Correction and transformation of normative neurophysiological data: Is there added value in the diagnosis of distal symmetrical peripheral neuropathy?

Abstract

Despite theoretical advantages, the practical impact of mathematical correction of normative electrodiagnostic data is poorly quantified. One hundred five healthy volunteers had clinical and neurophysiological assessment. The effects of age, height, gender, weight, and body mass index were explored using stepwise regression modeling. Reference values were derived from raw and adjusted data, which were transformed to allow appropriate use of parametric statistics. The diagnostic accuracy of derived limits was tested in patients at risk of distal symmetric peripheral neuropathy (DSPN) from chemotherapy. The variability of our normative data was reduced by up to 69% through the use of regression modeling, but the overall benefits of mathematical correction were marginal. The most accurate reference limits were established using the 2.5th and 97.5th percentiles of the raw data. Stepwise statistical regression and mathematical transformation improve the distribution of normative data, but their practical impact for diagnosis of distal symmetrical polyneuropathy is small.