RP403 - COSI 2.0: What Do Today's Clients and Clinicians Need to Optimize Patient Outcomes?

Saturday, March 29, 2025

9:15 AM - 9:45 AM CST

Location: Rooms 220-222

Earn 0.05 CEUs

Instructional Level: Intermediate

Lead Presenter(s)

Taegan Young, MClinAud (she/her/hers)

Research Audiologist
National Acoustic Laboratories
Marsfield, New South Wales, Australia

Disclosure(s): No financial or nonfinancial relationships to disclose.

PDF Slides

Presenter(s)

Padraig Kitterick, PhD (he/him/his)

Head of Audiological Science
National Acoustic Laboratories
Macquarie University, New South Wales, Australia

Disclosure(s): Financial: GN: Grant/Research Support (Ongoing), Research Grant (includes principal investigator, collaborator or consultant and pending grants as well as grants already received) (Ongoing); Sonova: Research Grant (includes principal investigator, collaborator or consultant and pending grants as well as grants already received) (Ongoing), Speaker/Honoraria (Ongoing); Nonfinancial: No nonfinancial relationships to disclose

Abstract:
COSI has long been a gold standard for identifying hearing needs, but many clinicians either simplify the process or do not use it. This presentation covers a study evaluating an AI-driven tool (COSI 2.0) designed to autonomously assess clients' key hearing challenges, comparing its performance to the traditional clinician-led COSI method. The study also evaluates user responses on trust, ease of use, and satisfaction. Results will be presented to show whether the AI tool effectively identifies unmet needs on par with clinicians, laying the foundation for measuring outcomes with hearing device interventions and potentially improving clinical workflows.

Summary:
Rationale: COSI has long been a gold standard for identifying hearing needs and measuring outcomes centered on what is most important to the patient, yet many clinicians either don’t use it or improvise a simplified version due to time constraints, perceived lack of immediate benefit to the client, or insufficient training. As a result, care becomes less personalized, and opportunities to demonstrate the true benefits of hearing technology to the patient are missed. With the evolving landscape of patient needs, technology, and care models, there is a growing necessity to have a practical tool to deliver personalized care in a consistent and effective way. This research sought to reimagine COSI—making it more adaptable, efficient, and integrated into today’s healthcare environment to improve both patient outcomes and clinician workflows.

Objectives:
The objective of this study was to evaluate the effectiveness of an AI-driven tool (COSI 2.0) that autonomously identifies clients' most pressing hearing challenges. The study compared the AI tool’s performance and client experience against the traditional clinician-led process, focusing on measures of trust, engagement, empowerment, satisfaction, and efficiency. Our hypothesis was that participants who use the new tool will report a comparable level of trust, engagement, empowerment and satisfaction with the clinical process as those who receive the traditional clinician-led process.

Design:
Participants were adults with hearing difficulties who did not wear hearing devices and had not had a hearing assessment in the last three or more years.
All participants were invited to attend an in-person appointment designed to mimic a standard audiological assessment appointment in a typical Australian clinic with the only difference between each group being the method for how hearing needs were elicited.

Participants were randomly allocated into two groups:
Control group – Traditional clinician-led goal discussion as per the NAL COSI instructions.
Test group – Participants interact with an AI-driven tool, answering questions about their hearing needs and important listening situations to generate a personalized and prioritized list of hearing needs to be discussed with the clinician.

All participants completed a post-appointment survey to capture their impressions of the clinical experience and the tool, as well as a follow-up survey two weeks post-appointment, to obtain their reflections on the process and to understand what actions they intend to take, in light of their hearing assessment and clinical recommendations.

Results:
The tool was designed to assess and summarize clients' key hearing needs, performing on par with clinicians, and lays the foundation for measuring outcomes with hearing device intervention. Additionally, the tool highlights how hearing devices could address these needs and provides counsel on the measurable benefits in real-world situations.

Conclusions:
The use of artificial intelligence as part of the assessment process has great potential to support the identification and personalization of hearing needs, enhance patient-centered care, and maximize the efficiency of limited time with clinicians. It also shows promise as a method of embedding patient-centered care in care pathways that don't include face to face contact with clinicians such as remote care and with over-the-counter devices.

Brief Summary of Clinical Takeaways:
The clinical takeaway for this session is that an AI-driven tool for assessing hearing needs could offer a consistent, personalized care alternative to traditional methods, while reducing the burden on clinicians and optimizing the delivery of patient-centered care within the limited time available in clinical settings.

Learning Objectives:

Upon completion, participants will be able to describe the original purpose and intent of the Client Oriented Scale of Improvement
Upon completion, participants will be able to enumerate key elements of patient-centred needs assessment and outcomes measurement
Upon completion, participants will be able to understand how an AI-based tool could support the identification and personalization of hearing needs, enhance person-centred care, and maximise the efficiency of limited time with clinicians