Inter-trial Variation in the Sensitivity of Thermal Threshold Testing for the Diagnosis of Neuropathy in Type 2 Diabetes Mellitus

Background: Thermal threshold testing (TTT) is a simple non-invasive approach for diagnosing diabetic neuropathy earlier. Conventionally the TTT is done in all four limbs and at least 6 trials are done to obtain the mean threshold, which is time consuming.

Aim: We propose to assess the validity and reliability of reduced number of trials of TTT in the lower limbs.

Materials and methods: After obtaining ethics approval from the Institute Ethics Committee, 100 patients with type 2 Diabetes Mellitus of both gender between the ages of 35 to 65 years attending medicine OPD were recruited. Neuropathy assessment was done using Temperature threshold testing. At least 6 trials were performed for each site and the mean threshold obtained. The mean of 5 trials, 4 trials and 3 trials were noted for the comparison.

Results: On comparing hot tests of 3 trials with 6 trials had a sensitivity and specificity of 88.7% and 96.6 %. In cold threshold testing, 4 trials and 3 trials showed similar results of sensitivity of 77.8%, specificity of 98.8%. The measures of agreement between the hot trials 6 vs 5 had Kappa value of 0.953, 6vs 4 showed a Kappa value of 0.862 and 6 vs 3 showed Kappa value of 0.819.

Conclusion: Hot threshold tests of lower limb are more sensitive than cold thresholds. The 4 trial test is a reliable test and can be performed over 6 trial tests. When time is a factor, three trials are sufficient to diagnose small fibre neuropathy.