Search

Introduction Les donnees collectees pour une etude clinique et non utilisees, par la suite, pour les analyses statistiques peuvent representer une part importante de l’ensemble des donnees collectees. Le volume de donnees a recueillir dans les « Case Report Form » (CRFs) a un impact sur le temps passe par chacun des acteurs de la recherche dans le cadre de leurs missions respectives. L’objectif de cette etude etait d’estimer le temps passe pour developper, saisir et monitorer une page de CRF et d’evaluer les couts associes. Methodes Une enquete a ete menee en mai 2016 dans deux hopitaux (CHU de Nantes et CHD de Vendee) aupres de 213 professionnels de la recherche (chefs de projet, coordinateurs d’etudes cliniques, techniciens d’etudes cliniques, « data-managers », attaches de recherche clinique) impliques dans les differentes etapes de la gestion d’un CRF. Un auto-questionnaire a ete propose aux professionnels afin d’estimer le temps passe pour chaque etape : creation de la maquette du CRF, saisie des donnees, « monitoring » et « data management ». Un exemple de page de CRF leur a ete soumis pour realiser ces estimations ; le choix s’est porte sur une page de bilan biologique contenant 25 parametres biologiques a saisir. Cet exemple a ete choisi pour differentes raisons : les bilans biologiques sont utilises pour le suivi des patients dans la plupart des pathologies etudiees ; il s’agit de donnees quantitatives qui posent souvent des problemes de conversion et pour lesquelles l’estimation du temps passe pour chaque etape est generalisable pour tous les patients ; ces bilans sont souvent repetes a differentes visites de suivi. Resultats Le taux de participation a l’enquete a ete de 53 %. Les temps medians ont ete estimes pour chaque etape et un outil a ete cree afin de convertir le temps passe en couts de personnels, en fonction du nombre de patients inclus dans l’etude, du nombre de visites de suivi prevues et du pourcentage de donnees monitorees. L’extrapolation a une etude sur 100 patients, avec un monitoring a 100 % et pour cinq visites de suivi avec bilan biologique donne un cout de 15 € pour la creation de la maquette, 1069 € pour la gestion des donnees, 3267 € pour la saisie des donnees et 2389 € pour le « monitoring » (ces couts concernent uniquement les pages de bilans biologiques). La meme etude pour 10 visites de suivi avec bilan biologique, voit son cout global multiplie par deux. Conclusion L’enquete a permis de quantifier le temps passe et les couts engendres par la gestion de visites de suivi avec bilan biologique. Lorsque ces donnees ne sont pas analysees par la suite cela entraine des couts superflus, principalement pour la saisie et le « monitoring » de ces donnees. Par ailleurs, la qualite des donnees pourrait en etre affectee. En effet, le temps passe au traitement de ces donnees non analysees peut l’etre au detriment du suivi de la qualite des donnees indispensables aux objectifs initiaux de l’etude (criteres principal et secondaires) et a l’evaluation de la securite des patients (pharmacovigilance). L’outil de calcul qui a ete cree permet de communiquer aupres de l’ensemble des acteurs sur l’importance d’une reflexion approfondie sur les donnees a inclure dans un CRF.

Issues regarding recommendations on empiric antimicrobial therapy for ventilator-associated pneumonia (VAP) have emerged in specific populations.To develop and validate a score to guide empiric therapy in brain-injured patients with VAP, we prospectively followed a cohort of 379 brain-injured patients in five intensive care units. The score was externally validated in an independent cohort of 252 brain-injured patients and its extrapolation was tested in 221 burn patients.The multivariate analysis for predicting resistance (incidence 16.4%) showed two independent factors: preceding antimicrobial therapy ≥48 h (pLimited-spectrum empirical antimicrobial therapy has low risk of failure in brain-injured patients presenting with VAP before day 10 and when prior antimicrobial therapy lasts

