Background: Antibiotic prescribing is associated with the growth of antibiotic resistance in micro-organisms. Prescribing therefore needs to be restricted by well-designed antibiotic protocols. Audit of such protocols is only possible if clinical information about infection is properly recorded.
Objective: To assess the quality of clinical data recording in case notes of patients receiving antibiotics in hospital and to assess the effect of feedback of this information on clinical practice.
Design: Four point prevalence surveys.
Setting: Directorate of General Medicine, Dundee Teaching Hospitals NHS Trust.
Subjects: Inpatients receiving antibiotics.
Intervention: Feedback of results from the first two surveys to clinicians.
Main outcome measures: Proportion of patients in the Medical Directorate with the recording of an antibiotic indication and the temperature in the case notes before, immediately after, and three months after, feedback of survey results.
Results: In each survey one third of patients were receiving antibiotics. The antibiotic indication was recorded in 64% of patients prior to feedback and in 86.5% after feedback (p = 0.004). Three months later this had fallen to 74%. Temperature recording improved between 2nd and 3rd surveys (57% to 86.1%, p < 0.001) and declined to 61% in the 4th survey (p < 0.001). Respiratory infection was the commonest recorded indication for an antibiotic; recording of this indication improved between the 2nd and 3rd surveys but declined by the 4th survey.
Conclusions: Audit and feedback of information about clinical record keeping improves standards but improvements are not sustained. Clinical record keeping must be improved so compliance with antibiotic protocols can be properly audited. An "antibiotic sticker" in addition to immediate concurrent feedback may improve such deficiencies.