March 9, 1997
Media Contact: John Cramer
Phone: (410) 955-1534
The most common method for controlling blood-sugar levels in hospital patients with diabetes is ineffective and in some cases worsens their condition, a Johns Hopkins study suggests.
Researchers studying 171 hospital patients with diabetes mellitus found that sliding scale insulin regimens provided no benefit. Sliding scale regimens are insulin doses given four times a day according to changing levels of blood sugar. Results also show that when used alone, sliding scale regimens may increase the risk of hyperglycemia (high blood sugar). An estimated 75 percent of all diabetic patients in U.S. hospitals are put on sliding scale insulin regimens. Results are published in the March 10 issue of Archives of Internal Medicine.
"Glycemic control problems are common in hospitalized patients with diabetes mellitus, but sliding scale insulin regimens are one of those approaches that has been passed from generation to generation of physicians without evidence that they work," says William S. Queale, M.D., lead author and a resident in internal medicine. "Until further studies are done, we recommend keeping these patients on whatever standard insulin regimen has been working for them at home and modifying that regimen according to their response to treatment."
There are four insulin regimens: nutrition without insulin; a standing regimen, which generally delivers a fixed amount of intermediate-acting insulin twice a day; a sliding scale regimen, which four times a day delivers varying amounts of short-acting insulin when the blood-sugar level is abnormal; and a combination of standing and sliding scale regimens.
Results of the seven-week study show 76 percent of the patients were placed on sliding scale regimens. Patients on standing regimens fared better than those on sliding scale regimens and those on no insulin who managed their diabetes through nutrition. When used alone, sliding scale insulin regimens tripled the risk of hyperglycemia compared with those patients receiving no insulin.
The Hopkins team also identified several predictors of hypoglycemia (low blood sugar) and hyperglycemia among diabetic inpatients. Hypoglycemic episodes occurred in 23 percent of patients and hyperglycemic episodes in 40 percent. Those at highest risk for hypoglycemia were African Americans and those with a low body weight and low levels of the protein albumin in the blood. Use of corticosteroids reduced the risk of hypoglycemia. Those at highest risk for hyperglycemia were women and those with severe illnesses, severe diabetic complications, high blood-sugar level when admitted, infectious diseases and corticosteroid use.
An estimated three million people are admitted to U.S. hospitals annually for diabetes-related problems. Poor blood-sugar control weakens the immune system and increases the risk of nerve damage and other complications.
Other authors of the study, which was supported by the American Diabetes Association, were Alexander J. Seidler, Ph.D., and senior author Frederick L. Brancati, M.D.