Perspectives of audiologists and hearing screeners in the clinical use of AI in detecting otitis media – a qualitative study

Artificial Intelligence (AI) machine learning models to improve diagnostic decision making for otitis media are emerging as potential future tools to improve diagnostic accuracy and Ear Nose and Throat (ENT) service delivery. This study aimed to explore the perspectives of audiologists and hearing screeners on the opportunities and challenges of AI tools in diagnosing otitis media within the context of a public healthcare system. Twenty audiologists with paediatric experience and thirteen hearing screeners were purposively recruited to participate in three focus groups in New Zealand. Reflexive thematic analysis yielded four key themes: benefits; concerns and challenges of implementation; integration into clinical workflows; and future considerations and cautions. All participants had an optimistic outlook as to the benefits of AI in enhancing service delivery, patient engagement and professional advancement. All participants reported benefits related to AI technological capabilities that may improve accuracy and consistency of diagnoses. Hearing screeners reported few concerns beyond the cost of equipment being a barrier to implementation. Audiologists however, had a wide range of concerns, particularly related to clinical workflow, accountability and ethical considerations. All participants recognised that AI technology could have unintended consequences such as increasing demand on already strained ear and hearing services by detecting more cases of otitis media in the community without the necessary resources to manage them. Audiologists and hearing screeners are positive regarding implementation of AI diagnostic tools in the context of otitis media diagnosis. Novel benefits including potential for enhanced family engagement and personal professional growth were highlighted, but challenges and barriers need to be addressed for successful clinical translation.