To compare the actual with the reported incidence of pressure sores to determine the accuracy of data (classification errors) and completeness of data (differences between manual and computer generated figures), retrospective data was collected regarding pressure sore rates following primary elective total hip arthroplasty operations carried out in 2001. Pressure sores rates were noted by nursing staff and entered into a computer database.

Four consultant orthopaedic surgeons were involved, across 2 sites – 1 NHS (PRH) and 1 local private hospital.

Preliminary audit reports indicated an alarmingly high pressure sore rate across the two units – 17/172 (9.9%) PRH and 23/71 (32.4%) private hospital.

Two major errors were revealed. In terms of accuracy of data, grade 1 areas (erythema without active ulceration) were included at both sites. These are only potential sites of pressure sores and should not have been used to calculate actual pressure sore rate. In terms of completeness of data, manual verification of the number of operations performed revealed a discrepancy between the theatres’ logbook entries and private unit computer figures. 97 rather than 71 operations were performed. There was no such discrepancy at the NHS site.

The data was reanalysed to obtain the actual pressure sore rate. For the NHS unit, grade 1areas were subtracted, causing the rate to fall from 32.4% to 1.0%. The two errors caused a dramatic and significant difference between reported and actual pressure sore rate.

Poor data collection leads to inaccurate audit, leading to inappropriate management. The concern is that similar errors, accumulated across key complication targets and specialities, will have a profound impact on NHS star ratings.