Antimicrobial prescription in severe COVID-19 and CAP: a matched case-control study.

BACKGROUND: In severe coronavirus diseases 2019 (COVID-19), a high and potentially excessive use of antimicrobials for suspected bacterial co-infection and intensive care unit (ICU)-acquired infections has been repeatedly reported. OBJECTIVES: To compare an ICU cohort of community-acquired pneumonia (CAP) with a cohort of severe COVID-19 pertaining to co-infections, ICU-acquired infections and associated antimicrobial consumption. METHODS: We retrospectively compared a cohort of CAP patients with a cohort of COVID-19 patients matched according to organ failure, ICU length of stay (LOS) and ventilation days. Patient data such as demographics, infection focus, probability and severity, ICU severity scores and ICU and in-hospital mortality, days of antimicrobial therapy (DOT) and number of antimicrobial prescriptions, using an incremental scale, were registered and analysed. The total number of cultures (sputum, urinary, blood cultures) was collected and corrected for ICU LOS. FINDINGS: CAP patients (n = 148) were matched to COVID-19 patients (n = 74). Significantly less sputum cultures (68.2% versus 18.9%, P < 0.05) and bronchoalveolar lavages (BAL) (73.7% versus 36.5%, P < 0.05) were performed in COVID-19 patients. Six (8.1%) COVID-19 patients were diagnosed with a co-infection. Respectively, 58 of 148 (39.2%) CAP and 38 of 74 (51.4%) COVID-19 patients (P = 0.09) developed ICU-acquired infections. Antimicrobial distribution, both in the number of prescriptions and DOT, was similar in both cohorts. CONCLUSIONS: We found a low rate of microbiologically confirmed bacterial co-infection and a high rate of ICU-acquired infections in COVID-19 patients. Infection probabilities, antimicrobial prescriptions and DOT were comparable with a matched CAP cohort.

Machine learning in infection management using routine electronic health records: tools, techniques, and reporting of future technologies.

BACKGROUND: Machine learning (ML) is increasingly being used in many areas of health care. Its use in infection management is catching up as identified in a recent review in this journal. We present here a complementary review to this work. OBJECTIVES: To support clinicians and researchers in navigating through the methodological aspects of ML approaches in the field of infection management. SOURCES: A Medline search was performed with the keywords artificial intelligence, machine learning, infection∗, and infectious disease∗ for the years 2014-2019. Studies using routinely available electronic hospital record data from an inpatient setting with a focus on bacterial and fungal infections were included. CONTENT: Fifty-two studies were included and divided into six groups based on their focus. These studies covered detection/prediction of sepsis (n = 19), hospital-acquired infections (n = 11), surgical site infections and other postoperative infections (n = 11), microbiological test results (n = 4), infections in general (n = 2), musculoskeletal infections (n = 2), and other topics (urinary tract infections, deep fungal infections, antimicrobial prescriptions; n = 1 each). In total, 35 different ML techniques were used. Logistic regression was applied in 18 studies followed by random forest, support vector machines, and artificial neural networks in 18, 12, and seven studies, respectively. Overall, the studies were very heterogeneous in their approach and their reporting. Detailed information on data handling and software code was often missing. Validation on new datasets and/or in other institutions was rarely done. Clinical studies on the impact of ML in infection management were lacking. IMPLICATIONS: Promising approaches for ML use in infectious diseases were identified. But building trust in these new technologies will require improved reporting. Explainability and interpretability of the models used were rarely addressed and should be further explored. Independent model validation and clinical studies evaluating the added value of ML approaches are needed.

A comparison of propofol-to-BIS post-operative intensive care sedation by means of target controlled infusion, Bayesian-based and predictive control methods: an observational, open-label pilot study.