Introduction: According to the French health authorities’ guidelines relative to depression and anxiety disorder treatments, six Selective Serotonin Reuptake Inhibitors are available for prescription as a first-line of treatment. The guidelines suggest equivalence between these treatment options, but studies diverge regarding efficacy and safety profiles. Moreover, conditions in clinical trials are strictly controlled and do not truly reflect real life utilization. The objective of this study was to evaluate differences in efficacy and/or safety between these six selective serotonin reuptake inhibitors in real conditions of use. Methods: Efficacy and safety were evaluated using a regional database of the French national health insurance. Patients who received a selective serotonin reuptake inhibitor for a new depressive disorder and who were compliant to the treatment for a period of at least 6 months were included. Events indicative of a lack of efficacy and/or safety during the 12-month follow-up period were identified in the database (i.e., a dose increase, a switch to another antidepressant drug or an association with another antidepressant drug). A Cox model was used to compare the frequency and the delay to onset of each type of indicative event for each selective serotonin reuptake inhibitor. Results: Out of 3542 patients included, 1081 (30.5%) experienced an indicative event. The Cox model showed differences in terms of efficacy and safety. Patients treated with paroxetine, sertraline or citalopram as a first antidepressant were more likely to present a therapeutic failure than those treated by escitalopram or fluoxetine. Conclusion: A Cox model identified differences between selective serotonin reuptake inhibitors in terms of efficacy and/or safety profile. Our study positioned Escitalopram as the most efficient and/or safe treatment option. This study strategy can viably be used to evaluate the real life usage and effects of other drugs, an essential part of post approval evaluation.

International audience; PURPOSE: Stress cardiomyopathy is a common life-threatening complication after aneurysmal subarachnoid hemorrhage (SAH). We hypothesized that left ventricular (LV) longitudinal strain alterations assessed with speckle tracking could identify early systolic function impairment. METHODS: This was an observational single-center prospective pilot controlled study conducted in a neuro-intensive care unit. Forty-six patients with severe SAH with a World Federation of Neurological Surgeons grade (WFNS) ≥III were included. Transthoracic echocardiography (TTE) was performed on day 1, day 3, and day 7 after the patient's admission. A cardiologist blinded to the patient's management analyzed the LV global longitudinal strain (GLS). The control group comprised normal subjects matched according to gender and age. RESULTS: On day 1 median (25th-75th percentile) GLS was clearly impaired in SAH patients compared to controls [-16.7 (-18.7/-13.7) % versus -20 (-22/-19) %, p \textless 0.0001], whereas LVEF was preserved [65 (59-70) %]. GLS was severely impaired in patients with a WFNS score of V versus III-IV [-15.6 (-16.9/-12.3) % versus -17.8 (-20.6/-15.8) %, p = 0.008]. Seventeen (37 %) patients had a severe GLS alteration (\textgreater-16 %). In these patients, GLS improved from day 1 [-12.4 (-14.8/-10.9) %] to last evaluation [-16.2 (-19/-14.6) %, p = 0.0007] in agreement with the natural evolution of stress cardiomyopathy. CONCLUSIONS: On the basis of LV GLS assessment, we demonstrated for the first time that myocardial alteration compatible with a stress cardiomyopathy is detectable in up to 37 % of patients with severe SAH while LVEF is preserved. GLS could be used for sensitive detection of stress cardiomyopathy. This is critical because cardiac impairment remains a major cause of morbidity and mortality after SAH

Background Alcohol is the leading psychoactive substance consumed in France, with about 15 million regular consumers. The National institute on Alcohol Abuse and Alcoholism (NIAAA) considers alcohol abuse to be more than 14 units of alcohol a week for men and 7 units for women. The specific complication of alcoholism is the alcohol withdrawal syndrome. Its incidence reaches up to 30 % and its main complications are delirium tremens, restlessness, extended hospital stay, higher morbidity, and psychiatric and cognitive impairment. Without appropriate treatment, delirium tremens can lead to death in up to 50 % of patients. Methods/design This prospective, double-blind, randomised controlled study versus placebo will be conducted in twelve French intensive care units (ICU). Patients with an alcohol intake level higher than the NIAAA threshold, who are under mechanical ventilation, will be included. The primary objective is to determine whether baclofen is more efficient than placebo in preventing restlessness-related side effects in the ICU. Secondary outcomes include mechanical ventilation duration, length of ICU stay, and cumulative doses of sedatives and painkillers received within 28 days of ICU admission. Restlessness-related side effects in the ICU are defined as unplanned extubation, medical disposal removal (such as urinary catheter, venous or arterial line or surgical drain), falling out of bed, ICU runaway (leaving ICU without physician's approval), immobilisation device removal, self-aggression or aggression towards medical staff. Daily doses of baclofen/placebo will be guided by daily creatinine clearance assessment. Discussion Restlessness in alcoholic patients is a life-threatening issue in ICUs. BACLOREA is a randomised study assessing the capacity of baclofen to prevent agitation in mechanically ventilated patients. Enrolment of 314 patients will begin in June 2016 and is expected to end in October 2018. Trial registration ClinicalTrials.gov Identifier: NCT02723383, registered on 3 March 2016. Electronic supplementary material The online version of this article (doi:10.1186/s13063-016-1539-2) contains supplementary material, which is available to authorized users.

