February 19, 1997

Treatment of Patients With Non-insulin-dependent Diabetes With the Implantable Insulin Pump-Reply

Author Affiliations

Johns Hopkins University Baltimore, Md for the Department of Veterans Affairs Implantable Insulin Pump Study Group

JAMA. 1997;277(7):530. doi:10.1001/jama.1997.03540310028020

In Reply.  —Dr Pazart would like to define efficacy in terms of success rate in achieving perfectly normal glycemia. Of note, he quotes only half of our stated treatment objective, which was not only to achieve normal blood glucose levels but to do so "while avoiding clinical hypoglycemia." These were, in essence, treatment targets. If the success rate is measured in achievement of perfectly normal hemoglobin A1c (not, of course, assuring uniformly normal blood glucose levels), it was attained in 29 of our 121 subjects (16 who received the implantable insulin pump, 13 who received multiple-dose insulin). But achieving euglycemia was limited, as always, by the risk of hypoglycemia. Therefore, we reject dichotomizing the success rate or defining it in terms of normal blood glucose. To do so implies that diabetes treatment is unsuccessful in virtually every case, although there is ample reason to believe that less than normal glycemia is,