Interactive diabetes case 5: Insulin management in a 73-year-old patient with diabetes admitted to the hospital – Comment

COMMENT — The use of an insulin sliding scale as the sole source of insulin is not a satisfactory means of treating a patient with type 1 diabetes and many insulin-treated patients with type 2 diabetes. The persistent overuse of sliding scale insulin is a serious problem. Scheduled subcutaneous insulin with basal, nutritional, and correction (supplemental) components is the preferred method for achieving and maintaining glucose control in hospitalized patients who are not critically ill. The problems with the sliding scale are numerous:

●As usually written, an insulin sliding scale is a device for sustaining hyperglycemia in a patient with uncontrolled diabetes. It provides insulin only when the blood glucose is above a threshold value, typically 200 mg/dL (11.1 mmol/L), and then provides only small doses of insulin (eg, 4 to 6 units) for values just above that threshold. The consequence is that the patient often receives inadequate doses of insulin until the blood glucose level is unacceptably high, often over 300 mg/dL (16.7 mmol/L). If one thinks of the sliding scale as a servomechanism to maintain glucose homeostasis, the set-point is often in the range of 300 mg/dL (16.7 mmol/L), well above contemporary goals for glucose control in the hospital.

●An insulin sliding scale often generates a roller coaster pattern of glucose values in which very high blood glucose values alternate with very low blood glucose values. Very high blood glucose values are treated with insulin, sometimes with ensuing hypoglycemia. The hypoglycemia is then treated with (often excessive quantities of) carbohydrates and without insulin, with ensuing hyperglycemia. This cycle may occur repeatedly, as in the present case.

●An insulin sliding scale as used in this case is reactive, responding to the current blood glucose level, but it is not anticipatory or proactive. It does not anticipate carbohydrate intake, metabolic stress, or physical activity. It can be anticipatory if the dose takes into account both the pre-meal blood glucose level and the estimated carbohydrate intake of the meal. Information on an approach to estimating the dose of short- and very short-acting insulins can be found elsewhere. (See "Interactive diabetes case 3: Hypoglycemia in a patient with type 1 diabetes – B2".)

Nevertheless, a sliding scale may be useful in certain circumstances:

●In a patient not previously treated with insulin who enters the hospital with satisfactory blood glucose values but is acutely ill (eg, with an infection), in whom the possibility of deterioration in control is high. In this setting, an insulin sliding scale provides the opportunity to assess the needs of the patient and to determine whether a more aggressive treatment program is needed (eg, an insulin infusion or an intermediate-acting insulin).

●In a patient treated with an intermediate or long-acting insulin who enters the hospital with satisfactory blood glucose values but is acutely ill (eg, with an infection), in whom the possibility of deterioration in control is high. This provides the opportunity to assess the needs of the patient, to determine whether larger doses of the intermediate or long-acting insulin are indicated, and to decide whether a more aggressive treatment program is needed (eg, an insulin infusion).

When an insulin-treated patient is in the fasted state (ie, nil per os [NPO]) for a procedure, an appropriate approach is to administer a reduced dose of insulin and to provide intravenous (IV) glucose. Ordinarily, one gives one-half the usual dose of insulin as intermediate or long-acting insulin, assuming that the patient is in satisfactory metabolic control. In our patient, one could give 11 units of NPH insulin in the morning of the procedure and provide D5W at 80 to 100 cc/hour starting at approximately 7 AM and continuing until the patient has eaten a full meal. If the patient had been treated with a mixed dose of intermediate and short-acting insulin (eg, NPH and regular insulin), one might take one-half the total dose of insulin and give that as the intermediate-acting insulin. As an example, if the patient's morning insulin dose had been NPH 20 units and regular 10 units, one might give one-half the total insulin as NPH insulin (ie, 15 units of NPH). While some clinicians prefer to administer one-half of each type of insulin, this consultant prefers to avoid the administration of short-acting insulin to patients who are NPO, even when receiving IV glucose. This is especially a concern when the patient will be sedated or under anesthesia and unable to respond to the symptoms of hypoglycemia, or to inform a medical person of the symptoms.

To explore the consequences of the other actions, return to the case at the beginning of this sequence. (See "Interactive diabetes case 5: Insulin management in a 73-year-old patient with diabetes admitted to the hospital".)

[1-3]