BACKGROUND AND AIM
Clinical governance is important. Clinical audit is part of clinical governance. The aim of this study was to perform a clinical governance exercise, and the reporting arrangements at an independent hospital provided the opportunity to do this over two phases between 1999 and 2002. Six physicians from four different UK National Health Service (NHS) trusts participated.
METHODS
Reports were shown anonymously to between two and five of the physicians who had not produced the report. Reports with at least one disagreement were reviewed by the group in order to reach concensus as to whether the disagreement was non-sustainable (NS), trivial (T) or non-trivial (NT), the last two, respectively, judged to make an insignificant or potentially significant impact on patient management.
RESULTS
In phase 1,239 audits were produced on 83 reports (2.9 per report), and in phase 2, 636 on 137 reports (4.6 per report). In phase 1, 14 (17%) reports attracted at least one disagreement (NS, five; T, four; NT, five). Of 239 audits, there were 20 disagreements of which five were NS. Moreover, nine audits agreed with a report with a NT disagreement, giving 14 suboptimal audits (5.9%). In phase 2, 80 (58%) reports attracted at least one disagreement (NS, 31 (P<0.003 vs phase 1); T, 35 (P<0.001); NT, 14 (P>0.05)). Of 636 audits, there were 153 disagreements, of which 37 were NS (P<0.05 vs phase 1). Twenty-five audits agreed with a report with a NT disagreement, giving 62 suboptimal audits (9.7%) (P>0.05). Overall, 19/220 reports (8.6%) were thought NT, an error rate comparable to reporting elsewhere in radiology. After phase 1, auditors became more aggressive but the quality of auditing tended to decline, as did the quality of reporting (although not significantly).
CONCLUSION
This study provides a useful framework for monitoring performance.