Diagnostic Challenges in Sepsis

Improving sepsis classification performance with artificial intelligence algorithms: A comprehensive overview of healthcare applications

PURPOSE

This study investigates the potential of machine learning (ML) algorithms in improving sepsis diagnosis and prediction, focusing on their relevance in healthcare decision-making. The primary objective is to contribute to healthcare decision-making by evaluating the performance of various supervised and unsupervised models.

MATERIALS AND METHODS

Through an extensive literature review, optimal ML models used in sepsis research were identified. Diverse datasets from relevant sources were employed, and rigorous evaluation metrics, including accuracy, specificity, and sensitivity, were applied. Innovative techniques were introduced, such as a Stacked Blended Ensemble Model and Skopt Optimization with Blended Ensemble, incorporating Bayesian optimization for hyperparameter tuning.

RESULTS

ML algorithms demonstrate efficacy in sepsis diagnosis, presenting an improved balance between specificity and sensitivity, critical for effective clinical decision-making. Classifier ensemble models show enhanced accuracy and efficiency, with novel optimization techniques contributing to improved adaptability.

CONCLUSION

The study emphasizes the potential benefits of ML algorithms in sepsis management, advocating for ongoing research to optimize performance and ensure ethical utilization in healthcare decision-making. Ethical considerations, interpretability, and transparency are crucial factors in implementing these algorithms in clinical practice.

Artificial Intelligence to Close the Gap between Pharmacokinetic/Pharmacodynamic Targets and Clinical Outcomes in Critically Ill Patients: A Narrative Review on Beta Lactams

Antimicrobial dosing can be a complex challenge. Although a solid rationale exists for a link between antibiotic exposure and outcome, conflicting data suggest a poor correlation between pharmacokinetic/pharmacodynamic targets and infection control. Different reasons may lead to this discrepancy: poor tissue penetration by β-lactams due to inflammation and inadequate tissue perfusion; different bacterial response to antibiotics and biofilms; heterogeneity of the host's immune response and drug metabolism; bacterial tolerance and acquisition of resistance during therapy. Consequently, either a fixed dose of antibiotics or a fixed target concentration may be doomed to fail. The role of biomarkers in understanding and monitoring host response to infection is also incompletely defined. Nowadays, with the ever-growing stream of data collected in hospitals, utilizing the most efficient analytical tools may lead to better personalization of therapy. The rise of artificial intelligence and machine learning has allowed large amounts of data to be rapidly accessed and analyzed. These unsupervised learning models can apprehend the data structure and identify homogeneous subgroups, facilitating the individualization of medical interventions. This review aims to discuss the challenges of β-lactam dosing, focusing on its pharmacodynamics and the new challenges and opportunities arising from integrating machine learning algorithms to personalize patient treatment.

The clinical application value of multi-site mNGS detection of patients with sepsis in intensive care units

BACKGROUND

Sepsis remains a leading cause of mortality in intensive care units, and rapid and accurate pathogen detection is crucial for effective treatment. This study evaluated the clinical application of multi-site metagenomic next-generation sequencing (mNGS) for the diagnosis of sepsis, comparing its performance against conventional methods.

METHODS

A retrospective analysis was conducted on 69 patients with sepsis consecutively admitted to the Department of Intensive Care Medicine, Meizhou People's Hospital. Samples of peripheral blood and infection sites were collected for mNGS and conventional method tests to compare the positive rate of mNGS and traditional pathogen detection methods and the distribution of pathogens. The methods used in this study included a comprehensive analysis of pathogen consistency between peripheral blood and infection site samples. Additionally, the correlation between the pathogens detected and clinical outcomes was investigated.

RESULTS

Of the patients with sepsis, 57.97% experienced dyspnea, and 65.2% had underlying diseases, with hypertension being the most common. mNGS demonstrated a significantly higher pathogen detection rate (88%) compared to the conventional method tests (26%). The pathogen consistency rate was 60% between plasma and bronchoalveolar lavage fluid samples, and that of plasma and local body fluid samples was 63%. The most frequently detected pathogens were gram-negative bacteria, and Klebsiella pneumonia. There were no significant differences in the clinical features between the pathogens.

