Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department: a synthesis of current evidence

Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department: a synthesis of current evidence - NCBI Bookshelf An official website of the United States government Here's how you know The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site. The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely. Log in Show account info Close Account Logged in as: username Dashboard Publications Account settings Log out Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Bookshelf Search database Books All Databases Assembly Biocollections BioProject BioSample Books ClinVar Conserved Domains dbVar Gene Genome GEO DataSets GEO Profiles GTR Identical Protein Groups MedGen MeSH NLM Catalog Nucleotide OMIM PMC Protein Protein Clusters Protein Family Models PubChem BioAssay PubChem Compound PubChem Substance PubMed SNP SRA Structure Taxonomy ToolKit ToolKitAll ToolKitBookgh Search term Search Browse Titles Advanced Help Disclaimer NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health. Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department: a synthesis of current evidence Health Technology Assessment Mari Imamura , Sinead N Duggan , Thenmalar Vadiveloo , Jamie G Cooper , Callum T Kaye , Paul Manson , Gianni Virgili , Lorna Aucott , Mike Clarke , and Miriam Brazzelli . Author Information and Affiliations Authors Mari Imamura , 1 Sinead N Duggan , 2 Thenmalar Vadiveloo , 1 Jamie G Cooper , 3 Callum T Kaye , 4,5 Paul Manson , 1 Gianni Virgili , 6 Lorna Aucott , 1 Mike Clarke , 2 and Miriam Brazzelli 1 ,* . Affiliations 1 Aberdeen Centre for Evaluation, Institute of Applied Health Sciences, University of Aberdeen, Aberdeen, UK 2 Centre for Public Health, Queen’s University Belfast, Belfast, UK 3 Emergency Department, Aberdeen Royal Infirmary, Aberdeen, UK 4 Anaesthetics and Intensive Care Medicine, NHS Grampian, Aberdeen Royal Infirmary, Aberdeen, UK 5 School of Medicine, Medical Sciences and Nutrition, University of Aberdeen, Aberdeen, UK 6 NEUROFARBA Department, University of Florence, Florence, Italy * Corresponding author ; Email: [email protected] Southampton (UK): National Institute for Health and Care Research ; 2026 Mar 25 . Copyright and Permissions Copyright © 2026 Imamura et al. This work was produced by Imamura et al . under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/ . For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited. Abstract Background: Sepsis is characterised as life-threatening organ dysfunction due to a dysregulated host response to infection. It carries high mortality and is a major public health issue globally. Among adults presenting to the emergency department with features of suspected sepsis, rapid and accurate differentiation of those at high risk for prompt delivery of key therapies and escalation of care may improve outcomes. Objectives: To conduct a comprehensive evidence synthesis assessing the clinical utility of established and novel biomarkers – used individually or in combination – for predicting the risk of death or clinical deterioration in adults with clinically suspected sepsis presenting to the emergency department. Design and methods: Search strategies were designed to identify English-language studies published between 2013 and 2023 evaluating biomarker performance to predict mortality or clinical deterioration in adults with suspected sepsis. Eligible studies had to evaluate biomarker performance in patients with clinically suspected sepsis (as defined by study authors) at emergency department presentation. Outcomes of interest included all-cause mortality, critical care admission, septic shock and organ failure. A comprehensive search was conducted across multiple databases, including MEDLINE, EMBASE, the Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trials (search date 5 July 2023). For each biomarker, used alone or in combination, risk ratios, hazard ratios, odds ratios and area under the receiver-operating characteristics curve values were extracted. Subgroup analyses were planned to compare patients in terms of age, presence of comorbidities, National Health Service or other health systems, and biomarker assessment timing. Methodological quality was evaluated using the Quality in Prognostic Factor Studies tool. Results: Of 1986 citations identified, 884 duplicates were removed, and 1102 were screened by title and abstract. Of the 430 full-text articles assessed, 377 were excluded as they did not meet the eligibility criteria, and 53 reports, related to 51 studies, were deemed suitable for inclusion. A total of 107 unique biomarkers or biomarker combinations were assessed across the included studies. However, due to limited data and clinical heterogeneity, only biomarkers commonly used in clinical practice (lactate, C-reactive protein and procalcitonin) were analysed through meta-analyses, and none effectively predicted adverse outcomes. Although novel biomarkers could not be pooled, several (Inflammatix Severity 2, mid-regional proadrenomedullin, neutrophil gelatinase-associated lipocalin, tyrosine kinase with immunoglobulin-like and epidermal growth factor-like domains 2, monocyte distribution width and neutrophil-to-lymphocyte ratio) showed potential for predicting mortality and admission to critical care. The predictive accuracy of these biomarkers improved when combined with other biomarkers and/or clinical scores. Limitations: There was considerable heterogeneity across included studies and insufficient data for meta-analysis of novel biomarkers. Conclusions: In emergency department patients with clinically suspected sepsis, lactate, C-reactive protein and procalcitonin did not effectively predict mortality or need for clinical care admission. However, some novel biomarkers showed promise for the prediction of these outcomes, particularly when combined with other biomarkers and/or commonly used clinical scores. Future research: Focusing on prospectively designed studies, there should be standardisation of biomarker thresholds, measurement timing and identification of participants to facilitate comparability across studies. Researchers should apply a clear and consistent definition of suspected sepsis. Adjusted and unadjusted analyses should be reported, accounting for critical variables. Funding: This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR159912. Plain language summary Sepsis is a common, serious and often fatal clinical condition that happens when the body responds in the wrong way to infection. People with sepsis do better when it is found early so that they can get antibiotics and fluids, and, when possible, surgical management of the infection source. However, patients often present to the emergency department with vague and non-specific symptoms and signs of possible sepsis. Therefore, identifying patients who are high risk and need intervention can be very difficult. Biomarkers are biological molecules that may increase or decrease in the blood, or other body fluids when a person is unwell. The best biomarkers would only appear when a specific disease was present, helping clinicians to diagnose the disease. Ideally, biomarker levels would be in line with the severity of the disease, which would guide treatment decisions. However, no perfect biomarker exists. When patients present to the emergency department with possible sepsis, we do not know whether biomarkers can reliably identify which patients are at risk of becoming seriously unwell or dying. We reviewed clinical studies from the past 11 years to find out if biomarkers measured in adults presenting to the emergency department with symptoms suspicious of sepsis can help recognition of those at a higher risk of becoming seriously ill or dying. Our review found that in patients with suspected sepsis, common biomarkers, such as C-reactive protein, procalcitonin and lactate, did not help to identify high-risk patients. However, newer biomarkers were promising, especially when biomarkers were combined or when biomarkers were used alongside clinical score. Nevertheless, there are still too few studies to make any recommendations for clinical practice. High-quality research is urgently needed to understand how these biomarkers work and how they could help in treating patients with possible sepsis in the emergency department. Background Description of the health problem Sepsis, which is defined as a ‘ life-threatening organ dysfunction characterised by a dysregulated host response to infection ’, 1 carries a high mortality and is a major public health issue worldwide. 2 The Global Burden of Sepsis study reported 48.9 million cases of sepsis worldwide in 2017, resulting in 11 million deaths. 3 In the UK, recent estimates suggest that there are about 245,000 cases of sepsis per year, responsible for around 48,000 deaths. 4 Sepsis is a heterogeneous disease with high variability between individuals regarding causative microorganisms, site of infection, host response, comorbidities and response to treatment. Early and accurate identification of individuals with suspected sepsis who are at high risk of deterioration to critical illness may enable earlier delivery of key therapies, prompt escalation of care and initiation of organ support measures where appropriate, with the potential to improve patient outcomes. The identification of a causative organism through microbiological testing takes time and may sometimes be unfruitful, even in the presence of strong clinical evidence to the contrary so cannot aid decisions in unwell patients with features of systemic infection who present to the emergency department (ED). Furthermore, identification and application of terminology regarding which patients do and do not have ‘sepsis’ are always somewhat subjectively and inconsistently applied. Clinical risk stratification of sepsis The 1991 International Consensus Definition Task Force (Sepsis-1) defined ‘sepsis’ as the presence of clinical suspicion of infection in association with the presence of at least two of the four systemic inflammatory response syndrome (SIRS) criteria: (1) body temperature > 38 o C or < 36 o C; (2) heart rate > 90 beats per minute; (3) respiratory rate > 20 breaths per minute or arterial partial pressure of CO 2 (PaCO 2 ) of < 32 mmHg; (4) white blood cell count < 4 × 10 9 /l or > 12 × 10 9 /l (see Appendix 1 ). Further, definitions of ‘severe sepsis’ or ‘septic shock’ require the additional presence of organ dysfunction or fluid-resistant hypotension, respectively. 5 The Task Force repeated its work in 2001 and although limitations of the existing definitions were acknowledged and detail was added to clinical recognition of sepsis, the Sepsis-2 diagnostic definitions did not change. 6 In February 2016, in response to significant advances in sepsis epidemiology and management, the third international consensus reported the Sepsis-3 definitions. 1 The SIRS criteria, recognised as inadequately sensitive or specific for a diagnosis of sepsis, were removed, as was the confusing term, ‘severe sepsis’. The Sepsis-3 definition, which remains current, focuses on the early diagnosis of sepsis through the prompt identification of organ dysfunction and quick initiation of treatments. Sepsis-3 recommends the use of the Sequential Organ Failure Assessment (SOFA) score in patients with presumed infection, noting that a 2-point increase correlates with a > 10% in-hospital mortality. 1 However, the SOFA score was developed for use in intensive care units (ICUs), and not validated for use in the ED , or other receiving areas, at the first hospital assessment. Consequently, the quick SOFA (qSOFA) score was created, comprising three clinical measurements: systolic blood pressure (SBP), conscious level and respiratory rate, which are all immediately available to a clinician at the bedside at presentation (see Appendix 1 ). 1 The qSOFA score allocates 1 point each for hypotension ( SBP ≤ 100 mmHg); altered mental status (Glasgow Coma Score < 15) and tachypnoea (respiratory rate ≥ 22/minute). When applied to patients with systemic evidence of infection, a qSOFA score of ≥ 2 quickly identifies those with organ dysfunction (and, hence, likely ‘sepsis’) who are more likely to need intensive care or to die in hospital. If such patients have persistent hypotension [mean arterial blood pressure (BP) ≤ 65 mmHg or vasopressor requirement despite adequate volume replacement] and blood lactate ≥ 2 mmol/l, they are diagnosed with septic shock, with an associated 40% in-hospital mortality. In the UK, the second iteration of the National Early Warning Score 2 (NEWS2) is widely used to provide a detailed evaluation of a patient’s physiological status. It assesses six parameters (outlined in Appendix 1 ) to generate an aggregate score ranging from 0 to 20, indicating the severity of illness. 