International audience; BACKGROUND: The use of analgesics can lead to cases of drug abuse and dependence. It can also cause pseudo-addiction in patients suffering from pain. What is the actual situation in patients suffering from severe sickle-cell disease, exposed to acute pain during vaso-occlusive crises? Evaluation of the use of analgesics, on the basis of Diagnostic and Statistical Manual of Mental Disorders criteria for substance abuse and dependence, makes it possible to differentiate the symptoms occurring only in a context of pain, in the aim of managing the pain, and thus describing pseudo-addiction, from symptoms also occurring when there is no pain, and more in favour of true addiction. Currently there is no data available in France on this problem, and no studies have been carried out in children or adolescents with sickle-cell disease. The purpose of the study is to evaluate the prevalence of problematic use of equimolar mixture of oxygen and nitrous oxide and other analgesic drugs in a population of subjects with severe sickle-cell disease in France. METHODS/DESIGN: PHEDRE (Pharmacodépendance Et DREpanocytose-drug dependence and sickle-cell disease) is an observational, descriptive and transversal study. Patients under the age of 26 with sickle-cell disease are included in the study by the doctors looking after them in sickle-cell disease centres. The patients are then contacted by a trained researcher for a telephone interview, including an evaluation of the Diagnostic and Statistical Manual of Mental Disorders criteria for abuse and dependence to equimolar mixture of oxygen and nitrous oxide and for each of the analgesic drugs taken by the patient. The data are also completed using the subject's medical record. DISCUSSION: This study will make it possible to provide an initial quantitative and qualitative evaluation of problematic use of equimolar mixture of oxygen and nitrous oxide and analgesic drugs in the sickle-cell disease population. The results will be used firstly to provide additional data essential for monitoring the risk of overdose, abuse, dependence and misuse of these products, and to begin awareness-raising and to provide information for health care professionals, in order to significantly improve the management of sickle-cell disease-related pain. TRIAL REGISTRATION: Clinical Trials.gov ID: NCT02580565 registered 16 October 2015 Unique Protocol ID: RC14\₀344.

Summary Introduction Delivering practices of opioid substitution treatment (OST) in community pharmacies are heterogeneous. This study aims at drawing up an inventory and at characterizing different populations of pharmacists, regarding their practices and perceptions. Methods We distributed a questionnaire to pharmacists, which was divided into two parts: socio-demographic questions, and 49 binary questions collecting pharmacists’ perceptions. Statistical analyses were performed using SAS 9.3 and SPAD7 software. The categorical variables were expressed as numbers and percentages. We characterized the different profiles of pharmacists using a multivariate analysis method. Results We analyzed 303 questionnaires; 60.8% of our cross-section of pharmacists consider that OSTs are a treatment for chronic disease, a treatment in which they consider that they play a key role. Regarding OSTs, 89.3% felt “comfortable” with current regulations, while 77.3% have mastered possible interactions with other drugs. 16.8% of pharmacists equate the act of delivering OSTs as a “legal narcotic deal”. Regarding the patients, 49.2% of pharmacists consider them as “different”, and 39.1% fear that they cause relationship problems. Most pharmacists perceive the role they are meant to play, although a lack of training and the need for recommendations have been highlighted. PCA allowed us to characterize 4 different profiles of pharmacists. Conclusion Continuous training adapted to the different profiles we drew up will be offered in order to allow pharmacists to deepen their knowledge about drug dependence and related care support.

