Candidemia after cardiac surgery in the intensive care unit: an observational study.

Candidemia is a well-recognized complication of hospital stay, especially in critically ill patients. There is not a general consensus that predictors for candidemia in cardiosurgical intensive care unit (cICU) are different from a general ICU and it has been reported that cardiopulmonary bypass time is a specific risk factor in the cICU. We performed a prospective study to evaluate the main predictors for candidemia in patients admitted to the cICU. Included patients were adults admitted between July 2005 and December 2007 with an ICU-length of stay (ICU-LOS) ≥48 hours after cardiac surgery. Exclusion criteria were solid organ or bone marrow transplants, previous diagnosis of candidemia or other invasive infections and ICU stay before surgery. A multiple regression analysis was performed to identify the risk factors. Among 1955 patients admitted to the cICU, 345 were enrolled. Only 26 patients (1.3%) had candidemia after an ICU-LOS of 20 days (inter-quartile range, IQR 8-49 days). Total parenteral nutrition [odds ratio (OR)=9.56; confidence interval (CI)=1.741-52.534], severe sepsis (OR=4.20; CI=1.292-13.667), simplified acute physiology score II (OR=1.16; CI=1.052-1.278) and ICU-LOS >20 days (OR=6.38; CI=1.971-20.660) were independent predictors of candidemia. Patients undergoing cardiac surgery developed candidemia late after cICU admission and the independent predictors were similar to the general ICU.

High tidal volume is associated with the development of acute lung injury after severe brain injury: an international observational study.

OBJECTIVE: Although a significant number of patients with severe brain injury develop acute lung injury, only intracranial risk factors have previously been studied. We investigated the role of extracranial predisposing factors, including hemodynamic and ventilatory management, as independent predictors of acute lung injury in brain-injured patients. DESIGN: Prospective multicenter observational study. SETTING: Four European intensive care units in university-affiliated hospitals. PATIENTS: Eighty-six severely brain-injured patients enrolled in 13 months. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: All patients with severe brain injury (Glasgow Coma Scale score <9) were studied for 8 days from admission. Ventilatory pattern, respiratory system compliance, blood gas analysis, and hemodynamic profile were recorded and entered in a stepwise regression model. Length of stay in the intensive care unit, ventilator-free days, and mortality were collected. Eighteen patients (22%) developed acute lung injury on day 2.8 +/- 1. They were initially ventilated with significantly higher tidal volume per predicted body weight (9.5 +/- 1 vs. 10.4 +/- 1.1), respiratory rate, and minute ventilation and more often required vasoactive drugs (p < .05). In addition to a lower Pao2/Fio2 (odds ratio 0.98, 95% confidence interval 0.98-0.99), the use of high tidal volume (odds ratio 5.4, 95% confidence interval 1.54-19.24) and relatively high respiratory rate (odds ratio 1.8, 95% confidence interval 1.13-2.86) were independent predictors of acute lung injury (p < .01). After the onset of acute lung injury, patients remained ventilated with similar tidal volumes to maintain mild hypocapnia and had a longer length of stay in the intensive care unit and fewer ventilator-free days (p < .05). CONCLUSIONS: In addition to a lower Pao2/Fio2, the use of high tidal volume and high respiratory rate are independent predictors of acute lung injury in patients with severe brain injury. In this patient population, alternative ventilator strategies should be considered to protect the lung and guarantee a tight CO2 control.