CONCLUSION

mNGS is significantly superior to conventional methods in pathogen detection. There was a notable high pathogen consistency detection between blood and local body fluid samples, supporting the clinical relevance of mNGS. This study highlights the superiority of mNGS in detecting a broad spectrum of pathogens quickly and accurately.

TRIAL REGISTRATION

Not applicable.

Improving Antibiotic Use for Ventilator-Associated Pneumonia Through Diagnostic Stewardship: A Proof-of-Concept Mixed Methods Study

Background

Overtreatment of ventilator-associated pneumonia (VAP) in the intensive care unit is driven by positive respiratory tract cultures in the absence of a clinical picture of pneumonia. We evaluated the potential for diagnostic stewardship at the respiratory culture reporting step.

Methods

In this mixed methods study, we conducted a baseline evaluation of lower respiratory tract (LRT) culture appropriateness and antibiotic prescribing, followed by a nonrandomized intervention in 2 adult intensive care units. The intervention was a comment in the report to indicate potential colonization instead of organism identification when LRT cultures were inappropriate-that is, not meeting criteria for pneumonia as adjudicated by a physician using a standard algorithm.

Results

At baseline, among 66 inappropriate LRT cultures, antibiotic treatment for VAP was more frequent with identification of potential pathogens in the index culture when compared with no growth/normal flora (16/35 [46%] vs 7/31 [23%], P = .049). In the intervention period, 28 inappropriate cultures with growth of potential pathogens underwent report modification. The proportion of episodes for which antibiotic therapy for VAP was completed was significantly lower in the intervention group vs the baseline group (5/28 [18%] vs 16/35 [46%], P = .02).

Conclusions

Diagnostic stewardship for VAP could be facilitated by modification of LRT culture reporting guided by clinical features of pneumonia.

Comparative Analysis of C-reactive Protein and Procalcitonin as Biomarkers for Prognostic Assessment in Pediatric Sepsis

Background Sepsis poses a critical medical challenge due to its profound systemic inflammatory response, which frequently results in organ dysfunction and high mortality rates, especially in pediatric patients. The condition requires prompt recognition and aggressive management to mitigate its severe outcomes. Methods This prospective study enrolled 248 pediatric patients admitted with sepsis to the pediatric intensive care unit (PICU) at our tertiary care center. Patients were randomly assigned to either the C-reactive protein (CRP) or procalcitonin (PCT) groups, with biomarker levels measured upon admission (hour zero) and again at 72 hours post-admission. Clinical parameters such as the need for ionotropic support, use of steroids, incidence of acute kidney injury (AKI), requirement for invasive ventilation, patient outcomes, and changes in antibiotic management were assessed based on these biomarker levels. Results Procalcitonin-positive sepsis cases demonstrated notable clinical severity compared to their C-reactive protein counterparts, showing significantly lower systolic blood pressure (p = 0.012), heightened need for ionotropic support (p < 0.0001), and more pronounced liver and renal dysfunction as indicated by elevated serum bilirubin (p = 0.001) and creatinine levels (p = 0.0058). The incidence of AKI was also higher in procalcitonin-positive cases. Despite these severe clinical parameters, there were no significant differences in the length of the PICU stay or in patient outcomes concerning discharge and mortality rates. Procalcitonin levels effectively guided antibiotic management, resulting in therapy adjustments in a substantial proportion of cases, with 67 (54%) experiencing downgrades and 33 (27%) requiring upgrades based on procalcitonin levels measured 72 hours post-admission. Conclusion Procalcitonin proves to be a valuable biomarker in assessing the severity and management of sepsis in pediatric patients. It correlates significantly with clinical parameters such as blood pressure, the need for ionotropic support, and markers of organ dysfunction.

Assessment of hemostatic profile in neonates with necrotizing enterocolitis using Rotational Thromboelastometry (ROTEM)

BACKGROUND

This study aimed to explore the hemostatic profile of neonates with necrotizing enterocolitis (NEC) using Rotational Thromboelastometry (ROTEM) and to investigate if ROTEM parameters have the capacity to play a role in the differentiation of NEC from sepsis at the disease onset.

