Self-audit: its effect on quality of care.

Using a crossover design, it is shown that an individual criteria setting followed by immediate feedback of previous performance produced a sustained and continuing improvement in recording for two common conditions (cystitis and vaginitis). The intervention, which is simple and could easily be applied in other settings, produced improvements significant at the P = .001 level. The study controlled for overall improvement in record keeping. Further testing of this method of influencing physician performance is warranted.