We read with interest the paper by Launey and colleagues [1] regarding the effect of fever control on the incidence of ventilator-associated pneumonia (VAP) in brain-injured patients. We commend the authors for addressing this relationship, given that fever and nosocomial infections are frequent in these patients. However, we have concerns regarding the interpretation of data. Firstly, independent risk factors for VAP, namely disease severity and lung contusion [2,3], are unevenly distributed between groups. In addition, other risk factors, such as chronic obstructive pulmonary disease (COPD) and ICU readmission [2,4], were not taken into account, suggesting any effect of fever control is likely subject to confounding. Secondly, mortality in the control group was higher (34% versus 23%), lowering the observation period if death occurred within 28 days. Thus, patients in the fever-control group were observed for a longer period of time, increasing their time at risk for VAP. Thirdly, VAP incidence was higher in patients who were subject to fever control for longer than 3 days. However, duration of fever control was likely determined by factors also affecting the patients’ risk of acquiring VAP; that is, death. Not surprisingly, those who died within 3 or fewer days of initiating fever control did not develop VAP as frequently as those who did not. In summary, the intervention group and historical control group in this study do not seem to be optimally matched on crucial parameters. Additionally, decreased mortality and longer follow-up in the intervention group likely resulted in uncontrolled lead time bias/attrition bias. Decreased mortality also seems contradictory to the claim that VAP, a deadly condition for ICU patients, is more frequent among fever-control patients.

Objectifs En France, le reseau des Centres d’evaluation et d’information sur la pharmacodependance-addictovigilance” (CEIP-A) est responsable de la collecte et de l’evaluation des notifications rapportant une dependance a des substances psychoactives medicamenteuses ou non. Ces notifications sont evaluees a l’aide d’un score qui permet d’evaluer le potentiel de pharmacodependance des substances. Deux medicaments sous surveillance par ce reseau sont la zopiclone et le zolpidem, deux apparentes aux benzodiazepines indiques dans le traitement de l’insomnie occasionnelle ou transitoire. Le but de cette etude est d’evaluer le profil de dependance de ces deux medicaments. Methodes La base de donnees utilisee est celle du CEIP-A de Nantes. L’analyse a ete realisee sur les notifications recues entre le 1er janvier 2008 et le 31 mars 2015. Ces notifications ont ete evaluees a l’aide d’un score valide et compose de 8 criteres binaires repartis en deux dimensions. La premiere evalue les signes physiques et compulsifs de la dependance avec 4 criteres : la tolerance, les symptomes de sevrage, des doses ou duree superieures a ce qui etait initialement prevu et le desir ou la tentative d’arret de la consommation. La deuxieme dimension evalue les consequences dommageables de la dependance avec 4 criteres : temps passe pour obtenir le produit ou le consommer, soucis personnels ou professionnels lies a la consommation, la poursuite de la consommation malgre des problemes de sante et un comportement de transgression. Ces criteres permettent d’obtenir un score global sur 8 et deux sous-scores sur 4. Un outil informatique appele PROMESS permet d’evaluer l’ensemble des scores et aussi de comparer des substances a l’aide d’analyses quantitatives, qualitatives et d’homogeneite. Resultats Quarante notifications ont ete analysees pour le zolpidem et 21 pour la zopiclone. L’analyse de la premiere dimension du score montre que 40 % des notifications pour le zolpidem ont un score superieur a 3 versus 29 % pour la zopiclone. L’analyse de la deuxieme dimension du score montre que 40 % des notifications pour le zolpidem ont un score superieur a 3 versus 19 % pour la zopiclone. La comparaison du score global met en evidence que 40 % des notifications de zolpidem ont un score superieur a 6 contre 19 % pour la zopiclone. Conclusion Ces resultats sont en faveur d’une plus importante prevalence de signes physiques et compulsifs de la dependance, ainsi que de consequences dommageables suite a un traitement par le zolpidem plutot que par la zopiclone. Ces resultats sont en accord avec les informations disponibles dans la litterature ainsi que ceux identifies par les autres outils epidemiologiques des CEIP-A. A notre connaissance, PROMESS est le premier outil qui permet de realiser rapidement une comparaison clinique de profils de pharmacodependance.

