Integrated clustering of multiple immune marker trajectories reveals different immunotypes in severely injured patients.

BACKGROUND: The immune response of critically ill patients, such as those with sepsis, severe trauma, or major surgery, is heterogeneous and dynamic, but its characterization and impact on outcomes are poorly understood. Until now, the primary challenge in advancing our understanding of the disease has been to concurrently address both multiparametric and temporal aspects. METHODS: We used a clustering method to identify distinct groups of patients, based on various immune marker trajectories during the first week after admission to ICU. In 339 severely injured patients, we initially longitudinally clustered common biomarkers (both soluble and cellular parameters), whose variations are well-established during the immunosuppressive phase of sepsis. We then applied this multi-trajectory clustering using markers composed of whole blood immune-related mRNA. RESULTS: We found that both sets of markers revealed two immunotypes, one of which was associated with worse outcomes, such as increased risk of hospital-acquired infection and mortality, and prolonged hospital stays. This immunotype showed signs of both hyperinflammation and immunosuppression, which persisted over time. CONCLUSION: Our study suggest that the immune system of critically ill patients can be characterized by two distinct longitudinal immunotypes, one of which included patients with a persistently dysregulated and impaired immune response. This work confirms the relevance of such methodology to stratify patients and pave the way for further studies using markers indicative of potential immunomodulatory drug targets.

A computational framework to improve cross-platform implementation of transcriptomics signatures.

The emergence of next-generation sequencing technologies and computational advances have expanded our understanding of gene expression regulation (i.e., the transcriptome). This has also led to an increased interest in using transcriptomic biomarkers to improve disease diagnosis and stratification, to assess prognosis and predict the response to treatment. Significant progress in identifying transcriptomic signatures for various clinical needs has been made, with large discovery studies accounting for challenges such as patient variability, unwanted batch effects, and data complexities; however, obstacles related to the technical aspects of cross-platform implementation still hinder the successful integration of transcriptomic technologies into standard diagnostic workflows. In this article, we discuss the challenges associated with integrating transcriptomic signatures derived using high-throughput technologies (such as RNA-sequencing) into clinical diagnostic tools using nucleic acid amplification (NAA) techniques. The novelty of the proposed approach lies in our aim to embed constraints related to cross-platform implementation in the process of signature discovery. These constraints could include technical limitations of amplification platform and chemistry, the maximal number of targets imposed by the chosen multiplexing strategy, and the genomic context of identified RNA biomarkers. Finally, we propose to build a computational framework that would integrate these constraints in combination with existing statistical and machine learning models used for signature identification. We envision that this could accelerate the integration of RNA signatures discovered by high-throughput technologies into NAA-based approaches suitable for clinical applications.

Immunosuppression at ICU admission is not associated with a higher incidence of ICU-acquired bacterial bloodstream infections: the COCONUT study.

BACKGROUND: Immunosuppression at intensive care unit (ICU) admission has been associated with a higher incidence of ICU-acquired infections, some of them related to opportunistic pathogens. However, the association of immunosuppression with the incidence, microbiology and outcomes of ICU-acquired bacterial bloodstream infections (BSI) has not been thoroughly investigated. METHODS: Retrospective single-centered cohort study in France. All adult patients hospitalized in the ICU of Lille University-affiliated hospital for > 48 h between January 1st and December 31st, 2020, were included, regardless of their immune status. Immunosuppression was defined as active cancer or hematologic malignancy, neutropenia, hematopoietic stem cell and solid organ transplants, use of steroids or immunosuppressive drugs, human immunodeficiency virus infection and genetic immune deficiency. The primary objective was to compare the 28-day cumulative incidence of ICU-acquired bacterial BSI between immunocompromised and non-immunocompromised patients. Secondary objectives were to assess the microbiology and outcomes of ICU-acquired bacterial BSI in the two groups. RESULTS: A total of 1313 patients (66.9% males, median age 62 years) were included. Among them, 271 (20.6%) were immunocompromised at ICU admission. Severity scores at admission, the use of invasive devices and antibiotic exposure during ICU stay were comparable between groups. Both prior to and after adjustment for pre-specified baseline confounders, the 28-day cumulative incidence of ICU-acquired bacterial BSI was not statistically different between immunocompromised and non-immunocompromised patients. The distribution of bacteria was comparable between groups, with a majority of Gram-negative bacilli (~ 64.1%). The proportion of multidrug-resistant bacteria was also similar between groups. Occurrence of ICU-acquired bacterial BSI was associated with a longer ICU length-of-stay and a longer duration of invasive mechanical ventilation, with no significant association with mortality. Immune status did not modify the association between occurrence of ICU-acquired bacterial BSI and these outcomes. CONCLUSION: The 28-day cumulative incidence of ICU-acquired bacterial BSI was not statistically different between patients with and without immunosuppression at ICU admission.

