Improving Indirect Methods for Calculating Reference Limits for Nerve Conduction Studies From Historical Data

Area of Science:

• Neurology
• Biomedical Engineering
• Clinical Diagnostics

Background:

• High-quality reference limits for nerve conduction studies (NCS) are crucial for diagnosing neuromuscular disorders.
• Directly calculating reference limits from healthy controls is costly and time-intensive.
• Indirect methods (E-norms, E-Ref, MeRef) utilize historical hospital data but have limitations.

Purpose of the Study:

• To enhance indirect methods for calculating NCS reference limits by modifying S-curve selection algorithms.
• To improve the efficiency and accuracy of establishing reference limits from historical patient data.

Main Methods:

• Modified E-norms, E-Ref, and MeRef algorithms were applied to a large historical NCS database.
• Reference limits were calculated using the modified indirect methods.
• Results were validated against limits derived from 680 healthy subjects using Youden's J statistic and z-score deviation.

Main Results:

• Modified methods achieved Youden's J > 0.8 and z-score deviation < 0.9 for most NCS measurements.
• Performance was comparable or superior to unmodified methods.
• Optimal performance required at least 500 measurements and <20% abnormal data.

Conclusions:

• Modifying S-curve selection algorithms significantly improves indirect NCS reference limit calculation.
• The enhanced methods require adequate numbers and proportions of normal measurements in historical data.
• A combination of these improved indirect methods is viable for developing reference limits from historical patient data.