Lung abscess following ventilator-associated pneumonia during COVID-19: a retrospective multicenter cohort study.

BACKGROUND: Patients undergoing mechanical ventilation (MV) for COVID-19 exhibit an increased risk of ventilator-associated pneumonia (VAP). The occurrence of lung abscesses following VAP in these patients has been poorly studied. We aimed to describe the incidence, characteristics, risk factors and prognosis of lung abscesses complicating VAP after COVID-19. METHODS: We conducted an observational, retrospective study in three French intensive care units. Patients admitted for acute respiratory failure with a confirmed SARS-CoV-2 PCR and requiring MV for more than 48 h were included. RESULTS: Among the 507 patients included, 326 (64%) had a documented VAP. Of these, 23 (7%) developed a lung abscess. Enterobacterales (15/23, 65%) were the main documentation, followed by non-fermenting Gram-negative bacilli (10/23, 43%) and Gram-positive cocci (8/23, 35%). Lung abscesses were mainly plurimicrobial (15/23, 65%). In multivariate analysis, a plurimicrobial 1st VAP episode (OR (95% CI) 2.93 (1.16-7.51); p = 0.02) and the use of hydrocortisone (OR (95% CI) 4.86 (1.95-12.1); p = 0.001) were associated with lung abscess development. Intensive care unit (ICU) mortality of patients with lung abscesses reached 52%, but was not significantly higher than for patients with VAP but no lung abscess. Patients with lung abscesses had reduced ventilator-free days at day 60, a longer duration of MV and ICU stay than patients with VAP but no lung abscess (respectively, 0 (0-3) vs. 16 (0-42) days; p < 0.001, 49 (32-73) vs. 25 (11-41) days; p < 0.001, 52 (36-77) vs. 28 (16-47) days; p < 0.001). CONCLUSIONS: Lung abscessing pneumonia is not uncommon among COVID-19 patients developing VAP. A plurimicrobial first VAP episode and the use of hydrocortisone are independently associated with this complication. In COVID-19 patients with persistent VAP, a chest CT scan investigating the evolution toward lung abscess should be considered.

Implementation of antibiotic stewardship programmes in French ICUs in 2018: a nationwide cross-sectional survey.

BACKGROUND: Antibiotic stewardship programmes have a pivotal role in ICUs, but the level of implementation of these programmes at the regional or national level is not well known. OBJECTIVES: The aim of our study was to assess the level of implementation of antibiotic stewardship programmes in French ICUs. METHODS: We conducted a nationwide cross-sectional survey from January to March 2018 using an online questionnaire sent as an E-mail link to ICU specialists (one questionnaire per ICU). RESULTS: Overall, 113 out of 206 (55%) ICUs participated. Access to local epidemiology regarding bacterial resistance and antibiotic consumption data was reported in 84% and 65% of ICUs, respectively. Local guidelines for antibiotic use were available in 54% of ICUs. The duration of empirical antibiotic therapy was limited in 46% of cases, following the recommendation of an external expert in 33%. An antibiotic stewardship programme leader was reported at the hospital level by 94% of respondents, being an infectious disease physician in 80%. His/her role in the ICU was mostly to discuss specific cases (50%) and to provide advice on antibiotic prescriptions (26%). Regarding microbiological diagnosis, blood cultures were not processed at night or during weekends in 57%. Molecular biology and MS techniques were available in 62% and 59% of cases, respectively. Therapeutic drug monitoring of ß-lactams was available in 46% of cases. Forty-three percent of respondents knew the expression 'antimicrobial/antibiotic stewardship'. CONCLUSIONS: Antibiotic stewardship programmes are not optimally implemented in French ICUs. Improvement efforts and regular monitoring of the level of implementation are needed.

Early infectious acute respiratory distress syndrome is characterized by activation and proliferation of alveolar T-cells.