ICU-acquired infections in immunocompromised patients.

Immunocompromised patients account for an increasing proportion of the typical intensive care unit (ICU) case-mix. Because of the increased availability of new drugs for cancer and auto-immune diseases, and improvement in the care of the most severely immunocompromised ICU patients (including those with hematologic malignancies), critically ill immunocompromised patients form a highly heterogeneous patient population. Furthermore, a large number of ICU patients with no apparent immunosuppression also harbor underlying conditions altering their immune response, or develop ICU-acquired immune deficiencies as a result of sepsis, trauma or major surgery. While infections are associated with significant morbidity and mortality in immunocompromised critically ill patients, little specific data are available on the incidence, microbiology, management and outcomes of ICU-acquired infections in this population. As a result, immunocompromised patients are usually excluded from trials and guidelines on the management of ICU-acquired infections. The most common ICU-acquired infections in immunocompromised patients are ventilator-associated lower respiratory tract infections (which include ventilator-associated pneumonia and tracheobronchitis) and bloodstream infections. Recently, several large observational studies have shed light on some of the epidemiological specificities of these infections-as well as on the dynamics of colonization and infection with multidrug-resistant bacteria-in these patients, and these will be discussed in this review. Immunocompromised patients are also at higher risk than non-immunocompromised hosts of fungal and viral infections, and the diagnostic and therapeutic management of these infections will be covered. Finally, we will suggest some important areas of future investigation.

Smart-Plexer: a breakthrough workflow for hybrid development of multiplex PCR assays.

Developing multiplex PCR assays requires extensive experimental testing, the number of which exponentially increases by the number of multiplexed targets. Dedicated efforts must be devoted to the design of optimal multiplex assays ensuring specific and sensitive identification of multiple analytes in a single well reaction. Inspired by data-driven approaches, we reinvent the process of developing and designing multiplex assays using a hybrid, simple workflow, named Smart-Plexer, which couples empirical testing of singleplex assays and computer simulation to develop optimised multiplex combinations. The Smart-Plexer analyses kinetic inter-target distances between amplification curves to generate optimal multiplex PCR primer sets for accurate multi-pathogen identification. In this study, the Smart-Plexer method is applied and evaluated for seven respiratory infection target detection using an optimised multiplexed PCR assay. Single-channel multiplex assays, together with the recently published data-driven methodology, Amplification Curve Analysis (ACA), were demonstrated to be capable of classifying the presence of desired targets in a single test for seven common respiratory infection pathogens.

ICU-Acquired Colonization and Infection Related to Multidrug-Resistant Bacteria in COVID-19 Patients: A Narrative Review.

