Advice on antibiotic therapy: a case study
teaser
Frank L Vancoillie
PharmD
Hospital Pharmacist
Stedelijk Ziekenhuis Roeselare
Belgium
E:frank.vancoillie@szr.be
The study analysed in this article was carried out by
F Vancoillie, M Depamelaere (Pharm) and R De Smedt (MD).
The Belgian government intends to pass a law mandating the creation of an antibiotic therapy panel in each acute hospital in the country. An expert in antibiotic management should also be appointed.
The project developed by our team emphasises the results of different initiatives to improve antibiotic therapy. If an antibiotic therapy is not in agreement with the directive set out by the Belgian edition of the Sanford Guide (which is used as a basic therapeutic guideline at our hospital in Roeselare), our role is to find out whether a direct intervention from the hospital pharmacist on the decisions made by the medical practitioner could influence the therapeutic process, and, if so, the form this intervention should take.
A method to reduce the cost of antibiotic therapy has been described,(1) which follows two different strategies:
- Substituting parenteral for oral therapy as early as possible.
- Focusing on the third day of antibiotic treatment, as that is the most crucial time for switching from empiric to targeted therapy. Targeted therapy can be initiated after identifying and testing the susceptibility of the infective organism, and in agreement with the clinical diagnosis. The choice of an antibiotic regimen is based on the smallest possible spectrum with the lowest risk of inducing antibiotic resistance, and on low cost.
Methods
In our hospital, there is a small stock of the drugs most frequently dispensed in the different nursing wards. The first step to control antibiotic consumption was to remove all the intravenous (IV) antibiotics used in empirical settings from the wards (except in the intensive and coronary care unit). Oral antibiotic therapy and IV prophylaxis were not part of the investigation. The dispensing of antibiotics to each individual patient is based on a medical practitioner’s preprinted prescription document, with a listing of the different antibiotics available. In this document, the medical practitioner indicates the required dose of the antibiotic(s) and its/their medical indication(s) according to the standard classification of the Belgian edition of the Sanford Guide.
The antibiotic formulary list indicates the cost per dose and the total possible minimum and maximum cost per day. This cost is calculated based on the number of vials necessary to provide the adequate therapy for a patient of average bodyweight, treated with the minimum and the maximum dosing schedule per kg.
This project was approved by the medical board of the hospital, and all medical practitioners were informed of the actions that would be undertaken by the pharmacist. The investigation programme was divided into two periods, with succession being restricted to parenteral antibiotics.
In the first period (from 01/11/2002 to 28/02/2003), pharmacists only contacted a medical practitioner when the type of infection or the appropriate indication was unclear or missing, according to the Sanford Guide classification. In the second period (from 01/03/2003 to 31/05/2003), therapy advice took place only if the medical practitioner’s therapy did not conform to:
- The Sanford Guide’s therapeutic directives for a given indication.
- The results of the microbiological laboratory analysis.
The pharmacist informed the medical practitioner about the therapy rationale from day 3 onwards, and suggested changes to the therapy for a more appropriate choice of antibiotic, although the medical practitioner could still make his/her final decision according to the patient’s clinical condition.
Collecting data
The prescription document offers the following data (collected in a spreadsheet):
- Starting date of therapy.
- Patient’s name and admission code.
- Name of medical practitioner responsible for the prescription.
- Name of medical practitioner responsible for the admission.
- Antibiotic(s).
- Dosing schedule.
- Indication (according to Sanford Guide classification).
These data were completed with:
- Microbiological laboratory analysis (Yes/No).
- Date of laboratory analysis.
- Type of sample for analysis.
- Laboratory results: determination and susceptibility of the organism.
- Total of units of antibiotics dispensed.
- Total cost per day (= cost of antibiotic ¥ dose/24h).
- Total cost of the therapy.
- Date of discharge.
- Condition of the patient at time of discharge (cured, deceased or discharged with antibiotics).
- Type of antibiotic used at time of discharge.
- ICD-9 code (International Classification of Diseases).
Because of the complexity of managing data from patients with more than one indication (infection site), these data were not included in the study.
Problems
The major problems encountered were missing data on the prescription and a prescription written by a medical practitioner not familiar with the patient’s pathology. In this event, the medical practitioner was contacted by phone, although communication can sometimes be a problem. The most time-consuming activities were contacting medical practitioners and collecting additional data, such as laboratory analysis. Another problem was the difficulty to collect data within three days, in particular during weekends and holidays. The results of the study are summarised in Table 1 and Figure 1.
[[HPE17_fig1_67]]
[[HPE17_table1_67]]
Conclusions
This study demonstrates that advice given by a hospital pharmacist has a positive effect, bringing the medical practitioner’s prescription in accordance with the guidelines. During the second period of the study, a higher percentage of therapies were found to be in agreement with bacterial cultures and/or Sanford guidelines. However, the size of this study prevents it from being transformed into a statistical model.
Reference
- Rifenburg RP, Paladino JA, Hanson SC, et al. Benchmark analysis of strategies hospitals use to control antimicrobial expenditures. Am J Health-Syst Pharm 1996;53:2054-62.
