A need for Action: Results from the Australian General Practice National Antimicrobial Prescribing Survey (GP NAPS)

Problem

The use of antibiotics in the Australian community is much higher than most comparable countries, with 80% of antibiotics used in human medicine prescribed in general practice. Australia lacks a systematic approach to drive improved antibiotic prescribing in this setting and there is an urgent need to better understand antibiotic prescribing. The aim of this study was to conduct a pilot audit of antibiotic prescribing in general practice to determine compliance with the Australian Therapeutic Guidelines (TG) and prescription appropriateness.

Approach

Building on Hospital National Antimicrobial Prescribing Survey (NAPS) methodology, an antibiotic prescribing audit was conducted in 11 general practices in four Australian states between September 2017 and January 2018. We collected data from 50 patient encounters that had generated one or more antibiotic prescriptions per practice. Information collected included antibiotic prescribed; prescription indication; duration, dosage and frequency; and location of documentation of indication in the clinical record. A multidisciplinary team assessed the data to determine the antibiotic indication, guideline compliance and overall prescription appropriateness. Descriptive statistics were calculated using Excel 2016.

Findings

A total of 572 antibiotic prescriptions from 550 patient encounters were audited. The most common reasons for prescribing an antibiotic included respiratory tract (26.7%); ear, nose and throat (18.7%); skin and soft tissue (17.5%); and urinary tract (10.7%) infections. The most frequently prescribed antibiotics were cefalexin (19.4%), amoxicillin (18.5%), and amoxicillin-clavulanic acid (11.5%).Of the 572 antibiotics prescribed, 20.8% complied with TG, 66.6% were considered non-compliant and 5% were not assessable. A small number were consistent with local guidelines (0.3%) or directed therapy (3%), but not TG; guidelines were not available for 4% of indications. 38% of prescriptions were classified as inappropriate. Most common reasons for prescriptions being deemed inappropriate were excessive treatment duration (17%), incorrect dose/frequency (16%), and spectrum too broad (15%). 22% of scripts were assessed as not requiring an antibiotic. Using the entire GP record, the auditor could ascertain the clinical indication for 93% of prescriptions, however it was only documented in a field that can be extracted by current audit software in 61% of cases.

Consequences

Two-thirds of antibiotic prescriptions in audited practices were not compliant with national guidelines and over one-third were deemed inappropriate. This may be reflective of a lack of antimicrobial stewardship strategy for general practice. The finding that over one-third of indications are not recorded in extractable fields means that free text analysis of clinical notes will be required to facilitate electronic auditing of antibiotic prescribing. Based on these results, we are working with TG to develop electronic decision support for antimicrobial prescribing, and passive surveillance of infection management in primary care. These initiatives will contribute to Australia's National Antimicrobial Resistance Strategy and lead to improved prescribing quality and safety.