Acute respiratory distress syndrome (ARDS) in humans is characterized by the infiltration of polymorphonuclears in the alveolar spaces. However, the role of T-cells in ARDS is unknown. Our aim was to characterize the T-cell phenotype in bronchoalveolar lavage (BAL) during the early phase of acute lung infection(ALI)/ARDS-infected patients in comparison to a control group (CG). BAL lymphocyte phenotypes of two ALI, 16 ARDS, and eight CG were examined by flow cytometry. ALI/ARDS showed a significant increase in CD4 and CD8 T-cell activation as compared to CG. Moreover, a significant level of proliferation was observed using the Ki67 marker in ARDS patients as compared to controls (median): 37 versus 6 % for CD4 T-cells (p = 0.022) and 34 versus 2 % for CD8 T-cells (p = 0.009). In contrast, the percentage of T-regulatory cells and apoptotic T-cells were similar in both groups. Among costimulatory molecules, we observed an overexpression of CTLA-4/CD152 on CD4 T-cells in ALI/ARDS as compared to CG: 30 versus 7 %, respectively (p = 0.063). In further characterizing T-cell subsets expressing high levels of CD152, we found the presence of IL-17 secreting CD4 T-cells in ALI/ARDS. In humans, ALI/ARDS due to infection is associated with a high level of T-cell activation and proliferation, along with the presence of Th17 cells, which are known to attract polymorphonuclears.

Severe community-acquired pneumonia and positive urinary antigen test for S. pneumoniae: amoxicillin is associated with a favourable outcome.

Positive urinary antigen tests (UAT) for pneumococcal infection in community-acquired pneumonia (CAP) may lead to targeted antibiotic therapy. We report an audit aimed at defining the link between mortality and targeted therapy. We conducted a retrospective multicentre audit of patients with severe CAP for whom a UAT was positive for S. pneumoniae. Patients admitted from January 2010 to December 2013 to 8 medical centres (from A to H) were included. Co-morbidities were defined by the specific treatment administered before hospital care, or if the diagnosis was newly established during the hospital stay. We used the Pneumonia Severity Index (PSI) to assess disease severity. Only patients with PSI > 90 were included. Antibiotic treatments and the PSI were extracted from patients' charts. Amoxicillin had to be prescribed as a targeted antibiotic treatment or at the time of antibiotic reassessment. A total of 389 patients were included. The mean (±STD) PSI score was 128 ± 29; 38.9% of the patients had a class 5 PSI score. Intensive care was required for 36.6% of the patients. Amoxicillin was initially prescribed in 47 cases (12.1%) and in 34 cases after reassessment (8.7%). In logistic regression analysis, we found three parameters associated with mortality: being hospitalised in institution D, class 5 PSI score, and metastatic cancer. In contrast, three antibiotic regimens were protective factors, including targeted therapy: OR = 0.09, p < 0.001. In the context of severe CAP with positive UAT for S. pneumoniae, targeted therapy was associated with a reduction in mortality.

Enhanced T-cell apoptosis in human septic shock is associated with alteration of the costimulatory pathway.

T-cell apoptosis during septic shock (SS) has been associated with deleterious outcome, but the mechanisms of apoptosis are not well understood. As T-cells are not infected in bacterial infection, our hypothesis was that deleterious interactions between lymphocytes and monocytes could be involved. This is a cross-sectional study of 27 patients presenting with community-acquired SS, 23 infected patients without SS and 18 controls. Cytofluorometric techniques were used to study apoptosis, the costimulatory pathway and cytokine synthesis. Apoptosis was increased in SS compared to infected patients without SS and controls: the median values were 18, 2 and 3%, respectively, for CD4(+) T-cells (P < 0.001), and 12, 5 and 2%, respectively, for CD8(+) T-cells (P < 0.001). Patients with SS exhibited significant CD152 over-expression on T-cells, while CD86 expression was decreased on monocytes (P = 0.004). The synthesis of interleukin-2 was decreased in patients with SS compared to the other groups, while secretions of interferon-gamma and TNF-alpha were not altered. Ten surviving patients with SS showed a trend towards the normalisation of these parameters on day 7. In SS, T-cell apoptosis is related, at least in part, to the alteration of the costimulatory pathway, which, in turn, leads to significant modification of the cytokine network.

