Adjuvant selenium supplementation in the form of sodium selenite in postoperative critically ill patients with severe sepsis.

INTRODUCTION: Plasma selenium (Se) concentrations are reduced in critically ill surgical patients, and lower plasma Se concentrations are associated with worse outcomes. We investigated whether adjuvant Se supplementation in the form of sodium selenite could improve outcomes in surgical patients with sepsis. METHODS: In this retrospective study, all adult patients admitted to a 50-bed surgical ICU with severe sepsis between January 2004 and April 2010 were included and analysed according to whether they had received adjuvant Se supplementation, which was given at the discretion of the attending physician. When prescribed, Se was administered in the form of sodium selenite pentahydrate (Na2SeO3∙5H2O), in which 100 µg of Se corresponds to 333 µg of sodium selenite. A bolus of sodium selenite corresponding to 1,000 µg of Se was injected intravenously through a central venous line for 30 minutes, followed by infusion of 1,000 µg/day for 24 hours for 14 days until ICU discharge or death. We performed logistic regression analysis to investigate the impact of adjuvant Se supplementation on hospital mortality. RESULTS: Adjuvant Se was administered to 413 (39.7%) of the 1,047 patients admitted with severe sepsis. Age and sex were similar between patients who received adjuvant Se and those who did not. Compared with patients who did not receive adjuvant Se supplementation, patients who did had higher scores on the Simplified Acute Physiology Score II, a greater prevalence of cancer upon admission to the ICU and were more commonly admitted after abdominal surgery. Compared with patients who did not receive adjuvant Se, patients who did had higher hospital mortality rates (46% versus 39.1%; P = 0.027), and longer median (interquartile range (IQR)) ICU stays (15 days (6 to 24) versus 11 days (4 to 24); P = 0.01) and hospital lengths of stay (33 days (21 to 52) versus 28 days (17 to 46); P = 0.001). In multivariable analysis, adjuvant Se supplementation was not independently associated with favourable outcome (odds ratio = 1.19, 95% confidence interval = 0.86 to 1.65; P = 0.288). CONCLUSIONS: In this retrospective analysis of a large cohort of surgical ICU patients with severe sepsis, adjuvant Se supplementation in the form of sodium selenite had no impact on in-hospital death rates after adjustment for confounders.

SAPS 3 is not superior to SAPS 2 in cardiac surgery patients.

OBJECTIVES: Cardiac surgery patients are excluded from SAPS2 but included in SAPS3. Neither score is evaluated for this exclusive population; however, they are used daily. We hypothesized that SAPS3 may be superior to SAPS2 in outcome prediction in cardiac surgery patients. DESIGN: All consecutive patients undergoing cardiac surgery between January 2007 and December 2010 were included in our prospective study. Both models were tested with calibration and discrimination statistics. We compared the AUC of the ROC curves by DeLong's method and calculated OCC values. RESULTS: A total of 5207 patients with mean age of 67.2 ± 10.9 years were admitted to the ICU. The mean length of ICU stay was 4.6 ± 7.0 days and the ICU mortality was 5.9%. The two tested models had acceptable discriminatory power (AUC: SAPS2: 0.777-0.875; SAPS3: 0.757-893). SAPS3 had a low AUC and poor calibration on admission day. SAPS2 had poor calibration on Days 1-6 and 8. CONCLUSIONS: Despite including cardiac surgery patients, SAPS3 was not superior to SAPS2 in our analysis. In this large cohort of ICU cardiac surgery patients, performance of both SAPS models was generally poor. In this subset of patients, neither scoring system is recommended.

Glasgow Coma Scale for outcome prediction after cardiac surgery: is it applicable?

