Common use of Surgical procedures Clause in Contracts

Surgical procedures. The indication for surgery was assessed by the multi-disciplinary heart team and was consisted with the institutional MISSION! heart failure protocol.2 Surgical left ventricular restoration according to the technique described by ▇▇▇,3 CorCap (Acorn Cardiovascular Inc, St ▇▇▇▇, Minnesota) implantation1 and LVAD (HeartWare Inc, Framingham, Massachusetts) implantation16 were performed as previously described. All operations were performed using cardiopulmonary bypass, aortic cross-clamping and intermittent warm blood cardioplegia, except for the majority of LVAD patients. In those LVAD patients aortic cross-clamping was not necessary and implantation was performed on the beating heart with the use of cardiopulmonary bypass. Patients received an arterial line and a pulmonary artery 2 catheter for intra- and postoperative monitoring of blood pressure, cardiac output and pulmonary pressure. These data were used to calculate the cardiac index and systemic vascular resistance. Patients did not receive ACE inhibitors, ARBs and diuretics on the day of surgery. Continuous variables are expressed as mean ± standard deviation (SD) when normally distributed, or otherwise as median and interquartile range (IQR). Categorical variables are presented as numbers and percentages. Missing values for N-terminal fragment of pro-hormone of brain natriuretic peptide (NT-ProBNP) (N=66, 29%) and thyroxine (N=69, 31%) were replaced using multiple imputation (R package MICE, version 2.22), which was repeated a hundred times. Vasoplegic and non-vasoplegic patients were compared. Comparison of continuous data was performed using two-tailed unpaired Student t test for normally distributed variables or otherwise the ▇▇▇▇-▇▇▇▇▇▇▇ U test. The ▇▇▇▇▇▇ ▇▇▇▇▇ method was used to assess 90-day survival in vasoplegic and non-vasoplegic patients. The survival distributions were compared using the log-rank test. To explore the association of variables with the occurrence of vasoplegia, univariable logistic regression analysis was performed. Odds ratios (OR) with 95% confidence intervals (CI) were reported. Next, all variables were entered one by one in a multivariable logistic regression, to assess their independent association with vasoplegia after adjusting for clinically relevant variables (age, sex and surgical procedure). Furthermore, to assess whether thyroxine levels were influenced by amiodarone use and/or thyroid hormonal replacement, these were entered in a separate multivariable logistic regression analysis. Subsequently, Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression with leave-one-out cross-validation was used to identify and to calibrate the best risk prediction model.17 This is a state-of-the-art model fitting and variable selection methodology which can jointly select candidate predictors and optimise the resulting model for prediction. Age, sex and surgical procedure were forced into the model. The model fitting procedure was repeated for each of the hundred imputations. For the final model, a single model was selected from the imputations containing all variables which were used in >85% of the models. The imputation linked to this model was used to re-place the missing values for NT-ProBNP and thyroxine. The performance of the final model was assessed by computing the area under the receiver operating characteristic (ROC) curve. Next, patients were divided into 3 risk categories: low (predicted probability of <25%), intermediate (25-50%) and high risk (≥50%), after which the observed incidence per risk group was calculated. P-values <0.05 were considered statistically significant. Statistical analysis was performed using SPSS for Windows (version 20.0, Chicago, Illinois) and R (version 3.2.1, Vienna, Austria). The preliminary data of this study were presented at the American College of Cardiology 2015 scientific session.18 Between 2006 and 2015, 260 heart failure patients with a LVEF ≤35% underwent surgical left ventricular restoration, CorCap device implantation, or LVAD implantation. A total of 35 patients (22 left ventricular restoration and 13 CorCap patients) were excluded. Accordingly, the final study population consisted of 225 patients (166 (74%) men, age 62±10 years, LVEF 24±6%). Baseline data are summarised in Table 1. A total of 178 (79%) patients were admitted for elective surgery. The remaining 47 patients (21%) required surgery during an ongoing admission for heart failure. Surgical left ventricular restoration, CorCap implantation or LVAD implantation was performed in 121 (54%), 71 (32%) and 33 (15%) patients respectively. Concomitant cardiac procedures were coronary artery bypass grafting in 82 (36%), mitral valve surgery in 137 (61%), triscuspid valve surgery in 115 (51%) and aortic valve surgery in 15 (7%) patients. Mean cross clamp and cardiopulmonary bypass duration were 127±50 and 193±69 minutes, respectively.