7 However, similar to SOFA and qSOFA scoring systems, NEWS2 readily identifies patients with organ dysfunction but is not specific for determining the cause of infection and the need for antimicrobial treatment, which still requires clinical interpretation. 8 Though the diagnosis and definition of sepsis are often challenging, mortality rates associated with sepsis syndromes remain high. Initiatives like the Surviving Sepsis Campaign have highlighted the importance of promptly administering sepsis care bundles (including parenteral antibiotic, particularly in patients with suspected infection with significant physiological derangement), 9 and consideration of rapid senior clinical review regarding escalation of care. However, many patients at risk of deterioration from sepsis present with clinical symptoms and signs that do not portray this possibility and progression to critical illness may remain underappreciated and important clinical interventions, therefore, delayed. There remains considerable room for improvement in identifying which patients with suspected sepsis will most benefit from early and targeted interventions, senior clinical input and prompt escalation to critical care facilities. Prognostic biomarkers In the ED , clinical biomarkers play a key role in diagnosing, risk stratifying and determining the prognosis of diverse patient groups, for example using cardiac troponin to assess patients with suspected myocardial infarction. For ED patients with clinically suspected sepsis, integrating biomarkers into diagnostic and risk-stratification assessment could enhance early recognition of sepsis and identify those at higher risk of deterioration, enabling more targeted interventions and potentially improving clinical outcomes. Well-established biomarkers, such as procalcitonin (PCT), C-reactive protein (CRP), lactate and interleukin (IL) 6, have been extensively studied in patients with proven or suspected sepsis in a variety of clinical environments and, in many healthcare systems are part of routine care of these patients. Although various novel biomarkers for the diagnosis and prognosis of sepsis, such as blood leucocyte transcriptomic markers and genetic markers, have been proposed, current evaluations have yet to provide definite conclusions about their effectiveness. Further, the current literature does not provide clear guidance on the optimal use of established, or novel, biomarkers in the diagnosis, risk stratification and prognosis of patients who present to the ED with clinical features of suspected sepsis. Aims and objectives Our objective was to conduct a comprehensive evidence synthesis to evaluate the clinical utility of both established and emerging biomarkers, used individually or in combination, in the identification of adult patients presenting to the hospital ED with clinically suspected sepsis who are at high risk of deterioration. The risk of deterioration was defined by critical outcomes, including admissions to critical/intensive care, occurrence of septic shock, organ failure and in-hospital mortality. We addressed the following research question: ‘In ED patients with suspected sepsis, do biomarker concentrations measured at presentation improve identification of those at higher risk of mortality or other adverse clinical outcomes?’ Table 1 presents the review question using the Population, Intervention, Comparator, Outcome, Timing and Setting (PICOTS) format based on the CHecklist for critical Appraised and data extraction for systematic Reviews of prediction Modelling Studies and of Prognostic Factors studies (CHARMS-PF), a modified version of CHARMS, which includes key items relevant to Prediction Factors studies. 10 – 12 TABLE 1 PICOTS criteria for this systematic review Methods Search methods for identification of studies A sensitive literature search strategy was developed by an Information Specialist to identify published, peer-reviewed studies. The search strategy included index terms and free text to encompass the various facets of sepsis, selected biomarkers, the ED setting and prognosis. A range of databases were searched including MEDLINE, EMBASE, the Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trials (CENTRAL). There were no restrictions on study type or language at the search stage, but to capture recent emerging biomarkers, the selected results were limited to articles published in English from January 2013 to July 2023. All references were exported to EndNote [Clarivate Analytics (formerly Thomson Reuters), Philadelphia, PA, USA] for recording and deduplication. The reference lists of all articles selected for full-text appraisal were screened for additional studies. Outline searches for Ovid MEDLINE and EMBASE are shown in Appendix 2 . Inclusion and exclusion criteria Types of studies We included studies of any design that assessed the value of biomarkers in the prediction of clinically relevant outcomes in adults with suspected sepsis presenting to the ED . We focused on studies examining predictive factors rather than those assessing or validating prediction models. Conference abstracts were excluded because they were not considered to provide sufficient details. We had planned to seek more complete information from other sources if potentially relevant conference abstracts were identified; however, in practice this was not possible due to time and resource constraints. Targeted population We only included studies that assessed adult patients who presented to the ED with clinically suspected sepsis. We excluded studies that restricted analysis of prognostic factors solely to those later confirmed with sepsis. When it was unclear whether the population represented those with clinically suspected sepsis or just confirmed sepsis, reviewers sought consensus expert clinical opinion regarding eligibility. We recorded each study’s definition of ‘suspected’ and ‘confirmed’ sepsis as provided by the authors. We accepted each study’s definition of ‘adults’ as reported by the authors. Studies involving patients with postoperative or hospital-acquired infection, trauma, or burn injuries were not deemed suitable for inclusion. Mixed population studies were considered eligible for inclusion if at least 80% of the participants met the pre-specified criteria for inclusion. Target condition For each study, we scrutinised the author’s definition of sepsis. We accepted definitions based on Sepsis-3 criteria ( qSOFA score ≥ 2), 1 a NEWS2 score ≥ 5, 7 as well as comparable physiological definitions that are broadly equivalent. For studies published before 2016, we considered the earlier SIRS and Sepsis-2 criteria to be appropriate. 9 Biomarkers of interest The biomarkers of interest, assessed individually or in combination, included the following: PCT CRP Soluble urokinase-type plasminogen activator receptor (SUPAR) Lactate Cytokines [granulocyte-macrophage colony-stimulating factor (GM-CSF); interferon‐gamma ( IFN ‐γ); IL-1 beta; IL-2; IL-4; IL-5; IL-6; IL-8 (C-X-C motif chemokine ligand 8) IL-10 (chemokine interferon-γ inducible protein 10 kDa); IL-12p70; IL-13; IL-17A (cytotoxic T-lymphocyte associated protein 8); monocyte-chemotactic protein 3; chemokine (C-C motif) ligand; and tumour necrosis factor alpha (TNF-α)]. Monocyte distribution width (MDW). Blood leucocyte transcriptomic markers – eight neutrophils [cluster of differentiation antigens (CD) CD15; CD24; CD35; CD64; CD312; CD11b; CD274; CD279], seven monocyte [(CD35; CD64; CD312; CD11b; human leucocyte antigen (HLA)-DR; CD274; CD279) and a CD8 T-lymphocyte biomarker (CD279)]. Genetic markers. Any other biomarkers deemed important by the reviewers. Outcome measures Outcomes of interest were: Mortality In-hospital mortality at any time point 30-day mortality Overall survival rate Critical care admission Septic shock Organ failure Timing of biomarker measurement The timing of biomarker measurement was a key consideration in the conduct of this systematic review. We focused on biomarker performance in patients with suspected sepsis identified within 12 hours of ED arrival. Studies were excluded if biomarker measurement occurred more than 24 hours after ED arrival. However, to be comprehensive, we included studies that reported biomarker measurement in the ED without specifying the exact timing. Selection of studies One review author (MI) screened all titles and abstracts identified by the search strategies and a second review author (SD) independently screened a random sample (20%) to ensure consistency. A cursory examination of full-text articles deemed potentially relevant was conducted by a single review author to exclude studies that clearly did not meet the pre-specified inclusion and exclusion criteria. The remaining full-text studies were then thoroughly examined independently by two review authors (MI, SD). We recorded the number of excluded studies and documented the main reasons for exclusion. Data extraction For each study, we extracted key information including the publication date; study design; study period; demographic and baseline characteristics of participants; number, type and definition of outcome measures; and details of biomarker measurements (e.g. manufacturer of the biomarker assay as outlined by the CHARMS-PF checklist). Data were extracted by one reviewer using a bespoke data extraction form (MI or SD) and verified by a second reviewer (MI or SD). Risk-of-bias assessment We assessed the risk of bias in studies eligible for meta-analysis using the QUIPS (Quality in Prognostic Factor Studies) tool. 13 The QUIPS tool evaluates six domains: study participation (selection bias) and study attrition (assessed at the study level), prognostic factor measurement, outcome measurement, confounding, and statistical analysis and reporting (assessed at the outcome level). For the risk-of-bias assessment, we classified mortality and critical care admission as objective outcomes, while septic shock and organ failure were considered subjective outcomes. Studies were rated as having a ‘ low ’, ‘ moderate ’ or ‘ high ’ risk of bias. While studies rated at high risk were not excluded, we considered them to be less reliable than those at low or moderate risk-of-bias studies when synthesising the evidence. The risk of bias was assessed by one reviewer (MI) and checked by a second reviewer (SD). Any disagreements between review authors regarding study selection, data extraction or risk-of-bias assessment were resolved by consensus or referred to a third author for final adjudication. Data synthesis From each included study, we recorded the relevant estimate data for each relevant prognostic factor. When a study reported multiple biomarker thresholds, we selected the estimates corresponding to the best-performing threshold. In cases where biomarker thresholds were defined by the study authors, we noted these for later comparison. We extracted all relevant risk factor summary estimates along with measures of uncertainty [95% confidence intervals (CIs) or standard errors (SEs)] for both raw and any adjusted estimates, noting any adjusting factors. When available, we extracted the area under the receiver-operating characteristic curve (AUROC) and its associated measures of uncertainty. Following current standards we categorised AUROCs as follows: < 0.70 as poor, 0.70–0.79 as fair, 0.80–0.89 as good and 0.90–1 as excellent. 14 When possible, we converted AUROC values into odd ratio (OR) estimates provided the necessary assumptions were met. 15 , 16 If the AUROC was also not available, we recorded sensitivity and specificity estimates instead. Due to time constraints, we did not contact the study authors for missing data. Meta-analyses According to how data were reported in the included studies, we performed separate meta-analyses for each biomarker or combination of biomarkers using appropriate estimates of effect: Risk ratios (RRs), ORs, and hazard ratios (HRs) (unadjusted and adjusted estimates). Prognostic factors assessed at similar thresholds. Meta-analyses of HRs, RRs and ORs were conducted using the extracted estimates along with SEs or CIs. For studies that did not report RRs or ORs, we calculated these estimates and their SEs using the extracted frequencies and denominators. All estimates and 95% CIs were transformed to the logit scale, and the variance of the logit was calculated. Pooled logit estimates and 95% CIs were then back-transformed to the original scale. Similarly, when appropriate data were available, we performed meta-analyses of AUROC values. AUROC values and accompanying 95% CIs were transformed to the logit scale and the variance of logit AUROC was computed. Pooled logit AUROC and 95% CIs were then back-transformed to the original scale. To account for heterogeneity among studies, in all analyses, summary effect estimates and their 95% CIs were calculated using the Der Simonian and Laird random-effects model. Meta-analyses were performed using the metan command in STATA (StataCorp. Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC; 2023). The I 2 -statistic was used to describe the percentage of variation across included studies due to heterogeneity. We applied the following thresholds: < 30% for low heterogeneity, 30–60% for moderate heterogeneity and > 60% for high heterogeneity. 17 To investigate potential small study bias, we visually inspected funnel plots and performed the Egger’s bias test. 18 In the presence of such bias, we conducted a trim-and-fill adjusted analysis to impute missing studies and re-calculate the effect size. 19 Subgroup analysis Where sufficient data were available, to explore potential sources of between-study heterogeneity, we had planned to perform subgroup analyses including: Patients of different age groups. Patients with specific comorbidities. Studies performed in the NHS versus studies not performed in the NHS. Timing of biomarker measurement. All statistical analyses were performed using STATA software (version 18 or the latest version). Patient and public involvement The use of biomarkers to identify those patients who present to the ED with symptoms of sepsis and may benefit from early treatments was recognised as a top 10 priority for UK emergency medicine by the James Lind Alliance Emergency Medicine Priority Setting Partnership Refresh Steering Group in 2023. 20 The research prioritisation process was a collaboration between patients, healthcare professionals and researchers with 39 responses received from patients or carers. In addition, one member of the Aberdeen Centre for Evaluation Public Involvement Partnership Group provided extensive comments on this manuscript, enhancing its readability and two other members contributed to the development of the Plain language summary to ensure it was suitable for different, non-expert audiences. To enhance patient involvement in the process, we also conducted a separate, concurrent Study Within a Review (SWAR) in which we investigated how best to communicate the finding of the review to the general public. Participants ( n = 105) either read or listened to the Plain language summary and answered predefined questions designed to determine their understanding of the findings. Several weeks later, we assessed their retention of the information. Findings of the SWAR will be published separately. Equality, diversity and inclusion In conducting this review, we adhered to the principles outlined in the PRO-EDI tool ( www.trialforge.org/trial-diversity/pro-edi-improving-how-equity-diversity-and-inclusion-is-handled-in-evidence-synthesis/ ), with a focus on extracting data related to age, sex, gender, race/ethnicity and ancestry, socioeconomic status and location (where reported in the included studies). In writing this manuscript, we have made a concerted effort to use clear and accessible language, providing definitions as necessary to ensure the content is understandable to a broad audience. Results This review is reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. 21 Quantity of the evidence The electronic searches initially identified a total of 1986 citations. Of these, 884 were removed before screening comprising 754 duplicates, 80 paediatric studies, 43 reports of ongoing trials and 7 conference abstracts. This left a total of 1102 citations that were screened by title and abstract. As a result of this screening process, 672 records were excluded (including 5 duplicates) and 430 reports were retrieved for full-text assessment. After in-depth scrutiny, 377 were subsequently excluded, based on eligibility criteria and 53 reports of 51 unique studies were deemed suitable for inclusion. The selection process is illustrated in Figure 1 . Details of the included studies are presented in Appendix 3 , while Appendix 4 lists the excluded studies and provides the main reasons for exclusion. FIGURE 1 Preferred Reporting Items for Systematic Reviews and Meta-Analysis flow diagram for identification of the quantitative studies. Reproduced with permission from Page et al . This is an Open Access article distributed in accordance with the terms of the (more...) Description of studies Key baseline demographic characteristics of study participants in the 51 included studies are summarised in Appendix 5 and additional study characteristics are presented in Report Supplementary Material 1 . The included studies were conducted across 19 countries: 17 (33%) in Europe (2 from the UK), 22 (43%) in Asia and 8 (16%) in the USA. Population sample sizes varied widely, ranging from 36 22 to 8698, 23 with a median of 349 participants. Thirteen studies recruited 200 or fewer participants. Recruitment periods spanned from as early as 2003–5 24 to 2021–2, 25 but four studies did not report recruitment dates. 26 – 29 Definitions of suspected sepsis and inclusion and exclusion criteria varied considerably across studies. Fourteen studies did not report any exclusion criteria. A summary of inclusion criteria and definitions of suspected sepsis adopted in the included studies is provided in Appendix 6 . Biomarker performance In total, 107 different biomarkers or biomarker combinations were investigated across the 51 studies. The most studied biomarkers were lactate ( n = 29), PCT ( n = 21) and CRP or high-sensitivity CRP (hsCRP) ( n = 15). Relevant prognostic data for the remaining biomarkers were available from fewer than three studies, which precluded their inclusion in a meta-analysis. There were three studies each of IL-6 and neutrophil-to-lymphocyte ratio (NLR), and two studies each of MDW , IL-10, neutrophil gelatinase-associated lipocalin (NGAL) and presepsin (PSEP). The remaining biomarkers were investigated in only one study each, either individually or in combination with other biomarkers or clinical scores: osteopontin (OPN), arterial pH, ischaemia-modified albumin (IMA), News-L, angiopoietin-1 (Ang-1), angiopoietin-2 (Ang-2), tyrosine kinase with immunoglobulin-like and epidermal growth factor-like domains 2 (Tie-2), mid-regional proadrenomedullin (MR-proADM), platelet-to-lymphocyte ratio (PLR), respiratory adjusted shock index (RASI), heparin-binding protein (HBP), delta neutrophil index (DNI), IL-5, resistan, vascular cell adhesion molecule-1 (VCAM-1), inter-cellular adhesion molecule-1 (ICAM-1), calprotectin, microRNA (miR)-140, miR-223, miR-150, N-terminal prohormone of brain natriuretic peptide (NT-pro BMP), pentraxin 3 (PTX3), neutrophil-to-white blood cell ratio (NWR), endothelial cell-specific molecule-1 (ENDOCAN), von Willebrand factor (vWF) and disintegrin‑like and metalloprotease with thrombospondin type 1 motif (ADAMTS-13). Lists of prognostic biomarkers evaluated individually or in combination in the included studies are given in Appendix 7 , Tables 6 and 7 , respectively. Risk of bias We assessed the risk of bias in all studies included in the meta-analyses of established biomarkers, as well as in those assessing novel biomarkers for which results were displayed graphically using forest plots ( n = 36 in total). Study participation (selection bias) Of these 36 studies, 23 were rated as high risk for study participation/selection bias due to: Inclusion was restricted to patients with complete biomarker data available, potentially distilling the population retrospectively to only those with confirmed sepsis ( n = 13 studies). 30 – 42 Lack of a clear definition of ‘suspected sepsis’ population ( n = 7). 43 – 49 Use of non-consecutive recruitment methods such as voluntary or convenience sampling ( n = 3). 50 – 52 The number and reason for patients lost to follow-up were not fully described (three studies). 44 , 53 , 54 Six studies were assessed as moderate risk of bias due to incomplete description of baseline characteristics 54 , 55 or lack of information on the participation rate, resulting in insufficient information to assess whether participants and non-eligible participants were similar. 25 , 53 , 56 , 57 Seven studies were assessed as having low risk of bias. 23 , 58 – 63 Study attrition Nine studies had a low risk of bias for study attrition. 47 , 51 , 52 , 55 , 57 – 59 , 61 , 62 However, in 27 studies, the risk was unclear because: Analyses included only cases for which there were no missing values with no description of the excluded population ( n = 24) 23 , 25 , 30 – 43 , 45 , 46 , 48 , 50 , 56 , 60 , 63 , 64 The number and reason for patients lost to follow-up were not fully described ( n = 3). 44 , 53 , 54 Prognostic factor (biomarker measurement) For biomarker measurement, studies were rated as having a moderate risk of bias if cases with incomplete prognostic factor data were excluded from the analysis ( n = 19) 30 , 32 – 34 , 36 – 43 , 45 , 46 , 48 , 50 , 56 , 62 – 64 or if there was a limited description of incomplete or missing prognostic factor data ( n = 7). 44 , 46 , 47 , 52 , 57 , 61 Five studies were judged to be at high risk of bias due to the use of a dichotomised variable without appropriate, non-data-dependent cut-points. 23 , 25 , 31 , 35 , 52 The remaining studies, which use either continuous or dichotomous variables, were considered to be at low risk of bias for this domain. Outcome measurement Most studies were rated at a low risk of bias for outcome measurement. However, five studies were assessed as having a high risk of bias due to a lack of a clear outcome definition 23 , 36 , 46 , 55 , 57 and a further eight studies were judged at moderate risk of bias due to the report of a subjective outcome with no description of blinding. 30 , 32 , 37 , 38 , 41 , 45 , 52 , 58 We considered studies with objective outcomes to be at lower risk of bias with regard to the lack of blinding. Confounding We assessed the risk of bias from confounding based on five factors identified by our clinical review authors: age, pre-morbid functional status or frailty score, care status (proportion of participants who are in care), comorbidities and clinical score (e.g. NEWS). Studies were considered: low risk of bias if outcomes were adjusted for four or five factors moderate risk if adjusted for between one and three factors, and high risk if studies were not adjusted for any of the pre-specified potential confounders. These risks are presented for each outcome in the forest plots ( Figures 2 – 11 ). FIGURE 2 Forest plots presenting pooled estimates of the utility of CRP as a predictor of mortality in patients presenting to the ED with clinically suspected sepsis. (A) Unadjusted HRs of CRP (continuous) for predicting mortality. (B) Adjusted HRs of CRP (continuous) (more...) Statistical analysis and reporting Statistical analysis and reporting in the included studies were generally considered adequate and at low risk of bias, on the basis that all analyses appeared to be presented in the results section. Meta-analysis of common biomarkers Due to limited data and clinical heterogeneity across studies, only biomarkers already widely used in clinical practice such as lactate, CRP and PCT were suitable to be combined in meta-analysis (see Figures 2 – 5 ). Even for these biomarkers, the data pool was relatively small, and the analyses were constrained by the variability in the confounders adjusted for in the included studies, limiting our ability to draw meaningful clinical conclusions. Our analyses suggest that an initial lactate, CRP or PCT measurement in the ED has limited prognostic value in predicting all-cause mortality in patients with suspected sepsis. Data on the need for critical care admission were particularly sparse ( n = 3). Likewise, we were unable to perform a meta-analysis of results for the prediction of septic shock or organ failure. FIGURE 3 Forest plots presenting pooled estimates of the utility of lactate as a predictor of mortality in patients presenting to the ED with clinically suspected sepsis. (A) Adjusted HRs of lactate (continuous) for predicting mortality (adjusted by various factors). (more...) FIGURE 4 