Papers by Thierry Verplancke
Critical Care, 2007
Introduction Tacrolimus is an important immunosuppressive drug for organ transplantation patients... more Introduction Tacrolimus is an important immunosuppressive drug for organ transplantation patients. It has a narrow therapeutic range, toxic side effects, and a blood concentration with wide intra- and interindividual variability. Hence, it is of the utmost importance to monitor tacrolimus blood concentration, thereby ensuring clinical effect and avoiding toxic side effects. Prediction models for tacrolimus blood concentration can improve clinical care by optimizing monitoring of these concentrations, especially in the initial phase after transplantation during intensive care unit (ICU) stay. This is the first study in the ICU in which support vector machines, as a new data modeling technique, are investigated and tested in their prediction capabilities of tacrolimus blood concentration. Linear support vector regression (SVR) and nonlinear radial basis function (RBF) SVR are compared with multiple linear regression (MLR). Methods Tacrolimus blood concentrations, together with 35 other relevant variables from 50 liver transplantation patients, were extracted from our ICU database. This resulted in a dataset of 457 blood samples, on average between 9 and 10 samples per patient, finally resulting in a database of more than 16,000 data values. Nonlinear RBF SVR, linear SVR, and MLR were performed after selection of clinically relevant input variables and model parameters. Differences between observed and predicted tacrolimus blood concentrations were calculated. Prediction accuracy of the three methods was compared after fivefold cross-validation (Friedman test and Wilcoxon signed rank analysis). Results Linear SVR and nonlinear RBF SVR had mean absolute differences between observed and predicted tacrolimus blood concentrations of 2.31 ng/ml (standard deviation [SD] 2.47) and 2.38 ng/ml (SD 2.49), respectively. MLR had a mean absolute difference of 2.73 ng/ml (SD 3.79). The difference between linear SVR and MLR was statistically significant (p Conclusion Prediction of tacrolimus blood concentration with linear and nonlinear SVR was excellent, and accuracy was superior in comparison with an MLR model.
Resuscitation, 2008
Introduction: Good quality basic life support (BLS) results in better survival. BLS is a core com... more Introduction: Good quality basic life support (BLS) results in better survival. BLS is a core competence of nurses but despite regular refresher training, the quality of BLS is often poor and the reasons for this are not well known. We therefore investigated the relation between BLS quality and some of its potential determinants. Materials and methods: During a BLS refresher course, 296 nurses from non-critical care wards completed a questionnaire including demographic data and a ''self-confidence'' score. Subsequently, they performed a BLS test on a manikin connected to a PC using Skillreporting System software (Laerdal, Norway). The following variables were recorded: number of ventilations/min, tidal volume, number of compressions/min, compression rate, compression depth, ''good ventilation'' (n ≥ 4 min −1 and tidal volume = 700-1000 ml) and ''good compression'' (n ≥ 40 min −1 and rate = 80-120 min −1 and compression depth = 40-50 mm). To detect independent determinants of BLS quality, associations between the demographic data and the objective variables of BLS quality were examined. Results: Forty-three percent of the nurses rated their confidence as good or very good. Male gender was associated with good compression (P < 0.001). Greater self-confidence was also associated with good ventilation (P < 0.03) and with good compression (P < 0.001). A short time since last BLS training was associated with a higher number of ventilations/min (P = 0.01). A short time since last experience of CPR was associated with a higher number of compressions (P < 0.01). Conclusions: Male gender, greater self-confidence, recent BLS training and recent CPR were associated with better quality of BLS.
Critical Care, 2008
all 10 cities including the rural areas of the province of Kerman. All data were finally analyzed... more all 10 cities including the rural areas of the province of Kerman. All data were finally analyzed by SPSS software (version 11.5). Results On the basis of recorded statistical analysis, the mortality cases of human rabies in the province of Kerman during one decade was 10 persons (eight males and two females). One-half of them (50%) were bitten by dogs and the others (50%) by foxes. Among the reported deaths, 40% were from Kahnooj county (Jiroft region). The reported data indicated that 21,546 persons were bitten by animals during 10 years in the province of Kerman. The mean of age of the people who were bitten by dogs was 24.80 years (SD = ±14.6), while the mean age of the people who were bitten by foxes was 57.25 years (SD = ±1.50). There was a significant difference between the mean age of these two groups of the people (P < 0.05). The most frequent rate of injured people was reported in the age group 10-19 years old and the frequency rate of males (76.00%) was more than females (24.00%). Therefore, there was a statistically significant difference between males and females in this study (P < 0.01). About 60% of all persons that were bitten by animals were from rural areas and 40% of them were from urban areas (P < 0.05). Among the people who were bitten and injured by animals during one decade in the province of Kerman, 85.70% of them were not treated by the rabies prophylaxis treatment regimen. Among all of them who were bitten by animals, 50% were injured through hands and feet, 40%
Critical Care, 2009
There is considerable uncertainty about the reproducibility of the various instruments used to me... more There is considerable uncertainty about the reproducibility of the various instruments used to measure dyspnea, their ability to reflect changes in symptoms, whether they accurately reflect the patient's experience and if its evolution is similar between acute heart failure syndrome patients and nonacute heart failure syndrome patients. URGENT was a prospective multicenter trial designed to address these issues. Methods Patients were interviewed within 1 hour of first physician evaluation, in the emergency department or acute care setting, with dyspnea assessed by the patient using both a five-point Likert scale and a 10-point visual analog scale (VAS) in the sitting (60º) and then supine (20º) position if dyspnea had not been considered severe or very severe by the sitting versus decubitus dyspnea measurement. Results Very good agreements were found between the five-point Likert and VAS at baseline (0.891, P <0.0001) and between changes (from baseline to hour 6) in the five-point Likert and in VAS (0.800, P <0.0001) in acute heart failure (AHF) patients. Lower agreements were found when changes from baseline to H6 measured by Likert or VAS were compared with the seven-point comparative Likert (0.512 and 0.500 respectively) in AHF patients. The worse the dyspnea at admission, the greater the amplitude of improvement in the first 6 hours; this relationship is stronger when dyspnea is measured with VAS (Spearman's rho coefficient = 0.672) than with the five-point Likert (0.272) (both P <0.0001) in AHF patients. By the five-point Likert, only nine patients (3% (1% to 5%)) reported an improvement in their dyspnea, 177 (51% (46% to 57%)) had no change, and 159 (46% (41% to 52%)) reported worse dyspnea supine compared with sitting up in AHF patients. The PDA test with VAS was markedly different between AHF and non-AHF patients. Conclusions Both clinical tools five-point Likert and VAS showed very good agreement at baseline and between changes from baseline to tests performed 6 hours later in AHF patients. The PDA test with VAS was markedly different between AHF and non-AHF patients. Dyspnea is improved within 6 hours in more than threequarters of the patients regardless of the tool used to measure the change in dyspnea. The greater the dyspnea at admission, the greater the amplitude of improvement in the first 6 hours.