A large proportion of ICU-acquired infections are related to multidrug-resistant bacteria (MDR). Infections caused by these bacteria are associated with increased mortality, and prolonged duration of mechanical ventilation and ICU stay. The aim of this narrative review is to report on the association between COVID-19 and ICU-acquired colonization or infection related to MDR bacteria. Although a huge amount of literature is available on COVID-19 and MDR bacteria, only a few clinical trials have properly evaluated the association between them using a non-COVID-19 control group and accurate design and statistical methods. The results of these studies suggest that COVID-19 patients are at a similar risk of ICU-acquired MDR colonization compared to non-COVID-19 controls. However, a higher risk of ICU-acquired infection related to MDR bacteria has been reported in several studies, mainly ventilator-associated pneumonia and bloodstream infection. Several potential explanations could be provided for the high incidence of ICU-acquired infections related to MDR. Immunomodulatory treatments, such as corticosteroids, JAK2 inhibitors, and IL-6 receptor antagonist, might play a role in the pathogenesis of these infections. Additionally, a longer stay in the ICU was reported in COVID-19 patients, resulting in higher exposure to well-known risk factors for ICU-acquired MDR infections, such as invasive procedures and antimicrobial treatment. Another possible explanation is the surge during successive COVID-19 waves, with excessive workload and low compliance with preventive measures. Further studies should evaluate the evolution of the incidence of ICU-acquired infections related to MDR bacteria, given the change in COVID-19 patient profiles. A better understanding of the immune status of critically ill COVID-19 patients is required to move to personalized treatment and reduce the risk of ICU-acquired infections. The role of specific preventive measures, such as targeted immunomodulation, should be investigated.

Relationship between COVID-19 and ICU-Acquired Bloodstream Infections Related to Multidrug-Resistant Bacteria.

A bloodstream infection (BSI) is a severe ICU-acquired infection. A growing proportion is caused by multidrug-resistant bacteria (MDRB). COVID-19 was reported to be associated with a high rate of secondary infections. However, there is a lack of data on the relationship between COVID-19 and the incidence of MDRB ICU-acquired BSI. The aim of this study was to evaluate the relationship between COVID-19 and ICU-acquired BSI related to MDRB. This retrospective study was conducted in a single-center ICU during a one-year period. All adult patients admitted for more than 48 h were included. The cumulative incidence of ICU-acquired BSI related to MDRB was estimated using the Kalbfleisch and Prentice method. The association of COVID-19 status with the risk of ICU-acquired BSI related to MDRB was assessed using cause-specific Cox's proportional hazard model. Among the 1320 patients included in the analysis, 497 (37.65%) had COVID-19. ICU-acquired BSI related to MDRB occurred in 50 patients (36 COVID patients (7%) and 14 non-COVID patients (1.6%)). Extended-spectrum beta-lactamase Enterobacteriacae (46%) and carbapenem-resistant Acinetobacter baumannii (30%) were the most commonly isolated MDRB. COVID-19 was significantly associated with a higher risk of MDRB ICU-acquired BSI (adjusted cHR 2.65 (1.25 to 5.59) for the whole study period). However, this relationship was only significant for the period starting at day 15 after ICU admission. ICU-acquired BSI related to MDRB was significantly associated with ICU mortality (HR (95%CI) 1.73 (1-3)), although COVID-19 had no significant impact on this association (p het 0.94). COVID-19 is significantly associated with an increased risk of ICU-acquired BSI related to MDRB, mainly during the period starting at day 15 after ICU admission.

Relationship between COVID-19 and ICU-acquired colonization and infection related to multidrug-resistant bacteria: a prospective multicenter before-after study.

PURPOSE: Patients presenting the most severe form of coronavirus disease 2019 (COVID-19) pneumonia, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), have a prolonged intensive care unit (ICU) stay and are exposed to broad-spectrum antibiotics, but the impact of COVID-19 on antimicrobial resistance is unknown. METHODS: Observational prospective before-after study in 7 ICUs in France. All consecutive patients with an ICU stay > 48 h and a confirmed SARS-CoV-2 infection were included prospectively and followed for 28 days. Patients underwent systematic screening for colonization with multidrug-resistant (MDR) bacteria upon admission and every week subsequently. COVID-19 patients were compared to a recent prospective cohort of control patients from the same ICUs. The primary objective was to investigate the association of COVID-19 with the cumulative incidence of a composite outcome including ICU-acquired colonization and/or infection related to MDR bacteria (ICU-MDR-col and ICU-MDR-inf, respectively). RESULTS: From February 27th, 2020 to June 2nd, 2021, 367 COVID-19 patients were included, and compared to 680 controls. After adjustment for prespecified baseline confounders, the cumulative incidence of ICU-MDR-col and/or ICU-MDR-inf was not significantly different between groups (adjusted sub-hazard ratio [sHR] 1.39, 95% confidence interval [CI] 0.91-2.09). When considering both outcomes separately, COVID-19 patients had a higher incidence of ICU-MDR-inf than controls (adjusted sHR 2.50, 95% CI 1.90-3.28), but the incidence of ICU-MDR-col was not significantly different between groups (adjusted sHR 1.27, 95% CI 0.85-1.88). CONCLUSION: COVID-19 patients had an increased incidence of ICU-MDR-inf compared to controls, but the difference was not significant when considering a composite outcome including ICU-MDR-col and/or ICU-MDR-inf.

