Feature / A measured revolution
In the future tariffs may be based on patient-level costs, but in the short term reference costs will continue to play a key role in tariff calculation. Steve Brown reports
Perhaps it is overstating things to talk about a revolution in NHS costing practice. It will certainly not register with the wider public and many within the NHS will remain blissfully unaware. However, the current transfer of responsibilities for pricing from the Department of Health to Monitor and the Commissioning Board – and Monitor’s interest in the costing data to inform this task – is likely to mean a much more rapid adoption of costing at the patient level.
The HFMA is involved, having worked with the Department and subsequently taken on the task of developing the clinical costing standards to drive improvement in patient costing. The future is potentially very exciting. Clinicians engaging with cost data for individual patients – a level that makes sense to them. Prices set on the basis of a deeper understanding of how the costs are built up across theatres, medical staff, drugs and wards, for example. And rich data supporting detailed benchmarking to help drive improvement and efficiency.
But regardless of the potential, the here and now remains very much about the average costs of healthcare resource groups (HRGs), without the greater understanding provided by underpinning patient level costs.
These averages continue to be captured and published as part of the annual reference costs exercise and the Department published the latest telephone directory-like tome of cost data in early November.
Despite Monitor’s patient-level vision for costing, reference costs will be the bedrock of tariff pricing for the near future – a point highlighted in Monitor’s new costing guidance and vision document.
Costing connections
Of course, HRG costing undertaken for reference costs and patient-level costing are not – or at least do not have to be – completely separate activities. Some 89 out of 93 trusts that have implemented patient-level costing claim to be using patient-level systems to inform some or all of their most recent reference costs return.
However, the continued use of reference costs underlines the need to ensure that the completion of the reference costs return is not simply seen as an irrelevant sideshow.
Reference costs rubbishing is a favourite pastime for many finance managers. They claim – with some grounds – that as an average of an average, they are meaningless and a turn-off for clinicians. But reference costs do have uses (not least within the tariff setting process) and even the crude headlines make for interesting reading.
For example, the £53.4bn of total expenditure collected for 2011/12 – up
£0.5bn from 2010/11 – represents more than 50% of the total revenue spend in the NHS (£101.6bn). Costs were provided for more than 1,600 treatments or procedures covering more than 15 million episodes within admitted patient care alone.
And the overall crude average costs – in many ways an average of an average of an average, if you like – would provide interesting reading on any accident and emergency or
GP waiting room wall. The average elective inpatient (excluding excess bed days) cost £3,215 (compared with £3,093 in 2010/11).
A day case cost £682 on average (£668).
Non-elective admissions (excluding excess bed days) cost an average of £1,436 (£1,399), with excess bed days above HRG-specific trim points adding an extra £264 per day (£258). Meanwhile, each outpatient appointment clocked up a further £106 and, in the case of A&E attendance, £108.
More meaningful for practitioners are the data sets showing average costs for the different HRGs (the schedule) and the relative cost performances of different providers (the index) – see box right.
Nearly 10% of the £53.4bn was spent on elective inpatient care, while a quarter was on non-elective inpatients. A further 7% went on day cases and 14% on outpatient attendances. Nearly 40% of the total spend was in non-acute settings. Looking at the £22.5bn spent on HRGs in the acute system, three accounted for nearly 40% of total spend over the 20 chapters:
- The musculoskeletal system (HRG chapter H) – £3.8bn
- The digestive system (chapter F) – £2.8bn
- Cardiac surgery and primary cardiac conditions (chapter E) – £2.3bn.
Spell-based schedule
For the first time, the Department mandated the collection of spell-based costs alongside the normal finished consultant episode (FCE) costs. This has led to the first separate publication of cost schedules using spells.
A spell covers the period from admission to discharge within a single provider and may include more than one FCE. The tariff pays on the basis of spells not FCEs, so a complicated conversion is needed between FCE-based reference costs and spell-based tariff prices. The move towards spell-based cost collection is seen as a key step in the move towards a more transparent tariff calculation.
The spell data must be viewed with care alongside the FCE data as the two cost sets are not directly comparable. For example, spell costs include excess bed days over the HRG spell trim points and so need to be compared with the FCE costs with excess bed days and not the more frequently quoted ‘without excess bed days’. Although 90%-95% of spells comprise a single episode, on average each spell includes 1.16 FCEs, making its unit cost higher than the related FCE-based cost. The Department stresses spell costs that appear lower than FCE costs are either the result of grouper issues or poor data quality.