International audience; The high consumption of psychotropic drugs is a public health prob- lem. Rigorous statistical methods are needed to identify consumption characteristics in post-marketing phase. Agglomerative hierarchical clustering (AHC) and latent class analysis (LCA) can both provide clusters of subjects with similar characteristics. The objective of this study was to compare these two methods in pharmacoepidemiology, on several criteria: number of clusters, concordance, interpretation and stability over time. From a data set on bromazepam consump- tion, the two methods present a good concordance. AHC is a very stable method and provides homogeneous classes. LCA is an inferen- tial approach and seems to allow identifying more accurately extreme deviant behaviour.

International audience; IntroductionFever is associated with a poor outcome in severely brain-injured patients, and its control is one of the therapies used in this condition. But, fever suppression may promote infection, and severely brain-injured patients are frequently exposed to infectious diseases, particularly ventilator-associated pneumonia (VAP). Therefore, we designed a study to explore the role of a fever control protocol in VAP development during neuro-intensive care.MethodsAn observational study was performed on severely brain-injured patients hospitalized in a university ICU. The primary goal was to assess whether fever control was a risk factor for VAP in a prospective cohort in which a fever control protocol was applied and in a historical control group. Moreover, the density of VAP incidence was compared between the two groups. The statistical analysis was based on a competing risk model multivariate analysis.ResultsThe study included 189 brain-injured patients (intervention group, n¿=¿98, and historical control group, n¿=¿91). The use of a fever control protocol was an independent risk factor for VAP (hazard ratio 2.73, 95% confidence interval [1.38, 5.38; P¿=¿0.005]). There was a significant increase in the incidence of VAP in patients treated with a fever control protocol (26.1 versus 12.5 VAP cases per 1000 days of mechanical ventilation). In cases in which a fever control protocol was applied for >3 days, we observed a higher rate of VAP in comparison with the rate among patients treated for ¿3 days.ConclusionFever control in brain-injured patients was a major risk factor for VAP occurrence, particularly when applied for >3 days.

International audience; BACKGROUND: Benzodiazepines (BZD) and nonbenzodiazepines hypnotics (z-drugs) are recognized as one of the most widely prescribed medications in the world. OBJECTIVES: The purpose of the study was to assess the BZD and z-drugs dependence in young to middle-aged outpatients who were taking BZD/z-drugs on a chronic basis, and to characterize their profile. METHODS: This is a forward-looking cross-sectional epidemiological study. Data were collected through a semi-structured interview within a network of partner pharmacies from the Nantes area, in France. All data were obtained exclusively through patients' declarations. 212 patients (19-64 years old) were included: they were considered dependent when they answered positively to at least three items of the DSM IV. A multivariate logistic regression and a principal component analysis (PCA) were carried out to determine their profile. RESULTS: Almost half of the patients met criteria for BZD/z-drugs dependence. The risk to develop BZD/z-drugs dependence is significantly associated with psychiatric history and with the quantity of BZD/z-drugs that is taken. A two factor concept of dependence could be identified according to the PCA: one axis with items of "tolerance" and "long term administration or higher doses", and a second axis with "concerned by treatment" and "somatic consequences". Conclusions/Importance: Among this BZD/z-drug dependent population, the two axes identified in the PCA represent two profiles of dependence: being in positive conditioning or suffering from negative consequences. Clinicians need to know them: these two clinical profiles may have an influence in terms of decision-making, especially to manage discontinuation.

International audience; PURPOSE: Frailty is a recent concept used for evaluating elderly individuals. Our study determined the prevalence of frailty in intensive care unit (ICU) patients and its impact on the rate of mortality. METHODS: A multicenter, prospective, observational study performed in four ICUs in France included 196 patients aged ≥65 years hospitalized for >24 h during a 6-month study period. Frailty was determined using the frailty phenotype (FP) and the clinical frailty score (CFS). The patients were separated as follows: FP score