Deep Domain Adaptation Enhances Amplification Curve Analysis for Single-Channel Multiplexing in Real-Time PCR.

Data-driven approaches for molecular diagnostics are emerging as an alternative to perform an accurate and inexpensive multi-pathogen detection. A novel technique called Amplification Curve Analysis (ACA) has been recently developed by coupling machine learning and real-time Polymerase Chain Reaction (qPCR) to enable the simultaneous detection of multiple targets in a single reaction well. However, target classification purely relying on the amplification curve shapes faces several challenges, such as distribution discrepancies between different data sources (i.e., training vs testing). Optimisation of computational models is required to achieve higher performance of ACA classification in multiplex qPCR through the reduction of those discrepancies. Here, we proposed a novel transformer-based conditional domain adversarial network (T-CDAN) to eliminate data distribution differences between the source domain (synthetic DNA data) and the target domain (clinical isolate data). The labelled training data from the source domain and unlabelled testing data from the target domain are fed into the T-CDAN, which learns both domains' information simultaneously. After mapping the inputs into a domain-irrelevant space, T-CDAN removes the feature distribution differences and provides a clearer decision boundary for the classifier, resulting in a more accurate pathogen identification. Evaluation of 198 clinical isolates containing three types of carbapenem-resistant genes (blaNDM, blaIMP and blaOXA-48) illustrates a curve-level accuracy of 93.1% and a sample-level accuracy of 97.0% using T-CDAN, showing an accuracy improvement of 20.9% and 4.9% respectively. This research emphasises the importance of deep domain adaptation to enable high-level multiplexing in a single qPCR reaction, providing a solid approach to extend qPCR instruments' capabilities in real-world clinical applications.

Next-generation molecular diagnostics: Leveraging digital technologies to enhance multiplexing in real-time PCR.

Real-time polymerase chain reaction (qPCR) enables accurate detection and quantification of nucleic acids and has become a fundamental tool in biological sciences, bioengineering and medicine. By combining multiple primer sets in one reaction, it is possible to detect several DNA or RNA targets simultaneously, a process called multiplex PCR (mPCR) which is key to attaining optimal throughput, cost-effectiveness and efficiency in molecular diagnostics, particularly in infectious diseases. Multiple solutions have been devised to increase multiplexing in qPCR, including single-well techniques, using target-specific fluorescent oligonucleotide probes, and spatial multiplexing, where segregation of the sample enables parallel amplification of multiple targets. However, these solutions are mostly limited to three or four targets, or highly sophisticated and expensive instrumentation. There is a need for innovations that will push forward the multiplexing field in qPCR, enabling for a next generation of diagnostic tools which could accommodate high throughput in an affordable manner. To this end, the use of machine learning (ML) algorithms (data-driven solutions) has recently emerged to leverage information contained in amplification and melting curves (AC and MC, respectively) - two of the most standard bio-signals emitted during qPCR - for accurate classification of multiple nucleic acid targets in a single reaction. Therefore, this review aims to demonstrate and illustrate that data-driven solutions can be successfully coupled with state-of-the-art and common qPCR platforms using a variety of amplification chemistries to enhance multiplexing in qPCR. Further, because both ACs and MCs can be predicted from sequence data using thermodynamic databases, it has also become possible to use computer simulation to rationalize and optimize the design of mPCR assays where target detection is supported by data-driven technologies. Thus, this review also discusses recent work converging towards the development of an end-to-end framework where knowledge-based and data-driven software solutions are integrated to streamline assay design, and increase the accuracy of target detection and quantification in the multiplex setting. We envision that concerted efforts by academic and industry scientists will help advance these technologies, to a point where they become mature and robust enough to bring about major improvements in the detection of nucleic acids across many fields.