BMC Medical Informatics and Decision Making, 2010
Background: Echo-state networks (ESN) are part of a group of reservoir computing methods and are ... more Background: Echo-state networks (ESN) are part of a group of reservoir computing methods and are basically a form of recurrent artificial neural networks (ANN). These methods can perform classification tasks on time series data. The recurrent ANN of an echo-state network has an 'echo-state' characteristic. This 'echo-state' functions as a fading memory: samples that have been introduced into the network in a further past, are faded away. The echostate approach for the training of recurrent neural networks was first described by Jaeger H. et al. In clinical medicine, until this moment, no original research articles have been published to examine the use of echo-state networks.
BMC Medical Informatics and Decision Making, 2008
Background Several models for mortality prediction have been constructed for critically ill patie... more Background Several models for mortality prediction have been constructed for critically ill patients with haematological malignancies in recent years. These models have proven to be equally or more accurate in predicting hospital mortality in patients with haematological malignancies than ICU severity of illness scores such as the APACHE II or SAPS II [1]. The objective of this study is to compare the accuracy of predicting hospital mortality in patients with haematological malignancies admitted to the ICU between models based on multiple logistic regression (MLR) and support vector machine (SVM) based models. Methods 352 patients with haematological malignancies admitted to the ICU between 1997 and 2006 for a life-threatening complication were included. 252 patient records were used for training of the models and 100 were used for validation. In a first model 12 input variables were included for comparison between MLR and SVM. In a second more complex model 17 input variables were used. MLR and SVM analysis were performed independently from each other. Discrimination was evaluated using the area under the receiver operating characteristic (ROC) curves (± SE). Results The area under ROC curve for the MLR and SVM in the validation data set were 0.768 (± 0.04) vs. 0.802 (± 0.04) in the first model (p = 0.19) and 0.781 (± 0.05) vs. 0.808 (± 0.04) in the second more complex model (p = 0.44). SVM needed only 4 variables to make its prediction in both models, whereas MLR needed 7 and 8 variables in the first and second model respectively. Conclusion The discriminative power of both the MLR and SVM models was good. No statistically significant differences were found in discriminative power between MLR and SVM for prediction of hospital mortality in critically ill patients with haematological malignancies.
Critical Care, 2007
Hydrogen sulfide is produced endogenously by a variety of enzymes involved in cysteine metabolism... more Hydrogen sulfide is produced endogenously by a variety of enzymes involved in cysteine metabolism. Clinical data indicate that endogenous levels of hydrogen sulfide are diminished in various forms of cardiovascular diseases. The aim of the current study was to investigate the effects of hydrogen sulfide supplementation on cardiac function during reperfusion in a clinically relevant experimental model of cardiopulmonary bypass. Twelve anesthetized dogs underwent hypothermic cardiopulmonary bypass. After 60 minutes of hypothermic cardiac arrest, reperfusion was started after application of either saline vehicle (control, n = 6), or the sodium sulfide infusion (1 mg/kg/hour, n = 6). Biventricular hemodynamic variables were measured by combined pressure-volume-conductance catheters. Coronary and pulmonary blood flow, vasodilator responses to acetylcholine and sodiumnitroprusside and pulmonary function were also determined. Administration of sodium sulfide led to a significantly better recovery of left and right ventricular systolic function (P < 0.05) after 60 minutes of reperfusion. Coronary blood flow was also significantly higher in the sodium sulfide-treated group (P < 0.05). Sodium sulfide treatment improved coronary blood flow, and preserved the acetylcholine-induced increases in coronary and pulmonary blood (P < 0.05). Myocardial ATP levels were markedly improved in the sulfide-treated group. Thus, supplementation of sulfide improves the recovery of myocardial and endothelial function and energetic status after hypothermic cardiac arrest during cardiopulmonary bypass. These beneficial effects occurred without any detectable adverse hemodynamic or cardiovascular effects of sulfide at the dose used in the current study.
Uploads
Papers by Thierry Verplancke