PCSK9 (proprotein convertase subtilisin kexin type 9) is a secreted protease that modulates cholesterol homeostasis by decreasing low-density lipoprotein receptor expression. Low levels of plasma lipoproteins are related to severity of illness and survival in patients of intensive care units (ICU).The aim of the study was to investigate the regulation of plasma PCSK9 and its association with plasma lipid parameters and clinical markers of severity during critical illness.The plasma biobank from the previously published HYPOLYTE prospective study was used to measure PCSK9 concentrations by ELISA at days 0 and 8 in 111 patients admitted to surgical ICU for severe multiple trauma. Patients were randomly assigned to hydrocortisone therapy or placebo.Plasma PCSK9 levels were increased by 2-fold between days 0 and 8 (231 ± 116 vs 481 ± 227 ng/ml; P = .0001). Hydrocortisone therapy did not alter PCSK9 concentrations (451 ± 216 vs 511 ± 239 ng/ml in placebo group; P = .33). PCSK9 was positively associated with low-density lipoprotein-cholesterol (Pearson coefficient, 0.26; P = .007) at day 0, but not at day 8. At day 8, an inverse correlation was found between PCSK9 and high-density lipoprotein-cholesterol (β = -653; P = .004). Although baseline PCSK9 concentrations were not associated to severity scores, PCSK9 values at day 8 were related to injury severity score (β = 6.17; P = .0007), length of stay in ICU (β = 6.14; P = .0001), and duration of both mechanical ventilation (β = 8.26; P = .0001) and norepinephrine infusion (β = 18.57; P = .015).Plasma PCSK9 appears as a late biomarker of illness severity in patients with severe multiple trauma.

Introduction Un tiers des patients admis en reanimation necessite une aide respiratoire avec une ventilation mecanique invasive (VM). Une duree de VM trop importante est associee a une augmentation de la morbi-mortalite. En recherche clinique, la duree de VM est souvent utilisee comme critere de jugement principal. Cependant un tiers des patients ventiles vont deceder, il a donc ete propose un critere combinant le moment de l’arret de la VM (extubation) et le deces : le ventilator-free days (VFD). A j90 par exemple, ce score augmente de 1 point par jour sans VM a j90 post-intubation ; mais en cas de deces avant j90, le score est de zero. Plusieurs problemes sont associes a ce critere : sa definition varie entre les etudes, son interpretation est cliniquement non-intuitive, et sa distribution non-symetrique rendant l’analyse statistique difficile. Nous proposons une methode d’analyse des donnees de VM avec un modele de melange parametrique de type illness-death. Methodes Pour valider l’approche illness-death et comparer ses performances a celles de tests statistiques bases sur le VFD (test de Student et test de Wilcoxon), nous avons simule des donnees selon plusieurs scenarios, avec ou sans difference d’effet entre les deux interventions. Nous avons aussi re-analyse les donnees d’une etude interventionnelle avant-apres (Roquily et al. AJRCCM, 2013). Cette etude evaluait l’efficacite de la mise en place d’un bundle de mesures pour favoriser l’extubation par rapport a une strategie courante de soins. Resultats En comparaison avec les tests bases sur le VFD, les simulations ont montre de bonnes performances du modele illness-death, en particulier en termes de risque de seconde espece. L’analyse de l’essai clinique a aussi permis de montrer l’interet du modele illness-death permettant une interpretation plus aisee des resultats. Conclusion Alors que le VFD represente aujourd’hui un indicateur incontournable pour l’analyse d’une etude clinique en reanimation, il apparait essentiel de changer ces habitudes et d’utiliser un modele illness-death, en particulier pour augmenter la puissance de ces etudes et obtenir des resultats non biaises et interpretables.