Immune Profiling Panel Gene Set Identifies Critically Ill Patients With Low Monocyte Human Leukocyte Antigen-DR Expression: Preliminary Results From the REAnimation Low Immune Status Marker (REALISM) Study.

OBJECTIVES: There is a crucial unmet need for biomarker-guided diagnostic and prognostic enrichment in clinical trials evaluating immune modulating therapies in critically ill patients. Low monocyte expression of human leukocyte antigen-DR (mHLA-DR), considered as a reference surrogate to identify immunosuppressed patients, has been proposed for patient stratification in immunostimulation approaches. However, its widespread use in clinic has been somewhat hampered by technical constraints inherent to flow cytometry technology. The objective of the present study was to evaluate the ability of a prototype multiplex polymerase chain reaction tool (immune profiling panel [IPP]) to identify immunosuppressed ICU patients characterized by a low mHLA-DR expression. DESIGN: Retrospective observational cohort study. SETTING: Adult ICU in a University Hospital, Lyon, France. PATIENTS: Critically ill patients with various etiologies enrolled in the REAnimation Low Immune Status Marker study (NCT02638779). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: mHLA-DR and IPP data were obtained from 1,731 blood samples collected from critically ill patients with various etiologies and healthy volunteers. A partial least square regression model combining the expression levels of IPP markers was trained and used for the identification of samples from patients presenting with evidence of immunosuppression, defined here as mHLADR less than 8,000 antibodies bound per cell (AB/C). The IPP gene set had an area under the receiver operating characteristic curve (AUC) of 0.86 (95% CI 0.83-0.89) for the identification of immunosuppressed patients. In addition, when applied to the 123 patients still in the ICU at days 5-7 after admission, IPP similarly enriched the number of patients with ICU-acquired infections in the immunosuppressed group (26%), in comparison with low mHLA-DR (22%). CONCLUSIONS: This study reports on the potential of the IPP gene set to identify ICU patients presenting with mHLA-DR less than 8,000 AB/C. Upon further optimization and validation, this molecular tool may help in the stratification of patients that could benefit from immunostimulation in the context of personalized medicine.

Ventilator-Associated Pneumonia in Immunosuppressed Patients.

Immunocompromised patients-including patients with cancer, hematological malignancies, solid organ transplants and individuals receiving immunosuppressive therapies for autoimmune diseases-account for an increasing proportion of critically-ill patients. While their prognosis has improved markedly in the last decades, they remain at increased risk of healthcare- and intensive care unit (ICU)-acquired infections. The most frequent of these are ventilator-associated lower respiratory tract infections (VA-LTRI), which include ventilator-associated pneumonia (VAP) and tracheobronchitis (VAT). Recent studies have shed light on some of the specific features of VAP and VAT in immunocompromised patients, which is the subject of this narrative review. Contrary to previous belief, the incidence of VAP and VAT might actually be lower in immunocompromised than non-immunocompromised patients. Further, the relationship between immunosuppression and the incidence of VAP and VAT related to multidrug-resistant (MDR) bacteria has also been challenged recently. Etiological diagnosis is essential to select the most appropriate treatment, and the role of invasive sampling, specifically bronchoscopy with bronchoalveolar lavage, as well as new molecular syndromic diagnostic tools will be discussed. While bacteria-especially gram negative bacteria-are the most commonly isolated pathogens in VAP and VAT, several opportunistic pathogens are a special concern among immunocompromised patients, and must be included in the diagnostic workup. Finally, the impact of immunosuppression on VAP and VAT outcomes will be examined in view of recent papers using improved statistical methodologies and treatment options-more specifically empirical antibiotic regimens-will be discussed in light of recent findings on the epidemiology of MDR bacteria in this population.

Epidemiology, clinical presentation, and outcomes of 620 patients with eosinophilia in the intensive care unit.