PURPOSE: We evaluated the feasibility and robustness of three methods for propofol-to-bispectral index (BIS) post-operative intensive care sedation, a manually-adapted target controlled infusion protocol (HUMAN), a computer-controlled predictive control strategy (EPSAC) and a computer-controlled Bayesian rule-based optimized control strategy (BAYES). METHODS: Thirty-six patients undergoing short lasting sedation following cardiac surgery were included to receive propofol to maintain a BIS between 40 and 60. Robustness of control for all groups was analysed using prediction error and spectrographic analysis. RESULTS: Although similar time courses of measured BIS were obtained in all groups, a higher median propofol effect-site concentration (CePROP) was required in the HUMAN group compared to the BAYES and EPSAC groups. The time course analysis of the remifentanil effect-site concentration (CeREMI) revealed a significant increase in CeREMI in the EPSAC group compared to BAYES and HUMAN during the case. Although similar bias and divergence in control was found in all groups, larger control inaccuracy was observed in HUMAN versus EPSAC and BAYES. Spectrographic analysis of the system behavior shows that BAYES covers the largest spectrum of frequencies, followed by EPSAC and HUMAN. CONCLUSIONS: Both computer-based control systems are feasible to be used during ICU sedation with overall tighter control than HUMAN and even with lower required CePROP. EPSAC control required higher CeREMI than BAYES or HUMAN to maintain stable control. Clinical trial number: NCT00735631.

Outcome in patients perceived as receiving excessive care across different ethical climates: a prospective study in 68 intensive care units in Europe and the USA.

PURPOSE: Whether the quality of the ethical climate in the intensive care unit (ICU) improves the identification of patients receiving excessive care and affects patient outcomes is unknown. METHODS: In this prospective observational study, perceptions of excessive care (PECs) by clinicians working in 68 ICUs in Europe and the USA were collected daily during a 28-day period. The quality of the ethical climate in the ICUs was assessed via a validated questionnaire. We compared the combined endpoint (death, not at home or poor quality of life at 1 year) of patients with PECs and the time from PECs until written treatment-limitation decisions (TLDs) and death across the four climates defined via cluster analysis. RESULTS: Of the 4747 eligible clinicians, 2992 (63%) evaluated the ethical climate in their ICU. Of the 321 and 623 patients not admitted for monitoring only in ICUs with a good (n = 12, 18%) and poor (n = 24, 35%) climate, 36 (11%) and 74 (12%), respectively were identified with PECs by at least two clinicians. Of the 35 and 71 identified patients with an available combined endpoint, 100% (95% CI 90.0-1.00) and 85.9% (75.4-92.0) (P = 0.02) attained that endpoint. The risk of death (HR 1.88, 95% CI 1.20-2.92) or receiving a written TLD (HR 2.32, CI 1.11-4.85) in patients with PECs by at least two clinicians was higher in ICUs with a good climate than in those with a poor one. The differences between ICUs with an average climate, with (n = 12, 18%) or without (n = 20, 29%) nursing involvement at the end of life, and ICUs with a poor climate were less obvious but still in favour of the former. CONCLUSION: Enhancing the quality of the ethical climate in the ICU may improve both the identification of patients receiving excessive care and the decision-making process at the end of life.

Long-term outcome and health-related quality of life in difficult-to-wean patients with and without ventilator dependency at ICU discharge: a retrospective cohort study.

BACKGROUND: Long-term outcome and quality of life (QOL) in patients requiring prolonged mechanical ventilation after failure to wean in the ICU is scarcely documented. We aimed to evaluate long-term survival and QOL in patients discharged from the ICU with a tracheostomy for difficult weaning, and with or without ventilator dependency at ICU discharge. METHODS: We retrospectively investigated post-ICU trajectories and survival in patients requiring tracheostomy for difficult weaning admitted to the medical ICU of a tertiary center between 1999 and 2013, discriminating between patients who were ventilator dependent or were weaned at ICU discharge. In 2014, a QOL assessment was done in survivors with the use of the Short Form Health Survey (SF-36) and the Severe Respiratory Insufficiency questionnaire. RESULTS: A total of 114 patients was included, of whom 59 were ventilator dependent and 55 were weaned at ICU discharge. One-year survival rates were 73 % and 69 %, respectively. Overall QOL scores for physical functioning were low, and not significantly different between patients ventilated and those weaned at ICU discharge; scores for social functioning and mental health were less below norm and similar between both groups. CONCLUSIONS: Long-term survival in patients discharged from the ICU with tracheostomy and ventilator dependency after failure to wean was not significantly different from that of patients with tracheostomy and weaned at ICU discharge. Despite the physical QOL scores being low in both groups, mental QOL was acceptable. Given the intrinsic limitations of this retrospective study, prospective and preferentially multicenter studies are required to confirm these preliminary results.

