The Audit Commission is to undertake a national programme of clinical coding audit in the NHS in England that will affect every PCT, trust and foundation trust.

The results of this national audit – due to start in 2007 and covering 74,000 episodes of care – could be put into a league table of error rates, with trusts shown to have poor data potentially facing fines.

The move by the Department of Health to hand policing of coding to the Audit Commission follows a pilot audit study that found a “high level” of coding error leading to inaccurate payments under Payment by Results.

Although this level of error is not sufficient to destabilise payment by results, it is significant, said Steve Bundred, chief executive of the Audit Commission.

"Accuracy of payments under Payment by Results depends upon high-quality data,” he said. “We are pleased that the department has agreed to the proposed independent national clinical coding audit programme to help assure and improve the quality of the data on which payments are determined.”

The Audit Commission’s pilot project, which took place in 12 trusts in South Yorkshire and Avon, Gloucestershire and Wiltshire on data from the last quarter of 2005, found a health resource group (HRG) error rate averaging 11.9 per cent and ranging from 3.5 to 28 per cent.

The absolute impact on payments ranged from £7,091 to £67,698 for the 200 episodes per trust sampled. The Audit Commission said it could not accurately extrapolate this to provide a figure for the 50,000 episodes of care undertaken in a typical large acute trust. There were cases of underpayment and overpayment by commissioners.

Some of the errors were down to coding issues including low numbers of accredited coders, insufficient training and insufficient guidance for coders. The report says: “Error rates are noticeably higher where there is a low proportion of accredited coders or a poor coding process.”

There were also non-coding errors including the quality of clinical documentation.

The national audit programme will attempt to take a statistically significant sample. The results should allow trusts to benchmark themselves. There will also be national data analysis regularly fed back to PCTs and trusts.

The national overview is needed to provide objective and independent results, says the report. “Even where there are strong local monitoring arrangements in place at commissioner and provide level, there is a need for an external and independent audit to supplement and provide additional impetus to improve data quality.”

Andy McKeon, head of health at the Audit Commission, said the National Programme for IT would boost coding. “The job of clinical coding is clearly going to become much more important,” he said.

The NHS Alliance and the Professional Association of Clinical Coders UK welcomed the move.

Sue Eve-Jones, co-director of PACC-UK said: ‘We are absolutely delighted. It is the first time there will be a large scale and comprehensive audit of coding. It’s long overdue and good that it’s going to be done.”

She said the level of error found in the pilot study did not surprise her. “We have always said that we have got serious problems in terms of error levels. There are hospitals where coders are brilliantly supported but others were they are really struggling. This will make the problems transparent.”

Dr Michael Dixon, NHS Alliance chair, said: "Payment by Results is the new currency of NHS commissioning and it is absolutely crucial that it works for PCTs, the new practice based commissioners and their patients. NHS Alliance warmly congratulates the Audit Commission on exposing weaknesses and inaccuracies in the present system and believes it has richly deserved its new role in leading the independent national coding audit programme, which will ensure a level playing field for both commissioners and providers."