PURPOSE: Although eosinophil-induced manifestations can be life-threatening, studies focusing on the epidemiology and clinical manifestations of eosinophilia in the intensive care unit (ICU) are lacking. METHODS: A retrospective, national, multicenter (14 centers) cohort study over 6 years of adult patients who presented with eosinophilia ≥ 1 × 109/L on two blood samples performed from the day before admission to the last day of an ICU stay. RESULTS: 620 patients (0.9% of all ICU hospitalizations) were included: 40% with early eosinophilia (within the first 24 h of ICU admission, ICU-Eo1 group) and 56% with delayed (> 24 h after ICU admission, ICU-Eo2 group) eosinophilia. In ICU-Eo1, eosinophilia was mostly due to respiratory (14.9%) and hematological (25.8%) conditions, frequently symptomatic (58.1%, mainly respiratory and cardiovascular manifestations) requiring systemic corticosteroids in 32.2% of cases. In ICU-Eo2, eosinophil-related organ involvement was rare (25%), and eosinophilia was mostly drug-induced (46.8%). Survival rates at day 60 (D60) after ICU admission were 21.4% and 17.2% (p = 0.219) in ICU-Eo1 and ICU-Eo2 patients, respectively. For ICU-Eo1 patients, in multivariate analysis, risk factors for death at D60 were current immunosuppressant therapy at ICU admission, eosinophilia of onco-hematological origin and the use of vasopressors at ICU admission, whereas older age and the use of vasopressors or mechanical ventilation at the onset of eosinophilia were associated with a poorer prognosis for ICU-Eo2 patients. CONCLUSION: Eosinophilia ≥ 1 × 109/L is not uncommon in the ICU. According to the timing of eosinophilia, two subsets of patients requiring different etiological workups and management can be distinguished.

Relationship between immunosuppression and intensive care unit-acquired colonization and infection related to multidrug-resistant bacteria: a prospective multicenter cohort study.

PURPOSE: The impact of immunosuppression on intensive care unit (ICU)-acquired colonization and infection related to multidrug-resistant (MDR) bacteria (ICU-MDR-col and ICU-MDR-inf, respectively) is unknown. METHODS: We carried out an observational prospective cohort study in 8 ICUs in France (all with single-bed rooms and similar organizational characteristics). All consecutive patients with an ICU stay > 48 h were included, regardless of immune status, and followed for 28 days. Patients underwent systematic screening for colonization with MDR bacteria upon admission and every week subsequently. Immunosuppression was defined as active cancer or hematologic malignancy, neutropenia, solid-organ transplant, use of steroids or immunosuppressive drugs, human immunodeficiency virus infection and genetic. The primary endpoint was the incidence rate of a composite outcome including ICU-MDR-col and/or ICU-MDR-inf. RESULTS: 750 patients (65.9% males, median age 65 years) were included, among whom 264 (35.2%) were immunocompromised. Reasons for ICU admission, severity scores and exposure to invasive devices and antibiotics during ICU stay were comparable between groups. After adjustment for center and pre-specified baseline confounders, immunocompromised patients had a lower incidence rate of ICU-MDR-col and/or ICU-MDR-inf (adjusted incidence ratio 0.68, 95% CI 0.52-0.91). When considered separately, the difference was significant for ICU-MDR-col, but not for ICU-MDR-inf. The distribution of MDR bacteria was comparable between groups, with a majority of Enterobacteriacae resistant to third-generation cephalosporins (~ 74%). CONCLUSION: Immunocompromised patients had a significantly lower incidence rate of a composite outcome including ICU-MDR-col and/or ICU-MDR-inf. This finding points to the role of contact precautions and isolation measures, and could have important implications on antibiotic stewardship in this population.

A 9-mRNA signature measured from whole blood by a prototype PCR panel predicts 28-day mortality upon admission of critically ill COVID-19 patients.

Immune responses affiliated with COVID-19 severity have been characterized and associated with deleterious outcomes. These approaches were mainly based on research tools not usable in routine clinical practice at the bedside. We observed that a multiplex transcriptomic panel prototype termed Immune Profiling Panel (IPP) could capture the dysregulation of immune responses of ICU COVID-19 patients at admission. Nine transcripts were associated with mortality in univariate analysis and this 9-mRNA signature remained significantly associated with mortality in a multivariate analysis that included age, SOFA and Charlson scores. Using a machine learning model with these 9 mRNA, we could predict the 28-day survival status with an Area Under the Receiver Operating Curve (AUROC) of 0.764. Interestingly, adding patients' age to the model resulted in increased performance to predict the 28-day mortality (AUROC reaching 0.839). This prototype IPP demonstrated that such a tool, upon clinical/analytical validation and clearance by regulatory agencies could be used in clinical routine settings to quickly identify patients with higher risk of death requiring thus early aggressive intensive care.