Feasibility and safety of ultra-low tidal volume ventilation without extracorporeal circulation in moderately severe and severe ARDS patients.

PURPOSE: Mechanical ventilation with ultra-low tidal volume (VT) during ARDS may reduce alveolar strain, driving pressure and hence ventilator-induced lung injury, with the main drawback of worsening respiratory acidosis. We hypothesized that VT could be reduced down to 4 ml/kg, with clinically significant decrease in driving pressure, without the need for extracorporeal CO2 removal, while maintaining pH > 7.20. METHODS: We conducted a non-experimental before-and-after multicenter study on 35 ARDS patients with PaO2/FiO2 ≤ 150 mmHg, within 24 h of ARDS diagnosis. After inclusion, VT was reduced to 4 ml/kg and further adjusted to maintain pH ≥ 7.20, respiratory rate was increased up to 40 min-1 and PEEP was set using a PEEP-FiO2 table. The primary judgment criterion was driving pressure on day 2 of the study, as compared to inclusion. RESULTS: From inclusion to day 2, driving pressure decreased significantly from 12 [9-15]  to 8 [6-11] cmH2O, while VT decreased from 6.0 [5.9-6.1] to 4.1 [4.0-4.7] ml/kg. On day 2, VT was below 4.2 ml/kg in 65% [CI95% 48%-79%], and below 5.25 ml/kg in 88% [CI95% 74%-95%] of the patients. 2 patients (6%) developed acute cor pulmonale after inclusion. Eleven patients (32%) developed transient severe acidosis with pH < 7.15. Fourteen patients (41%) died before day 90. CONCLUSION: Ultra-low tidal volume ventilation may be applied in approximately 2/3 of moderately severe-to-severe ARDS patients, with a 4 cmH2O median reduction in driving pressure, at the price of transient episodes of severe acidosis in approximately 1/3 of the patients.

Sevoflurane but not propofol increases interstitial glycolysis metabolites availability during tourniquet-induced ischaemia-reperfusion.

BACKGROUND: Ischaemia/reperfusion (I/R) is one of the main pathophysiological phenomena involved in the anaesthetic practice. The authors hypothesized that anaesthetic regimens can influence skeletal muscle tolerance to tourniquet-induced I/R that should be reflected by the interstitial metabolite levels of anaerobic glycolysis. METHODS: Microdialysis probes were implanted in three groups of 10 patients each receiving either sevoflurane (SEVO), propofol (PRO), or spinal (SA) anaesthesia (for induction and maintenance). SA group was considered as a control group. Interstitial fluid was obtained during tourniquet-induced I/R and was analysed for interstitial glucose, lactate, pyruvate, and glycerol. RESULTS: The microdialysis flow rate was 0.5 microl min(-1). Compared with the control group, the SEVO group had a higher level of both lactate and pyruvate and an increase in glucose during ischaemia. In contrast, the PRO group had a lower level of pyruvate, resulting in a significant higher increase (eight times from baseline) of the lactate pyruvate ratio. Glucose level remained low in this group. During reperfusion, lactate, pyruvate, and glucose remained at a significantly higher level in the SEVO group. In the PRO group, there was no difference in lactate, pyruvate, and glucose levels compared with the control group. The interstitial level of glycerol exhibits only few and comparable changes during I/R between the groups. CONCLUSIONS: Our results indicate that there is a better availability of interstitial glycolysis metabolites (glucose, lactate, and pyruvate) in the skeletal muscle during ischaemia and reperfusion after sevoflurane exposure than after propofol, suggesting a potential preconditioning effect of sevoflurane on tourniquet-induced skeletal muscle I/R.

Abrupt oxygen decrease influences thrombosis and bleeding in stenosed and endothelium-injured rabbit carotid arteries.

