Search

Severe hyperkalemia (serum potassium level >6 mmol per liter [mEq per liter]), often with electrocardiographic disturbances, was noted at presentation in 30% of 73 hyperglycemic episodes (serum glucose concentration >25 mmol per liter [455 mg per dl]) observed in 15 in-hospital patients with insulin-dependent diabetes mellitus who were receiving long-term hemodialysis or peritoneal dialysis. Serum glucose concentration and total carbon dioxide content correlated significantly with the presenting serum potassium concentration. Treatment with parenteral insulin alone resulted in a decrease of the serum glucose value from 41 +/- 14 (standard deviation) to 11 +/- 5 mmol per liter (P <.001) and of serum potassium level from 5.2 +/- 1.2 to 4.0 +/- 0.6 mmol per liter (P <.001). The changes in serum glucose concentration and in carbon dioxide content and the serum potassium concentration at hyperglycemia were found to be independent correlates of the decrease in potassium concentration during treatment. Insulin alone resulted in correction of hyperkalemia in all instances. Posttreatment hypokalemia was noted in only two instances, each associated with both ketoacidosis and low-normal serum potassium concentration at hyperglycemia. Giving insulin is the only treatment usually needed for the hyperkalemia of hyperglycemia in patients on ongoing dialysis.

We measured by affinity chromatography glycosylated hemoglobin levels in the blood of 43 diabetic and nondiabetic patients (139 measurements) on long-term dialysis therapy (continuous ambulatory peritoneal dialysis and hemodialysis) to determine the usefulness of this method of estimating glycemic control in diabetic persons on dialysis therapy. In nondiabetic patients, glycosylated hemoglobin levels were within the normal range (4.0% to 6.8% of total blood hemoglobin levels) for both continuous ambulatory peritoneal dialysis and hemodialysis. Glycosylated hemoglobin values correlated significantly with fasting blood glucose levels, serum urea levels, and serum total carbon dioxide content. By stepwise regression, fasting blood glucose values accounted statistically for .54 of the variability (R2) in glycosylated hemoglobin. The contribution of the other variables to this variability was minimal. In 9 diabetic patients (3 on hemodialysis), glycosylated hemoglobin levels correlated significantly with average daily blood glucose levels. Regression of the fasting blood glucose value on glycosylated hemoglobin was similar between continuous ambulatory peritoneal dialysis and hemodialysis. Measuring glycosylated hemoglobin levels by affinity chromatography is a suitable method for assessing glycemia in dialysis patients.