Forest plots presenting pooled estimates of the utility of PCT as a predictor of mortality in patients presenting to the ED with clinically suspected sepsis. (A) Unadjusted HRs of PCT (continuous) for predicting mortality. (B) Unadjusted ORs (continuous) (more...) FIGURE 5 Forest plot presenting pooled estimates of the utility of lactate as a predictor of need for critical care in patients presenting to the ED with clinically suspected sepsis. Adjusted ORs of lactate (continuous) for predicting the need for critical care (more...) Most of the meta-analyses for the common biomarkers showed considerable heterogeneity ( I 2 > 75%), 65 further limiting the utility and generalisability of the findings. Although we had planned to perform subgroup analyses to investigate potential sources of between-study heterogeneity such as age groups, specific comorbidities, studies performed in NHS settings versus studies not performed in the UK, and timing of biomarker measurement, the available data were insufficient to support these analyses. This additionally constrains the clinical applicability of our findings. Studies of common biomarkers that could not be included in the meta-analysis are summarised in Report Supplementary Material 2 . Novel biomarkers Descriptive results of emerging or novel biomarkers are presented graphically in forest plots ( Figures 6 – 11 ) and summarised in Appendix 8 . Each forest plot presents the predictive utility of novel biomarkers used independently or in combination with other novel or common biomarkers or clinical scores. Novel biomarkers for which there were insufficient data (they did not provide a point estimate of effect with accompanied CIs) are not included in the forest plots but are summarised in Appendix 9 . FIGURE 6 Unadjusted HRs and ORs of novel biomarkers for predicting mortality. (A) Unadjusted HRs and ORs for individual novel biomarkers. (B) Unadjusted HRs and ORs for combined biomarkers. (C) Unadjusted HRs for novel biomarkers combined with clinical scores. (more...) Utility of biomarkers used independently or in combination to predict mortality Unadjusted odds ratio/hazard ratio For the prediction of all-cause mortality, six studies reported unadjusted ORs or HRs for five individual biomarkers: HBP , MR-proADM , NGAL , PSEP and PTX3 ( Figure 6 , A). Reported unadjusted ORs range from 1.76 (95% CI 1.12 to 2.78) for PTX3 52 to 15.40 (95% CI 6.20 to 38.40) for MR-proADM . 53 Unadjusted HRs range from 1.00 (95% CI 1.00 to 1.00) for PSEP 30 to 4.10 (95% CI 2.60 to 6.50) for MR-proADM . 53 Two studies provided unadjusted ORs or HRs for a novel biomarker combined with another biomarker or a clinical score ( Figure 6 , B/C). Katsaros et al . (2022) found that patients with elevated levels in both HBP and PCT at ED presentation had greater odds of 28-day mortality (OR 3.53, 95% CI 1.62 to 7.67) ( Figure 6 , B) compared with those with an increase in HBP level alone (OR 2.20, 95% CI 0.92 to 5.30). 35 Gonzalez del Castilllo et al . (2019) reported unadjusted HRs for MRproADM combined with another biomarker ( CRP , lactate or PCT) or a clinical score [modified version of CURB-65 (confusion, uraemia, respiratory rate, blood pressure, age ≥ 65 years) (CRB-65), NEWS, qSOFA , SIRS , or SOFA]. 53 They found that 28-day mortality was higher in patients with raised MRproADM and raised PCT on ED presentation (HR 5.70, 95% CI 2.80 to 11.60). The same study also reported that patients with elevated levels of MRproADM also had a higher risk of death in association with a raised CRP or lactate, or an elevated clinical score (HRs ranging from 2.40 to 4.10, see Figure 1 , A). 53 Adjusted odds ratio/hazard ratio Adjusted ORs and HRs for the prediction of all-cause mortality are reported for 18 single biomarkers within 9 studies ( Figure 7 , A). The adjusted ORs range from 0.01 (95% CI 0.00 to 1.05) for NWR 44 to 4.00 (95% CI 0.83 to 19.32) for MDW , 23 except for one study reporting an OR of 48.59 (95% CI 1.41 to 167.90) for HBP , which seems clinically implausible. 35 The adjusted HRs ranged from 0.50 (95% CI 0.20 to 1.20) for PSEP 58 to 3.80 (95% CI 2.20 to 6.50) for MRproADM. 53 Notably, for most biomarkers, the adjusted effect estimates were not statistically significant at the 95% CIs. FIGURE 7 Adjusted HRs and ORs of novel biomarkers for predicting mortality. (A) Adjusted HRs and ORs of mortality for individual novel biomarkers. (B) Adjusted ORs of mortality for combined biomarkers. RoB, risk of bias: Six letters represent assessments of six (more...) Area under the receiver-operating characteristic curve Mortality prediction based on the AUROC analysis was available for 19 single biomarkers across 10 studies ( Figure 8 , A), 15 novel biomarker combinations with another biomarker in 4 studies ( Figure 8 , B) and 10 combinations of a novel biomarker with a clinical score in 5 studies ( Figure 8 , C). FIGURE 8 Area under the receiver-operating characteristic curve values of novel biomarkers for predicting mortality. (A) AUROC values of individual novel biomarkers (continuous) for predicting mortality. (B) AUROC values of combined biomarkers (continuous) for (more...) Among individual biomarkers, MR-proADM , Inflammatix Severity 2 (IMX-SEV-2) and Tie-2 demonstrated relatively high AUROCs for mortality prediction (0.84, 0.82 and 0.80, respectively) ( Figure 8 , A; Appendix 8 ). Several other biomarkers showed modest discriminatory ability, including Ang-2 (0.79), Ang-1 (0.78), ADAMTS-13 (0.76) and vWF (0.75). NGAL was examined in two studies: Hong et al . (2016) reported an AUROC of 0.80 (95% CI 0.76 to 0.83) for a cohort of 470 individuals, indicating good discriminative capability for predicting mortality, 43 while MacDonald et al . (2017) reported a lower AUROC of 0.65 (95% CI 0.57 to 0.72) for a sample of 186 participants. 37 When evaluated independently for prediction of mortality, IL-6 exhibited moderate discrimination, but when combined with other biomarkers such as lactate, NWR , PCT , or combinations thereof, the AUROC improved to above 0.7. This finding was seen in a small population described by Xie et al ., 44 and a larger study of 440 participants reported by Fang et al .. 54 We also observed that the predictive utility of Ang-1 , Ang-2 and Tie-2 for mortality improved when these biomarkers were used in combination ( Figure 8 , B). Integrating clinical scores with novel biomarkers also appeared to improve mortality prediction ( Figure 8 , C). For example, while PSEP alone yielded an AUROC of 0.68, this increased to 0.78 when combined with PCT and the Rapid Emergency Medicine Score (REMS) clinical score. Similarly, when Ang-1 , Ang-2 and Tie-2 were combined with PCT and the Mortality Emergency Department Sepsis (MEDS) clinical score, the AUROC improved to 0.93 from 0.78, 0.79 and 0.80, respectively. This enhancement was also observed for vWF , and ADAMTS-13. Although their single AUROCs were below 0.80, combining these biomarkers with the MEDS clinical score increased their AUROC to 0.86. Similarly, the combination of IMX-SEV-2 with qSOFA enhanced the AUROC to 0.89 (95% CI, 0.84 to 0.94), compared to 0.82 (95% CI, 0.74 to 0.90) when IMX-SEV-2 was used alone. Further, in a large study of 1480 participants, a model based on the SOFA score and constructed with NLR , PLR and MDW achieved an AUROC for prediction of mortality of 0.91 (95% CI 0.87 to 0.96). Utility of biomarkers for the prediction of other outcomes (critical care admission, septic shock, organ failure) Critical care admission: unadjusted and adjusted odds ratios and area under the receiver-operating characteristic curve Limited data were available for the prediction of critical care admission. This may be due to the variation in resources and access to critical care facilities in different healthcare systems. For example, the UK, with relatively fewer critical care beds per capita than Europe and the USA, has stricter admission criteria than many other countries. 66 Two studies (Uusitano-Seppala 2013; Gonzalez del Castillo 2019) 52 , 53 reported unadjusted ORs for PTX3 and MR-proADM used alone ( Figure 9 , A), and in combination with other biomarkers ( Figure 9 , B) or clinical scores ( Figure 9 , C). PTX3 predicted critical care admission with an unadjusted OR of 3.54, though with wide 95% CIs ( Figure 9 , A). MR-proADM was assessed in a large cohort of 684 subjects (Gonzalez del Castillo, 2019) and showed an unadjusted OR of 4.10 (95% CI 2.30 to 7.10) for predicting critical care admission. 53 When combined with SOFA or SIRS criteria, its predictive accuracy showed a modest improvement, with unadjusted ORs of 4.40 (95% CI 2.40 to 7.80) and 4.50 (95% CI 2.20 to 9.30), respectively, but the 95% CIs remained wide. However, combining MR-proADM with other scoring systems or established biomarkers did not enhance its performance further ( Figure 9 , A/B/C). Instead, when results were analysed using an adjusted model, the OR for MR-proADM , used alone, increased to 5.8 (95% CI 3.1 to 10.8). Nevertheless, the wide CIs indicated a considerable degree of uncertainty around the effect estimate ( Figure 10 , A). FIGURE 9 Unadjusted ORs of novel biomarkers for predicting other outcomes (critical care admission and septic shock). (A) Unadjusted ORs of individual novel biomarkers (continuous) for predicting critical care admission. (B) Unadjusted ORs of combined biomarkers (more...) FIGURE 10 Adjusted ORs of individual novel biomarkers for predicting other outcomes (critical care admission and septic shock). (A) Unadjusted ORs of individual novel biomarkers (continuous) for predicting critical care admission. (B) Unadjusted ORs of individual (more...) Two studies evaluating IMX-SEV-2 and MR-proADM for predicting critical care admission reported moderately good AUROC values [0.85 (95% CI 0.79 to 0.92) and 0.79 (95% CI 0.7 to 0.88), respectively – Figure 11 , A]. FIGURE 11 Area under the receiver-operating characteristic curve values of novel biomarkers for predicting other outcomes (critical care admission and septic shock). (A) AUROC values of individual novel biomarker (continuous) for predicting critical care admission. (more...) Septic shock: unadjusted and adjusted odds ratios and area under the receiver-operating characteristic curve The performance of novel biomarkers for the prediction of septic shock was evaluated in five studies. 30 , 52 , 58 One study evaluated several biomarkers (MRproADM, PTX3 , ICAM-1 , IL-6, IL-10, NGAL , Resistan, VAM-1) and reported low adjusted predictive estimates for septic shock for ICAM-1 , IL-6, IL-10, and moderate estimates for VCAM-1 , NGAL and Resistan ( Figure 10 , B). 37 Another study evaluated IMA in fewer than 200 participants and found it of little value for the prediction of septic shock (see Figure 10 , B). 57 Two studies reported unadjusted ORs for individual use of PSEP but showed inconsistent results with one study showing an increased risk of septic shock among patients with elevated PSEP measurement [OR of 3.40 (95% CI 1.80 to 6.50)] and the other indicating that PSEP was not a useful biomarker for predicting septic shock (see Figure 9 , E). 30 , 52 , 58 PTX3 was evaluated in a study of 537 participants that reported an unadjusted OR of 3.69 (95% CI 1.46 to 9.28). 52 For the prediction of septic shock, AUROC values were available in five studies evaluating 11 individual biomarkers ( Figure 11 , B) and 8 combinations of biomarkers in total ( Figure 11 , C/D). 37 , 38 , 41 , 57 , 58 In general, the AUROC values increased when novel biomarkers were combined with established biomarkers or clinical scores, but the magnitude of improvement varied across biomarker combinations ( Figure 11 , B/C/D). Organ failure One study assessing PTX3 for the prediction of organ failure reported a relatively strong OR of 3.25 (95% CI 1.12 to 9.45) using an adjusted model ( Figure 9 , D). Other novel biomarkers Some novel biomarkers, notably, micro-RNA, calprotectin, IL-5 and OPN , were identified by our search strategies but could not be incorporated into the forest plots, as they did not provide sufficient data (shown in Appendix 9 ). One study ( n = 69 participants) reported that micro-RNA 150 (miR150) was associated with mortality in unadjusted and adjusted logistic regression models ( p = 0.000 and p = 0.003, respectively). In the same study, miR146a and miR223 were found to be statistically inconclusive ( p -value reported to be not significant). 