Feature / The plus factor

A ground-breaking new version of healthcare resource groups (HRG4+) has been released, bringing with it greater recognition of different levels of patient complexity and comorbidity, writes Paula Monteith of the Health and Social Care Information Centre.

The tariff this year – 2013/14 – will be the fifth tariff to be based on the version of healthcare resource groups known as HRG4. In fact, HRG4 has been around a while longer than that, with its first use in the collection of reference costs for 2006/07. It has not been static since then. Each year, there have been enhancements that see new groups added, existing groups split and revised grouping logic that determines how underpinning diagnosis and procedure codes map to which HRG. However, this year heralds a major development in the world of patient classification. Say hello to HRG4+.

We start with reference costs. Costs for 2012/13 will be collected for HRGs described in the new HRG4+ system. In total, this will involve some 2,100 groupings – a 25% increase on the 1,657 HRGs used in last year’s cost collection. Although the increase is not insignificant, the benefits of HRG4+ to ensure appropriate acknowledgement of services, and their funding requirements in the longer term, should outweigh any short-term costs.

However, what is clear is that it opens up radically enhanced opportunities for costing, cost comparisons and tariff setting.

Aim of HRGs

The underlying purpose of HRGs – or of similar diagnosis-related groups (DRGs) used across the world – is to group together clinically similar treatments that consume similar levels of resources. Put simply, HRG4+ builds on the work of its predecessors to do this more effectively than ever before. Specifically it provides better recognition of the resources consumed in treating the sickest patients and those undergoing multiple complex procedures.

It also takes a more responsive approach to the issue of age. Age can clearly be a factor in the resources needed to treat two patients with ostensibly the same illness or condition. Treating a child may well involve more – staff, specialist equipment, time – than treating an adult. But a simplistic presumption that this different resource consumption kicks in at the same age across all specialties is becoming less realistic. HRG4+ unpicks this issue.

For example, in skin surgery, the age split is now at 13, not 19. Building on an approach started last year, other chapters recognise three different age ranges – infant, child and adult. It does all this using standard nationally defined data items that NHS providers collect as routine via their commissioning data sets – so no additional data will be needed.

HRG4+ also removes some of the discrepancies that arose as a consequence of the complex sequence of steps that occur in grouping a patient event into the relevant HRG.

So, for example, in the previous system, if a patient had more than one procedure code recorded on their record, and multiple procedure logic wasn’t used in HRG grouping for that clinical area, a hierarchy list was used to identify the dominant procedure and use that to assign the HRG. This did not always lead to the most appropriate HRG being generated, and in certain cases, could require local contracts to be renegotiated and workaround solutions put in place to deliver a fair result across commissioners and providers.

Part of this issue may have been too narrow a hierarchy scale to reflect all the different procedures. In response, this range is broadened in HRG4+, increasing the potential hierarchies from 10 levels to more than 30.

But the most significant change is definitely around the way that HRG4+ recognises complexity. HRG4 basically took a binary approach to the additional complexity (and higher resource consumption) involved in treating patients with complications or comorbidities (CC). So generally, even if there were multiple CCs, it would only recognise one of them, albeit the most significant, and then allocate to an HRG split either into ‘with major’, ‘with’ or ‘without’ CC (or sometimes, more simply, 'with' or 'without').

HRG4+ takes a more sophisticated approach using summative logic to reflect the interactive nature of these CCs. The presence of a single comorbidity may well increase complexity and costs, but equally this complexity and cost could rise further for patients with multiple comorbidities or complications. So, under HRG4+, each recorded CC is assigned a score and these scores are then summed to derive HRGs based on stratified values of summed CCs.

Case study

Take an adult patient admitted to hospital with an acute myocardial infarction, unspecified primary diagnosis (ICD-10 code I21.9), who has no significant procedures (in HRG grouping terms) during their hospital stay. Under HRG4 this would generate HRG EB10Z (Actual or suspected myocardial infarction).

Under HRG4+, the HRG generated will depend upon the secondary diagnoses in the patient record, and this patient will group to one of five different HRGs, depending on what other complications and comorbidities the patient has (see box).

A CC score of 10 may be reached via 10 separate secondary diagnoses, each with a CC value of 1 (generally primary diagnoses do not count towards the summed score), or five secondary major CC diagnoses that each have a value of 2.

In the interests of improved transparency, full details of all CC values at the HRG subchapter level are available in the code to group Excel workbook that accompanies each grouper release. And detailed worked examples of how the new grouping logic works can be found in the HRG4+ chapter summaries, available to download from the casemix website*.

What we end up with is much greater flexibility to match episodes or spells of care with the right level of resources consumed. That will pay off in supporting local decision-making and service redesign and national funding policies. All international systems rely on a mix of payment models for funding services with different ranges of complexity and specialism. Top-ups, block and infrastructure payments can be found all across the globe. However, the greater recognition of complexity in HRG4+ should open up more potential options for policymakers to ensure that funding is targeted in the right place.

Phased introduction

HRG4+ will be brought in in phases. While most HRG chapters and subchapters have been revised, some have not. For a start, the summative approach to CCs is not appropriate for subchapters that rely solely on OPCS procedure codes to assign an HRG, as CCs are defined by ICD-10 diagnosis codes. This includes unbundled HRGs for radiotherapy and chemotherapy cancer treatments.

There are also chapters waiting in the wings for their 4+ makeover – largely a result of the desire to introduce a concept as fully and effectively as possible.

For paediatric medicine (subchapter PA), interventional radiology (RB), orthopaedics (HA and HB) and ENT surgery (CZ), the HRG4+ changes are planned for implementation 2013/14 reference costs. This is not to say that the HRG4+ concept is not suitable for these subchapters, but rather that the design enhancements that are being pursued by the individual expert working groups (EWGs) extend further than those originally envisaged by HRG4+.

For example, the paediatric medicine EWG wants to reorganise the subchapter structure to better align with the specialisms recognised by Prescribed specialised services for children from the NHS?Commissioning Board (now NHS England). It also wants to refine some of the existing HRGs that err on the generic side. Also, before appropriate summation can

begin in earnest, a thorough overhaul of the content of the current CC list is required, and indeed is under way.

No classification system can be perfect. Healthcare isn’t straightforward and there are lots of different demands. Nor is HRG4 broken. It is clearly working and currently being used to reimburse the NHS to the tune of more than £30bn. But in the same way that HRG4 was an evolution, improving on the existing HRG version 3.5, HRG4+ takes similar strides forward and better reflects the changing landscape and content of healthcare provision.

It offers a basis on which to develop future iterations of the casemix classification. Recognition of the impact on resource use of providing for patients who are multiply comorbid, and who may well undergo multiple significant interventions, will provide an effective foundation from which patient pathways can be developed. And not just for those patients with chronic long-term conditions, but also in support of the identification of, and funding for, specialist and specialised care.

This evolution demands a team approach. The National Casemix Office, within the Health and Social Care Information Centre, relies on NHS feedback, and experts from the NHS to take forward its work on classification. We work closely with the respective royal colleges and academies to ensure our classifications remain current – keeping step with policy and clinical practice. This engagement has been vital and will need to continue.

HRG4+ has come in with neither a bang nor a whimper – it is the next, necessary step on the casemix classification evolution ladder. To put it simply, it moves us even closer to a system of classification that accurately represents what the NHS actually does, and to whom.

Further information at www.hscic.gov.uk, including a summary of HRG changes at root level and subchapter summaries (follow links to casemix)

Paula Monteith is principal casemix consultant for design at the Health and Social Care Information Centre’s