PURPOSE

Because of the immune-suppressive effect of cerebral damage, stroke patients are at high risk for infections. These might result in sepsis, which is the major contributor to intensive care unit (ICU) mortality. Although there are numerous studies on infections in stroke patients, the role of sepsis as a poststroke complication is unknown.

METHODS

We retrospectively analyzed incidence of and risk factors for sepsis acquisition as well as outcome parameters of 238 patients with ischemic or hemorrhagic strokes consecutively admitted to the neurologic ICU in a tertiary university hospital between January 1, 2009, and December 31, 2010. Basic demographic and clinical data including microbiological parameters as well as factors describing stroke severity (eg, lesion volume and National Institute of Health stroke scale score) were recorded and included into the analysis. The diagnosis of sepsis was based on the criteria of the German Sepsis Society.

RESULTS

We identified 30 patients (12.6%) with sepsis within the first 7 days from stroke onset. The lungs were the most frequent source of infection (93.3%), and gram-positive organisms were dominating the microbiologic spectrum (52.4%). Comorbidities (chronic obstructive pulmonary disease and immunosuppressive disorders) and Simplified Acute Physiology Score II but none of the factors describing stroke severity were independent predictors of sepsis acquisition. Sepsis was associated with a significantly worse prognosis, leading to a 2-fold increased mortality rate during in-hospital care (36.7% vs 18.8%) and after 3 months (56.5% vs 28.5%), but only in the subgroup of supratentorial hemorrhages, it was an independent predictor of in-hospital and 3-month mortality. Other factors significantly associated with death in a multivariate analysis were chronic obstructive pulmonary disease, malignancies (in-hospital mortality only), and Simplified Acute Physiology Score II (3-month mortality only) for ischemia and heart failure (in-hospital mortality only), National Institute of Health stroke scale score (in-hospital mortality only), and stroke volume for hemorrhages, respectively.

CONCLUSIONS

Sepsis seems to be a frequent complication of stroke patients requiring neurologic ICU treatment. Predictors of sepsis acquisition in our study were comorbidities and severity of deterioration of physiological status, but not stroke severity. A better understanding of risk factors is important for prevention and early recognition, whereas knowledge of outcome may help in prognosis prediction. Further studies are needed to clarify the optimal preventive treatment for these patients.