The Audit Commission has called for “co-ordinated effort” by Royal Colleges, NHS organisations and regulators to improve record keeping in the NHS.

In a report reviewing the three years of its assurance programme for Payment by Results, the Audit Commission says there have been improvements in clinical coding, with the average error rate at trusts falling from 16% to 11%.

However, it says that coding auditors “continue to raise concerns over the quality of records they review.” Last year, the Commission excluded £600,000 or 1.7% of activity from its inpatient audit because the source documentation was too poor to review.

This is in line with NHS Connecting for Health’s audit methodology. But the Commission says “we regard these as errors” and that it will count them as errors in the future.

The Commission points out that notes and discharge summaries that are in poor condition, illegible or incomplete may “represent a clinical as well as a financial risk.”

They may also impact on the Healthcare Standardised Mortality Ratios that are reported to patients by NHS Choices and Dr Foster. The Commission says it will investigate the impact of coding errors on HSMRs next year.

The Commission has run its assurance programme since 1997. It initially focused on coding and payments for inpatient care, but last year it assessed outpatients for the first time.

Its three year report says that over the past three years, diagnosis and procedure coding has improved, although error rates at trusts varied from 0% to 28%. The Commission says that in 2010-11, it will focus on the trusts “that need to improve the most.”

The Commission also analyses the impact of coding errors on the derivation of the Healthcare Resource Groups that underpin payment under PbR. It says HRG errors actually increased last year, to an average of 9.1%, because of the introduction of HRG4, which is more susceptible to coding errors.

However, it says there is no evidence that hospitals are ‘upcoding’ (or assigning activity to more expensive HRGs to increase revenue), and that they are, if anything, being slightly underpaid under HRG4.

The new outpatient audits reveal a similar picture, with errors in recording attendances running at an average of 5.2%, with a range of 0% to 45%.

However, errors in recording attendance and procedure led to PCTs being charge an average of 1.4% than they should have been for the work done.

The Commission points out that there are wide variations between the processes for collecting the data that underpins PbR for outpatient activity, and sets out four areas in which trusts should be seeking to adopt best practice.

It says trusts need good accountability, policies and procedures, data entry and IT systems, and says that while most trusts are doing well in these areas some are not, with 10% being weak in one or more of these areas.

No trusts were felt to have ‘excellent’ IT systems and a small number were judged t have IT that was ‘below minimum requirements’, although the vast majority were judged to have ‘good’ or ‘adequate’ systems.

Link: Audit Commission: Improving Data Quality in the NHS: annual report on the PbR assurance programme.

The Audit Commission’s National Benchmarker, which enables trusts to assess how they are doing on clinical coding, data quality and PbR, was a winner in the BT E-Health Insider Awards 2008.