Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor
To the Editor: I disagree with 3 of the assertions
that Dr Holmboe makes in his Clinical Applications article1
about treatment of type 2 diabetes. First, he describes a patient with a random
plasma glucose concentration of 480 mg/dL (27 mmol/L) and states that such
a patient would require insulin. I believe that this patient would almost
certainly respond to high doses of a sulfonylurea agent. I have treated more
than 100 symptomatic patients with glucose levels of this magnitude, many
of whom have had ketosis, a few with slightly lowered bicarbonate levels (down
to 16 meq/L), and a fair number with significant weight loss. More than 90%
of them do not require insulin. After 4 months, 6 of 55 patients2
were lost to follow-up (4 patients had lost their health maintenance organization
insurance and 2 would not comply with the recommended follow-up). Of the remaining
49 patients, 6 continued taking a maximal dose of glyburide, 29 were taking
a submaximal dose, 11 were treated with diet alone, and 3 were taking insulin.
The insulin was started several weeks to several months later, when goal levels
of glycemia were unmet, not as an emergency to treat the initial hyperglycemia.
At the time this study was carried out, no other oral antidiabetes drugs were
available in the United States. It is likely that the addition of another
oral drug would have avoided the need for these 3 patients to take insulin
during the 4 months of the study. Thus, one can spare the patient the rigors
of immediate insulin therapy to see if he or she will respond to a high dose
of a sulfonylurea agent.
Davidson MB. Treatment Decisions for Type 2 Diabetes. JAMA. 2002;287(13):1646-1648. doi:10.1001/jama.287.13.1645