BACKGROUND AND OBJECTIVE: Blood oxygen concentration decrease may be associated with haemostatic impairments. We aimed to study the effect of oxygen decrease in a rabbit model of thrombosis and bleeding. METHODS: A total of 44 rabbits were anaesthetized, ventilated and monitored for blood pressure, blood arterial gas, temperature and carotid blood flow. The Folts model was used: a stenosis (75%) and an injury were carried out on the carotid artery, inducing thrombosis. Blood flow decreased as thrombus size increased until the pressure gradient was such that the thrombus was released and local arterial blood flow was suddenly restored. This is known as a cyclic flow reduction. After counting baseline cyclic flow reductions during a 20-min period (P1), rabbits were randomized blindly to one of three groups: hyperoxic, FiO2=100%; normoxic, FiO2 was decreased to obtain a PaO2 between 80 and 120 mmHg; hypoxic, PaO2 < 80 mmHg. Then CFRs were recorded over a second 20-min period (P2). At the end of the experiment, a hepatosplenic section was done and the amount of blood loss was recorded. After each period, the following parameters were measured: blood gas, ear-immersion bleeding time, haemoglobin, platelet count, prothrombin time, activated partial thromboplastin time and fibrinogen. RESULTS: Oxygen decrease during hypoxic and normoxic periods was associated with a decrease in cyclic flow reductions. Bleeding time increased in the hypoxic group unless hepatosplenic bleeding remained stable. A slight increase in activated partial thromboplastin time was observed in the normoxic and hypoxic groups. CONCLUSION: An abrupt decrease in oxygen administration was responsible for an antithrombotic effect. Increase in bleeding time occurred during hypoxia. No clinically relevant variation of any haemostasis parameters was observed.

Impact of walking epidural analgesia on obstetric outcome of nulliparous women in spontaneous labour.

BACKGROUND: To explore the effects of walking epidural analgesia on obstetric and neonatal outcomes, we performed a case-control study. METHOD: Each nulliparous woman receiving walking epidural analgesia using 0.0625% bupivacaine (n = 44) was matched to two nulliparous historical controls receiving 0.125% or 0.25% bupivacaine (n = 88 each) for epidural analgesia while recumbent. RESULTS: Maternal and obstetric parameters, fetal status and presentation, and oxytocin use were comparable among groups. Those receiving walking epidural analgesia walked for a mean of 60 min (range: 20-75 min). In the control groups the mean total durations of labour were shorter (58 min in the 0.125% group and 99 min in the 0.25% group, P < 0.05). Significantly fewer walking epidural analgesia cases than controls required instrumental vaginal delivery (P < 0.05). No other differences in obstetric or fetal outcome were observed and no mother fell or stumbled while walking. CONCLUSION: Although it was associated with a prolonged first stage of labour, walking epidural analgesia appeared safe for nulliparous women and their babies.

The ANTIBIOPERF study: a nationwide cross-sectional survey about practices for ß-lactam administration and therapeutic drug monitoring among critically ill patients in France.

Our objective was to assess current practices about the administration (intermittent, extended, or continuous infusions) and therapeutic drug monitoring (TDM) of ß-lactam antibiotics and vancomycin in France. We conducted a nationwide cross-sectional survey in May-August 2015, using an online questionnaire, sent as an e-mail link to infectious disease specialists and intensive care specialists through national mailing lists. We used clinical vignettes of critically ill patients to assess physicians' practices about administration and TDM practices for amoxicillin, cloxacillin, piperacillin/tazobactam, cefotaxime, ceftazidime, cefepime, meropenem and vancomycin. In all, 507 physicians participated (507/1200, response rate 42%). TDM was rarely available for ß-lactams (from 16.5% (81/490) for cloxacillin to 30% (145/490) for ceftazidime), whereas vancomycin TDM was available in 97% (477/490) of the cases. In the clinical vignettes, ceftazidime and piperacillin/tazobactam were the ß-lactams administered most frequently by extended or continuous infusions (76% (336/440) and 57% (252/444), respectively). Gaps in knowledge about the duration of stability of intravenous ß-lactams were common (correct answers ranged from 8% (35/432) for cloxacillin to 33% (146/438) for ceftazidime). Most physicians (77%, 339/442) were convinced of the value of extended or continuous infusions for ß-lactams in critically ill patients, but 48% (211/442) did not have access to practical guidelines. Our survey found that most infectious disease and intensive care specialists are favourable to optimized administration of ß-lactams in critically ill patients. But the lack of guidelines and limited TDM availability for ß-lactams in hospitals are potential barriers to its implementation.

