Evaluation of audit of medical inpatient records in a district general hospital.

OBJECTIVE To evaluate an audit of medical inpatient records. DESIGN Retrospective comparison of the quality of recording in inpatients' notes over three years (1988, 1989, 1990). SETTING Central Middlesex Hospital. MATERIALS Random sample of 188 notes per year drawn systematically from notes from four selected one month periods and audited by two audit nurses and most hospital physicians. MAIN MEASURES General quality of routine clerking, assessment, clinical management, and discharge, according to a standardised, criterion based questionnaire developed in the hospital. RESULTS 1988 was the year preceding the start of audit in the hospital, 1989 the year of active audit with implicit and loosely defined criteria, and 1990 the year after introduction and circulation of explicit criteria for note keeping. There was a significant trend over the three years in 21/56 items of the questionnaire, including recording of alcohol intake (x2 = 8.4, df = 1, p = 0.01), ethnic origin (x2 = 57, df = 1, p = 0.001), allergies and drug reactions (x2 = 10, df = 1, p = 0.01) at admission and of chest x ray findings (x2 = 8, df = 1, p = 0.01), final diagnosis (x2 = 5.6, df = 1, p = 0.025), and signed entries (x2 = 11.3, df = 1, p = 0.001). Documentation of discharge and notification of discharge to general practitioners was not significantly improved. CONCLUSIONS Extended audit of note keeping failed to sustain an initial improvement in practice; this may be due to coincidental decline in feedback to doctors about their performance.