29 In another study of 351 participants, calprotectin exhibited a moderate AUROC of 0.65 for predicting admission to ICU or high dependency unit (HDU). However, its performance was reported to be statistically significantly better than that of NLR ( AUROC 0.47), another novel biomarker (see Appendix 9 ), and of PCT ( AUROC 0.46), an established biomarker, and also numerically larger than that of CRP ( AUROC 0.55) another established biomarker (see Report Supplementary Material 2 ). 67 In a small study of 36 patients, IL-5, IL-6 and IL-6 were reported to be effective in predicting ICU admission ( AUROC values of 0.81, 0.71 and 0.72, respectively), but not mortality. 22 Osteopontin ( n = 92) 59 was evaluated in one study and showed little prognostic value for the prediction of mortality (see Appendix 9 ). Discussion Summary of main results Any systematic review aiming to evaluate biomarkers for predicting symptom deterioration in patients with suspected sepsis is inherently ambitious, particularly given the complexity of sepsis as a clinical entity. In this systematic review, we collected evidence from 51 studies on the performance of established and emerging biomarkers in adults presenting to the ED with suspected sepsis, to predict mortality and other recognised clinically important outcomes: the need for critical care admission and the development of septic shock and organ failure. Although we initially identified several well-established biomarkers of interest when developing our search strategy, our inclusive approach aiming at comprehensively evaluating all available biomarkers (alone, in combination, or in addition to a clinical score) significantly increased the complexity of our data extraction process, resulting in examination of 107 different biomarker strategies. Across included studies, we extracted a range of statistical measures of effect – ORs, HRs, RRs and AUROC values. Many studies reported multiple statistical analyses for each biomarker, combination and outcome, resulting in a database of over 500 data points. Our analyses encompass data from studies that assessed well-established biomarkers (such as lactate, CRP and PCT) and a range of novel biomarkers (such as IMX-SEV-2 , MR-proADM , MDW , PSEP , angiopoietins, and others). Despite our extensive data set and the wide variety of biomarkers investigated, only a limited subset, specifically those already in use in clinical practice, provided data sufficient for performing meaningful meta-analyses. However, even within this subset, predictive accuracy for adverse outcomes, including mortality and critical care admission, was limited, highlighting the need for more refined predictive measures in clinical practice. This aligns with current clinical practice where lactate, CRP and PCT are already accepted as being of limited prognostic value in this group of patients, and outlined in current UK consensus guidance. 68 While such biomarkers are valuable indicators of systemic inflammation, they may not adequately capture the multifactorial risk profile associated with sepsis-related deterioration in the ED setting. Moreover, it is worth noting that our analyses were also constrained by substantial between-study heterogeneity ( I 2 > 75% for several biomarkers). While this poses challenges to the generalisability of our findings, it also reveals the inherent difficulties researchers face in conducting consistent, high-quality research in this complex clinical field. Sepsis is not merely a single diagnosis, but a multifaceted syndrome that manifests as a result of the body’s pathological host response to numerous distinct pathologies, each presenting with different symptoms and clinical signs. Moreover, a wide range of factors – including genetic, geographical, economic, cultural, and healthcare resource disparities, as well as patient-specific elements like comorbidities and concurrent drug therapies – can influence the likelihood of sepsis being suspected and the consequent use of biomarker tests. Moving forward, improving the standardisation and objectivity of measurements that account for the diversity of the studied population will be essential. Such advancements will facilitate the synthesis of findings across studies and the harmonisation of recommendations. In contrast, our review highlights promising, though preliminary, evidence for several novel biomarkers including IMX-SEV-2 , MR-proADM , NGAL , Tie-2 for prediction of mortality; IMX-SEV-2 for prediction of critical care admission and MDW and NLR for prediction of septic shock. Interestingly, these biomarkers along with others that showed only modest discriminatory ability when used alone – such as angiopoietins, PSEP , IMA , vWF and ADAMTS-13 – demonstrated improved predictive accuracy when combined with other biomarkers or clinical scoring systems like SOFA , qSOFA , REMS and MEDS . This likely reflects the significant complexity and variability of the pathophysiological responses in sepsis. It also highlights the potential benefits of adopting a multifaceted approach to identify those at increased risk. This would involve targeting different pathways of inflammation in combination and interpreting results alongside objective clinical scoring systems. Due to the limited data available for these novel biomarkers, the lack of meta-analysis results, and the wide CIs around most estimates of effect, these findings should be interpreted tentatively. It is worth noting that, to date, these emerging biomarkers have almost exclusively been used within research environments and their practical utility in routine clinical practice, and ideally to enable decision-making at the point of care, has yet to be established. In the ED setting, there is a critical need to improve the identification of patients with systemic infection who are likely to clinically deteriorate. This includes recognising those whose condition may not initially appear serious but require escalation of care and prompt initiation of key treatments. While it is evident that further extensive research is needed to determine which of the many proposed novel biomarkers confer the best discriminatory performance – in terms of optimal combinations, thresholds and cost-effectiveness – there is scope for optimism. Some promising candidates might have the potential to perform well and be effective. Strength and limitations This review has several strengths. The search strategy was pre-specified, comprehensive and rigorously planned with reference to a clear PICO format research question. Strict eligibility criteria were set at the outset and the review was performed in line with current methodological standards and reported transparently. There are, however, some limitations to acknowledge. First, there was substantial heterogeneity across studies, particularly regarding patient populations, biomarkers, thresholds and definition of suspected sepsis. Second, many studies exhibited a high risk of bias, largely due to selective recruitment and attrition, thereby limiting the robustness of the synthesised effects. Third, the size of study participant populations varied widely and even though the median was 349 participants, a quarter of the included studies enrolled fewer than 200 participants, which reduced the predictive ability of the observed findings. Fourth, the limited data on subgroups also prevented us from investigating differential biomarker performance across key clinical and demographic subgroups, such as age or comorbidity status. Fifth, while the search was limited to studies published since 2013 to capture emerging biomarkers and current practices, several studies recruited participants up to a decade earlier. Sixth, our study selection process involved double screening of only a subset of records at both the title/abstract and full-text stage, with these records screened independently by two reviewers. This was a pragmatic decision made in the context of available resources. Patient and public involvement While this study arose from the identification of sepsis as a top-10 priority by the James Lind Alliance Emergency Medicine Priority Setting Partnership Refresh Steering Group in 2023, 20 we did not involve patients in the data analysis or interpretation stage of this complex topic. However, clinicians with direct patient involvement and expertise guided the clinical interpretation and context throughout. Equality, diversity and inclusion This evidence synthesis was developed with close attention to the National Institute for Health and Care Research’s commitment to embedding equality, diversity and inclusion across all research activities. We actively considered the principles outlined in the PRO-EDI tool throughout the planning and conduct of this project. Sepsis is a highly heterogeneous condition that disproportionately affects vulnerable groups, including older adults, people from socioeconomically deprived backgrounds, and those with pre-existing comorbidities. Ethnic and geographic disparities in sepsis outcomes have also been documented, including differences in access to timely care, recognition of symptoms and ICU admission. Where available, we extracted and reported data on age, sex, gender, race/ethnicity, socioeconomic status and healthcare setting. Unfortunately, many of the included studies did not stratify outcomes or biomarker performance across these demographic characteristics. Most included studies were conducted in Europe, Asia and the USA; only two were based in the UK NHS setting. As such, caution should be taken when extrapolating findings to UK-specific populations, particularly underserved or minoritised groups. Our multidisciplinary team included individuals with diverse professional and academic backgrounds, including clinicians, methodologists and health services researchers. We aimed to foster an inclusive research environment and adopted consensus approaches to resolve any clinical or methodological disagreements. Recommendations for future studies Future research should prioritise robust study designs, precise definitions, appropriate descriptions of patient populations, and comprehensive reporting. To enhance the identification of patients at high risk of clinical deterioration among those presenting to ED with signs of systemic infection, the following priorities are recommended: Adopt a Prospective Study Design: Focus on prospective study designs with consecutive enrolment of participants to minimise selection bias and improve generalisability. Harmonise Research Protocols: Standardise protocols, particularly concerning biomarker thresholds, the timing of measurements and how participants are identified, to facilitate comparability across studies. Define Suspected Sepsis Clearly: Establish and utilise clear, precise and consistent criteria for identifying suspected sepsis to ensure uniformity in the population being studied. Analyse Outcomes Across All Suspected Sepsis Cases: Analyse outcomes for the entire population with suspected sepsis, including both confirmed sepsis cases and those where sepsis is subsequently ruled out. To facilitate meaningful comparisons, it is important to apply a formal and consistent sepsis diagnosis across studies. For instance, reliance solely on culture-positive cases may exclude a substantial number of patients with sepsis. Evaluate Relevant Clinical Outcomes: Inclusion of key clinical deterioration outcomes, such as admission to critical care, organ failure, septic shock and mortality. Report Adjusted and Unadjusted Analyses: Include both unadjusted and adjusted analyses, accounting for critical variables such as: Age Premorbid functional status or frailty (for participants aged > 60 years) Presence of ‘do not attempt cardiopulmonary resuscitation’ (DNACPR) decisions Care home residency Major comorbidities Clinical scores Present Prognostic Factor Measures Transparently: Provide continuous measures of prognostic factors and their dichotomised counterparts. Continuous measures should include metrics like the AUROC to determine the discriminatory ability of biomarkers to correctly classify individuals at risk of deterioration. Supplementary data on all cut-off points should be made available to enhance interpretability and replication. These recommendations would be useful to improve the methodological quality and clinical applicability of future studies in identifying and managing ED patients with features of systemic infection. Conclusions In adult patients presenting to the ED with suspected sepsis, the measurement of established biomarkers such as lactate, CRP and PCT do not confer important prognostic information regarding the likelihood of clinical deterioration or death. Though data on emerging, or novel, biomarkers like