Deutscher Rheumatologiekongress 2025
The Rheuma-VOR App: Optimizing the early diagnosis of inflammatory arthritis
2Department of Pharmacolgy, University Medical Center of the Johannes Gutenberg University Mainz, Mainz
3Institute for Medical Biostatistics, Epidemiology and Informatics (IMBEI), University Medical Center of the Johannes Gutenberg University Mainz, Mainz
4Department of Internal Medicine I, Division of Rheumatology and Clinical Immunology, University Medical Center of the Johannes Gutenberg University Mainz, Mainz
5Rheumatology Center Rhineland-Palatinate, Bad Kreuznach
Text
Introduction: The Rheuma-VOR network study developed a smartphone app to support primary care providers in the task of identifying RA, PSA, axSpA as early as possible [1]. The Rheuma-VOR Screening-App study aims to proof and validate the smartphone app diagnosis (AD) and assess its predictive accuracy against, 15-minutes short examination without laboratory (DWL) and the final clinical diagnosis (FD).
Methods: Based on the multi-stage Delphi Procedure, 17 questions for a differential screening of the three diseases were identified, including four laboratory parameters. The questions are read to the patient and offer the answer options “YES”, “NO” and “I DO NOT KNOW”. 13 questions aimed at clinical symptoms and 4 questions sampling out laboratory criteria. The questions cover a range of indicators, such as joint swelling, CCP and RF levels, HLA-B27 status, inflammatory markers (ESR and CRP), and specific symptoms like morning stiffness and dactylitis. Each response is scored, and the app generates a cumulative score to suggest potential diagnosis.
Results: The internal consistency of the app’s 17-item questionnaire was excellent, with a Cronbach’s Alpha of 0.929. A total of 815 responses were collected, between November 2018 and October 2020. Among the 815 participants, 388 cases were concordant, with both the AD and DWL indicating a rheumatologic condition. 76 cases showed agreement in the absence of a rheumatic diagnosis. Discrepant cases included 308 false positives (positive AD contradicted by a negative DWL) and 43 false negatives (negative AD and positive DWL). From these findings, the app demonstrated a sensitivity of 0.900 and a specificity of 0.198 in comparison with the DWL outcomes. The positive predictive value was 0.557, while the negative predictive value reached 0.639. False positive rate and false negativ rate were 0.802 and 0.100, respectively. The sensitivity and specificity values were to 0.883 and 0.156, respectively for the final diagnosis.
Conclusion: The Rheuma-VOR App demonstrates its capability to enhance primary care providers’ ability to screen for potential rheumatic diseases, effectively supporting specialist referrals. Detailed, disease specific data will be shown at the conference.
