Fanconi Anemia: Examining Guidelines for Testing All Patients with Hand Anomalies Using a Machine Learning Approach

Area of Science:

• Pediatric Genetics
• Hematology
• Clinical Diagnostics

Background:

• Current UK guidelines recommend genetic testing for Fanconi anemia (FA) in all children with radial ray dysplasia or thumb anomalies.
• This approach involves extensive and costly investigations.
• The necessity of these guidelines for FA diagnosis in children with hand anomalies was questioned.

Purpose of the Study:

• To evaluate the necessity of genetic testing for Fanconi anemia in pediatric patients presenting with hand anomalies.
• To analyze referral patterns, financial implications, and service provision related to current UK guidelines.
• To determine if specific clinical features can predict Fanconi anemia in this patient group.

Main Methods:

• A three-year prospective study involving all patients with thumb or radial ray anomalies referred for Fanconi Anemia testing.
• Application of Classification and Regression Trees (CART) analysis and machine learning (WEKA) to assess predictive clinical features.
• Evaluation of Youden Index and Predictive Summary Index (PSI) scores for clinical significance.

Main Results:

• Hand anomalies alone were found to have no clinical significance in predicting Fanconi anemia.
• CART analysis and WEKA revealed no single clinical feature predictive of Fanconi anemia.
• All Fanconi anemia patients identified in the study presented with additional symptoms beyond isolated upper limb anomalies.

Conclusions:

• The study does not support the routine genetic testing for Fanconi anemia based solely on isolated hand abnormalities.
• Current UK guidelines for Fanconi anemia screening in children with hand anomalies may warrant revision.
• Clinical evaluation for other associated features is crucial before considering Fanconi anemia testing.