It has been suggested that recent national hospital job cuts in nursing, medical and administrative areas may be a direct result of a loss of expected revenue from PbR due to inaccuracies in coding. This could result in further job cuts, trust downsizing, and even hospital closure.

Payment by results was introduced into the NHS to establish a fair financial infrastructure, reward work volume and create a cost efficient organisation through competition.

It currently covers elective and non-elective admissions, accident and emergency and outpatient appointments, and£22b worth of services will be funded by PbR during 2006-07. This will rise to 90 per cent of NHS hospital activity by 2008 and trusts will have few alternative sources of income once the system is fully implemented.

It is a fixed national payment system based on healthcare resource groups similar to systems used in most of Europe, the US and Australia. As poor documentation and coding can result in an incorrectly low tariff and loss of trust revenue, the need for accurate coding is essential..

Hospital coding departments generate the HRG code from the primary medical diagnosis, co-morbidity and inpatient complications (ICD-10 codes), the surgical procedure (OPCS-4 codes), patient age and length of stay. Additional costs per day are incurred if the average anticipated stay is exceeded and emergency work is more costly than elective. Each HRG carries an average cost which is agreed nationally, with regional variations to reflect cost of living..

The importance of accurate data coding
Methods of documentation and data collection vary in different trusts. Coding departments derive tariffs from basic discharge summaries, medical case notes or a combination of the two..

Of the nine acute trusts in the northern region, five code from case notes, two use discharge summaries and two use a combination. There have been reports of 40 per cent error rates in coding. Our audit work has shown that case notes produce accurate tariffs, but a loss of income of 20 per cent was reported when discharge summaries alone were used..

To illustrate this, an orthopaedic department with an income of£30m per year (based on local trust figures) may lose£6m due to coding error. Across all departments (assuming similar levels of inaccuracies) this may represent a huge loss of income per year for the trust involved.

Many trust are currently forecasting heavy deficits for this financial year. It has been suggested that recent national hospital job cuts in nursing, medical and administrative areas may be a direct result of a loss of expected revenue from PbR due to inaccuracies in coding. This could result in further job cuts, trust downsizing, and even hospital closure.

Problems with PbR
PbR will create competition between providers to drive down costs, possibly to the detriment of patient care. Providers unable to compete will attract fewer patients and less money. In addition, the current HRG classification is limited and cannot appreciate the many differences in case-mix complexity. Fixed tariffs may be inappropriate in some cases but several US studies have shown that elective procedure tariffs do not reflect actual costs and often result in financial loss to the unit performing them..

The DoH report, Good Doctors, Safer Patients, put forward the use of hospital episode statistics in the process of clinical governance, appraisal and revalidation of hospital consultants. However, reports have shown that data routinely collected for these episodes is not good enough for monitoring performance of individual doctors. Over 80 per cent of physicians had little or no confidence in centrally held data concerning their practice. Reasons for this were inaccuracies and questionable relevance of data collected. If the accuracy of this data improves, statistics and league tables may be a reliable representation of a clinician's practice.

Improving the system
The Audit Commission reported that PbR may increase administrative costs by up to£180,000 for each trust. Some of these costs are IT investment, but most represent long-term staffing increases in information, coding and finance departments to ensure better extraction of patient data from case notes and to audit the process..

Clinical and administrative staff must work together to make improvements in poorly performing trusts. Medical staff should be aware of the implications of PbR and the need for accurate documentation. Tariffs should be derived from case notes rather than discharge summaries, despite the costs involved. Coding departments require adequate numbers of well-trained staff who can interact with clinicians and feedback on a regular basis. The accuracy of data needs to be verified by clinical staff to identify coding errors.

PbR aims to improve efficiency in the NHS. Accuracy of information collection and coding is essential to ensure sound trust finances and prevent job cuts and hospital downsizing. Reliable central data for monitoring consultant performance will be a spin-off benefit. Poorly performing trusts may need to implement changes to their coding practices. However, in an effort to drive down costs and increase profits, patient care may suffer. It is of the utmost importance that other, non-financially driven performance indicators are developed to counteract this.

Simon Jameson is a clinical fellow in trauma and orthopaedics and Mike Reed a consultant in trauma and orthopaedics at Northumbria Healthcare NHS trust. The authors would like to thank Kim Oliver and Margaret Flynn for their coding expertise in compiling information for this article.