[Glycemic control in sepsis and septic shock: friend or foe?].

INTRODUCTION: Intensive care patients commonly suffer from hyperglycemia. Evidence is growing that strictly maintaining normoglycemia by intensive insulin therapy (IIT) ameliorates outcome in these patients. Whether or not this also holds true for patients with sepsis and septic shock is the issue of this post-hoc analysis of the database (2,748 patients) of 2 recent prospective clinical trials. MATERIAL AND METHODS: A total of 950 patients suffering from sepsis were identified and of these 462 fulfilled the diagnostic criteria of septic shock upon admission to the intensive care unit (ICU). Patients were treated by either IIT [mean glycemia 5.88 mmol/l (106 mg/dl)] or conventional glucose management [mean glycemia 8.44 mmol/l (152 mg/dl)]. RESULTS: Under IIT the mortality of patients treated for more than 3 days in the ICU was lowered by 7.6% (p=0.03) in septic patients and by 8.7% (p=0.08) in septic shock patients. Polyneuropathy occurred less frequently under IIT compared to conventional glucose management (sepsis -9.8%, septic shock -14%; p<0.001). The incidence of acute renal failure was not affected by either treatment regimen (sepsis -3.3%, septic shock -3.1%; p<0.25). Intensive insulin therapy was associated with an increased risk of hypoglycemia (sepsis +16.7%, septic shock +18.8; p<0.0001) which did not, however, directly affect morbidity nor mortality. CONCLUSIONS: These data suggest that IIT improves outcome of patients with sepsis or septic shock. Hypoglycemia is a frequent complication, but its clinical relevance remains to be defined.