Platform for efficient switching between multiple devices in the intensive care unit.

INTRODUCTION: This article is part of the Focus Theme of METHODS of Information in Medicine on "Managing Interoperability and Complexity in Health Systems". OBJECTIVES: Handheld computers, such as tablets and smartphones, are becoming more and more accessible in the clinical care setting and in Intensive Care Units (ICUs). By making the most useful and appropriate data available on multiple devices and facilitate the switching between those devices, staff members can efficiently integrate them in their workflow, allowing for faster and more accurate decisions. This paper addresses the design of a platform for the efficient switching between multiple devices in the ICU. The key functionalities of the platform are the integration of the platform into the workflow of the medical staff and providing tailored and dynamic information at the point of care. METHODS: The platform is designed based on a 3-tier architecture with a focus on extensibility, scalability and an optimal user experience. After identification to a device using Near Field Communication (NFC), the appropriate medical information will be shown on the selected device. The visualization of the data is adapted to the type of the device. A web-centric approach was used to enable extensibility and portability. RESULTS: A prototype of the platform was thoroughly evaluated. The scalability, performance and user experience were evaluated. Performance tests show that the response time of the system scales linearly with the amount of data. Measurements with up to 20 devices have shown no performance loss due to the concurrent use of multiple devices. CONCLUSIONS: The platform provides a scalable and responsive solution to enable the efficient switching between multiple devices. Due to the web-centric approach new devices can easily be integrated. The performance and scalability of the platform have been evaluated and it was shown that the response time and scalability of the platform was within an acceptable range.

Validity analysis of a unique infection surveillance system in the intensive care unit by analysis of a data warehouse built through a workflow-integrated software application.

BACKGROUND: An electronic decision support programme was developed within the intensive care unit (ICU) that provides an overview of all infection-related patient data, and allows ICU physicians to add clinical information during patient rounds, resulting in prospective compilation of a database. AIM: To assess the validity of computer-assisted surveillance (CAS) of ICU-acquired infection performed by analysis of this database. METHODS: CAS was compared with prospective paper-based surveillance (PBS) for ICU-acquired respiratory tract infection (RTI), bloodstream infection (BSI) and urinary tract infection (UTI) over four months at a 36-bed medical and surgical ICU. An independent panel reviewed the data in the case of discrepancy between CAS and PBS. FINDINGS: PBS identified 89 ICU-acquired infections (13 BSI, 18 UTI, 58 RTI) and CAS identified 90 ICU-acquired infections (14 BSI, 17 UTI, 59 RTI) in 876 ICU admissions. There was agreement between CAS and PBS on 13 BSI (100 %), 14 UTI (77.8 %) and 42 RTI (72.4 %). Overall, there was agreement on 69 infections (77.5%), resulting in a kappa score of 0.74. Discrepancy between PBS and CAS was the result of capture error in 11 and 14 infections, respectively. Interobserver disagreement on probability (13 RTI) and focus (two RTI, one UTI) occurred for 16 episodes. The time required to collect information using CAS is less than 30% of the time required when using PBS. CONCLUSION: CAS for ICU-acquired infection by analysis of a database built through daily workflow is a feasible surveillance method and has good agreement with PBS. Discrepancy between CAS and PBS is largely due to interobserver variability.

Extended versus bolus infusion of meropenem and piperacillin: a pharmacokinetic analysis.

