Influence of blood culture results on antimicrobial prescribing in a private hospital in North West, South Africa

Abstract

Blood culture laboratory testing and the results thereof can be seen as a cornerstone for optimised antimicrobial treatment and has the power to influence and change how effectively patients are treated with antibiotics, ultimately influencing patient treatment outcomes. Its ability to identify causative organisms and their resistance patterns make it very useful and yet it is underutilised. This study aimed to determine the prevalence and impact of blood culture testing before or during antibiotic treatment, as well as the prescribing patterns that impact the patient outcomes. This was done by putting a focus on the prevalence of blood culture sampling within specific age groups or diagnostic groups (looking at admission diagnoses without the diagnosis or suspected infection being the determinant of whether a blood culture sample should have been taken or not), and if it led to changes in antibiotic treatment. A retrospective quantitative observational study was embarked on at a private hospital situated in the North West province of South Africa and approval was sought from the Heath Research Ethics Committee (HREC), with Ethics number: NWU-00001-07-S1, as well as the private hospital where data was collected from. Specific inclusion and exclusion criteria were selected for this study. The data relevant to the study was recorded on a data collection tool that included: (i) demographic information; (ii) medical information and (iii) antibiotic information. Guidelines set by Ampath® laboratory (Titled: Guidelines to antimicrobial therapy and the laboratory diagnosis of infectious diseases) was used as to evaluate the appropriateness of antibiotic prescriptions (whether the correct empirical antibiotic was selected according to the diagnosis, as well as the correct dose, duration and interval). Lastly, the data were analysed by using the Statistical Analysis System®, SAS 9.3® (SAS Institute Inc., 2009). The study is presented in article format; two manuscripts intended for the Southern African Journal of Infectious Diseases (SAJID). Manuscript one focused on the prevalence of patients from whom a blood culture sample was obtained before or after antibiotic treatment was given, as well as the effect of blood culture testing on antibiotic treatment changes (stratified by patient age groups and diagnosis). The overall utilisation of blood cultures, whether obtained before or after antibiotic treatment, was extremely low - only 8.77% (n = 120) of blood culture samples were obtained before empirical treatment started and lead to a change in 14.17% (n = 17) of these empirical prescriptions. In patients where blood culture samples were taken after the empirical treatment commenced, blood culture samples were collected in 12.05% (n = 165) of patients which lead to a change in 13.33% (n = 22) of these prescriptions. Blood culture samples were obtained more readily from patients in the age groups 30 ≤ 60 years and 60+ years and where a diagnosis of central nervous system infections, cancer related infections and pneumonia were made. Manuscript two focused on investigating the prevalence of empirical antibiotic treatment, as well as the antibiotic treatment that was prescribed, following positive blood culture results. The appropriateness of antibiotic treatment, whether empirical or definitive, was furthermore compared to the guidelines set by Ampath® laboratory, since the private healthcare sector of South Africa is not governed by a specific antibiotic treatment protocol. These guidelines set by Ampath® laboratory is distributed to physicians and is the most generally used guideline; therefore, it was selected for use during this study. Inappropriate empirical antibiotic selection was observed in 52.08% (n = 713) of the prescriptions reviewed and the appropriateness of the dose, interval and duration could not be determined, whereas the dose, interval and duration were inappropriate in 37.40% (n = 236) of the empiric prescriptions where an appropriate antibiotic was selected. Specific antibiotic classes were prescribed inappropriately more often and included carbapenems, fluoroquinolones, aminoglycosides and cephalosporin. Lastly, 51.28% (n = 20) of patients of whom the prescription was changed according to blood culture results, were discharged. The prevalence of blood culture testing, regardless whether the blood culture sample was obtained before or after antibiotic initiation, was determined and indicated that blood culture testing during antibiotic treatment was seldom used. The prevalence of blood culture testing was more likely to be used during the antibiotic treatment of a patient in a specific age group or when a specific diagnosis was made. In patients where blood culture testing was done (before or after the initiation of antibiotic treatment), the results led to a change in less than 15% of prescriptions. Inappropriate antibiotic treatment selection linked to a specific diagnosis was more prevalent. Often the appropriate antibiotic class was selected but the inappropriate dose, duration and interval, was prescribed. The successfulness of antibiotic treatment, implicating the patient’s outcomes, can be directly linked to appropriateness of the initial antibiotic treatment selection, as well as the dose, interval and duration of the selection.