The collection was not without problems. Some trusts reported the same number of FCEs as spells, which seems unlikely. Some reported significantly higher numbers of FCEs per spell than the average and a small number of trusts reported all their non-elective short stay FCEs as non-elective long-stay spells. All of these would have an impact on trusts’ spell-based cost index.
However, the Department had more to say about quality in general. Monitor may be looking ahead to how patient-level costing data could be validated – perhaps using a materiality and quality score (MAQS) self-assessment tool described in the HFMA costing standards. But the continued importance of reference costs means the improvement in reference costs quality is vital.
This year finance directors were required to sign off their reference costs, confirming not just that guidance was followed and they represented a true and fair view but also that clinicians were engaged in the process and that the Audit Commission’s quality checklist had been used to check the quality of the return.
The costing workbooks also included a number of mandatory and non-mandatory validations to highlight potentially spurious data. So, for example, flags were raised on day-case unit costs that were more than double the elective inpatient cost (with organisations asked to investigate the reasons).
Similarly checks were made in cases where reported costs were less than an expected minimum cost for a high-cost device used in the activity.
Where there were problems, they could not always be fixed – for example, problems with apportionment tables that could not be fixed in the time available – but the Department says the overall result of the extra validation was better quality data. All submissions were completed by the 27 July deadline and no resubmissions were required after that date.
Monitor assumes accountability for reference costs from the 2012/13 collection, but it is now clear the Department will continue to manage the collection on Monitor’s behalf.
The Department’s role as Monitor’s agent will make the changes around costing feel even less like a revolution. But with the continuing push on quality from the Department, real interest from Monitor in driving up standards, and backing and support for improvement from the HFMA, there is a definite mood change around costing.
|
Facts and figures The latest reference costs set out how the 248 NHS trusts and foundation trusts spent £53bn delivering healthcare in 2011/12. This is £500m more than the previous year and represents more than 50% of NHS expenditure. The cost information covers more than 1,600 procedures and treatments, including more than 15 million episodes of admitted patient care alone. Reference costs include services outside the payment by results regime – in 2011/12 PBR spending was around £29bn, while the value of services covered by reference costs was £53bn. The reference cost index (RCI) for 2011/12 – where 100 equals the national average cost – ranges from 74 to 167. However, it is traditional to strip out mental health and community trusts, where the collection of activity data is known to be problematic, and ambulance trusts. Setting these aside, together with single specialty trusts, the range is 87 (Blackpool Teaching Hospitals NHS Foundation Trust) to 118 (Mid Staffordshire NHS Foundation Trust). Last year, the spread was 86 to 117. While many trusts’ RCI remained relatively static, some had significant movements in both directions. Weston Area Health NHS Trust, which had the highest RCI for general acute and teaching hospitals in 2010/11, decreased from 117 to 100 – with the trust putting the change down to ‘improvement in data capture and recording of clinical activity across acute, specialist and community services’. The RCI of Gloucestershire Hospitals NHS Foundation Trust, which last year was the lowest for this group, rose from 86 to 97. RCI ranges*
* includes general acute and teaching hospitals but excludes mental health, community, single specialty and ambulance trusts Earlier in the year, Monitor and PricewaterhouseCoopers highlighted that over 40% of tariff prices (which are based on NHS reference costs) change by 10% or more each year. While this has already led to work looking at how to improve the underlying cost data, it is also likely to mean more scrutiny of the swings in the new data for 2011/12. Changes – whether large or small – could be indicative of real changes in cost (up or down) that may be related to volume or changes in practice. But they could also reflect data issues or changes in costing practice. For example, HB12C (major hip procedures for non-trauma, category 1 with no complications or comorbidities) is one of the high-volume elective orthopaedic inpatient procedures. The average cost for 2011/12 was £5,942 (£6,009 a year earlier), while interquartile costs ranged from £5,146 to £6,523 (£5,078 to £6,618). This represents a modest fall in unit costs, and there certainly appears to be a reduction in average length of stay down to 4.4 days from 4.8 days in 2010/11 to back up this up. (This adds up to 30,000 fewer bed days for this HRG alone). But there are other things going on that might influence unit costs, including a significant change in overall volume – down to 34,400 from 