BACKGROUND: Extended infusion of beta-lactam antibiotics has been advocated as a method for optimizing antibiotic exposure in critically ill patients. The objective of this study was to compare the pharmacokinetics/pharmacodynamics of extended infusion versus bolus infusion of piperacillin and meropenem in critically ill patients with normal renal function. METHODS: A prospective study of 3 hours extended infusion of meropenem and piperacillin in critically ill patients without renal dysfunction. Results from the extended infusion cohort were compared to previously published bolus infusion data in critically ill patients. RESULTS: Twenty extended infusion patients (15 piperacillin, 5 meropenem) were compared with 13 bolus infusion patients (8 piperacillin, 5 meropenem). The demographic and clinical characteristics between both groups were not statistically different. Significant pharmacokinetic differences were observed in median (interquartile range) Cmax for both meropenem (extended infusion 17 [12.6-21.9] vs. bolus 85.2 [66.7-140.3]; P=0.01) and piperacillin (extended infusion 76.2 [57.7-92.6] vs. bolus 240.2 [168.5-275.4]; P=0.001). Considerable pharmacokinetic variability existed in each group for both drugs. Compared to bolus infusion, fT>MIC using extended infusion was higher for both drugs: 96% (IQR 71-100%) compared to 77% (IQR 41-93%) for piperacillin (P=0.05) and 82% (IQR 63-89%) compared to 51% (IQR 48-63%) for meropenem (P=0.095); assuming a MIC of 16 mg/L and 2 mg/L respectively. CONCLUSION: This study confirms that extended infusion in critically ill patients result in advantageous pharmacokinetic profiles by increasing the fT>MIC for piperacillin and meropenem. In a significant subpopulation of critically ill patients with normal renal function, a 100% fT>MIC target is not reached, even with 3-hour extended infusions.

Therapeutic drug monitoring-based dose optimisation of piperacillin and meropenem: a randomised controlled trial.

PURPOSE: There is variability in the pharmacokinetics (PK) of antibiotics (AB) in critically ill patients. Therapeutic drug monitoring (TDM) could overcome this variability and increase PK target attainment. The objective of this study was to analyse the effect of a dose-adaption strategy based on daily TDM on target attainment. METHODS: This was a prospective, partially blinded, and randomised controlled trial in patients with normal kidney function treated with meropenem (MEM) or piperacillin/tazobactam (PTZ). The intervention group underwent daily TDM, with dose adjustment when necessary. The predefined PK/pharmacodynamic (PK/PD) target was 100% fT>4MIC [percentage of time during a dosing interval that the free (f) drug concentration exceeded 4 times the MIC]. The control group received conventional treatment. The primary endpoint was the proportion of patients that reached 100% fT>4MIC and 100 % fT>MIC at 72 h. RESULTS: Forty-one patients (median age 56 years) were included in the study. Pneumonia was the primary infectious diagnosis. At baseline, 100% fT>4MIC was achieved in 21% of the PTZ patients and in none of the MEM patients; 100% fT>MIC was achieved in 71% of the PTZ patients and 46 % of the MEM patients. Of the patients in the intervention group, 76 % needed dose adaptation, and five required an additional increase. At 72 h, target attainment rates for 100% fT>4MIC and 100% fT>MIC were higher in the intervention group: 58 vs. 16%, p = 0.007 and 95 vs. 68%, p = 0.045, respectively. CONCLUSIONS: Among critically ill patients with normal kidney function, a strategy of dose adaptation based on daily TDM led to an increase in PK/PD target attainment compared to conventional dosing.

Factors associated with inadequate early vancomycin levels in critically ill patients treated with continuous infusion.