Neurally adjusted ventilatory assist as an alternative to pressure support ventilation in adults: a French multicentre randomized trial.

PURPOSE: Neurally adjusted ventilatory assist (NAVA) is a ventilatory mode that tailors the level of assistance delivered by the ventilator to the electromyographic activity of the diaphragm. The objective of this study was to compare NAVA and pressure support ventilation (PSV) in the early phase of weaning from mechanical ventilation. METHODS: A multicentre randomized controlled trial of 128 intubated adults recovering from acute respiratory failure was conducted in 11 intensive care units. Patients were randomly assigned to NAVA or PSV. The primary outcome was the probability of remaining in a partial ventilatory mode (either NAVA or PSV) throughout the first 48 h without any return to assist-control ventilation. Secondary outcomes included asynchrony index, ventilator-free days and mortality. RESULTS: In the NAVA and PSV groups respectively, the proportion of patients remaining in partial ventilatory mode throughout the first 48 h was 67.2 vs. 63.3 % (P = 0.66), the asynchrony index was 14.7 vs. 26.7 % (P < 0.001), the ventilator-free days at day 7 were 1.0 day [1.0-4.0] vs. 0.0 days [0.0-1.0] (P < 0.01), the ventilator-free days at day 28 were 21 days [4-25] vs. 17 days [0-23] (P = 0.12), the day-28 mortality rate was 15.0 vs. 22.7 % (P = 0.21) and the rate of use of post-extubation noninvasive mechanical ventilation was 43.5 vs. 66.6 % (P < 0.01). CONCLUSIONS: NAVA is safe and feasible over a prolonged period of time but does not increase the probability of remaining in a partial ventilatory mode. However, NAVA decreases patient-ventilator asynchrony and is associated with less frequent application of post-extubation noninvasive mechanical ventilation. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT02018666.

[Repeated thromboembolism during pregnancy with constitutional antithrombin deficiency]. / Thromboses récidivantes gravidiques et déficit en antithrombine.

We report the case of a twenty-three-year old woman with constitutional antithrombin deficiency, who had oral anticoagulation since she was four years old. During her first pregnancy, after the introduction of unfractionated heparin prophylactic therapy, she presented a first venous thromboembolism at nine weeks, and a second one with low-molecular-weight heparin therapy at nineteen weeks. Because of a severe antithombin deficiency, regular infusions of antithrombin concentrates were necessary until delivery to ensure effective anticoagulation by heparin. Patients with antithrombin deficiency have a very high risk of venous thromboses during the pregnancy and post-partum. We discuss the significant points of management for this period.

[Assessment of 44 heat and moisture exchange filters. What to choose?]. / Evaluation de 44 filtres échangeurs de chaleur et d'humidité. Que choisir?

OBJECTIVES: If the use of heat and moisture exchange filter (HMEF) in anaesthesia is recommended by the French Society of Anaesthesia and Intensive Care (SFAR), the criteria's choice are not clearly defined. Many HMEF are proposed by manufacturers which technical characteristics are different. STUDY DESIGN: The aim of this study was to evaluate the HMEF using items of the American Association of Respiratory Care (AARC) and technical dossiers. METHODS: All manufactures producing filters have been contacted to give their technical dossiers. Forty-eight filters have been analyzed (13 mechanical filters, 31 electrostatic filters). Each item has been scaled 0, 5 or 10. The final result was on 10. RESULTS: Seventeen filters had a note superior to 5. There were 8 mechanical filters and 9 electrostatic filters. The difference between the filters was the size of the micro-organisms tested and the duration of the test. Some filters were not tested by independent laboratories (N = 8). There were differences between the commercial documentation and on Internet and the technical dossiers (N = 12). DISCUSSION: We noted the good quality of the filters particularly concerning criters recommended by the Sfar (filter medium, filtration efficiency, microbial challenge number and duration of the test). The electrostatic filters recently used in anaesthesia have high performance concerning filtration efficiency. To supprime the moisture output criteria did not change the results. CONCLUSIONS: Criteria's used by manufactures to evaluate there filters are not always precised or too restrictive. The technical tests, the international norms, the certificates of validation, the ergonomic qualities and the definition of our needs are the main elements of choice of a filter.