METHODS

This observational study included 62 neonates (mean gestational age 31.6 weeks and mean birth weight 1620g) hospitalized in a neonatal intensive care unit. The neonates were categorized in three groups: neonates with NEC (Bell stage II and above), neonates with sepsis and healthy neonates and they were matched 1:1:1 with regards to gestational age, delivery mode, and sex. Clinical, laboratory data as well as measurements of ROTEM parameters at disease onset were recorded.

RESULTS

ROTEM parameters differed between neonates with NEC and neonates with sepsis, indicating that NEC results in accelerated clot formation and higher clot strength compared to sepsis. The EXTEM CFT and A10 parameters demonstrated the highest diagnostic performance for NEC in terms of discrimination between NEC and sepsis (AUC, 0.997; 95% CI: 0.991-1.000 and 0.973; 95% CI: 0.932-1.000, respectively).

CONCLUSIONS

Neonates with NEC manifested accelerated clot formation and higher clot strength compared to septic and healthy neonates, as these were expressed by ROTEM parameters.

IMPACT

This work reports data on the hemostatic profile of neonates with necrotizing enterocolitis (NEC) using Rotational Thromboelastometry (ROTEM) and the capacity of ROTEM parameters in differentiating of NEC from sepsis at the disease onset. Neonates with NEC present acceleration of coagulation and exhibit a hypercoagulable profile, as this is expressed by ROTEM parameters, in comparison to septic and healthy neonates. ROTEM parameters demonstrated a good diagnostic capacity in differentiating NEC from sepsis at the disease onset.

Evaluation of Direct Antimicrobial Susceptibility Testing from Positive Flagged Blood Cultures in Sepsis Patients

Background

Presently, many laboratories are equipped with automated system for antimicrobial susceptibility testing (AST) for minimum inhibitory concentration-based reporting which enables the clinician to choose the right antimicrobial for timely treatment of sepsis. The study aimed to assess performance of direct AST from blood culture positive broth using automated AST system for accuracy and time taken to release the report.

Materials and methods

The present study conducted in a 25-bedded ICU in North India for 12 months. Single morphotype of bacteria on gram stain from positively flagged blood culture bottles were included, which was directly identified (using an in-house protocol) with MALDI-TOF-MS from positive blood culture broths. DAST was carried out from 200 such blood culture broths and results were compared with reference AST (RAST) which was also done using VITEK-2 using overnight grown bacterial colonies as per standard protocol.

Results

Among 60 isolates of Enterobacterales, 99% categorical agreement for both E. coli and K. pneumoniae observed by two methods were tested for AST. Among non-fermenters, Pseudomonas aeruginosa showed a categorical agreement of 99.6%, as compared with Acinetobacter spp. and exotic GNBs, which showed 95-96% agreement. A significant difference of 18-24 hours was noted in time to release the report between DAST and RAST, for GNB and GPC both.

Conclusion

Direct AST from positive flagged blood culture bottles can significantly reduce the time to release the bacterial susceptibility report by up to 24 hours, at the same time maintaining the accuracy.

How to cite this article

Singh V, Agarwal J, Nath SS, Sharma A. Evaluation of Direct Antimicrobial Susceptibility Testing from Positive Flagged Blood Cultures in Sepsis Patients. Indian J Crit Care Med 2024;28(4):387-392.

Clinical value of the red blood cell distribution width to albumin ratio in the assessment of prognosis in critically ill patients with sepsis: a retrospective analysis

Background

Red blood cell (RBC) distribution width (RDW) to albumin ratio is a novel biomarker and its prognostic effect on critically ill patients with sepsis has not been extensively investigated. The objective of this study was to identify the prognostic value of the RDW to albumin ratio in these patients.

Methods

Data were extracted from the Medical Information Mart for Intensive Care III (MIMIC-III) database. A Cox proportional hazards model and restricted cubic spline model were used to determine the association of RDW to albumin ratio with mortality. Receiver operating characteristic (ROC) curves and Kaplan-Meier survival curves were applied, and the area under the curve (AUC) was used to compare the predictive value.