Vancomycin administration using a loading dose and continuous infusion (CI) results in more rapid attainment of adequate concentrations. The aim of this retrospective study of ICU patients receiving vancomycin was to determine the efficacy of a vancomycin dosing protocol using a weight-based loading dose and to identify factors associated with inadequate concentrations. Patients received a loading dose (<65 kg, 1000 mg; ≥65 kg, 1500 mg), and 2000 mg/24 h CI with subsequent dose adaptation. Adequate levels were defined as concentrations ≥15 mg/L. In total, 227 patients (154 males) were included in the study (mean age 56.5 ± 16.1 years; mean APACHE II score 19.30 ± 7.7). The mean loading dose was 1129 ± 369 mg (15.07 ± 4.99 mg/kg). The dosing protocol was applied in 126 patients (55.5%). Mean vancomycin levels were 19.32 mg/L and 21.08 mg/L on Days 2 and 3, respectively. Vancomycin levels on Day 2 were adequate in 70.5% of patients, increasing to 84.1% on Day 3. Patients who received an appropriate loading dose more often had adequate vancomycin levels on Day 2. Older age, female sex, higher creatinine concentration, lower body temperature and use of a loading dose according to the vancomycin dosing protocol were independently associated with adequate vancomycin levels. A weight-based loading dose plus CI of vancomycin resulted in adequate concentrations in most patients and was superior compared with a non-standardised loading dose. Some patients may require higher doses, and factors other than weight, such as kidney function, age and sex, play a role.

A survey of beta-lactam antibiotics and vancomycin dosing strategies in intensive care units and general wards in Belgian hospitals.

Extended and continuous infusions with beta-lactam antibiotics have been suggested as a means of pharmacokinetic and pharmacodynamic optimisation of antimicrobial therapy. Vancomycin is also frequently administered in continuous infusion, although more for practical reasons. A survey was undertaken to investigate the recommendations by the local antibiotic management teams (AMTs) in Belgian acute hospitals concerning the administration (intermittent, extended or continuous infusion) and therapeutic drug monitoring of four beta-lactam antibiotics (ceftazidime, cefepime, piperacillin-tazobactam, meropenem) and vancomycin for adult patients with a normal kidney function. A structured questionnaire survey comprising three domains was developed and approved by the members of the Belgian Antibiotic Policy Coordination Committee (BAPCOC). The questionnaire was sent by e-mail to the official AMT correspondents of 105 Belgian hospitals, followed by two reminders. The response rate was 32 %, with 94 %, 59 %, 100 %, 100 % and 100 % of the participating Belgian hospitals using ceftazidime, cefepime, piperacillin-tazobactam, meropenem and vancomycin, respectively. Comparing intensive care unit (ICU) with non-ICU wards showed a higher implementation of extended or continuous infusions for ceftazidime (81 % vs. 41 %), cefepime (35 % vs. 10 %), piperacillin-tazobactam (38 % vs. 12 %), meropenem (68 % vs. 35 %) and vancomycin (79 % vs. 44 %) on the ICU wards. A majority of the hospitals recommended a loading dose prior to the first dose. For vancomycin, the loading dose and the trough target concentration were too low based on the current literature. This survey shows that extended and continuous infusions with beta-lactams and vancomycin are widely implemented in Belgian hospitals.

Long-term outcomes and quality of life in critically ill patients with hematological or solid malignancies: a single center study.

PURPOSE: Data concerning long-term outcomes and quality of life (QOL) in critically ill cancer patients are scarce. The aims of this study were to assess long-term outcomes and QOL in critically ill patients with hematological (HM) or solid malignancies (SM) 3 months and 1 year after intensive care unit (ICU) discharge, to compare these with QOL before ICU admission, and to identify prognostic indicators of long-term QOL. METHODS: During a 1 year prospective observational cohort analysis, consecutive patients with HM or SM admitted to the medical or surgical ICU of a university hospital were screened for inclusion. Cancer data, demographics, co-morbidity, severity of illness, organ failures, and outcomes were collected. The QOL before ICU admission, 3 months, and 1 year after ICU discharge was assessed using standardized questionnaires (EuroQoL-5D, Medical Outcomes Study 36-item Short Form Health Survey). Statistical significance was attained at P < 0.05. RESULTS: There were 483 patients (85 HM, 398 SM) (64% men) with a median age of 62 years included. Mortality rates of HM compared to SM were, respectively: hospital (34 vs. 13%), 3 months (42 vs. 17%), and 1 year (66 vs. 36%) (P < 0.001). QOL declined at 3 months, but improved at 1 year although it remained under baseline QOL, particularly in HM. Older age (P = 0.007), severe comorbidity (P = 0.035), and HM (P = 0.041) were independently associated with poorer QOL at 1 year. CONCLUSIONS: Long-term outcomes and QOL were poor, particularly in HM. Long-term expectations should play a larger role during multidisciplinary triage decisions upon referral to the ICU.