[Rapid sequence intubation in emergency: is there any place for fentanyl?]. / Induction à séquence rapide en urgence: quelle est la place du fentanyl?

OBJECTIVES: Rapid sequence intubation (RSI) with the association of etomidate and succinylcholine is the French "Gold standard" for urgent "full stomach" endotracheal intubations. The aim of this study is to assess the fentanyl as a co-induction agent to take over the sedation between the RSI and the keeping of sedation, which is a critical period in which harmful neuro-vegetatives events, and awakening signs are frequently seen. STUDY DESIGN: Randomized, double blind controlled prospective study, after acceptation by the local ethical committee. PATIENTS AND METHODS: Three groups of patients undergoing RSI in the intensive care unit and by the out-of-hospital medical team were compared: group A patients received fentanyl 3 micrograms kg-1 during RSI, before paralysis was induced. Group B patients received the same dose of fentanyl immediately after endotracheal intubation. Group C patients did not received fentanyl (control group). Outcome measures were awakening signs arrival (respiratory movements, eyes opening, spontaneous limb movements), Ramsay score assessment, and haemody namics. Attempt at intubation and vomiting incident were also measured. Discrete data were compared by chi-2 analysis, continuous data were compared with two-way analysis of variance. A p value < 0.05 was the significant threshold. RESULTS: Thirty-six patients were enrolled and completed the study. All the included patients presented awakening signs. The use of fentanyl did not prevent the recourse of other sedative medications. Ten minutes after endotracheal intubation, significant differences has been noticed for the awakening signs arrival between fentanyl groups (A: 42% and B: 36%) and control group (C: 77%). The Ramsay score evolution follows the same variation. All the patients were intubated on the first attempt, there was no vomiting incident noticed. CONCLUSION: The use of fentanyl, as a co-induction agent with etomidate and succinylcholine during RSI, allows a significant delay of the awakening signs arrival and attenuate the neurovegetative response during the minutes after endotracheal intubation after RSI, without deleterious haemodynamic effects.

[Secondary Streptococcus pyogenes peritonitis following necrotizing fasciitis]. / Péritonite à Streptococcus pyogenes et fasciite nécrosante.

We report the case of a 66-year-old woman with diabetes mellitus and disseminated lupus treated with immunosupressive drug. She was admitted for an inflammatory oedema of the right lower limb associated to diffuse abdominal pain and vomiting. The occurrence of septic shock with multiple organ failure and surgical abdominal picture led us to perform urgent laparotomy without taking in account the evolutive signs of cutaneous lesions i.e. purpuric elements and haemorrhagic phlycthena. The abdominal cavity exploration showed peritonitis without digestive tracts perforation. The patient died few hours after surgical procedure. All bacteriological samples i.e. peritoneal effusion, blood cultures and phlycthena liquid were positive for Streptocoque pyogenes (group A) and peritonitis was considered to be related to haematogen diffusion from rapid course necrotizing fasciitis.

Effect of different seated positions on lung volume and oxygenation in acute respiratory distress syndrome.

