Arising from: Mary Ann Banerji and Shefali Dham (2007) A comparison of classification schemes for ketosis-prone diabetes. Nat Clin Pract Endocrinol Metab 3: 506–507 doi:10.1038/ncpendmet0516

Banerji and Dham1 have performed an excellent and thoughtful analysis of our paper on the accuracy and predictive value of the Aβ classification scheme for ketosis-prone diabetes.2

They make the important point that “...it is essential to demonstrate that C-peptide levels measured 2 days after the index DKA predict insulin independence and good glycemic control (and not simply the C-peptide level at 12 months) otherwise the Aβ approach becomes tautological.” In fact we have already demonstrated both these points of clinical utility in earlier publications. The baseline fasting C-peptide level is indeed a strong predictor of good glycemic control.3 We have also found that the rate of achieving insulin independence is predicted by measures of β-cell function; specifically, a higher C-peptide to glucose ratio 6 months after the episode of ketoacidosis is a significant predictor of insulin discontinuation in patients who have preserved β-cell functional reserve at baseline.4

These data indicate that significance of the Aβ classification system for ketosis-prone diabetes extends beyond scientific accuracy to utility and reliability in clinical practice.