Evidence-based medicine audit as a tool for improving emergency ophthalmology

To audit the proportion of interventions in emergency ophthalmology that are evidence based and to determine whether the quality of care can be improved.

Audit of diagnosis-intervention pairs was carried out retrospectively in March 2003. The outcomes were assessed for evidence level reached in the Medline database 1966-2003 and the Cochrane Database of Systematic Reviews. Locally agreed guidelines were issued and the study repeated prospectively in March 2004, when new medical staff were at a similar level of experience. The participants had no prior knowledge of the study to avoid prescribing bias (Hawthorne's phenomenon).

In the first part of the audit in 2003, 71% of interventions were evidence based, with 36% derived from systematic reviews, meta-analysis or randomised controlled trials (evidence levels 1-3). After guidelines for care were implemented in 2004, there was an improvement in the number of evidence-based interventions to 82% (P=0.04), and levels 1-3 were reached in 60% (P=0.02). The proportion with no evidence or against evidence dropped from 29 to 18% (P=0.04). An additional benefit was to reduce the number of re-attendances required.

Evidence-based medicine can be used to improve the quality of care in the acute ophthalmic setting, both in refining the standard of interventions and in reducing the number of hospital visits.