37,800 – a nearly 10% fall in volume. This could indicate real changes in demand, patients being channelled to revised pathways or patients being grouped to other HRGs. As ever reference costs can only provide signposts for areas worth scrutinising further – both in terms of looking for service and cost improvement and for improvements in allocation methods. The real value of much of these high-level cost differences would come from being able to drill down into the data to see ranges across individual patients and how costs add up across different cost pool groups, which may be a short-term development for reference costs even before a switch-over to patient-level costing. The organisational data in the schedule also shows wide variations in costs for some HRGs. For example, in its commentary on the reference costs the Department points out the range of costs in a maternity HRG – normal delivery with complications and comorbidities (NZ11A) for non-elective inpatient (long stay). Though the national average unit cost for this HRG was £2,218, it found the cost varied from around £500 to just under £4,000. The schedule throws up some fascinating titbits of information. For example, 13,505 respiratory sleep studies (DZ50Z) in elective inpatients, which are used to identify sleep breathing disorders, were carried out in 2011/12. This is 5,000 more than in 2010/11. Clearly, while some high-volume procedures, particularly in orthopaedics, are falling, other areas are showing growth. Reference costs were collected for new currencies – mental healthcare clusters, cystic fibrosis year of care currency and the ambulance service currency. The 2011/12 costs will be used to inform development of benchmark costs for mental health and a tariff for cystic fibrosis in 2013/14. |
|
PLICS implementation grows Trusts have been shifting steadily towards implementing patient-level information and costing systems (PLICS) over the past seven years, but 20% of trusts are not planning to implement the systems, according to a Department of Health survey. Although the Department has been undertaking a PLICS survey alongside reference costs for the past couple of years, this was the first time the survey was mandatory for all trusts. Some 145 trusts of a total 248 (58%) have implemented or are in the process of implementing PLICS, while a further 53 trusts (21%) are planning to implement the systems. Most of the implementations are in acute trusts (88). Five mental health trusts have PLICS in place, but no community or ambulance trust has done so. Nearly all of the 93 trusts that have PLICS used their system to inform some of their 2011/12 reference cost return. PLICS data was used most in clinical areas with good data flows, such as admitted patient care and outpatients, and least in community services. Across all services, 42% of the reference costs (£22.4bn) were underpinned by PLICS. Acute trusts again dominate the group that are implementing PLICS (41), though eight mental health trusts, one ambulance and two community trusts are also doing so. No ambulance trust said it is planning to implement PLICS, but five community, 33 mental health and 15 acute trusts are doing so. This represents a marked increase in the mental health trusts planning to introduce PLICS over the previous year. Despite the move towards PLICS, 21 acute trusts (13%) are not currently planning to implement a patient-level costing system. This proportion has remained constant between 2010/11 and the 2011/12. Nine community trusts, 10 ambulance trusts and 10 mental health trusts said they were not planning to introduce PLICS. The survey also asked trusts to self-assess levels of clinical-financial engagement within their organisations – seen as integral to the costing process. Adopting an approach first used in a pair of HFMA clinical engagement surveys, trusts were asked to rate themselves between level 1 (engagement only at board level) and level 4 (collaborative working across all clinical specialties). Forty-nine trusts (20%) rated themselves at the highest level of engagement. Dr Mahmood Adil (pictured), who leads in this area for the Department, described the results as encouraging. ‘Clinicians need to be fully engaged in the costing process. That is the only way we can ensure costs are allocated in a meaningful way and that clinicians subsequently trust the data reported back to them. Establishing robust data everyone buys into is key to making progress towards improved value for patients,’ he says. He adds that the Department was planning to bring level 4 trusts together to help understand how engagement had been achieved. As part of this exercise, a guide will be developed to help other trusts to objectively review their level of engagement and its impact on service improvement and efficiency gains. Dr Adil says there was clear room for improvement with only 49 out of 248 trusts putting themselves at the highest level of engagement and learning from those trusts will help to spread best practice in the NHS. |
Related content
The Institute’s annual costing conference provides the NHS with the latest developments and guidance in NHS costing.
The value masterclass shares examples of organisations and systems that have pursued a value-driven approach and the results they have achieved.