Results

A total of 3,969 eligible patients were enrolled. The median RDW to albumin ratio was significantly higher in non-survivors than in survivors at 30 and 90 days. Patients were divided into groups according to the RDW to albumin ratio, and the risk of 30- and 90-day mortality markedly increased in the group with a higher ratio. The relationship between the RDW to albumin ratio as a continuous variable and 30-day mortality also showed an upward trend in the restricted cubic spline. The AUC of the RDW to albumin ratio was 0.633 in discriminating 30-day mortality which was similar to that of the lactate to albumin ratio (AUC =0.617; P=0.133) and higher than that of the neutrophil percentage to albumin ratio (AUC =0.559; P<0.001).

Conclusions

The RDW to albumin ratio is a promising biomarker for assessing the prognosis of critically ill patients with sepsis. Its predictive value in determining mortality was found to be similar to that of the lactate to albumin ratio and superior to that of the neutrophil percentage to albumin ratio.

The Impact of Pathogens on Sepsis Prevalence and Outcome

Sepsis, a severe global healthcare challenge, is characterized by significant morbidity and mortality. The 2016 redefinition by the Third International Consensus Definitions Task Force emphasizes its complexity as a "life-threatening organ dysfunction caused by a dysregulated host response to infection". Bacterial pathogens, historically dominant, exhibit geographic variations, influencing healthcare strategies. The intricate dynamics of bacterial immunity involve recognizing pathogen-associated molecular patterns, triggering innate immune responses and inflammatory cascades. Dysregulation leads to immunothrombosis, disseminated intravascular coagulation, and mitochondrial dysfunction, contributing to the septic state. Viral sepsis, historically less prevalent, saw a paradigm shift during the COVID-19 pandemic, underscoring the need to understand the immunological response. Retinoic acid-inducible gene I-like receptors and Toll-like receptors play pivotal roles, and the cytokine storm in COVID-19 differs from bacterial sepsis. Latent viruses like human cytomegalovirus impact sepsis by reactivating during the immunosuppressive phases. Challenges in sepsis management include rapid pathogen identification, antibiotic resistance monitoring, and balancing therapy beyond antibiotics. This review highlights the evolving sepsis landscape, emphasizing the need for pathogen-specific therapeutic developments in a dynamic and heterogeneous clinical setting.

Identifying biomarkers deciphering sepsis from trauma-induced sterile inflammation and trauma-induced sepsis

Objective

The purpose of this study was to identify a panel of biomarkers for distinguishing early stage sepsis patients from non-infected trauma patients.

Background

Accurate differentiation between trauma-induced sterile inflammation and real infective sepsis poses a complex life-threatening medical challenge because of their common symptoms albeit diverging clinical implications, namely different therapies. The timely and accurate identification of sepsis in trauma patients is therefore vital to ensure prompt and tailored medical interventions (provision of adequate antimicrobial agents and if possible eradication of infective foci) that can ultimately lead to improved therapeutic management and patient outcome. The adequate withholding of antimicrobials in trauma patients without sepsis is also important in aspects of both patient and environmental perspective.

Methods

In this proof-of-concept study, we employed advanced technologies, including Matrix-Assisted Laser Desorption/Ionization (MALDI) and multiplex antibody arrays (MAA) to identify a panel of biomarkers distinguishing actual sepsis from trauma-induced sterile inflammation.

Results

By comparing patient groups (controls, infected and non-infected trauma and septic shock patients under mechanical ventilation) at different time points, we uncovered distinct protein patterns associated with early trauma-induced sterile inflammation on the one hand and sepsis on the other hand. SYT13 and IL1F10 emerged as potential early sepsis biomarkers, while reduced levels of A2M were indicative of both trauma-induced inflammation and sepsis conditions. Additionally, higher levels of TREM1 were associated at a later stage in trauma patients. Furthermore, enrichment analyses revealed differences in the inflammatory response between trauma-induced inflammation and sepsis, with proteins related to complement and coagulation cascades being elevated whereas proteins relevant to focal adhesion were diminished in sepsis.