Objective. Pro-protein convertase subtilisin/kexin type 9 (PCSK9) is a circulating protein that impairs LDL clearance by promoting the LDL receptor (LDLR) degradation. Plasma lipid parameters (LDL-C and HDL-C) are related to severity of illness and survival in intensive care patients. Here, we aimed to determine whether circulating PCSK9 concentrations were correlated to the severity of illness in patients with multiple trauma. Methods/Results. Plasma PCSK9 were measured, at day 0 and 8 after admission, by ELISA in 111 patients hospitalized in ICU for severe trauma. Patients were included in the HIPOLYTE study and were randomly assigned to hydrocortisone therapy or placebo. Plasma PCSK9 levels were significantly increased by 161% at day 8 compared to day 0 (481 ± 227 vs 231 ± 117 ng/ml, P=0.0001). Hydrocortisone therapy did not alter PCSK9 concentrations at day 8 compared to placebo (451 ± 216 vs 511 ± 239ng/ml, P=0.33). In the whole population of the study, PCSK9 was positively associated with LDL-C (Pearson coefficient : 0.26, p=0.007) at day 0, but not with markers of severity illness. At day 8, an inverse correlation was found between PCSK9 and HDL-C (β =-653 ; P=0.004). Interestingly, PCSK9 concentrations at day 8 were related to markers of severity illness as injury severity score (β =6.17 ; P=0.0007), length of stay in ICU (β =6.14 ; P=0.0001), duration of both mechanical ventilation (β =8.26 ; P=0.0001) and cathecolamines infusion (β =18.57; P=0.015). Conclusion. PCSK9 appears as a late biomarker of severity illness in patients hospitalized for multiple trauma in ICU. These results open new perspectives for a role of PCSK9 in critical illness.

France has one of the highest recorded rates of psychotropic use of drugs compared with other European countries, especially for anxiolytics, hypnotics and antidepressants. The aim of this study was to characterize the use of three psychotropic drugs among the most prescribed in France (bromazepam, paroxetine, zolpidem) using reimbursement databases in real-life conditions. Individuals from a region affiliated to the French General Health Insurance Scheme, who had received at least two dispensings of bromazepam, paroxetine or zolpidem reimbursed between 1 January and 30 June 2008, were included. We used a latent class analysis to identify different subgroups of users for these three psychotropic drugs. A total of 40,644 patients were included for bromazepam, 36,264 for zolpidem and 31,235 for paroxetine. Using latent class analysis, four clinical subtypes of users of bromazepam and zolpidem were identified: nonproblematic users, at-risk users, users with a probable mental disorder and compulsive users. Three subgroups were identified for paroxetine that differed rather by the prescription patterns. Users of anxiolytics and hypnotics with at-risk behaviours represented a significant proportion in the studied population. This original method could be extended to other prescription databases to identify populations at risk of abuse or dependence to psychotropic drugs.

Introduction En France, le reseau des Centres d’evaluation et d’information sur la pharmacodependance (CEIP) a pour mission d’evaluer le potentiel de dependance des medicaments et substances. Ils disposent pour cela d’un outil permettant de recueillir les notifications des cas d’abus et de dependance emanant des professionnels de sante. Un score, renseigne pour chaque cas de notification, a ete valide lors d’une etude anterieure. Il permet de definir un profil de gravite de la dependance lie a l’utilisation de la substance etudiee. Ce score de gravite repose sur les sept items de la definition de la dependance du DSM IV auquel a ete ajoute un item evaluant le mesusage. Objectif Proposer, a partir du score de gravite, des indices statistiques adaptes et des representations graphiques originales permettant de comparer les profils de dependance de differentes substances et de distinguer differents profils de consommateurs pour une meme substance. Methodes Le score de gravite est compose de huit items binaires (positif/negatif), repartis en deux dimensions : « signes physiques et compulsifs de la dependance » et « consequences dommageables de la dependance ». Chaque item est exprime en taux de positivite. Un graphique en 3D permet d’illustrer les profils de dependance : les axes X et Y representent chacun une dimension du score, l’axe Z represente le taux de notifications. Ces informations sont egalement representees sur une surface, le taux de notifications etant alors defini selon une echelle de couleur. Afin de determiner l’homogeneite d’un profil de dependance, une courbe est definie a partir des differences entre les taux de notifications observes et les valeurs moyennes de ces taux (methodologie CUSUM, « Cumulativ Sum »). La valeur de l’aire sous la courbe CUSUM permet de classer un profil de dependance par rapport aux profils d’autres substances, en fonction de leur homogeneite. Une classification ascendante hierarchique est appliquee a partir des deux dimensions du score. Cette methode permet de mettre en evidence l’existence, ou non, de populations avec des caracteristiques distinctes concernant la dependance a une meme substance. Resultats Actuellement, la base de donnees du CEIP de Nantes contient 1276 notifications avec un score complet (huit items remplis). Un outil informatique a ete developpe afin d’automatiser et mettre a jour regulierement les rapports de resultats. Une comparaison visuelle est realisee a partir des representations graphiques 3D et surface. A titre d’exemple, pour la molecule zopiclone, on observe une population globale peu homogene (AUC = 830). Deux populations differentes peuvent etre definies : (1) les notifications pour lesquelles les items de la dimension « consequences dommageables » ont un taux de positivite eleve (100 % pour 3 des 4 items), (2) les notifications pour lesquelles le taux de positivite des items est en moyenne moins eleve. Comparee a d’autres substances psychoactives, la zopiclone semble etre une molecule avec une pharmacodependance moyenne. Conclusion La creation de cet outil novateur supplementaire du reseau des CEIP permet d’apporter une reponse rapide en ce qui concerne le profil de dependance de substances d’interet ou de substances emergentes. Ceci permettrait d’aider a la decision en sante publique lors des evaluations des substances et de reperer les alertes en comparant les profils des medicaments apparaissant dans la base au profil de substances deja connues.