OBJECTIVES: The Glasgow Coma Scale (GCS) is used commonly for assessing patients' neurologic condition and outcome in intensive care units (ICUs); however, its reliability in cardiac surgical patients has been questioned. It has been claimed that active sedation is the cause of its unsuitability for these patients. This study aimed to compare the accuracy of GCS in cardiac surgical patients with and without active sedation to find out if the inapplicability of GCS in surgical patients is related to active sedation. DESIGN: This was an observational cohort study. SETTING: The study was conducted in a cardiac surgical intensive care unit between January 1, 2007 and December 31, 2009. PARTICIPANTS: All consecutive adult cardiac surgical patients were included in this study. INTERVENTIONS: All types of cardiac surgical procedures performed during the study period were included without any exceptions. The study population was divided into 2 groups: sedated and non-sedated. MEASUREMENTS AND MAIN RESULTS: GCS was calculated daily for the first 7 postoperative days. The authors developed a new 4-point neurologic descriptor (ND): (1) neurologically free, (2) ICU psychosis, (3) actively sedated, and (4) documented focal neurologic deficits. The accuracy of both scales (GCS and ND) at predicting ICU mortality was compared by replacing the GCS in the Sequential Organ Failure Assessment (SOFA) score with the new ND, producing a modified SOFA. GCS was not an accurate outcome predictor in non-sedated or sedated patients. The ND was superior to GCS. Correspondingly, the modified SOFA showed a significantly higher accuracy of ICU-mortality prediction than the original SOFA. CONCLUSIONS: Regardless of active sedation, GCS is not accurate at outcome prediction for cardiac surgical patients. The suggested ND is a simple and more accurate risk stratification variable in cardiac surgical ICUs.

Anticoagulation strategies in venovenous hemodialysis in critically ill patients: a five-year evaluation in a surgical intensive care unit.

Renal failure is a common complication among critically ill patients. Timing, dosage, and mode of renal replacement (RRT) are under debate, but also anticoagulation strategies and vascular access interfere with dialysis success. We present a retrospective, five-year evaluation of patients requiring RRT on a multidisciplinary 50-bed surgical intensive care unit of a university hospital with special regard to anticoagulation strategies and vascular access. Anticoagulation was preferably performed with unfractionated heparin or regional citrate application (RAC). Bleeding and suspected HIT-II were most common causes for RAC. In CVVHD mode filter life span was significantly longer under RAC compared to heparin or other anticoagulation strategies (P=0.001). Femoral vascular access was associated with reduced filter life span (P=0.012), especially under heparin anticoagulation (P=0.015). Patients on RAC had higher rates of metabolic alkalosis (P=0.001), required more transfusions (P=0.045), and showed higher illness severity measured by SOFA scores (P=0.001). RRT with unfractionated heparin represented the most common anticoagulation strategy in this study population. However, patients with bleeding risk and severe organ dysfunction were more likely placed on RAC. Citrate provided longer filter life spans regardless of vascular access site. Attention has to be paid to metabolic disturbances.

Perioperative fluid therapy with tetrastarch and gelatin in cardiac surgery--a prospective sequential analysis*.

OBJECTIVES: To determine clinical outcomes of synthetic colloids, tetrastarch, and gelatin, used before and after cardiac surgery. DESIGN: Prospective observational cohort study. SETTING: Fifty-bed cardiac ICU. PATIENTS: Six thousand four hundred seventy-eight consecutive patients with cardiopulmonary bypass surgery. INTERVENTIONS: Fluid therapy in the operating room and on the ICU directed at preset hemodynamic goals: 1) hydroxyethyl starch (predominantly 6% hydroxyethyl starch 130/0.4) in 2004-2006, n = 2,137; 2) 4% gelatin in 2006-2008, n = 2,324; and 3) only crystalloids in 2008-2010, n = 2,017. MEASUREMENTS AND MAIN RESULTS: Renal replacement therapy was more common during periods when patients received synthetic colloids compared to only crystalloids. Risk of renal replacement therapy was greater after hydroxyethyl starch (odds ratio, 2.29; 95% CI, 1.47-3.60) and gelatin (odds ratio, 2.75; 95% CI, 1.84-4.16; both p < 0.001) compared to crystalloid. Propensity score stratification confirmed greater use of renal replacement therapy in the hydroxyethyl starch and gelatin periods compared to the crystalloid period (odds ratio, 1.46 [1.08, 1.97]; p = 0.013 and odds ratio, 1.72 [1.33, 2.24]; p < 0.001, respectively). Time to vasopressor cessation, normalization of serum lactate, and mean arterial pressure did not differ among groups. Total fluid requirement was 163 mL/kg in the hydroxyethyl starch period, 207 mL/kg in the gelatin period, and 224 mL/kg in the crystalloid period. Fluid intake was higher in the crystalloid group only during the first 20 hours. CONCLUSIONS: In cardiac surgery patients, fluid therapy with perioperative administration of synthetic colloids carries a high risk of renal replacement therapy and is not more effective than treating with only crystalloids.

