Hyperglycemia is an obvious metabolic accompaniment of diabetic emergencies and therefore it is tempting to correct it to normal values rapidly. However, hyperglycemia itself is not life threatening, although the metabolic sequelae and fluid losses can be. The high blood glucose levels may have been reached over days and sometimes weeks. The body's osmolar and metabolic homeostasis has slowly adjusted to this new biochemical environment. When insulin is given, the glucose is encouraged to move into the cell, causing water, potassium, magnesium, phosphate, and other compounds to enter the cell also. If this occurs too rapidly, it can cause cell swelling, hypokalemia, hypomagnesemia, and hypophosphatemia, and exacerbate hypovolemia (Fig 1). For this reason it is not necessary, and may in fact be dangerous, to give an initial insulin bolus.
Fig. 1 Effects of insulin therapy: insulin not only reduces the blood glucose levels, but also causes water and electrolytes to move into cells, with potential adverse effects.
Insulin should be commenced as an intravenous infusion at a low level (1 unit/h). It is then easy to measure blood glucose levels regularly and adjust the rate accordingly, as with any other drug infusion. Although the initial blood glucose levels of patients with hyperosmolar non-ketotic coma are higher than those with diabetic ketoacidosis, the former are more sensitive to insulin than the latter.
As with fluid regimens, there are many inflexible recommendations about the amount of insulin required for diabetic emergencies. Some patients may only ever require a low infusion rate (e.g. 1 unit/h), whereas others may require higher infusion rates (e.g. 15 units/h). Insulin needs should be tailored to individual patients. An average dose is misleading and inappropriate. Blood glucose should be performed initially and then at least hourly until a trend becomes obvious. Remember that the initial decrease in blood glucose is related mainly to the dilutional effect of fluid resuscitation. Measurement of blood glucose with bedside reflectance devices is inaccurate above levels of approximately 20 mmol/l. Blood must be sent to the laboratory for measurement until levels return to values below 20 mmol/l. Glucose can also be measured on some blood gas analyzers.
Aim to achieve a slow and smooth reduction in blood glucose over at least 24 to 72 h. So-called insulin resistance may be related to inadequate resuscitation and failure to transport the insulin to the peripheral cells.
Insulin may adsorb to glass and plastic, prompting the use of carrier solutions to deliver it. This appears to be more a theoretical than a practical concern and ignores the aim of reducing blood glucose levels by titrating the rate of insulin delivery rather than achieving absolute insulin levels.
Was this article helpful?