Introduction La France est dotee d’un reseau de 13 Centres d’evaluation et d’information sur la pharmacodependance-addictovigilance (CEIP-A) coordonne par l’Agence nationale de securite des medicaments et des produits de sante. Une des missions des CEIP-A : recueillir des informations cliniques sur les cas de pharmacodependance notifies spontanement par les professionnels de sante. Les patients peuvent utiliser une ou plusieurs substances qui sont chacune evaluee a l’aide du score de gravite de pharmacodependance. Ce score de huit items est base sur ceux de la definition de la pharmacodependance du DSM IV, excepte pour la notion de comportement de transgression : – quatre items evaluent les signes physiques et compulsifs de la dependance : tolerance, sevrage, dose ou duree superieures a celles prevues, desir d’arret ; – quatre items evaluent les consequences dommageables de la dependance : temps passe, problemes relationnels, persistance de la consommation malgre des problemes de sante, comportement de transgression. Methodes L’objectif de ce travail est de proposer de nouvelles representations graphiques (3D, courbes des sommes cumulees des differences a la moyenne…) et indices permettant des comparaisons faciles a interpreter entre substances de leur profil de pharmacodependance. Resultats Ces representations graphiques et indices permettent de caracteriser les profils de pharmacodependance des substances et de bien les discriminer. Ces representations permettent une analyse fine des profils de pharmacodependance des differentes substances. Conclusion Ce travail ouvre des perspectives pour la comparaison en termes de profil de pharmacodependance de l’ensemble des substances ayant ete notifiees au CEIP-A.

International audience; PURPOSE: Respiratory complications constitute an important determinant of length of stay in tetraplegic patients. In a population of tetraplegic patients, we investigated the factors involved in the duration of mechanical ventilation (MV) and whether the duration of MV was associated with the long-term neurologic status. MATERIAL AND METHODS: In a retrospective study in 3 intensive care units (ICUs) (January 2001 to December 2009), consecutive patients (≥18 years) hospitalized for acute (≤24 hours) traumatic tetraplegia were included in the study. Patients with severe brain injury or who died in the first 48 hours were excluded. The primary outcome was the duration of MV. The secondary outcomes were the American Spinal Injury Association (ASIA) motor score on ICU discharge and at 1 year. RESULTS: A total of 164 consecutive adult patients with tetraplegia were analyzed. Median (interquartile range) ASIA motor scores were 15 (6-26) on admission, 22 (9-40) on ICU discharge (n = 145 survivors), and 37 (10-80) at 1 year (n = 52 complete follow-up). The median duration of MV was 11 (2-26) days. In multivariate analysis, MV duration increased with pneumonia (P < .0001), atelectasis (P = .0042), and tracheotomy (P < .0001). In exploratory analysis, an increased duration of MV was the only factor associated in multivariate analysis with a low ASIA motor score on ICU discharge (P = .0201) and at 1 year (P = .0003). CONCLUSIONS: Prevention of pneumonia and atelectasis is critical for the reduction of MV in tetraplegic patients. Prolonged MV was independently associated with poor neurologic status.