Mortality prediction after cardiac surgery: blood lactate is indispensible.

BACKGROUND: Blood lactate is accepted as a mortality risk marker in intensive care units (ICUs), especially after cardiac surgery. Unfortunately, most of the commonly used ICU risk stratification scoring systems did not include blood lactate as a variable. We hypothesized that blood lactate alone can predict the risk of mortality after cardiac surgery with an accuracy that is comparable to those of other complex models. We therefore evaluated its accuracy at mortality prediction and compared it with that of other widely used complex scoring models statistically. METHODS: We prospectively collected data of all consecutive adult patients who underwent cardiac surgery between January 1, 2007, and December 31, 2009. By using χ2 statistics, a blood lactate-based scale (LacScale) with only four cutoff points was constructed in a developmental set of patients (January 1, 2007, and May 31, 2008). LacScale included five categories: 0 (≤ 1.7 mmol/L); 1 (1.8-5.9 mmol/L), 2 (6.0-9.3 mmol/L), 3 (9.4-13.3 mmol/L), and 4 (≥ 13.4 mmol/L). Its accuracy at predicting ICU mortality was evaluated in another independent subset of patients (validation set, June 1, 2008, and December 31, 2009) on both study-population level (calibration analysis, overall correct classification) and individual-patient-risk level (discrimination analysis, ROC statistics). The results were then compared with those obtained from other widely used postoperative models in cardiac surgical ICUs (Sequential Organ Failure Assessment [SOFA] score, Simplified Acute Physiology Score II [SAPS II], and Acute Physiology and Chronic Health Evaluation II [APACHE II] score). RESULTS: ICU mortality was 5.8% in 4,054 patients. LacScale had a reliable calibration in the validation set (2,087 patients). It was highly accurate in predicting ICU mortality with an area under the ROC curve (area under curve [AUC]; discrimination) of 0.88. This AUC was significantly larger than that of all the other models (SOFA 0.83, SAPS II: 0.79 and APACHE II: 0.76) according to DeLong's comparison. Integrating the LacScale in those scores further improved their accuracy by increasing their AUCs (0.88, 0.81, and 0.80, respectively). This improvement was also highly significant. CONCLUSION: Blood lactate accurately predicts mortality at both individual patient risk and patient cohort levels. Its precision is higher than that of other commonly used "complex" scoring models. The proposed LacScale is a simple and highly reliable model. It can be used (at bedside without electronic calculation) as such or integrated in other models to increase their accuracy.

Effects of fluid resuscitation with synthetic colloids or crystalloids alone on shock reversal, fluid balance, and patient outcomes in patients with severe sepsis: a prospective sequential analysis.

