Gender Aspects during and after the Diagnostic Odyssey in M. Fabry: Machine Learning Diagnostic Support Tool Reveals Different Answer Patterns in Diagnostic Questionnaire

Background/Purpose: M. Fabry is a rare X-linked disease that affects women and men differently, impairing their physical and mental quality of life. It is often diagnosed several years after symptom onset.

We examined these differences in the diagnostic odyssey and in the quality of life as well as whether these differences need to be taken into account in order to improve a machine learning based diagnostic support tool.

Methods: We collected data sets of 33 patients with Fabry disease for a data mining based diagnostic support tool and the SF-36 health-related quality of life questionnaire. We analyzed the answer patterns in the tool's questionnaires with regards to gender differences and searched for correlations between its performance, quality of life and course of the disease. Considering the results, we adjusted the tool followed by a performance analysis.

Results: We found major differences in the disease perception and answer patterns of females and males with Fabry disease, mainly depicting differences in burden of disease and mental quality of life. After adjusting to these findings the diagnostic support tool achieved sensitivity for Fabry disease of 90% in the stratified 10-fold cross-validation. Quality of life, age, symptom onset, or education of patients with Fabry disease does not seem to impair the performance of the tested diagnostic support tool.

Conclusion: We were able to show that consideration of gender aspects in rare diseases can lead to major improvement of diagnostic support.

Publication History

Article published online:
28 October 2021