RATIONALE: Lung volume available for ventilation is markedly decreased during acute respiratory distress syndrome. Body positioning may contribute to increase lung volume and partial verticalization is simple to perform. This study evaluated whether verticalization had parallel effects on oxygenation and end expiratory lung volume (EELV). METHODS: Prospective multicenter study in 40 mechanically ventilated patients with ALI/ARDS in five university hospital MICUs. We evaluated four 45-min successive trunk position epochs (supine slightly elevated at 15°; semi recumbent with trunk elevated at 45°; seated with trunk elevated at 60° and legs down at 45°; back to supine). Arterial blood gases, EELV measured using the nitrogen washin/washout, and static compliance were measured. Responders were defined by a PaO2/FiO2 increase >20 % between supine and seated position. Results are median [25th-75th percentiles]. RESULTS: With median PEEP = 10 cmH2O, verticalization increased lung volume but only responders (13 patients, 32 %) had a significant increase in EELV/PBW (predicted body weight) compared to baseline. This increase persisted at least partially when patients were positioned back to supine. Responders had a lower EELV/PBW supine [14 mL/kg (13-15) vs. 18 mL/kg (15-27) (p = 0.005)] and a lower compliance [30 mL/cmH2O (22-38) vs. 42 (30-46) (p = 0.01)] than non-responders. Strain decreased with verticalization for responders. EELV/PBW increase and PaO2/FiO2 increase were not correlated. DISCUSSION: Verticalization is easily achieved and improves oxygenation in approximately 32 % of the patients together with an increase in EELV. Nonetheless, effect of verticalization on EELV/PBW is not predictable by PaO2/FiO2 increase, its monitoring may be helpful for strain optimization.

Impact of an intervention designed to improve the documentation of the reassessment of antibiotic therapies in an intensive care unit.

OBJECTIVES: The study objectives were: (i) to design an intervention to improve the written documentation of empiric antibiotic prescriptions' reassessment; (ii) and to assess the impact of this intervention on the quality of prescriptions. PATIENTS AND METHODS: A prospective before and after 7-month intervention study in a medical ICU in a French teaching hospital, using interrupted time-series analysis. The intervention was made to improve the documentation of four process measures in medical records: antibiotic plan, reviewing the diagnosis, adapting to positive microbiological results, and IV-per os switch. RESULTS: One hundred and fourteen antibiotic prescriptions were assessed, 62 before and 52 after the intervention. The reassessment of antibiotic prescriptions was more often documented in the ICU after the intervention (P=0.03 for sudden change). The prevalence of appropriate antibiotic prescriptions was not statistically different before and after the intervention, either for sudden change and/or linear trend. CONCLUSION: A better documentation of antibiotic prescriptions' reassessment was achieved in this ICU, but it did not improve the quality of antibiotic prescriptions.

Impact of an assisted reassessment of antibiotic therapies on the quality of prescriptions in an intensive care unit.

OBJECTIVES: The study's objective was to assess the impact of a professional multifaceted intervention designed to improve the quality of inpatient empirical therapeutic antibiotic courses at the time of their reassessment, i.e. 24 to 96 hours after treatment initiation. DESIGN: We conducted a 5-month prospective pre- and post-intervention study in a medical Intensive Care Unit (ICU) in a teaching hospital, using time-series analysis. The intervention was a multifaceted professional intervention combining systematic 3-weekly visits of an infectious diseases specialist to discuss all antibiotic therapies, interactive teaching courses, and daily contact with a microbiologist. RESULTS: Eighty-one antibiotic prescriptions were assessed, 37 before and 44 after the intervention. The prevalence of adequate antibiotic prescriptions was high and not statistically different before and after the intervention (73% vs. 80%, P=0.31), both for sudden change (P=0.67) and linear trend (P=0.055), using interrupted time-series analysis. The intervention triggered a more frequent reassessment of the diagnosis between day 2 and day 4 (11% vs. 32%, P=0.02) and slightly improved the adaptation of antibiotic therapies to positive microbiology (25% before vs. 50% after, P=0.18). CONCLUSIONS: Our multifaceted intervention may have improved the quality of antibiotic therapies around day 3 of prescription, but the difference did not reach statistical significance, possibly because of a ceiling effect.