OBJECTIVE: To assess shock reversal and required fluid volumes in patients with septic shock. DESIGN: Prospective before and after study comparing three different treatment periods. SETTING: Fifty-bed single-center surgical intensive care unit. PATIENTS: Consecutive patients with severe sepsis. INTERVENTIONS: Fluid therapy directed at preset hemodynamic goals with hydroxyethyl starch (predominantly 6% hydroxyethyl starch 130/0.4) in the first period, 4% gelatin in the second period, and only crystalloids in the third period. MEASUREMENTS AND MAIN RESULTS: Main outcome was time to shock reversal (serum lactate <2.2 mmol/L and discontinuation of vasopressor use). Hemodynamic goals were mean arterial pressure >70 mm Hg; ScvO2 <70%; central venous pressure >8 mm Hg. Safety outcomes were acute kidney injury defined by Risk, Injury, Failure, Loss, and End-stage kidney disease criteria and new need for renal replacement therapy. Hemodynamic measures, serum lactate, and creatinine were comparable at baseline in all study periods (hydroxyethyl starch n = 360, gelatin n = 352, only crystalloids n = 334). Severity scores, hospital length of stay, and intensive care unit or hospital mortality did not differ significantly among groups. All groups showed similar time to shock reversal. More fluid was needed over the first 4 days in the crystalloid group (fluid ratios 1.4:1 [crystalloids to hydroxyethyl starch] and 1.1:1 [crystalloids to gelatin]). After day 5, fluid balance was more negative in the crystalloid group. Hydroxyethyl starch and gelatin were independent risk factors for acute kidney injury (odds ratio, 95% confidence interval 2.55, 1.76-3.69 and 1.85, 1.31-2.62, respectively). Patients receiving synthetic colloids received significantly more allogeneic blood products. CONCLUSIONS: Shock reversal was achieved equally fast with synthetic colloids or crystalloids. Use of colloids resulted in only marginally lower required volumes of resuscitation fluid. Both low molecular weight hydroxyethyl starch and gelatin may impair renal function.

Renal effects of synthetic colloids and crystalloids in patients with severe sepsis: a prospective sequential comparison.

OBJECTIVES: Hydroxyethyl starch 200 is associated with renal impairment in sepsis, but hydroxyethyl starch 130/0.4 and gelatin are considered to be less harmful. We hypothesized that fluid therapy with only crystalloids would decrease the incidence of acute kidney injury. DESIGN: Prospective sequential comparison during intensive care unit stay. SETTING: Surgical intensive care unit. PATIENTS: Patients with severe sepsis. INTERVENTIONS: Changes in standard fluid therapy, with predominantly 6% hydroxyethyl starch from January 2005 to June 2005, 4% gelatin from January 2006 to June 2006, and only crystalloids from September 2008 to June 2009. MEASUREMENTS AND MAIN RESULTS: Acute kidney injury was defined by the presence of at least one RIFLE class; 118 patients received hydroxyethyl starch, 87 patients received gelatin, 141 patients received only crystalloids. Baseline serum creatinine values were similar. Patients received median cumulative doses of 46 (interquartile range, 18-92) mL/kg hydroxyethyl starch and 43 (interquartile range, 18-76) mL/kg gelatin. Total median fluid amounts were 649 (interquartile range, 275-1098) mL/kg in the hydroxyethyl starch group, 525 (237-868) mL/kg in the gelatin group, and 355 (173-911) mL/kg in the crystalloid group. The difference was statistically significant for hydroxyethyl starch after adjustment for multiple testing. Mean daily fluid intake and fluid balance were higher on days 0 and 1 in the crystalloid group. Acute kidney injury occurred in 70% of patients receiving hydroxyethyl starch (adjusted p = .002) and in 68% of patients receiving gelatin (adjusted p = .025) vs. 47% patients receiving crystalloids. Need for renal replacement therapy tended to be higher in the hydroxyethyl starch group (34%; adjusted p = .086) and in the gelatin group (34%; adjusted p = .162) in comparison to the crystalloid group (20%). Intensive care unit and hospital mortality were similar in each group (hydroxyethyl starch: 35% and 43%; gelatin: 26% and 31%; crystalloids: 30% and 37%). CONCLUSION: Fluid resuscitation with only crystalloids was equally effective, resulted in a more positive fluid balance only on the first 2 days, and was associated with a lesser incidence of acute kidney injury.

Continuous transesophageal echo-Doppler assessment of hemodynamic function during laparoscopic cholecystectomy.

