# Policy, practice, and prediction: model-based approaches to evaluating N. gonorrhoeae antibiotic susceptibility test uptake in Australia

## TL;DR

This study models how to improve antibiotic susceptibility testing for gonorrhea in Australia by examining clinician-patient interactions and policy changes.

## Contribution

The study introduces a novel model-based approach to evaluate and improve antibiotic susceptibility test uptake for N. gonorrhoeae.

## Key findings

- Clinician education and clinical support tools can improve antibiotic susceptibility testing initiation.
- Adherence to guidelines and patient factors like symptoms and risk behaviors significantly influence testing rates.
- Mandated testing could increase testing rates but requires addressing practical and laboratory constraints.

## Abstract

Antimicrobial resistance (AMR) represents a significant threat to global health with Neisseria gonorrhoea emerging as a key pathogen of concern. In Australia, the Australian Gonococcal Surveillance Program (AGSP) plays a critical role in monitoring resistance patterns. However, antibiotic susceptibility test (AST) uptake – a crucial component for effective resistance surveillance – remains to be a limiting factor. The study aims to model the processes involved in generating AST tests for N. gonorrhoea isolates within the Australian healthcare system and assess the potential impact of systematic and policy-level changes.

Two models were developed. The first model was a mathematical stochastic health systems model (SHSM) and a Bayesian Belief Network (BBN) to simulate the clinician-patient dynamics influencing AST initiation. Key variables were identified through systematic literature review to inform the construction of both models. Scenario analyses were conducted with the modification of model parameters.

The SHSM and BBN highlighted clinician education and the use of clinical support tools as effective strategies to improve AST. Scenario analysis further identified adherence to guidelines and changes in patient-level factors, such as persistence of symptoms and high-risk behaviours, as significant determinants. Both models supported the notion of mandated testing to achieve higher AST initiation rates but with considerations necessary regarding practicality, laboratory constraints, and culture failure rate.

The study fundamentally demonstrates a novel approach to conceptualising the patient-clinician dynamic within AMR testing utilising a model-based approach. It suggests targeted interventions to educational, support tools, and legislative framework as feasible strategies to improve AST initiation rates. However, the research fundamentally highlights substantial research gaps in the underlying understanding of AMR.

The online version contains supplementary material available at 10.1186/s12879-024-09393-y.