Conclusions

Our findings, therefore, suggest that a combination of biomarkers is needed for the development of novel diagnostic approaches deciphering trauma-induced sterile inflammation from actual infective sepsis.

The Rapid Phenotypic Susceptibility Testing in Real-Life Experience: How the MIC Values Impact on Sepsis Fast Diagnostic Workflow

The MIC value definition faithfully reflects antimicrobial sensitivity, profoundly impacting the infection's clinical outcome. Our study aimed to evaluate the Accelerate PhenoTM System in defining the importance of fast phenotypic susceptibility data. A number of 270 monomicrobial samples simultaneously underwent standard procedures and fast protocols after a contemporary Gram stain. Finally, we provided Turn-around Time (TAT) and statistical evaluations. The fast technology required a medium value of 7 h to complete ID and AST profiles. Although there were some spectrum limitations, it revealed an optimal success rate in microbial identification directly from positive blood cultures. The Gram-negative AST reached a 98.9% agreement between the Accelerate Pheno™ System and the standard method. In addition, the Gram-positive AST gathered a 98.7% agreement comparing the same systems. The chance to rapidly provide precise MIC values is one of the last frontiers in clinical microbiology, especially in high-prevalence antimicrobial resistance areas.

Uncovering hub genes in sepsis through bioinformatics analysis

In-depth studies on the mechanisms of pathogenesis of sepsis and diagnostic biomarkers in the early stages may be the key to developing individualized and effective treatment strategies. This study aimed to identify sepsis-related hub genes and evaluate their diagnostic reliability. The gene expression profiles of GSE4607 and GSE131761 were obtained from the Gene Expression Omnibus. Differentially co-expressed genes between the sepsis and control groups were screened. Single-sample gene set enrichment analysis and gene set variation analysis were performed to investigate the biological functions of the hub genes. A receiver operating characteristic curve was used to evaluate diagnostic value. Datasets GSE154918 and GSE185263 were used as external validation datasets to verify the reliability of the hub genes. Four differentially co-expressed genes, FAM89A, FFAR3, G0S2, and FGF13, were extracted using a weighted gene co-expression network analysis and differential gene expression analysis methods. These 4 genes were upregulated in the sepsis group and were distinct from those in the controls. Moreover, the receiver operating characteristic curves of the 4 genes exhibited considerable diagnostic value in discriminating septic blood samples from those of the non-septic control group. The reliability and consistency of these 4 genes were externally validated. Single-sample gene set enrichment analysis and gene set variation analysis analyses indicated that the 4 hub genes were significantly correlated with the regulation of immunity and metabolism in sepsis. The identified FAM89A, FFAR3, G0S2, and FGF13 genes may help elucidate the molecular mechanisms underlying sepsis and drive the introduction of new biomarkers to advance diagnosis and treatment.

Novel MAGL Inhibitors Alleviate LPS-Induced Acute Kidney Injury by Inhibiting NLRP3 Inflammatory Vesicles, Modulating Intestinal Flora, Repairing the Intestinal Barrier, and Interfering with Serum Metabolism

Acute kidney injury (AKI) is a complication of a wide range of serious illnesses for which there is still no better therapeutic agent. We demonstrated that M-18C has a favorable inhibitory effect on monoacylglycerol lipase (MAGL), and several studies have demonstrated that nerve inflammation could be effectively alleviated by inhibiting MAGL, suggesting that M-18C has good anti-inflammatory activity. In this study, we investigated the effect of M-18C on LPS-induced acute kidney injury (AKI), both in vivo and in vitro, by using liquid chromatography-mass spectrometry (LC-MS), 16S rRNA gene sequencing, Western blot, and immunohistochemistry. The results showed that both in vivo and in vitro M-18C reduced the release of TNF-α and IL-1β by inhibiting the expression of NOD-like receptor thermal protein domain-associated protein 3 (NLRP3) and apoptosis-associated speck-like protein containing a CARD (ASC) protein; in addition, M-18C was able to intervene in LPS-induced AKI by ameliorating renal pathological injury, repairing the intestinal barrier, and regulating gut bacterial flora and serum metabolism. In conclusion, this study suggests that M-18C has the potential to be a new drug for the treatment of AKI.