STUDY OBJECTIVE: The objective of this study was to examine the utility of the transesophageal echo-Doppler device in evaluating hemodynamic changes during laparoscopic cholecystectomy. DESIGN: This was a prospective, controlled, observational open study. SETTING: The study took place in a university hospital. PATIENTS: Twenty patients with ASA physical statuses II and III undergoing laparoscopic cholecystectomy were enrolled into the study. INTERVENTIONS AND MEASUREMENTS: A standardized general anesthetic and surgical technique was used for all patients. Similar depth of hypnosis (using bispectral index monitoring) was maintained in all patients. Hemodynamic parameters including mean arterial pressure (MAP), cardiac index (CI), left ventricular (LV) ejection time interval indexed to the heart rate, maximum acceleration, peak velocity, and systemic vascular resistance (SVR) were recorded at predetermined intervals: before incision, after peritoneal CO(2) insufflation and head-up tilt, every 10 minutes thereafter, and after deflation of the abdomen and return to supine position. MAIN RESULTS: The transesophageal echo-Doppler probe placement was achieved in 3 to 5 minutes in all patients, and the probe position was maintained after creation of pneumoperitoneum and change in positioning. Induction of pneumoperitoneum and head-up tilt resulted in a significant increase in MAP and SVR (P < .05) that remained higher until deflation. The CI, LV ejection time interval indexed to the heart rate (a measure of LV filling), and maximum acceleration (a measure of contractility and global ventricular function) remained stable. CONCLUSIONS: The transesophageal echo-Doppler device can be used during laparoscopic cholecystectomy. The LV function, as determined by measurement of CI and maximum acceleration, was preserved during laparoscopic cholecystectomy despite significant increases in afterload (ie, MAP and SVR).

Combination of European System for Cardiac Operative Risk Evaluation (EuroSCORE) and Cardiac Surgery Score (CASUS) to Improve Outcome Prediction in Cardiac Surgery.

BACKGROUND We hypothesized that the combination of a preoperative and a postoperative scoring system would improve the accuracy of mortality prediction and therefore combined the preoperative 'additive EuroSCORE' (European system for cardiac operative risk evaluation) with the postoperative 'additive CASUS' (Cardiac Surgery Score) to form the 'modified CASUS'. MATERIAL AND METHODS We included all consecutive adult patients after cardiac surgery during January 2007 and December 2010 in our prospective study. Our single-centre study was conducted in a German general referral university hospital. The original additive and the 'modified CASUS' were tested using calibration and discrimination statistics. We compared the area under the curve (AUC) of the receiver characteristic curves (ROC) by DeLong's method and calculated overall correct classification (OCC) values. RESULTS The mean age among the total of 5207 patients was 67.2 ± 10.9 years. Whilst the ICU mortality was 5.9% we observed a mean length of ICU stay of 4.6 ± 7.0 days. Both models demonstrated excellent discriminatory power (mean AUC of 'modified CASUS': ≥ 0.929; 'additive CASUS': ≥ 0.920), with no significant differences according to DeLong. Neither model showed a significant p-value (<0.05) in calibration. We detected the best OCC during the 2nd day (modified: 96.5%; original: 96.6%). CONCLUSIONS Our 'additive' and 'modified' CASUS are reasonable overall predictors. We could not detect any improvement in the accuracy of mortality prediction in cardiac surgery by combining a preoperative and a postoperative scoring system. A separate calculation of the two individual elements is therefore recommended.

An Early Warning Scoring System to Identify Septic Patients in the Prehospital Setting: The PRESEP Score.