Interpretation of hemolysis tests following administration of a second-generation hemoglobin-based oxygen carrier.

Hemoglobin released into the circulation during hemolysis or therapy with chemically modified hemoglobins, exert oxidative and NO-scavenging toxic effects. Pyridoxalated hemoglobin polyoxyethylene conjugate (PHP) is one of the second-generation hemoglobin-based oxygen carriers (HBOCs). We wanted to investigate the metabolism of PHP with a special focus on its consequences for interpreting hemolysis-related diagnostic parameters in PHP-treated patients. Clinical samples were analyzed from 3 patients, who received PHP (as part of the PHOENIX phase III trial) for treatment of catecholamine-resistant distributive shock. In contrast to expectations, clearance of PHP by hemopexin, instead of haptoglobin was documented by increased hemolysis indices, absence of decreased haptoglobin values, presence of free PHP-hemoglobin and exhausted hemopexin concentrations. The present case report is important for both clinicians and laboratorians since it nicely illustrates that a hemolytic aspect of plasma is not necessarely synonymous with hemolysis. A hemolytic aspect of plasma or serum (high hemolysis index) in combination with normal or increased haptoglobin values should draw the attention; additional determination of lactate dehydrogenase and hemopexin may then be useful to distinguish the condition from in vitro hemolysis and to monitor the in vivo elimination of the heme compounds.

Early recognition of malignant lactic acidosis in clinical practice: report on 6 patients with haematological malignancies.

BACKGROUND: Malignant lactic acidosis is a potentially overlooked but life-threatening complication in patients with haematological malignancies. The aim of this study is to describe the features of six patients with malignant lactic acidosis and to discuss how its initial presentation can be differentiated from that of severe sepsis. METHODS: We prospectively collected data of all consecutive patients with haematological malignancies, admitted to the Ghent University Hospital Intensive Care Unit (ICU) between 2000 and 2007. RESULTS: Of 372 patients with haematological malignancies admitted to the ICU for life- threatening complications, 58 presented with lactic acid levels > or = 5 mmol/L. Six were diagnosed with malignant lactic acidosis. All patients with malignant lactic acidosis had high-grade lymphoblastic malignancies and were referred with a tentative diagnosis of severe sepsis or septic shock; lactic acid levels exceeded 9.45 mmol/L and lactate dehydrogenase (LDH) levels were at least 1785 U/L. Two patients had hypoglycaemia. All had a pronounced polypnea. In all patients hepatic malignant involvement was suspected. Two of the six patients survived their episode thanks to the early recognition of malignant lactic acidosis and the prompt administration of chemotherapy. One patient was still alive 6 months after initiating chemotherapy. CONCLUSION: Malignant lactic acidosis is a rare and often rapidly fatal metabolic complication if not promptly recognized and treated. An elevated lactic acid concentration, in disproportion with the level of tissue hypoxia, together with high serum LDH are cornerstones in the diagnosis. In contrast to septic shock patients, pronounced polypnea (Kussmaul's breathing pattern) rather than the haemodynamic instability is prominent.

Towards automated generation and execution of clinical guidelines: engine design and implementation through the ICU Modified Schofield use case.

