Audit Commission

Download
(PDF, 246K)

Benchmarking NHS payment by results

Background

Data is the lifeblood of most organisations and nowhere more so than within the NHS. Approximately 12 million hospital admissions take place in England each year accounting for over 50 million bed-days across 172 Acute trusts. The accuracy of the data recorded for each hospital admission is imperative to ensure the smooth operation of the health service. Not only does it assist the planning and resourcing of future healthcare, but under the Payment by Results (PbR) system it also directly affects the way trusts are reimbursed for the patient care they provide, through a transparent, rules-based costing system with payments linked to a national tariff.

The Department of Health asked the Audit Commission to develop a Data Assurance Framework for PbR to support the accuracy of inpatient data underpinning the new financial regime. A key element of the framework would be a programme of audits focused on clinical coding, a term used to describe the process of translating a patient's medical case notes into a set of internationally recognised codes. The translation relies heavily on the skills of a team of clinical coders and is therefore a potential source of error in patient data.

IPL's Role

Benchmarking is a technique used to compare similar entities, looking for areas where an entity appears to lie outside the bounds of expected behaviour, known as an “outlier”. Following an initial pilot study, a senior IPL consultant was brought in under IPL’s supply contract with OGCbuying.solutions to assist the Audit Commission with putting together a formal benchmarking process for the Data Assurance Framework. The goal was to define and develop a robust methodology that would help target the clinical coding audits at those areas within each trust where data quality was most likely to be an issue.

The consultant worked closely with the Audit Commission's statistical experts and those of the Information Centre for Health and Social Care, for a period of several months. During this time, the consultant helped to identify suitable data feeds, develop a robust set of performance indicators and create a rigorous statistical process for identifying outliers. This involved acquiring a detailed understanding of health data and translating highly complex statistical theories into a set of clear, unambiguous business requirements.

Once the requirements analysis was complete, the consultant provided additional project management expertise, leading a team of IPL developers to produce a data warehouse and associated tools to support the benchmarking process.

Outcome

Six months after IPL began work, the Audit Commission's programme of clinical coding audits commenced, driven by a data warehouse containing over 250 million records. This represents the biggest and most complex dataset ever compiled by the Commission. Furthermore, the benchmarking methodology, with its 22 performance indicators and use of funnel plot analysis has now gained considerable respect throughout the health community and is widely recognised for its depth and rigour.

Four months later, the scope of the work was extended with the launch of the National Benchmarker, an online tool to allow NHS organisations to access the PbR dataset and associated analyses directly. This will encourage self-monitoring and help trusts identify issues independently of the clinical coding audits. In the future, the work of the Audit Commission will help drive further improvements in data quality through the analysis of outpatient data – an even larger dataset.

IPL has continued to provide invaluable expertise throughout and as the second year of the audit programme approaches, the original consultant is helping to refine the methodology to ensure it remains at the cutting edge of health data analysis.