I have prescribed insulin either alone or as an adjunct to oral medication for a number of my patients with type 2 diabetes. Most of them are receiving long-acting insulin, either glargine (Lantus, Sanofi-Aventis) or detemir (Levemir, Novo Nordisk), given once daily in the evening.
I have convinced them to bring their blood glucose diaries to each clinic visit to try to work out why sometimes their fasting blood glucose is fine and other times way too high. Almost without fail, the high fasting blood glucose follows a bedtime snack and what a wide variety of bedtime snacks they report! These bedtime snacks can be a habit or may be taken because the patient feels their blood glucose is too low and they need nourishment. In those patients who check, the value is never low, but they go ahead with the snack anyway, and there really is no reason why they shouldnt.
What I had been neglecting to do was empower the patient to take more once-nightly insulin when they snack at night. For this to work, it is best for the patient to take insulin at bedtime extra insulin if they snack, no extra insulin if they dont. This is a daunting prospect for many patients because they are very worried about taking too much extra insulin and having a hypoglycemic event in the middle of the night.
Our goal in treating diabetes is, among other things, to maintain the blood glucose close to normal as often as possible. Hypoglycemia must be avoided as should wide fluctuations in blood glucose. Since each patient is unique, there can be no set formula for how to best achieve this goal. While we strive for optimum glycemic control as quickly as possible, we need to be realistic during the learning phase with insulin.
If the patient has fasting hyperglycemia the morning after the bedtime snack, take the time to explain to them why that is so, although most of them have already worked this out for themselves. Then take the time to carefully explain that small post-snack increments (one to two units above the daily dose you have prescribed) are very unlikely to cause hypoglycemia and in fact might not be enough to have an appreciable effect on the fasting blood glucose. However, with patient trial and error, most seem to find the right extra dose that works for them and their snack.
Having patients self-adjust their medication dose is not something to be taken lightly, but there are clearly circumstances when it is appropriate. We feel comfortable when patients with type 1 diabetes using insulin pumps adjust their dose based on blood glucose, food intake and energy expenditure, but we also make certain they have had sufficient training. Likewise we can do the same in patients with type 2 diabetes using pre-meal rapid-acting insulin based on a sliding scale (where did that term come from?) that we design.