As the complexity and amount of medical information keeps increasing, it is difficult to maintain the same quality of care. Therefore, clinical guidelines are used to structure best practices and care, but they also support physicians and nurses in the diagnostic and treatment process. Currently, no standardized format exists to represent these guidelines. Translating guidelines into a computer interpretable format can overcome problems in the physicians' workflow and improve clinician's uptake. An engine is proposed to automatically translate and execute clinical guidelines. These guidelines are represented as flowcharts, expressed in either (i) a computer interpretable guideline format or (ii) a UML diagram. A detailed overview of the architecture is presented and algorithms, aiming at grouping several components and distributing the guidelines, are proposed to optimize the execution of the guidelines. The Modified Schofield guideline for the calculation of the calorie need for burn patients was used for evaluation. Results show that the execution of guidelines using the engine is very efficient. Using optimization algorithms the execution times can be lowered.

Outcome in critically ill patients with allogeneic BM or peripheral haematopoietic SCT: a single-centre experience.

Outcome in haematological patients who develop critical illness has significantly improved over the last two decades, but less so in allogeneic BMT recipients. We prospectively investigated the outcome of 44 haematological patients with allogeneic BM or haematopoietic SCT (ABMT/AHSCT) requiring admission to the intensive care unit (ICU) of Ghent University Hospital between January 2000 and December 2007. We related outcome to the cause of critical illness, which was categorized as documented or clinically suspected bacterial infection, non-bacterial infection and non-infectious disease. Mechanical ventilation was required in 32 patients, and 12 patients received renal replacement therapy. Overall ICU-mortality, in-hospital mortality and 6-month mortality rates were 61, 75 and 80%, respectively. Hospital mortality rates in patients with bacterial infection (n=14), non-bacterial infection (n=13) and non-infectious disease (n=17) were 43, 85 and 94% (P=0.003). After adjustment for severity of illness sequential organ failure assessment (SOFA) score, bacterial infection (odds ratio 0.06, 0.01-0.36, P=0.002) was associated with significantly lower odds for hospital mortality. On the basis of our experience, ICU referral of ABMT/AHSCT patients is justifiable, as an acceptable fraction of these patients have longer-term survival. Documented or clinically suspected bacterial infection as the cause of critical illness is associated with better prognosis in comparison with other causes.

A novel time series analysis approach for prediction of dialysis in critically ill patients using echo-state networks.

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 echo-state 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. METHODS: This study examines the possibility of using an echo-state network for prediction of dialysis in the ICU. Therefore, diuresis values and creatinine levels of the first three days after ICU admission were collected from 830 patients admitted to the intensive care unit (ICU) between May 31 th 2003 and November 17th 2007. The outcome parameter was the performance by the echo-state network in predicting the need for dialysis between day 5 and day 10 of ICU admission. Patients with an ICU length of stay <10 days or patients that received dialysis in the first five days of ICU admission were excluded. Performance by the echo-state network was then compared by means of the area under the receiver operating characteristic curve (AUC) with results obtained by two other time series analysis methods by means of a support vector machine (SVM) and a naive Bayes algorithm (NB). RESULTS: The AUC's in the three developed echo-state networks were 0.822, 0.818, and 0.817. These results were comparable to the results obtained by the SVM and the NB algorithm. CONCLUSIONS: This proof of concept study is the first to evaluate the performance of echo-state networks in an ICU environment. This echo-state network predicted the need for dialysis in ICU patients. The AUC's of the echo-state networks were good and comparable to the performance of other classification algorithms. Moreover, the echo-state network was more easily configured than other time series modeling technologies.

Aspiration pneumonia: an underestimated cause of severe respiratory failure in patients with haematological malignancies and severe oral mucositis?

