Background/Purpose: In private healthcare settings, patients often have direct access to specialists without prior referral. However, in rheumatology, this can lead to inefficiencies due to the wide range of conditions and a general lack of knowledge about the specialty. Initial consultations often don’t result in a definitive diagnosis or treatment plan, leading to additional tests and follow-up visits, which prolongs time to diagnosis and treatment, increases waiting lists, and creates inefficiencies for patients, insurers, and healthcare providers.

Objective: To evaluate the effectiveness of ReumAI, an artificial intelligence (AI)-based triage system, in optimizing rheumatology consultations in a private hospital setting.

Methods: A pilot study was conducted at the rheumatology department of Hospital Quirón Barcelona, where all new patient requests were managed using ReumAI. This system, operated by non-physician healthcare staff, conducts an AI-guided telephone interview to obtain a syndromic diagnostic orientation and determine necessary tests before the specialist visit. The first 300 patients managed with this system were analyzed, evaluating diagnostic concordance and the utility of requested tests.

Results: The mean patient age was 52 years (SD 14), with 72% being women. The most frequent reasons for consultation were hand pain (28%), nonspecific pain (25%), and foot pain (11%). The final syndromic diagnosis was included in ReumAI’s differential diagnosis in 85% of cases, with a moderate correlation (53%) between ReumAI’s syndromic diagnosis and the rheumatologist’s diagnosis. Tests ordered by ReumAI were useful in most cases (95%).

Conclusion: ReumAI shows promise as a tool for optimizing rheumatology triage, potentially improving consultation efficiency and reducing wait times. Larger-scale studies are needed to confirm these findings and assess the long-term impact on patient care.

« Back to ACR Convergence 2024

ACR Meeting Abstracts - https://acrabstracts.org/abstract/enhancing-rheumatology-consultations-with-ai-powered-triage-the-reumai-pilot-study/