Objective: Hyperglycemia is a recognized complication of parenteral nutrition (PN). We aimed to determine the impact of hyperglycemia during PN unaccompanied by tight blood glucose (BG) control on hospital complications and mortality.
Methods: We reviewed the medical records of 276 medical and surgical patients receiving PN to determine the impact of hyperglycemia on survival after adjusting for known prognostic factors, and to determine whether BG levels before initiation of PN, within 24 hours of PN initiation, or during PN therapy are predictive of adverse outcomes.
Results: A total of 276 medical (35%) and surgical (65%) patients receiving PN initiated 12 ± 12 days after admission for a mean of 15 ± 24 days. Deceased patients (27.2%) were older, had higher Acute Physiology and Chronic Health Evaluation II scores, and had higher BG levels during PN therapy versus survivors (all, P < 0.01). Deceased patients had higher BG levels within 24 hours of PN initiation (162 ± 55 mg/dL vs 139 ± 37 mg/dL; P = 0.003) and higher BG levels during days 2 to 10 of PN (161 ± 53 mg/dL vs 142 ± 34 mg/dL; P = 0.013) compared with survivors. Blood glucose levels were associated with increased odds ratio (OR) for mortality pre-PN (P = 0.008), within 24 hours of PN initiation (P < 0.001), and during days 2 to 10 of PN (P < 0.001). In multiple regression models adjusted for age, sex, and history of diabetes, mortality was independently associated with pre-PN BG levels 121 to 150 mg/dL (OR, 2.2; 95% confidence interval [CI], 1.1-4.4), 151 to 180 mg/dL (OR, 3.41; 95% CI, 1.3-8.7,), and > 180 mg/dL (OR, 2.2; 95% CI, 0.9-5.2), and with BG levels within 24 hours of PN initiation > 180 mg/dL (OR, 2.8; 95% CI, 1.2-6.8). A BG level > 180 mg/dL within 24 hours of PN initiation was associated with increased risk of pneumonia (OR, 3.1; 95% CI, 1.4-7.1) and acute renal failure (OR, 2.3; 95% CI, 1.1-5.0).
Conclusion: Hyperglycemia during PN without tight BG control is associated with increased risk of hospital complications and mortality. Randomized controlled trials are needed to determine benefits of intensified glycemic control on clinical outcomes in hospitalized subjects receiving PN.