IMX-SEV-2 , MR-proADM , MDW and Tie-2 are not extensive or conclusive, these biomarkers show promise in their ability to predict poor clinical outcomes in patients with suspected sepsis, particularly when integrated with other biomarkers and clinical scoring systems. There is an urgent need for high-quality, prospective studies in this field. Research protocols should be standardised, objective measurements prioritised wherever possible, and outcomes reported consistently to allow synthesis and harmonisation of results. This approach will help facilitate the translation of findings into clinical practice to manage patients with suspected sepsis. Additional information CRediT contribution statement Mari Imamura ( https://orcid.org/0000-0003-4871-0354 ) : Methodology, Investigation, Data curation, Visualisation, Writing – original draft; Writing – reviewing and editing. Sinead N Duggan ( https://orcid.org/0000-0002-4517-4384 ) : Methodology, Investigation, Data curation, Visualisation, Writing – original draft; Writing – reviewing and editing. Thenmalar Vadiveloo ( https://orcid.org/0000-0001-5531-6289 ) : Formal analysis, Visualisation, Writing – reviewing and editing. Jamie G Cooper ( https://orcid.org/0000-0003-3812-7026 ) : Methodology, Writing – reviewing and editing. Callum T Kaye ( https://orcid.org/0000-0002-6904-7048 ) : Methodology, Writing – reviewing and editing. Paul Manson ( https://orcid.org/0000-0002-1405-1795 ) : Methodology, Resources, Writing – reviewing and editing. Gianni Virgili ( https://orcid.org/0000-0002-9960-2989 ) : Methodology, Writing – reviewing and editing. Lorna Aucott ( https://orcid.org/0000-0001-6277-7972 ) : Formal analysis, Supervision, Writing – reviewing and editing. Mike Clarke ( https://orcid.org/0000-0002-2926-7257 ) : Funding acquisition, Conceptualisation, Methodology, Supervision, Writing – reviewing and editing. Miriam Brazzelli ( https://orcid.org/0000-0002-7576-6751 ) : Funding acquisition, Conceptualisation, Methodology, Supervision, Writing – reviewing and editing, Project administration. Other contributions: Ian Scragg (Aberdeen Centre for Evaluation, Public Involvement Partnership Group): advice and feedback on the manuscript. Acknowledging the use of Artificial Intelligence (AI) tools We acknowledge the use of the machine learning function in the Covidence software, and in particular, its Relevance Sorting feature to prioritise the titles and abstracts likely to be relevant during the screening of search results. All titles and abstracts were screened by the review authors and no studies were excluded by the automation tool. We also acknowledge the use of AI Assistant in Adobe Acrobat to ask PDF documents questions to assist in finding relevant information for data extraction. All data were manually extracted by the review authors. Data-sharing statement This is an evidence synthesis; all technical data are presented in the text or contained within tables, figures, appendices and supplementary material. All queries should be submitted to the corresponding author for consideration. Ethics statement This is a synthesis of published or publicly available evidence and no primary research data were collected as part of this project. Ethics approval was not required. Information governance statement This systematic review did not involve the collection and processing of personal data. Therefore, compliance with the Data Protection Act (2018) and the General Data Protection Regulation (EU GDPR) 2016/679 was not applicable. Disclosure of interests Full disclosure of interests : All authors have completed ICMJE disclosure forms. Completed ICMJE forms for all authors, including all related interests, are available in the toolkit on the NIHR Journals Library report publication page at https://doi.org/10.3310/GJMB1730 . Primary conflicts of interest : Miriam Brazzelli participates on the NIHR HTA Commissioning Funding Committee 2023–8. Callum Kaye is Lead Clinician on the Scottish Acquired Brain Injuries Network. The remaining authors have no competing interests to disclose. Department of Health and Social Care disclaimer This publication presents independent research commissioned by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, NIHR Coordinating Centre, the Health Technology Assessment programme or the Department of Health and Social Care. This article was published based on current knowledge at the time and date of publication. NIHR is committed to being inclusive and will continually monitor best practice and guidance in relation to terminology and language to ensure that we remain relevant to our stakeholders. Study registration This study is registered as PROSPERO CRD42023454465. Funding This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR159912. Box This article reports on one component of the research award Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department . For more information about this research, please view the award (more...) About this article The contractual start date for this research was in August 2023. This article began editorial review in December 2024 and was accepted for publication in August 2025. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article. Copyright Copyright © 2026 Imamura et al . This work was produced by Imamura et al . under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/ . For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited. Disclaimer Every effort has been made to obtain the necessary permissions for reproduction, to credit original sources appropriately and to respect copyright requirements. However, despite our diligence, we acknowledge the possibility of unintentional omissions or errors and we welcome notifications of any concerns regarding copyright or permissions. List of supplementary material Report Supplementary Material 1. Additional characteristics of included studies (not presented in the main table in Appendix 5) Report Supplementary Material 2. Additional outcome data for common biomarkers (not presented in common biomarker meta-analyses forest plots) Supplementary material can be found on the NIHR Journals Library article page ( https://doi.org/10.3310/GJMB1730 ). Supplementary material has been provided by the authors to support the article and any files provided at submission will have been seen by peer reviewers, but not extensively reviewed. Any supplementary material provided at a later stage in the process may not have been peer reviewed. The supplementary materials (which include but are not limited to related publications, patient information leaflets and questionnaires) are provided to support and contextualise the publication. Every effort has been made to obtain the necessary permissions for reproduction, to credit original sources appropriately, and to respect copyright requirements. However, despite our diligence, we acknowledge the possibility of unintentional omissions or errors and we welcome notifications of any concerns regarding copyright or permissions. List of abbreviations ADAMTS‑13 a disintegrin‑like and metalloprotease with thrombospondin type 1 motif Ang-1 angiopoietin-1 Ang-2 angiopoietin-2 AUROC area under the receiver-operating characteristic curve CD cluster of differentiation antigens CENTRAL Cochrane Central Register of Controlled Trials CHARMS-PF CHecklist for critical Appraised and data extraction for systematic Reviews of prediction Modelling Studies and of Prognostic Factors studies CRB-65 modified version of CURB-65 (confusion, uraemia, respiratory rate, blood pressure, age ≥ 65 years) CRP C-reactive protein DNACPR do not attempt cardiopulmonary resuscitation DNI delta neutrophil index ED emergency department EDI equality, diversity and inclusion ENDOCAN endothelial cell-specific molecule-1 GM-CSF granulocyte–macrophage colony-stimulating factor HBP heparin-binding protein HDU high dependency HLA human leucocyte antigen hsCRP high-sensitivity C-reactive protein ICAM-1 intercellular adhesion molecule-1 ICU intensive care unit IFN interferon IL interleukin IMA ischaemia-modified albumin IMX-SEV-2 Inflammatix Severity 2 MDW monocyte distribution width MEDS Mortality Emergency Department Sepsis score miR microRNA miR146a micro-RNA 146a miR150 micro-RNA 150 miR223 micro-RNA 223 MR-proADM mid-regional proadrenomedullin NEWS2 National Early Warning Score 2 NGAL neutrophil gelatinase-associated lipocalin NLR neutrophil-to-lymphocyte ratio NT-pro BNP N-terminal prohormone of brain natriuretic peptide NWR neutrophil-to-white blood cell ratio OPN osteopontin PCT procalcitonin PLR platelet-to-lymphocyte ratio PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analysis PSEP Presepsin PTX3 pentraxin 3 qSOFA quick Sepsis-related Organ Failure Assessment score QUIPS Quality in Prognostic Factor Studies RASI respiratory adjusted shock index REMS Rapid Emergency Medicine Score SBP systolic blood pressure SIRS systemic inflammatory response syndrome SOFA Sepsis-related Organ Failure Assessment score SUPAR soluble urokinase-type plasminogen activator receptor SWAR Study Within a Review Tie-2 tyrosine kinase with immunoglobulin-like and epidermal growth factor-like domains 2 TNF tumour necrosis factor VCAM-1 vascular cell adhesion molecule-1 vWF von Willebrand factor References 1. Singer M, Deutschman CS, Seymour C, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801–10. [ PMC free article : PMC4968574 ] [ PubMed : 26903338 ] 2. von Groote T, Meersch-Dini M. Biomarkers for the prediction and judgement of sepsis and sepsis complications: a step towards precision medicine? J Clin Med 2022;11:5782. [ PMC free article : PMC9571838 ] [ PubMed : 36233650 ] 3. Rudd KE, Johnson SC, Agesa KM, Shackelford KA, Tsoi D, Kievlan DR, et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. Lancet 2020;395:200–11. [ PMC free article : PMC6970225 ] [ PubMed : 31954465 ] 4. UK Sepsis Trust. References and Sources. 2020. URL: https://sepsistrust ​.org ​/about/about-sepsis ​/references-and-sources/ (accessed 15 March 2024). 5. Bone RC, Balk RA, Cerra FB, Dellinger P, Fein AM, Knaus WA, et al. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992;20:864–74. 6. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, et al.; SCCM/ESICM/ACCP/ATS/SIS. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003;31:1250–6. [ PubMed : 12682500 ] 7. Royal College of Physicians. National Early Warning Score (NEWS) 2. 2017. URL: www ​.rcplondon.ac.uk/projects ​/outputs/national-early-warning-score-news-2 (accessed 15 March 2024). 8. Redfern OC, Smith GB, Prytherch DR, Meredith P, Inada-Kim M, Schmidt PE. A comparison of the quick sequential (sepsis-related) organ failure assessment score and the national early warning score in non-ICU patients with/without infection. Crit Care Med 2018;46:1923–33. [ PubMed : 30130262 ] 9. Levy MM, Evans LE, Rhodes A. The Surviving Sepsis Campaign Bundle: 2018 update. Crit Care Med 2018;46:997–1000. [ PubMed : 29767636 ] 10. Moons KG, de Groot JA, Bouwmeester W, Vergouwe Y, Mallett S, Altman DG, et al. Critical appraisal and data extraction for systematic reviews of prediction modelling studies: the CHARMS checklist. PLOS Med 2014;11:e1001744. [ PMC free article : PMC4196729 ] [ PubMed : 25314315 ] 11. Debray TP, Damen JA, Snell KI, Ensor J, Hooft L, Reitsma JB, et al. A guide to systematic review and meta-analysis of prediction model performance. BMJ 2017;356:i6460. [ PubMed : 28057641 ] 12. Riley RD, Moons KGM, Snell KIE, Ensor J, Hooft L, Altman DG, et al. A guide to systematic review and meta-analysis of prognostic factor studies. BMJ 2019;364:k4597. [ PubMed : 30700442 ] 13. Hayden JA, van der Windt DA, Cartwright JL, Côté P, Bombardier C. Assessing bias in studies of prognostic factors. Ann Intern Med 2013;158:280–6. [ PubMed : 23420236 ] 14. Hosmer D, Lemeshow SJ, Sturdivant R. Assessing the Fit of the Model. In Hosmer D, Lemeshow SJ, Sturdivant R, editors. Applied Logistic Regression. 3rd edn. Hoboken, NJ: John Wiley & Sons Inc; 2013. 15. Salgado J. Transforming the area under the normal curve (AUC) in Cohen’s d, Pearson’s RPB, odds-ratio, and natural log odds-ratio: two conversion tables. Eur J Psychol Appl Legal Context 2018;10:35–47. 16. Walter SD, Sinuff T. Studies reporting ROC curves of diagnostic and prediction data can be incorporated into meta-analyses using corresponding odds ratios. J Clin Epidemiol 2007;60:530–4. [ PubMed : 17419965 ] 17. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [ PubMed : 12111919 ] 18. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [ PMC free article : PMC2127453 ] [ PubMed : 9310563 ] 19. Barr SC, O’Neill TJ. The analysis of group truncated binary data with random effects: injury severity in motor vehicle accidents. Biometrics 2000;56:443–50. [ PubMed : 10877302 ] 20. Cottey L, Shanahan TAG, Gronlund T, Whiting C, Sokunbi M, Carley SD, Smith JE; James Lind Alliance (JLA) Emergency Medicine (EM) Priority Setting Partnership (PSP) Refresh Steering Group. Refreshing the emergency medicine research priorities. Emerg Med J 2023;40:666–70. [ PMC free article : PMC10447359 ] [ PubMed : 37491155 ] 21. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. [ PMC free article : PMC8005924 ] [ PubMed : 33782057 ] 22. Lee WJ, Woo SH, Kim DH, Seol SH, Park SK, Choi SP, et al. Are prognostic scores and biomarkers such as procalcitonin the appropriate prognostic precursors for elderly patients with sepsis in the emergency department? Aging Clin Exp Res 2016;28:917–24. [ PubMed : 26643799 ] 23. Hou SK, Lin HA, Chen SC, Lin CF, Lin SF. Monocyte distribution width, neutrophil-to-lymphocyte ratio, and platelet-to-lymphocyte ratio improves early prediction for sepsis at the emergency. J Pers Med 2021;11:732. [ PMC free article : PMC8402196 ] [ PubMed : 34442376 ] 24. Henning DJ, Puskarich MA, Self WH, Howell MD, Donnino MW, Yealy DM, et al. An emergency department validation of the SEP-3 sepsis and septic shock definitions and comparison with 1992 consensus definitions. Ann Emerg Med 2017;70:544–52.e5. [ PMC free article : PMC5792164 ] [ PubMed : 28262318 ] 25. Guarino M, Perna B, Cesaro AE, Spampinato MD, Previati R, Costanzini A, et al. Comparison between capillary and serum lactate levels in predicting short-term mortality of septic patients at the emergency department. Int J Mol Sci 2023;24:9121. [ PMC free article : PMC10252479 ] [ PubMed : 37298080 ] 26. Reshmi K, Oommen M, Belgundi P, Paul T, Mehta A. Prognostic role of N-terminal prohormone of brain natriuretic peptide for patients in the medical intensive care unit with severe sepsis. Lung India 2021;38:438–41. [ PMC free article : PMC8509162 ] [ PubMed : 34472521 ] 27. Singer AJ, Taylor M, Domingo A, Ghazipura S, Khorasonchi A, Thode HC, Shapiro NI. Diagnostic characteristics of a clinical screening tool in combination with measuring bedside lactate level in emergency department patients with suspected sepsis. Acad Emerg Med 2014;21:853–7. [ PubMed : 25155163 ] 28. Magrini L, Gagliano G, Travaglino F, Vetrone F, Marino R, Cardelli P, et al. Comparison between white blood cell count, procalcitonin and C reactive protein as diagnostic and prognostic biomarkers of infection or sepsis in patients presenting to the emergency department. Clin Chem Lab Med 2014;52:1465–72. [ PubMed : 24803611 ] 29. Puskarich MA, Nandi U, Shapiro NI, Trzeciak S, Kline JA, Jones AE. Detection of microRNAs in patients with sepsis. J Acute Dis 2015;4:101–6. 30. Lee JH, Kim SH, Jang JH, Park JH, Jo KM, No TH, et al. Clinical usefulness of biomarkers for diagnosis and prediction of prognosis in sepsis and septic shock. Medicine (United States) 2022;101:E31895. [ PMC free article : PMC9726295 ] [ PubMed : 36482619 ] 31. Devia Jaramillo G, Ibanez Pinilla M. Quick Sequential Organ Failure Assessment, Sequential Organ Failure Assessment, and procalcitonin for early diagnosis and prediction of death in elderly patients with suspicion of sepsis in the emergency department, based on sepsis-3 definition. Gerontology 2022;68:171–80. [ PubMed : 33951628 ] 32. Suttapanit K, Wisan M, Sanguanwit P, Prachanukool T. Prognostic accuracy of VqSOFA for predicting 28-day mortality in patients with suspected sepsis in the emergency department. Shock 2021;56:368–73. [ PubMed : 33577246 ] 33. Noparatkailas N, Inchai J, Deesomchok A. Blood lactate level and the predictor of death in non-shock septic patients. Indian J Crit Care Med 2023;27:93–100. [ PMC free article : PMC9973057 ] [ PubMed : 36865504 ] 34. Covino M, Manno A, De Matteis G, Taddei E, Carbone L, Piccioni A, et al. Prognostic role of serum procalcitonin measurement in adult patients admitted to the emergency department with fever. Antibiotics (Basel, Switzerland) 2021;10:788. [ PMC free article : PMC8300691 ] [ PubMed : 34209605 ] 35. Katsaros K, Renieris G, Safarika A, Adami EM, Gkavogianni T, Giannikopoulos G, et al. Heparin binding protein for the early diagnosis and prognosis of sepsis in the emergency department: the PROMPT multicenter study. Shock 2022;57:518–25. [ PubMed : 34907118 ] 36. Jiang L, Caputo ND, Chang BP. Respiratory adjusted shock index for identifying occult shock and level of Care in Sepsis Patients. Am J Emerg Med 2019;37:506–9. [ PMC free article : PMC7296509 ] [ PubMed : 30674427 ] 37. Macdonald SPJ, Bosio E, Neil C, Arendts G, Burrows S, Smart L, et al. Resistin and NGAL are associated with inflammatory response, endothelial activation and clinical outcomes in sepsis. Inflamm Res 2017;66:611–9. [ PubMed : 28424824 ] 38. Malinovska A, Hinson JS, Badaki-Makun O, Hernried B, Smith A, Debraine A, et al. Monocyte distribution width as part of a broad pragmatic sepsis screen in the emergency department. J Am Coll Emerg Physicians Open 2022;3:e12679. [ PMC free article : PMC8886187 ] [ PubMed : 35252973 ] 39. Shim BS, Yoon YH, Kim JY, Cho YD, Park SJ, Lee ES, Choi SH. Clinical value of whole blood procalcitonin using point of care testing, quick sequential organ failure assessment score, C-reactive protein and lactate in emergency department patients with suspected infection. J Clin Med 2019;8:833. [ PMC free article : PMC6617302 ] [ PubMed : 31212806 ] 40. Sohn YW, Jang HY, Park S, Lee Y, Cho YS, Park J, et al. Validation of quick sequential organ failure assessment score for poor outcome prediction among emergency department patients with suspected infection. Clin Exp Emerg Med 2019;6:314–20. [ PMC free article : PMC6952629 ] [ PubMed : 31910502 ] 41. Kim SJ, Hwang SO, Kim YW, Lee JH, Cha KC. Procalcitonin as a diagnostic marker for sepsis/septic shock in the emergency department; a study based on Sepsis-3 definition. Am J Emerg Med 2019;37:272–6. [ PubMed : 29861371 ] 42. Kece E, Yaka E, Yilmaz S, Dogan NT, Alyesil C, Pekdemir M. Comparison of diagnostic and prognostic utility of lactate and procalcitonin for sepsis in adult cancer patients presenting to emergency department with systemic inflammatory response syndrome. Turk J Emerg Med 2016;16:1–7. [ PMC free article : PMC4882194 ] [ PubMed : 27239630 ] 43. Hong DY, Kim JW, Paik JH, Jung HM, Baek KJ, Park SO, Lee KR. Value of plasma neutrophil gelatinase-associated lipocalin in predicting the mortality of patients with sepsis at the emergency department. Clin Chim Acta 2016;452:177–81. [ PubMed : 26626454 ] 44. Xie Y, Li B, Lin Y, Shi F, Chen W, Wu W, et al. Combining blood-based biomarkers to predict mortality of sepsis at arrival at the emergency department. Med Sci Monit 2021;27:e929527. [ PMC free article : PMC7923396 ] [ PubMed : 33630815 ] 45. Zhang Q, Li CS. Risk stratification and prognostic evaluation of endothelial cell-specific molecule1, von Willebrand factor, and a disintegrin-like and metalloprotease with thrombospondin type 1 motif for sepsis in the emergency department: an observational study. Exp Ther Med 2019;17:4527–35. [ PMC free article : PMC6488990 ] [ PubMed : 31086585 ] 46. Jekarl DW, Lee S, Kim M, Kim Y, Woo SH, Lee WJ. Procalcitonin as a prognostic marker for sepsis based on SEPSIS-3. J Clin Lab Anal 2019;33:e22996. [ PMC free article : PMC6868407 ] [ PubMed : 31420921 ] 47. Musikatavorn K, Thepnimitra S, Komindr A, Puttaphaisan P, Rojanasarntikul D. Venous lactate in predicting the need for intensive care unit and mortality among nonelderly sepsis patients with stable hemodynamic. Am J Emerg Med 2015;33:925–30. [ PubMed : 25936479 ] 48. Ueno R, Masubuchi T, Shiraishi A, Gando S, Abe T, Kushimoto S, et al. Quick sequential organ failure assessment score combined with other sepsis-related risk factors to predict in-hospital mortality: post-hoc analysis of prospective multicenter study data. PLOS ONE 2021;16:e0254343. [ PMC free article : PMC8282038 ] [ PubMed : 34264977 ] 49. Klimpel J, Weidhase L, Bernhard M, Gries A, Petros S. The impact of the Sepsis-3 definition on ICU admission of patients with infection. Scand J Trauma Resusc Emerg Med 2019;27:98. [ PMC free article : PMC6829802 ] [ PubMed : 31685006 ] 50. Hunter CL, Silvestri S, Dean M, Falk JL, Papa L. End-tidal carbon dioxide is associated with mortality and lactate in patients with suspected sepsis. Am J Emerg Med 2013;31:64–71. [ PubMed : 22867820 ] 51. Shetty AL, Thompson K, Byth K, Macaskill P, Green M, Fullick M, et al. Serum lactate cut-offs as a risk stratification tool for in-hospital adverse outcomes in emergency department patients screened for suspected sepsis. BMJ Open 2018;8:e015492. [ PMC free article : PMC5780682 ] [ PubMed : 29306875 ] 52. Uusitalo-Seppala R, Huttunen R, Aittoniemi J, Koskinen P, Leino A, Vahlberg T, Rintala EM. Pentraxin 3 (PTX3) is associated with severe sepsis and fatal disease in emergency room patients with suspected infection: a prospective cohort study. PLOS ONE 2013;8:e53661. [ PMC free article : PMC3544919 ] [ PubMed : 23341967 ] 53. Gonzalez Del Castillo J, Wilson DC, Clemente-Callejo C, Román F, Bardés-Robles I, Jiménez I, et al.; INFURG-SEMES investigators. Biomarkers and clinical scores to identify patient populations at risk of delayed antibiotic administration or intensive care admission. Crit Care 2019;23:335. [ PMC free article : PMC6819475 ] [ PubMed : 31665092 ] 54. Fang Y, Li C, Shao R, Yu H, Zhang Q, Zhao L. Prognostic significance of the angiopoietin-2/angiopoietin-1 and angiopoietin-1/Tie-2 ratios for early sepsis in an emergency department. Crit Care 2015;19:367. [ PMC free article : PMC4604731 ] [ PubMed : 26463042 ] 55. Dudaryk R, Navas-Blanco JR, Ferreira TD, Epstein RH. Failure to clear intermediate lactate levels in ward patients with admission blood cultures did not increase the risk of intensive care unit transfer or in-hospital mortality: a retrospective cohort study. Cureus 2021;13:e13326. [ PMC free article : PMC7958552 ] [ PubMed : 33738169 ] 56. D’Onofrio V, Meersman A, Vijgen S, Cartuyvels R, Messiaen P, Gyssens IC. Risk factors for mortality, intensive care unit admission, and bacteremia in patients suspected of sepsis at the emergency department: a prospective cohort study. Open Forum Infect Dis 2021;8:ofaa594. [ PMC free article : PMC7813192 ] [ PubMed : 33511231 ] 57. Choo SH, Lim YS, Cho JS, Jang JH, Choi JY, Choi WS, Yang HJ. Usefulness of ischemia-modified albumin in the diagnosis of sepsis/septic shock in the emergency department. Clin Exp Emerg Med 2020;7:161–9. [ PMC free article : PMC7550814 ] [ PubMed : 33028058 ] 58. Ruangsomboon O, Panjaikaew P, Monsomboon A, Chakorn T, Permpikul C, Limsuwat C. Diagnostic and prognostic utility of presepsin for sepsis in very elderly patients in the emergency department. Clin Chim Acta 2020;510:723–32. [ PubMed : 32946797 ] 59. Castello LM, Baldrighi M, Molinari L, Salmi L, Cantaluppi V, Vaschetto R, et al. The role of osteopontin as a diagnostic and prognostic biomarker in sepsis and septic shock. Cells 2019;8:174. [ PMC free article : PMC6407102 ] [ PubMed : 30781721 ] 60. Hargreaves DS, de Carvalho JLJ, Smith L, Picton G, Venn R, Hodgson LE. Persistently elevated early warning scores and lactate identifies patients at high risk of mortality in suspected sepsis. Eur J Emerg Med 2019;27:125–31. [ PubMed : 31464702 ] 61. Bolanaki M, Mockel M, Winning J, Bauer M, Reinhart K, Stacke A, et al. Diagnostic performance of procalcitonin for the early identification of sepsis in patients with elevated qsofa score at emergency admission. J Clin Med 2021;10:3869. [ PMC free article : PMC8432218 ] [ PubMed : 34501324 ] 62. Caramello V, Beux V, de Salve AV, Macciotta A, Ricceri F, Boccuzzi A. Comparison of different prognostic scores for risk stratification in septic patients arriving to the emergency department. Ital J Med 2020;14:79–87. 63. Gunes Ozaydin M, Guneysel O, Saridogan F, Ozaydin V. Are scoring systems sufficient for predicting mortality due to sepsis in the emergency department? Turk J Emerg Med 2017;17:25–8. [ PMC free article : PMC5357089 ] [ PubMed : 28345070 ] 64. Cheng HH, Chen FC, Change MW, Kung CT, Cheng CY, Tsai TC, et al. Difference between elderly and non-elderly patients in using serum lactate level to predict mortality caused by sepsis in the emergency department. Medicine (Baltimore) 2018;97:e0209. [ PMC free article : PMC5895436 ] [ PubMed : 29595662 ] 65. Deeks J, Higgins J, Altman D. Chapter 10: Analysing Data and Undertaking Meta-Analyses. In Cochrane Handbook for Systematic Reviews of Interventions version 64. Cochrane; 2023. URL: www ​.training.cochrane.org/handbook (accessed 15 March 2024). 