Adaptive Filtering Framework to Remove Nonspecific and Low-Efficiency Reactions in Multiplex Digital PCR Based on Sigmoidal Trends.

Real-time digital polymerase chain reaction (qdPCR) coupled with machine learning (ML) methods has shown the potential to unlock scientific breakthroughs, particularly in the field of molecular diagnostics for infectious diseases. One promising application of this emerging field explores single fluorescent channel PCR multiplex by extracting target-specific kinetic and thermodynamic information contained in amplification curves, also known as data-driven multiplexing. However, accurate target classification is compromised by the presence of undesired amplification events and not ideal reaction conditions. Therefore, here, we proposed a novel framework to identify and filter out nonspecific and low-efficient reactions from qdPCR data using outlier detection algorithms purely based on sigmoidal trends of amplification curves. As a proof-of-concept, this framework is implemented to improve the classification performance of the recently reported data-driven multiplexing method called amplification curve analysis (ACA), using available published data where the ACA is demonstrated to screen carbapenemase-producing organisms in clinical isolates. Furthermore, we developed a novel strategy, named adaptive mapping filter (AMF), to adjust the percentage of outliers removed according to the number of positive counts in qdPCR. From an overall total of 152,000 amplification events, 116,222 positive amplification reactions were evaluated before and after filtering by comparing against melting peak distribution, proving that abnormal amplification curves (outliers) are linked to shifted melting distribution or decreased PCR efficiency. The ACA was applied to assess classification performance before and after AMF, showing an improved sensitivity of 1.2% when using inliers compared to a decrement of 19.6% when using outliers (p-value < 0.0001), removing 53.5% of all wrong melting curves based only on the amplification shape. This work explores the correlation between the kinetics of amplification curves and the thermodynamics of melting curves, and it demonstrates that filtering out nonspecific or low-efficient reactions can significantly improve the classification accuracy for cutting-edge multiplexing methodologies.

Relationship between corticosteroid use and incidence of ventilator-associated pneumonia in COVID-19 patients: a retrospective multicenter study.

BACKGROUND: Ventilator-associated pneumonia (VAP) is common in patients with severe SARS-CoV-2 pneumonia. The aim of this ancillary analysis of the coVAPid multicenter observational retrospective study is to assess the relationship between adjuvant corticosteroid use and the incidence of VAP. METHODS: Planned ancillary analysis of a multicenter retrospective European cohort in 36 ICUs. Adult patients receiving invasive mechanical ventilation for more than 48 h for SARS-CoV-2 pneumonia were consecutively included between February and May 2020. VAP diagnosis required strict definition with clinical, radiological and quantitative microbiological confirmation. We assessed the association of VAP with corticosteroid treatment using univariate and multivariate cause-specific Cox's proportional hazard models with adjustment on pre-specified confounders. RESULTS: Among the 545 included patients, 191 (35%) received corticosteroids. The proportional hazard assumption for the effect of corticosteroids on the incidence of VAP could not be accepted, indicating that this effect varied during ICU stay. We found a non-significant lower risk of VAP for corticosteroid-treated patients during the first days in the ICU and an increased risk for longer ICU stay. By modeling the effect of corticosteroids with time-dependent coefficients, the association between corticosteroids and the incidence of VAP was not significant (overall effect p = 0.082), with time-dependent hazard ratios (95% confidence interval) of 0.47 (0.17-1.31) at day 2, 0.95 (0.63-1.42) at day 7, 1.48 (1.01-2.16) at day 14 and 1.94 (1.09-3.46) at day 21. CONCLUSIONS: No significant association was found between adjuvant corticosteroid treatment and the incidence of VAP, although a time-varying effect of corticosteroids was identified along the 28-day follow-up.