RheumaTool, a novel clinical decision support system for the diagnosis of rheumatic diseases, and its first validation in a retrospective chart analysis

Abstract

AIMS RheumaTool is a clinical decision support system designed to support the diagnostic process in rheumatology by presenting a differential diagnosis list after the input of clinical information. The objective of this study was to evaluate the performance of RheumaTool in detecting the correct diagnosis in referrals to a rheumatology clinic. METHODS In this retrospective chart analysis, data were gathered from patients with musculoskeletal complaints and an uncertain diagnosis who were referred to a Swiss tertiary rheumatology outpatient clinic. Data were entered into RheumaTool in a standardised fashion, while the principal diagnoses in the medical reports were blinded. RheumaTool’s output was compared to the correct diagnoses, established either by widely accepted diagnostic criteria or through the expert consensus of independent rheumatologists. Diagnostic precision, the primary endpoint, was defined as the proportion of correctly diagnosed cases among all cases. RESULTS One hundred and sixty cases with 46 different diseases were included in this analysis. RheumaTool correctly diagnosed 40% (95% confidence interval 32.4–48.1) of all cases. In 63.8% (95% confidence interval 55.7–71.1), the correct diagnosis was present in a differential diagnosis list consisting of a median of two diagnoses. CONCLUSION In this first validation, RheumaTool provides a useful list of differential diagnoses. However, there is not sufficient diagnostic reliability for unfiltered data entry, especially in patients with multiple concomitant musculoskeletal disorders. This must be taken into account when using RheumaTool.

Keywords

• medical informatics applications
• clinical decision support systems
• rheumatology
• rheumatic diseases
• software validation