Aspiration pneumonia is rarely considered in the differential diagnosis of respiratory failure in patients suffering from haematologic malignancies in daily practice. We describe four patients who were admitted with severe respiratory failure in the ICU over a one-year-period prospective survey (a total of 72 patients with haematological malignancies of which 34 presented with respiratory failure). All of these patients had chemotherapy-induced severe oral mucositis (WHO grade ILL-IV) for which three of them received opioids. All had a history of cough after oral rinsing and two of them experienced sudden brief desaturation in the days before ICU referral. Two of these patients, both in allogeneic bone marrow transplant setting, died. With this data, we want to draw the attention to the diagnosis of aspiration pneumonia in this group of patients.

A 50-year-old man with severe hypercalcemia: a case report.

OBJECTIVE: We present this case to emphasize the importance of early diagnosis and treatment of an acute severe hypercalcemic syndrome due to primary hyperparathyroidism as a consequence of an undiagnosed adenoma of the parathyroid gland. CASE REPORT: A 50-year-old man presented at another hospital with non-specific symptoms such as anorexia, nausea, vomiting, polyuria, dehydration, abdominal pain, weight loss, fatigue, muscular weakness, irritability and lethargy. Serum levels of calcium and parathyroid hormone (PTH) were markedly increased to 23.6 mg/dL (reference values 8.6-10.2 mg/dL) and > 1900 ng/L (reference values 14-72 ng/L) respectively. After initial treatment, the patient was transferred to the intensive care unit (ICU) of a tertiary care university hospital for further stabilization and treatment because the typical signs of hypercalcemia were not resolving. A parathyroid adenoma was diagnosed and a few days later a parathyroidectomy was performed. The postoperative course was uneventful and the patient could be discharged from the hospital in a good general condition. CONCLUSION: Acute primary hyperparathyroidism, also known as parathyroid storm or parathyroid crisis, is a rare but potentially fatal endocrine emergency if unrecognized and untreated. Appropriate diagnosis and immediate adequate management of hypercalcemia are important in reducing mortality. Nevertheless, mortality remains high, even with surgical treatment which is the cornerstone of the definitive therapy.

90-day follow-up of patients treated with Drotrecogin Alfa (activated) for severe sepsis: a Belgian open label study.

INTRODUCTION: Severe sepsis is the major cause of mortality in intensive care units (ICUs). The BOOST study (= B (Belgian) OO (Open Label) ST (Study)) is a Belgian open-label trial designed to pragmatically assess the safety and efficacy of Drotrecogin Alfa (activated) (DAA), the only registered treatment in this indication with favourable ratio benefit/risk. METHODOLOGY: Adult patients with severe sepsis and 2 or more sepsis-induced organ dysfunctions (OD) within the 48-hour period preceding the treatment (DAA at 24 microg/kg/h for 96 hours), were included between January 2003 and October 2003. Platelet count < 30 000/mm3 and increased risk for bleeding were exclusion criteria. Mortality and location were evaluated at 28 and 90 days. RESULTS: Of the 100 included patients, 97 (median age: 66 years; men/women: 57/40) were treated and completed the study. The predominant infection sites were lung (49%) and abdomen (29%) and 35% had had recent surgery. The mean and median numbers of OD were 3.4 and 3.0, respectively, and most patients (80 %; 77/97) had 3 or more organ failures at baseline, predominantly respiratory (95%) and cardiovascular (87%). The mean APACHE II score was 25.3 (range: 6-53). The 28-day mortality rate was 32.0% (90% CI: 24.2-39.7) and increased with the number of OD: from 15% (1.9-28.1) for2 ODs, to 71% (52.4-88.8) for 5 ODs. At day 28, the 66 surviving patients were located in general ward (35%), in the ICU (32%) or at home (30%). The 90-day mortality rate was 42% (90% CI: 34.0-50.5), with most of the survivors (73%) staying at home. Eight serious adverse events, including 4 bleedings, were reported between study days 2 and 5, in 5 patients (5.2%) and led to death in 2 patients (2.1%). CONCLUSION: Despite a higher severity of illness at baseline, this phase IV open-label long-term study in Belgian ICUs shows consistent results with previous studies with DAA. Importantly, most of the surviving patients at day 90 were staying at home.