OBJECTIVES: The objective was to develop and evaluate an early sepsis detection score for the prehospital setting. METHODS: A retrospective analysis of consecutive patients who were admitted by emergency medical services (EMS) to the emergency department of the Jena University Hospital was performed. Because potential predictors for sepsis should be based on consensus criteria, the following parameters were extracted from the EMS protocol for further analysis: temperature, heart rate (HR), respiratory rate (RR), oxygen saturation (SaO2 ), Glasgow Coma Scale score, blood glucose, and systolic blood pressure (sBP). Potential predictors were stratified based on inspection of Loess graphs. Backward model selection was performed to select risk factors for the final model. The Prehospital Early Sepsis Detection (PRESEP) score was calculated as the sum of simplified regression weights. Its predictive validity was compared to the Modified Early Warning Score (MEWS), the Robson screening tool, and the BAS 90-30-90. RESULTS: A total of 375 patients were included in the derivation sample; 93 (24.8%) of these had sepsis, including 60 patients with severe sepsis and 12 patients with septic shock. Backward model selection identified temperature, HR, RR, SaO2 , and sBP for inclusion in the PRESEP score. Simplified weights were as follows: temperature > 38°C = 4, temperature < 36°C = 1, HR > 90 beats/min = 2, RR > 22 breaths/min = 1, SaO2 < 92% = 2, and sBP < 90 mm Hg = 2. The cutoff value for a possible existing septic disease based on maximum Youden's index was ≥4 (sensitivity 0.85, specificity 0.86, positive predictive value [PPV] 0.66, and negative predictive value [NPV] 0.95). The area under the receiver operating characteristic curve (AUC) of the PRESEP score was 0.93 (95% confidence interval [CI] = 0.89 to 0.96) and was larger than the AUC of the MEWS (0.93 vs. 0.77, p < 0.001). The PRESEP score surpassed MEWS and BAS 90-60-90 for sensitivity (0.74 and 0.62, respectively), specificity (0.75 and 0.83), PPV (0.45 and 0.51), and NPV (0.91 and 0.89). The Robson screening tool had a higher sensitivity and NPV (0.95 and 0.97), but its specificity and PPV were lower (0.43 and 0.32). CONCLUSIONS: The PRESEP score could be a valuable tool for identifying septic patients in the prehospital setting in the case of suspected infection. It should be prospectively validated.

Inclusion of 'ICU-Day' in a Logistic Scoring System Improves Mortality Prediction in Cardiac Surgery.

BACKGROUND Prolonged intensive care unit (ICU) stay is a predictor of mortality. The length of ICU stay has never been considered as a variable in an additive scoring system. How could this variable be integrated into a scoring system? Does this integration improve mortality prediction? MATERIAL AND METHODS The 'modified CArdiac SUrgery Score' (CASUS) was generated by implementing the length of stay as a new variable to the 'additive CASUS'. The 'logistic CASUS' already considers this variable. We defined outcome as ICU mortality and statistically compared the three CASUS models. Discrimination, comparison of receiver operating characteristic curves (DeLong's method), and calibration (observed/expected ratio) were analyzed on days 1-13. RESULTS Between 2007 and 2010, we included 5207 cardiac surgery patients in this prospective study. The mean age was 67.2 ± 10.9 years. The mean length of ICU stay was 4.6 ± 7.0 days and ICU mortality was 5.9%. All scores had good discrimination, with a mean area under the curve of 0.883 for the additive and modified, and 0.895 for the 'logistic CASUS'. DeLong analysis showed superiority in favor of the logistic model as from day 5. The calibration of the logistic model was good. We identified overestimation (days 1-5) and accurate (days 6-9) calibration for the additive and 'modified CASUS'. The 'modified CASUS' remained accurate but the 'additive CASUS' tended to underestimate the risk of mortality (days 10-13). CONCLUSIONS The integration of length of ICU stay as a variable improves mortality prediction significantly. An 'ICU-day' variable should be included into a logistic but not an additive model.

Renal outcome after vancomycin treatment and renal replacement therapy in patients with severe sepsis and septic shock: a retrospective study.

