Hyperglycemia prior to, and shortly after, initiation of total parenteral nutrition was strongly associated with poor clinical outcomes in critically ill hospitalized patients, whether they had a history of diabetes or not, Emory University researchers determined in a retrospective study.
Patients had an almost threefold risk of dying if their maximum blood glucose before or within 24 hours of starting total parenteral nutrition (TPN) was above 180 mg/dL, compared with those whose levels stayed below 120 mg/dL. Many other factors were taken into account for the statistical analysis, including age, sex, and diabetes status, Dr. Guillermo E. Umpierrez, professor of medicine at Emory University, Atlanta, said at the Southern regional meeting of the American Federation for Medical Research.
Although hyperglycemia is a common complication of TPN, its prevalence and impact on clinical outcomes have been uncertain. Dr. Umpierrez and his associates reviewed the records of 276 medical/surgery patients who required TPN a mean 11 days after admission. Most came from surgical or medical intensive care units or the burn unit, but nearly 25% came from non-ICU floors. Also, 23% had a history of diabetes. Patients received TPN for a mean duration of 15 days. In-hospital mortality was 27% in the study, funded by the American Diabetes Association and the National Institutes of Health.
Patients who died had a higher maximum blood glucose before TPN (mean 147 mg/dL) than those who survived (mean 131 mg/dL), as well as a higher maximum blood glucose within 24 hours of TPN initiation (mean 202 mg/dL vs. 160 mg/dL). The differences in blood glucose were highly statistically significant.
In a multivariate analysis, the risk of pneumonia and the risk of acute renal failure were independently related to maximum blood glucose above 180 mg/dL vs. below 120 mg/dL.
In a later interview, Dr. Umpierrez said that pre-TPN blood glucose levels could alert medical teams to the possibility of TPN-related hyperglycemia. “Hospitalists should pay attention to blood glucose levels, not only in those receiving TPN but in patients with hyperglycemia before TPN,” he said. “Frequent blood glucose monitoring is needed to prevent and/or correct hyperglycemia.”
At his institution, the findings prompted a change in protocol to initiate insulin infusion as TPN is begun or to start insulin infusion in patients on TPN whose blood glucose is “persistently elevated,” which he defined as a level over 140 mg/dL.