Highly multiplexed targeted sequencing strategy for infectious disease surveillance

BACKGROUND

Global efforts to characterize diseases of poverty are hampered by lack of affordable and comprehensive detection platforms, resulting in suboptimal allocation of health care resources and inefficient disease control. Next generation sequencing (NGS) can provide accurate data and high throughput. However, shotgun and metagenome-based NGS approaches are limited by low concentrations of microbial DNA in clinical samples, requirements for tailored sample and library preparations plus extensive bioinformatics analysis. Here, we adapted molecular inversion probes (MIPs) as a cost-effective target enrichment approach to characterize microbial infections from blood samples using short-read sequencing. We designed a probe panel targeting 2 bacterial genera, 21 bacterial and 6 fungi species and 7 antimicrobial resistance markers (AMRs).

RESULTS

Our approach proved to be highly specific to detect down to 1 in a 1000 pathogen DNA targets contained in host DNA. Additionally, we were able to accurately survey pathogens and AMRs in 20 out of 24 samples previously profiled with routine blood culture for sepsis.

CONCLUSIONS

Overall, our targeted assay identifies microbial pathogens and AMRs with high specificity at high throughput, without the need for extensive sample preparation or bioinformatics analysis, simplifying its application for characterization and surveillance of infectious diseases in medium- to low- resource settings.

Evaluation of the liquid colony for identification and antimicrobial susceptibility testing directly from positive blood cultures

BACKGROUND

Sepsis represents a time-sensitive disease requiring early therapeutical intervention to avoid adverse patient outcomes. Rapid microbiological diagnosis is essential to investigate sepsis aetiological agents. The FAST™ system (Qvella, ON, Canada) provides a concentrated microbial suspension, known as a Liquid Colony™ (LC), directly from positive blood samples (PBCs) in 30-40 min to perform rapid identification (ID) and antimicrobial susceptibility testing (AST).

METHODS

Qvella's FAST™ System and FAST PBC Prep cartridges were tested on PBCs from the Policlinico Hospital of Catania during a six-month study. Two millilitres of PBC were converted into an LC for rapid ID and AST using Bruker Biotyper Sirius MALDI and BD Phoenix systems. Standard of care (SOC) methods were used as a reference, requiring 48-72 h. Agreement between the innovative technology and the standard method was calculated.

RESULTS

FAST System processing was performed on 100 monomicrobial PBCs. Median turnaround times from blood cultures flagging positive to ID and AST completion were 2 and 26 h respectively. Therefore, the LC procedure was 24 h faster than the median turnaround times for SOC methods. 100% ID identification concordance was observed across 48 Gram-negative bacteria, 42 Gram-positive bacteria and 11 yeast for the genus level. 78% of Gram-negative and 95% of Gram-positive bacteria were resistant to ≥ 2 antimicrobial agents, including 45% (15/33) carbapenem-resistant enteric Gram-negative bacteria and 90% (28/31) oxacillin-resistant staphylococci. An AST essential agreement of 100% was observed due to the absence of MIC discrepancies > 1-fold dilution. Categorical errors were not observed due to the absence of MIC categorization discordances. Yeast AST was not performed.

CONCLUSIONS

The Qvella FAST System produces an LC that reliably reflects the MALDI spectra and phenotypic antimicrobial susceptibility profile of microbial cells growing in the blood culture. Timely processing of PBCs with the Qvella FAST System enables sepsis diagnostic confirmation 1 day sooner than the standard methods. In line with these results, it is vital now to focus attention on establishing best practices for incorporating this strategic tool into the clinical microbiology laboratory workflow.