PURPOSE: Acute kidney injury during systemic infections is common; however, renal outcome is poorly investigated. The increase of multiresistant pathogens leads to the use of potential nephrotoxic antibiotics as vancomycin. We investigated the impact of vancomycin and renal replacement therapy (RRT) for renal recovery during sepsis. MATERIALS AND METHODS: This is a retrospective data analysis of 1159 patients with severe sepsis or septic shock. Logistic regression models were performed. RESULTS: In total, 390 (33.6%) patients required RRT during intensive care unit (ICU) stay; 233 (20.1%), at discharge. Admission estimated glomerular filtration rate (eGFR) predicted the need of RRT during stay (odds ratio [OR] 0.969 [0.959-0.979] per increase of 1 mL/min, P<.001) and the prolonged need of RRT at ICU discharge (OR 0.979 [0.967-0.990], P<.001). Survivors without any RRT showed an improvement of eGFR at discharge, whereas patients after RRT did not (7.1 vs 0.8 mL/[min 1.73 m2], P<.001). The use (OR 1.648 [1.067-2.546], P<.05) and duration of vancomycin treatment (OR 1.043 [1.004-1.084] per each additional treatment day, P<.05) were predictors for ongoing RRT at discharge. CONCLUSIONS: Estimated GFR at ICU admission predicts renal outcome, whereas the use of vancomycin increases the probability of a prolonged need for RRT at discharge from ICU. The use of alternative antibiotics for certain patients, indicated by eGFR at admission, might be considered.

The impact of age on outcome after major surgical procedures.

PURPOSE: The aim of this study was to investigate the effects of age on outcome in a large cohort of surgical intensive care unit (ICU) patients. METHODS: In this retrospective analysis of prospectively collected data, all 11537 adult patients admitted directly from the operating room to our 50-bed surgical ICU between January 1, 2004, and January 31, 2009, were included. Patients were classified into 5 subgroups according to age (18-50 [reference category], 51-65, 66-75, 76-85, >85 years). RESULTS: Severity scores and the incidence of comorbid conditions on ICU admission increased steadily with age. Intensive care unit and hospital mortality rates were 4.4% and 8.7%, respectively, and increased with age to reach 12.4% and 28.2%, respectively, in patients older than 85 years. In multivariate logistic regression analysis, age was an independent risk factor for in-hospital death (odds ratio, 1.04; 95% confidence interval, 1.03-1.04, per year; P < .001). Gastrointestinal surgery was independently associated with a higher risk of in-hospital mortality in patients older than 50 years, whereas neurosurgery was associated with a higher risk of in-hospital death only in patients older than 65 years. CONCLUSIONS: Mortality rates increase with age, with an exponential increase in patients older than 65 years. Age is an independent risk factor for in-hospital death, irrespective of the type of surgical intervention.

Attributable costs of patients with candidemia and potential implications of polymerase chain reaction-based pathogen detection on antifungal therapy in patients with sepsis.

PURPOSE: The purposes of this study were to calculate attributable costs of candidemia in patients with severe sepsis and to obtain preliminary data regarding the potential effects of polymerase chain reaction-based pathogen detection on antifungal therapy for these patients. METHODS: Patients treated between 2004 and 2010 because of severe sepsis were included into this retrospective analysis. The hospital management provided annual fixed costs per patient-day; data for variable intensive care unit costs were taken from the literature. Multiplex polymerase chain reaction (PCR) was used (VYOO, SIRS-Lab, Jena, Germany) for pathogen detection in the blood. RESULTS: Thirty-two patients with candidemia were identified. Of 874 patients with sepsis, propensity score matching found 32 corresponding patients with sepsis but without candida infection but similar risk factors for developing candidemia. Attributable costs of candidemia were 7713.79 Euro (cost increase, 19.4%). Initiation of antifungal therapy was reduced from 67.5 (52.4, 90) hours in the group, where candida infection was determined by blood culture, to 31.0 (28.0, 37.5; P < .01) hours after detection by multiplex PCR. CONCLUSIONS: Candidemia increases costs of care in patients with septic shock. Polymerase chain reaction-based pathogen detection significantly reduces the time to initiation of antifungal therapy. This might impact on the clinical course of the disease but need to be confirmed in further trials.