The impact of microbiological cultures on antibiotic prescribing in general internal medicine wards: microbiological evaluation and antibiotic use

Antibiotics are frequently prescribed because of a clinical suspicion of infection while results of microbiological analysis are still pending. Empirical therapy may be influenced by microbiological culture results, either by discontinuation of therapy in the case of negative cultures, or by broadening or narrowing (streamlining) the spectrum of antibiotics. Several strategies have been suggested to improve antibiotic prescribing, such as surveillance of antibiotic resistance, monitoring and auditing of antibiotics use, the use of consensus guidelines or computer decision support and improving multidisciplinary co-operation [1–3]. Although appropriateness of antibiotic prescribing has been investigated by some [4–6], the impact of microbiological culture results on antibiotic prescribing has not been analysed frequently [7,8]. In two longitudinal studies in internal medicine wards, microbiological analyses were performed in about 70% of all episodes of infection, relevant micro-organisms were isolated in about 50% of episodes and this resulted in modification of antibiotic therapy in about 50% of the cases [7,9]. In another study, Arbo et al. reported that cultures obtained within the first 7 days after admission, and the isolation of Staphylococcus aureus or Gramnegative bacteria were associated with modification of antibiotic therapy [8]. We prospectively investigated the impact of microbiological analyses on antibiotic prescribing in three general internal medicine wards of a Dutch university hospital. Therapeutic antibiotic prescribing in the general internal medicine wards (69 beds) at the Maastricht University Hospital in the Netherlands was prospectively monitored over a 12month period, between June 1996 and June 1997. Patients with HIV-infections and kidney or marrow transplantation were excluded from the study. Demographic characteristics, types, dose and route of administration were monitored daily for all patients receiving antibiotics. The results of microbiological tests (growth of bacteria and susceptibility testing) and its impact on antibiotic prescribing was evaluated. The following definitions were employed. Infection episode and antibiotic episode: an episode of infection was defined as a documented or suspected episode of