IgG N-glycan Signatures as Potential Diagnostic and Prognostic Biomarkers

IgG N-glycans are an emerging source of disease-specific biomarkers. Over the last decade, the continued development of glycomic databases and the evolution of glyco-analytic methods have resulted in increased throughput, resolution, and sensitivity. IgG N-glycans promote adaptive immune responses through antibody-dependent cellular cytotoxicity (ADCC) and complement activation to combat infection or cancer and promote autoimmunity. In addition to the functional assays, researchers are examining the ability of protein-specific glycosylation to serve as biomarkers of disease. This literature review demonstrates that IgG N-glycans can discriminate between healthy controls, autoimmune disease, infectious disease, and cancer with high sensitivity. The literature also indicates that the IgG glycosylation patterns vary across disease state, thereby supporting their role as specific biomarkers. In addition, IgG N-glycans can be collected longitudinally from patients to track treatment responses or predict disease reoccurrence. This review focuses on IgG N-glycan profiles applied as diagnostics, cohort discriminators, and prognostics. Recent successes, remaining challenges, and upcoming approaches are critically discussed.

Predictors of outcomes in emergency department patients with suspected infections and without fulfillment of the sepsis criteria

BACKGROUND

Data on patient characteristics and determinants of serious outcomes for acutely admitted patients with infections who do not fulfill the sepsis criteria are sparse. The study aimed to characterize acutely admitted emergency department (ED) patients with infections and a composite outcome of in-hospital mortality or transfer to the intensive care unit without fulfilling the criteria for sepsis and to examine predictors of the composite outcome.

METHODS

This was a secondary analysis of data from a prospective observational study of patients with suspected bacterial infection admitted to the ED between October 1, 2017 and March 31, 2018. A National Early Warning Score 2 (NEWS2) ≥ 5 within the first 4 h in the ED was assumed to represent a sepsis-like condition with a high risk for the composite endpoint. Patients who achieved the composite outcome were grouped according to fulfillment of the NEWS2 ≥ 5 criteria. We used logistic regression analysis to estimate the unadjusted and adjusted odds ratio (OR) for the composite endpoint among patients with either NEWS2  < 5 (NEWS2-) or NEWS2  ≥ 5 (NEWS2+).

RESULTS

A total of 2055 patients with a median age of 73 years were included. Of these, 198 (9.6%) achieved the composite endpoint, including 59 (29.8%) NEWS2- and 139 (70.2%) NEWS2+ patients, respectively. Diabetes (OR 2.23;1.23-4.0), a Sequential Organ Failure Assessment (SOFA) score ≥ 2 (OR 2.57;1.37-4.79), and a Do-not-attempt-cardiopulmonary-resuscitation order (DNACPR) on admission (OR 3.70;1.75-7.79) were independent predictive variables for the composite endpoint in NEWS2- patients (goodness-of-fit test P = 0.291; area under the receiver operating characteristic curve for the model (AUROC) = 0.72). The regression model for NEWS2+ patients revealed that a SOFA score ≥ 2 (OR 2.79; 1.59-4.91), hypothermia (OR 2.48;1.30-4.75), and DNACPR order on admission were predictive variables for the composite endpoint (goodness-of-fit test P = 0.62; AUROC for the model = 0.70).

CONCLUSION

Approximately one-third of the patients with infections and serious outcomes during hospitalization did not meet the NEWS2 threshold for likely sepsis. Our study identified factors with independent predictive values for the development of serious outcomes that should be tested in future prediction models.

Longitudinal change of serum inter-alpha-trypsin inhibitor heavy chain H4, and its correlation with inflammation, multiorgan injury, and death risk in sepsis

BACKGROUND

Inter-alpha-trypsin inhibitor heavy chain H4 (ITIH4) inhibits infection-induced inflammation and multiorgan injury through several methods. The present study aimed to estimate the association of serum ITIH4 with inflammatory cytokines, multiorgan injury, and death risk in sepsis patients.

METHODS

Serum samples were collected to detect ITIH4 by enzyme-linked immunosorbent assay in 127 sepsis patients at admission (baseline), day (D)1, D3, and D7 after admission, as well as in 30 healthy controls (HCs). Additionally, 28-day mortality was recorded in sepsis patients.

