period – rather than switching it on overnight.

Last year may have been the first time all acute trusts were mandated to submit data, adhering to the new standards, but the change started well before. Many acute trusts had even taken part in voluntary submissions to test the new approach and the centre’s ability to cope with the significantly increased amount of data.

The data will also have been improved by factors beyond the new guidance – which is still bedding in and faces ongoing calls for simplification. ‘We’ve benefited from the work that costing practitioners are doing at their own organisations,’ says Mr Hardman.

‘Rather than doing the submission at the aggregate reference costs level, people can see their individual patient costs. This allows trusts to look at their own internal outliers and fix any obvious errors in cost apportionment before they submit. Practitioners are more aware of any errors.’

The schedule and national cost index provide some interesting insights into the cost of NHS activity (see Headline figures). But in reality, the added value should come from the ability to go beyond the averages and identify the causes of variations in costs.

Enter the acute PLICS portal (see Portal power), released a week or so after the average cost figures were published. According to Mr Hardman, the portal ‘puts the power of the data into users’ hands’ whether that is a costing practitioner, finance business partner or frontline clinician. It does not provide access all the way down to individual patient level, but it does enable data to be analysed by treatment function code, point of delivery and HRG. Users can compare their own trust’s performance with all providers or their own selection of peer organisations.

Within this, they can check they are benchmarking with an organisation doing a comparable volume of activity. Then they have the option to drill down to look at the component costs behind these organisational averages – ward costs, nursing costs or medical costs, for example. Using a key metrics screen, users could compare average length of stay across their chosen peers. More metrics will be added during the summer.

Some of the way in which data is displayed will remind users of the Model Hospital interface and there are longer term plans to link the two systems, enabling Model Hospital users to continue their analysis of improvement opportunities by accessing the PLICS-derived cost data.

‘It is a tool to inform and ask questions,’ says Mr Hardman, with the portal giving providers a key benefit – the ability to benchmark detailed costs – as a reward for the significant effort required in the submission process.

The work involved remains a key sticking point. Last year, the HFMA raised a number of costing practitioner concerns in a report to NHS England and NHS Improvement. While the association and the costing community remain supportive of the move to collect costs at the patient level, practitioners feel that the methodology and standards are unnecessarily complex. Providers are required to identify costs across a massive number of different resources and activities, creating a large matrix of costs for each patient – even though resource costs are then collected at a summarised level.

Providers said there was too much focus on collecting small costs – many of them immaterial in accounting terms. According to one practitioner, although there were around 260 resource IDs (covering all sectors), just 26 of them accounted for 80% of the cost quantum, and 40 covered 90%. There were also concerns about the requirement to map the whole general ledger to a cost ledger as a first step – seen as ‘overly bureaucratic, overly prescriptive and of questionable value’.

Ledger mapping

NHS Improvement has been listening and it is understood that a new option for this year’s collection is being introduced – although formal notification had not been issued as Healthcare Finance went to press.

As a minimum level of compliance, providers would be allowed to map their general ledger directly to the smaller group of collection resources – with the number of collection resources being extended slightly compared with last year.

The second – preferred – level of compliance would continue to involve full mapping of the general ledger to the standardised cost ledger and subsequently to the full list of allocation resources, before grouping these to the collection resources for submission.

The new Approved costing guidance for the 2020 collection reduces the maximum possible number of resources for an acute trust (including patient facing and type 1 and type 2 support costs) to closer to 200, while also slightly increasing the collection resources.

Some practitioners feel the cost ledger should be set aside, with the standards simply providing clear definitions for the prescribed resources. While this would involve a manual mapping exercise, costing practitioners say it would be more accurate. And if the resource definitions were left unchanged between years, the bulk of the mapping work would be one-off.

However, there are also plans to require trusts to generate new PLICS feed types – including new extracts for adult critical care and high-cost drugs.

These will be introduced as part of a live collection – albeit with certain fields ‘required’ rather than ‘mandatory’, according to a February cost update.

Practitioners spoken to by Healthcare Finance were worried about the destabilising effect of these changes and the further burden on already hard-pressed costing teams.

This was especially the case given that software suppliers had also raised concerns about the plan.

There are further concerns about the late publication of the national cost data. Reference costs data was typically published at the end of November or early December. Publishing the national cost data two months later reduces costing practitioners’ window to make use of the data before becoming preoccupied again in April with the subsequent year’s submission.

