149,988 members

Skip to content. | Skip to navigation


Counting the cost of coding errors


Blog headlines

Friday, 16 July 2010

Counting the cost of coding errors

According to a report by the Audit Commission this week, coding errors are costing the NHS £9 million a year. Whilst tackling this problem is hardly more than a drop in the ocean of the £20 billion savings the NHS is required to make, coding errors are nevertheless the tip of a very expensive iceberg.

A lot of the cases that turn up in hospital are miscoded. If no more accurate diagnosis is available, an episode might be coded as a chest infection or UTI. Getting a tick in the box is more important than getting the right box.

The consequences of bad coding, whether deliberate or accidental, are that we know less about why people are in hospital, we may be spending too much on the wrong care or too little on the right care and that future commissioning decisions are based on the wrong data. The unintended consequences of NHS targets are well known – patients kept in ambulances or needlessly admitted to avoid breaching four-hour wait times. Whatever else it does, getting rid of targets may have a positive effect on the data in the system because it will remove the temptation to bend the reporting rules.

No patients are harmed by fiddling the numbers. Discounting MRSA and the occasional homicidal nurse, a night in hospital is rarely fatal. But the indirect harm and waste of resources caused by bad and late data is huge.

Hospital episode statistics (HES) data provide commissioners with a picture of patterns of use, but it’s not a very accurate one. Check out the FAQ on the HES website and you’ll see why. Data about ethnicity, which relies on the patient to record it, is either not reliable or not collected at all. Year on year comparisons are difficult to make because the system of coding and counting is subject to frequent tweaks.

But the real problem with HES and all the big picture data is that it doesn’t go into enough detail. The NHS treats patients, not trends or patterns of use.

Most GPs know which patients are turning up regularly at their surgeries, who is routinely failing to take medicine and ending up in hospital, where the 20% of problems that consume 80% of the resources are occurring.

If GPs, community pharmacists, nurses, allied health professionals and other frontline staff can get to grips with the high consumers, the late presenters, the patients with several long-term conditions, learning disabilities or an aversion to doctors, we could save far more than £9m.

Dr James Kingsland has pointed out that if every practice made one less referral a day as the result of better informed commissioning, the annual saving would be in the order of £500m.

Preventing coding errors won’t provide savings of this magnitude, but it will help.