RESULTS

ITIH4 was reduced in sepsis patients versus HCs (median [interquartile range]: 147.9 [78.2-208.8] vs. 318.8 [237.2-511.4] ng/ml) (p < 0.001). In sepsis patients, ITIH4 was associated with the absence of cardiovascular and cerebrovascular disease history (p = 0.021). Additionally, ITIH4 was negatively correlated with tumor necrosis factor-α (p < 0.001), interleukin (IL)-1β (p < 0.001), IL-6 (p = 0.019), IL-17A (p = 0.002), and C-reactive protein (p = 0.001), but positively related to IL-10 (p = 0.007). Moreover, ITIH4 was also inversely associated with Acute Physiology and Chronic Health Evaluation II score (p = 0.002), Sequential Organ Failure Assessment (SOFA) score (p < 0.001), SOFA-respiratory system score (p = 0.023), and SOFA-renal system score (p = 0.007). Interestingly, ITIH4 gradually increased from baseline to D7 (p < 0.001); besides, ITIH4 at baseline (p = 0.009), D1 (p = 0.002), D3 (p < 0.001), and D7 (p = 0.015) were all decreased in sepsis deaths versus sepsis survivors.

CONCLUSION

Serum ITIH4 is raised from baseline to D7 after disease onset, and it reflects the reduction of systemic inflammation, disease severity, and 28-day mortality for sepsis. However, further verification is required.

Sepsis biomarkers and diagnostic tools with a focus on machine learning

Over the last years, there have been advances in the use of data-driven techniques to improve the definition, early recognition, subtypes characterisation, prognostication and treatment personalisation of sepsis. Some of those involve the discovery or evaluation of biomarkers or digital signatures of sepsis or sepsis sub-phenotypes. It is hoped that their identification may improve timeliness and accuracy of diagnosis, suggest physiological pathways and therapeutic targets, inform targeted recruitment into clinical trials, and optimise clinical management. Given the complexities of the sepsis response, panels of biomarkers or models combining biomarkers and clinical data are necessary, as well as specific data analysis methods, which broadly fall under the scope of machine learning. This narrative review gives a brief overview of the main machine learning techniques (mainly in the realms of supervised and unsupervised methods) and published applications that have been used to create sepsis diagnostic tools and identify biomarkers.

Evaluating the potency of blood long noncoding RNA PVT1 as candidate biomarker reflecting inflammation, multiple organ dysfunction, and mortality risk in sepsis patients

OBJECTIVE

Long noncoding RNA plasmacytoma variant translocation 1 (lnc-PVT1) promotes septic inflammation and organ injuries via multiple ways, while its clinical engagement in sepsis management is indistinct. This study aimed to investigate its relationship with inflammation, multiple organ dysfunction, and mortality risk in sepsis patients.

METHODS

Sepsis patients and age-/gender-matched healthy controls were enrolled; their lnc-PVT1 expression in plasma were detected by RT-qPCR. For sepsis patients only, the inflammatory cytokine levels (tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-17A) in plasma were detected by ELISA. According to the survival data during 28-day follow-up, sepsis patients were divided into sepsis survivors and sepsis deaths.

RESULTS

Lnc-PVT1 expression was increased in sepsis patients (N = 157) compared with healthy controls (N = 80) (p < 0.001). In sepsis patients, lnc-PVT1 was linked with higher acute physiology and chronic health evaluation II (APACHEII) score (p = 0.001), total sequential organ failure assessment (SOFA) score, and its most subitems (SOFA-respiratory system, SOFA-coagulation, SOFA-liver, SOFA-cardiovascular system, and SOFA-renal system scores) (all p < 0.01), but not SOFA-nervous system score (p = 0.091); it did not relate to primary infection sites either (p = 0.204). Furthermore, lnc-PVT1 correlated with increased C-reactive protein, TNF-α, IL-1β, and IL-17 in sepsis patients (all p < 0.01). Additionally, lnc-PVT1 expression was higher in sepsis deaths than that in sepsis survivors (p < 0.001), following receiver-operating characteristic curve disclosed that lnc-PVT1 predicted 28-day septic mortality risk (area under the curve: 0.789, 95% confidence interval: 0.702-0.875).

CONCLUSION

Circulating lnc-PVT1 exhibits the potential as a biomarker in sepsis patients to inform inflammation, multiple organ dysfunction, and mortality risk.