NHS England and NHS Improvement acknowledge the data needs to be out earlier.

Mr Hardman points out that there have been mitigating circumstances. This has been the first year of the mandated acute submission, which has involved joint working with NHS Digital for the actual collection.

‘It has been a huge learning curve. We have streamlined the process since last summer and we believe our partners have done the same,’ he says. The goal is to return to a December publication, although this could be challenging over the next couple of years as mental health trusts, ambulance services and then community services also come on board.

However, Mr Hardman insists: ‘Everyone wants to make this quicker.’ This will need to be the case. Currently, commercial benchmarking services deliver data – albeit often just based on a software supplier’s clients – around October.

Mr Hardman is also aware of the growing interest in being able to access the data outside the provider community. In a world of system working and open book accounting – a commitment to which is supposed to be included in new system agreements – it would seem appropriate to allow commissioners to have access.

And there are wide-ranging third-party organisations – the HFMA, thinktanks, royal colleges and economic research bodies among them – that would be keen to use the data as the foundation for research and study. This would arguably help to raise awareness of the data and promote its wider use. However, Mr Hardman says that information governance issues need to be addressed before this could even be thought about.

The national cost collection data does mark a key milestone on the journey to patient-level costing. There is continuing excitement about the potential applications for this detail – supporting the delivery of value alongside outcome data and enhancing the accuracy of improvement tools such as the Model Hospital and Getting it Right First Time. Progress is being made, but there continues to be a long way to go.

Headline figures

The costs collected for 2018/19 cover £69bn and represent 61% of the total NHS expenditure of £113bn. Acute costs accounted for £54.1bn of this amount, with mental health next at £7.5bn.

Within the acute data, non-elective inpatient activity cost nearly £18bn, with elective and day cases costing £5.4bn and £4.5bn respectively. Outpatient attendances cost a further £9.6bn. The NHS undertook just over 23,000 major hip procedures (with a complications and comorbidities score of 0 or 1) at an average cost of £6,040. This was about 1.3% cheaper than in the previous year even allowing for the extra costs of excess bed days (which were separately reported in 2017/18). There was a similar small rise in cost of the most straightforward day-case cataract operations on a slightly smaller volume of activity.

But these changes (and other more dramatic swings) could be a result of several factors – changes of age or complexity within the HRG; changes in clinical practice; changes in costing practice; or efficiency.

In mental health, 62% of total costs were costed against care clusters with an average cost per cluster day of £19 (total cost of cluster period divided by the days spent in the cluster). Care contacts in community services cost £46 for nurses, £67 for health visitors and midwives, and £73 for allied health professionals.

Costing chartJust over two-thirds of ambulance costs were incurred against the ‘see, treat and convey’ currency, with each such episode costing £257 on average.

Analysis of the first mandated acute PLICS data by NHS Digital shows 52% of admitted patient care related to females and 43% to males, with the gender unknown or unspecified in 6% of cases.

There are increases in costs and activity for women aged 15 to 49, representing maternity-related admissions.

Analysis of the PLICS data by deprivation (using the indices of multiple deprivation) shows that A&E attendances and admitted patient care episodes increase with deprivation.

However, there was little difference in the total cost and count of outpatient appointments across the deprivation categories.


Portal power

The acute PLICS portal is accessible by acute providers. Many data submissions in the NHS are one way – with trusts required to submit returns but seeing no benefits from the submission.

The PLICS portal aims to change this – giving trusts the ability to analyse and compare their own costs against those of all providers or selected peers. If done in a timely way, this could replace the need for trusts to buy in additional benchmarking services to inform improvement.

Having selected peers for comparison, which can take account of the quality of a provider’s costing submission as measured using the costing assessment tool (CAT), users can then explore their potential opportunities to reduce costs. An overall opportunity for the trust is broken down by treatment function code and healthcare resource group.

A new reporting functionality shows a breakdown for a selected HRG of the costs by collection activities and collection resources. So, if a trust is higher cost than its peers, it should be able to see if higher costs are driven by ward care or theatre costs. Equally, looking at resources, are the higher costs driven by consultant, nursing or device and implant costs?

The portal also provides comparisons across eight key metrics, including average inpatient length of stay and emergency department minutes and average pathology. More metrics will be added in a portal release this summer.  

Supporting documents
Costing - patient progress