We thank Dr. Rizvi (1) for his comments about our study (2). Unfortunately, due to space constraints, we were not able to include all of the details for insulin initiation and titration in the main body of the article. As indicated in the fifth paragraph of the research design and methods section, these details are contained in an online appendix for the article (available at http://care.diabetesjournals.org). We did not measure 2-h postprandial blood glucoses nor use any insulin-to-carbohydrate ratios or sensitivity factors. We are not sure that these are practical for busy general medical wards.
Our approach of using NPH and regular insulin twice daily (BID) is a basal-bolus approach to subcutaneous insulin therapy. BID NPH provides basal insulin and the bolus for lunch. BID regular insulin provides bolus insulin for breakfast and supper. This approach is completely different from sliding-scale regular insulin given every 6 h where basal insulin is never provided. Since we performed our study, the use of once-daily insulin glargine as basal insulin and three rapid-acting insulin analog mealtime boluses has become more commonly used. However, this regimen has been studied largely in type 1 diabetes. About 95% of our subjects and indeed most inpatients with hyperglycemia have type 2 diabetes.
Our blood glucose targets are described in the research design and methods section of our work. Dr. Rizvi misinterprets the use of HbA1c (A1C) in our inpatients. As shown in Table 3 of our original article, the mean A1C in our study patients was 8.7% (2). An important goal of the teaching program with our residents was to use the opportunity presented by the hospitalization to improve diabetic therapy in all inpatients with A1C >7%. With our approach, we increased the percentage of patients who were discharged on improved therapy from 32 to 80%. This was associated with an improved A1C after 1 year, as seen in Fig. 3 in our original article. If sliding-scale regular insulin were the only regimen used for inpatient control of hyperglycemia, it would be impossible to titrate a new improved diabetic regimen (oral agent or insulin) on which to safely discharge the patient.
We have recently completed two studies of insulin glargine use in inpatients (3,4), and its applicability is promising. We are currently comparing the use of aspart versus regular insulin in an ongoing study of medical inpatients.