66. Wunsch H, Angus DC, Harrison DA, Linde-Zwirble WT, Rowan KM. Comparison of medical admissions to intensive care units in the United States and United Kingdom. Am J Respir Crit Care Med 2011;183:1666–73. [ PubMed : 21471089 ] 67. Parke A, Unge C, Yu D, Sunden-Cullberg J, Stralin K. Plasma calprotectin as an indicator of need of transfer to intensive care in patients with suspected sepsis at the emergency department. BMC Emerg Med 2023;23:16. [ PMC free article : PMC9922172 ] [ PubMed : 36774492 ] 68. Academy of Medical Royal Colleges. Statement on the Initial Antimicrobial Treatment of Sepsis. London: Academy of Medical Royal Colleges; 2022. URL: www ​.aomrc.org.uk/wp-content ​/uploads/2022 ​/10/Statement_on_the ​_initial_antimicrobial ​_treatment_of_sepsis_V2_1022.pdf (accessed 15 March 2024). 69. Athan S, Athan D, Wong M, Hussain N, Vangaveti V, Gangathimmaiah V, Norton R. Pathology stewardship in emergency departments: a single-site, retrospective, cohort study of the value of C-reactive protein in patients with suspected sepsis. Pathology (Phila) 2023;55:673–9. [ PubMed : 37248118 ] 70. Baumann BM, Greenwood JC, Lewis K, Nuckton TJ, Darger B, Shofer FS, et al. Combining qSOFA criteria with initial lactate levels: improved screening of septic patients for critical illness. Am J Emerg Med 2020;38:883–9. [ PubMed : 31320214 ] 71. Rodriguez RM, Greenwood JC, Nuckton TJ, Darger B, Shofer FS, Troeger D, et al. Comparison of qSOFA with current emergency department tools for screening of patients with sepsis for critical illness. Emerg Med J 2018;35:350–6. [ PubMed : 29720475 ] 72. Boland LL, Hokanson JS, Fernstrom KM, Kinzy TG, Lick CJ, Satterlee PA, LaCroix BK. Prehospital lactate measurement by emergency medical services in patients meeting sepsis criteria. West J Emerg Med 2016;17:648–55. [ PMC free article : PMC5017855 ] [ PubMed : 27625735 ] 73. Contenti J, Corraze H, Lemoel F, Levraut J. Effectiveness of arterial, venous, and capillary blood lactate as a sepsis triage tool in ED patients. Am J Emerg Med 2015;33:167–72. [ PubMed : 25432592 ] 74. Dadeh AA, Kulparat M. Predictive performance of the NEWS-Lactate and NEWS towards mortality or need for critical care among patients with suspicion of sepsis in the emergency department: a prospective observational study. Open Access Emerg Med 2022;14:619–31. [ PMC free article : PMC9677920 ] [ PubMed : 36419573 ] 75. Kostaki A, Wacker JW, Safarika A, Solomonidi N, Katsaros K, Giannikopoulos G, et al. A 29-MRNA host response whole-blood signature improves prediction of 28-day mortality and 7-day intensive care unit care in adults presenting to the emergency department with suspected acute infection and/or sepsis. Shock 2022;58:224–30. [ PMC free article : PMC9512237 ] [ PubMed : 36125356 ] 76. Lucas G, Bartolf A, Kroll N, De Thabrew AU, Murtaza Z, Kumar S, et al. Procalcitonin (PCT) level in the emergency department identifies a high-risk cohort for all patients treated for possible sSepsis. EJIFCC 2021;32:20–6. [ PMC free article : PMC7941056 ] [ PubMed : 33753971 ] 77. Magrini L, Travaglino F, Marino R, Ferri E, De Berardinis B, Cardelli P, et al. Procalcitonin variations after emergency department admission are highly predictive of hospital mortality in patients with acute infectious diseases. Eur Rev Med Pharmacol Sci 2013;17:133–42. [ PubMed : 23436675 ] Appendix 1. Clinical criteria for sepsis TABLE 2 Components of SIRS , qSOFA and NEWS2 for sepsis View in own window SIRS criteria 5 Meet two or more of the following: Temperature > 38 °C or < 36 °C Heart rate > 90/minute Respiratory rate > 20/minute or PaCO 2 < 32 mmHg (4.3 kPa) White blood cell count > 12,000/mm 3 or < 4000/mm 3 or > 10% immature bands qSOFA score 1 Meet two or more of the following: Respiratory (rate ≥ 22/minute) Change in mental SBP ≤ 100 mmHg NEWS2 criteria 7 1. Respiration rate 2. Oxygen saturation 3. SBP 4. Pulse rate 5. Level of consciousness or new confusion a 6. Temperature. a The patient has new-onset confusion, disorientation and/or agitation, where previously their mental state was normal – this may be subtle. The patient may respond to questions coherently, but there is some confusion, disorientation and/or agitation. This would score 3 or 4 on the Glasgow Coma Scale (rather than the normal five for verbal response) and score 3 on the NEWS system. Appendix 2. Literature search strategies All searches were run on 5 July 2023 Search strategy Ovid MEDLINE (R) and Epub Ahead of Print, In-Process, In-Data-Review and Other Non-Indexed Citations, Daily and Versions <1946–July 03, 2023> Sepsis/ or Septic Shock/ Systemic Inflammatory Response Syndrome/ (sepsis or “septic shock” or SIRS or “Systemic Inflammatory Response Syndrome”).tw. 1 or 2 or 3 *Procalcitonin/ or Procalcitonin.tw. *Receptors, Urokinase Plasminogen Activator/ or SUPAR .tw. exp *Cytokines/ or (interleukin*, or TNF *, or IFN gamma or GM-CSF).tw. *C-Reactive Protein/ or “C-reactive protein”.tw. exp *Phenotype/ or *Phenomics/ or (phenotype? or sub-phenotype? or subphenotype? or clinicomolecular or metabolic? or metabolomic?).tw. Gene expression profiling/ exp Genetic structures/ transcriptome.tw. (“monocyte distribution width” or MDW).tw. lactate.tw. (CD15 or CD24 or CD35 or CD64 or CD312 or CD11b or CD274 or CD279 or CD35 or CD64 or CD312 or CD11b or HLA-DR or CD274 or CD279).tw. or/5-15 Emergency Service, Hospital/ (emergency adj5 (room? or service? or department? or ward? or admit* or admission? or triage or care or hospital? or physician?)).tw. 17 or 18 exp mortality/ or follow up studies/ exp risk/ exp cohort studies/ exp prognosis/ exp incidence/ exp survival analysis/ (prognos* or outcome? or predict* or risk or cohort or incidence or survival or causal factors or course).tw. or/20-26 4 and 16 and 19 and 27 limit 28 to yr=“2013 -Current” EMBASE <1974–2023 Week 26> sepsis/ or septic shock/ systemic inflammatory response syndrome/ (sepsis or “septic shock” or SIRS or “Systemic Inflammatory Response Syndrome”).tw. 1 or 2 or 3 *Procalcitonin/ or Procalcitonin.tw. urokinase receptor/ or SUPAR .tw. exp *cytokine/ or (interleukin*, or TNF *, or IFN gamma or GM-CSF).tw. *C-Reactive Protein/ or “C-reactive protein”.tw. exp *Phenotype/ or *Phenomics/ or (phenotype? or sub-phenotype? or subphenotype? or clinicomolecular or metabolic? or metabolomic?).tw. gene expression profiling/ exp gene structure/ transcriptome.tw. (“monocyte distribution width” or MDW).tw. lactate.tw. (CD15 or CD24 or CD35 or CD64 or CD312 or CD11b or CD274 or CD279 or CD35 or CD64 or CD312 or CD11b or HLA-DR or CD274 or CD279).tw. or/5-15 emergency ward/ (emergency adj5 (room? or service? or department? or ward? or admit* or admission? or triage or care or hospital? or physician?)).tw. 17 or 18 mortality/ or follow up/ risk/ cohort analysis/ prognosis/ exp incidence/ survival analysis/ (prognos* or outcome? or predict* or risk or cohort or incidence or survival or causal factors or course).tw. or/20-26 4 and 16 and 19 and 27 conference abstract.pt. 28 not 29 limit 30 to yr=“2013 -Current” Cochrane Library # 1. MeSH descriptor: [Sepsis] this term only # 2. MeSH descriptor: [Shock, Septic] this term only # 3. MeSH descriptor: [Systemic Inflammatory Response Syndrome] explode all trees # 4. sepsis or “septic shock” or SIRS or “Systemic Inflammatory Response Syndrome” # 5. #1 or #2 or #3 or #4 # 6. MeSH descriptor: [Procalcitonin] this term only # 7. Procalcitonin # 8. MeSH descriptor: [Receptors, Urokinase Plasminogen Activator] this term only # 9. SUPAR # 10. MeSH descriptor: [Cytokines] explode all trees # 11. interleukin*, or TNF *, or IFN gamma or GM-CSF # 12. MeSH descriptor: [C-Reactive Protein] this term only # 13. “C-reactive protein” # 14. MeSH descriptor: [Phenotype] this term only # 15. MeSH descriptor: [Phenomics] explode all trees # 16. phenotype? or sub-phenotype? or subphenotype? or clinicomolecular or metabolic? or metabolomic? # 17. MeSH descriptor: [Gene Expression Profiling] explode all trees # 18. transcriptome # 19. “monocyte distribution width” or MDW # 20. lactate # 21. CD15 or CD24 or CD35 or CD64 or CD312 or CD11b or CD274 or CD279 or CD35 or CD64 or CD312 or CD11b or HLA-DR or CD274 or CD279 # 22. #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 # 23. MeSH descriptor: [Emergency Service, Hospital] this term only # 24. emergency near/5 (room? or service? or department? or ward? or admit* or admission? or triage or care or hospital? or physician?) # 25. #23 or #24 # 26. MeSH descriptor: [Mortality] explode all trees # 27. MeSH descriptor: [Follow-Up Studies] this term only # 28. MeSH descriptor: [Risk] explode all trees # 29. MeSH descriptor: [Cohort Studies] explode all trees # 30. MeSH descriptor: [Prognosis] explode all trees # 31. MeSH descriptor: [Incidence] explode all trees # 32. MeSH descriptor: [Survival Analysis] explode all trees # 33. prognos* or outcome? or predict* or risk or cohort or incidence or survival or “causal factors” or course # 34. #26 or #27 or #28 or #29 or #30 or #31 or #32 or #33 # 35. #5 and #22 and #25 and #34 Appendix 3. List of included studies * denotes primary reference Ref ID = reference ID number assigned by the project Athan, 2023 69 Athan S, Athan D, Wong M, Hussain N, Vangaveti V, Gangathimmaiah V, et al . Pathology stewardship in emergency departments: a single-site, retrospective, cohort study of the value of C-reactive protein in patients with suspected sepsis. Pathology 2023; 55 :673–9. [Ref ID: 248] Baumann, 2020 70 * Baumann BM, Greenwood JC, Lewis K, Nuckton TJ, Darger B, Shofer FS, et al . Combining qSOFA criteria with initial lactate levels: improved screening of septic patients for critical illness. Am J Emerg Med 2020; 38 :883–9. [Ref ID: 698] Rodriguez RM, Greenwood JC, Nuckton TJ, Darger B, Shofer FS, Troeger D, et al . Comparison of qSOFA with current emergency department tools for screening of patients with sepsis for critical illness. Emerg Med J 2018; 35 :350–6. [Ref ID: 874] 71 Bolanaki, 2021 61 Bolanaki M, Mockel M, Winning J, Bauer M, Reinhart K, Stacke A, et al . Diagnostic performance of procalcitonin for the early identification of sepsis in patients with elevated qsofa score at emergency admission. J Clin Med 2021; 10 :3869. [Ref ID: 541] Boland, 2016 72 Boland LL, Hokanson JS, Fernstrom KM, Kinzy TG, Lick CJ, Satterlee PA, et al . Prehospital lactate measurement by emergency medical services in patients meeting sepsis criteria. West J Emerg Med 2016; 17 :648–55. [Ref ID: 1776] Caramello, 2020 62 Caramello V, Beux V, de Salve AV, Macciotta A, Ricceri F, Boccuzzi A. Comparison of different prognostic scores for risk stratification in septic patients arriving to the emergency department. Ital J Med 2020; 14 :79–87. [Ref ID: 700] Castello, 2019 59 Castello LM, Baldrighi M, Molinari L, Salmi L, Cantaluppi V, Vaschetto R, et al . The role of osteopontin as a diagnostic and prognostic biomarker in sepsis and septic shock. Cells 2019; 8 :174. [Ref ID: 725] Cheng, 2018 64 Cheng HH, Chen FC, Change MW, Kung CT, Cheng CY, Tsai TC, et al . Difference between elderly and non-elderly patients in using serum lactate level to predict mortality caused by sepsis in the emergency department. Medicine (United States) 2018; 97 :e0209. [Ref ID: 884] Choo, 2020 57 Choo SH, Lim YS, Cho JS, Jang JH, Choi JY, Choi WS, et al . Usefulness of ischemia-modified albumin in the diagnosis of sepsis/septic shock in the emergency department. Clin Exp Emerg Med 2020; 7 :161–9. [Ref ID: 669] Contenti, 2015 73 Contenti J, Corraze H, Lemoel F, Levraut J. Effectiveness of arterial, venous, and capillary blood lactate as a sepsis triage tool in ED patients. Am J Emerg Med 2015; 33 :167–72. [Ref ID: 1065] Covino, 2021 34 Covino M, Manno A, De Matteis G, Taddei E, Carbone L, Piccioni A, et al . Prognostic role of serum procalcitonin measurement in adult patients admitted to the emergency department with fever. Antibiotics 2021; 10 :788. [Ref ID: 334] D’Onofrio, 2021 56 D’Onofrio V, Meersman A, Vijgen S, Cartuyvels R, Messiaen P, Gyssens IC. Risk factors for mortality, intensive care unit admission, and bacteremia in patients suspected of sepsis at the emergency department: a prospective cohort study. Open Forum Infect Dis 2021; 8 :ofaa594. [Ref ID: 603] Dadeh, 2022 74 Dadeh AA, Kulparat M. Predictive performance of the NEWS-lactate and NEWS towards mortality or need for critical care among patients with suspicion of sepsis in the emergency department: a prospective observational study. Open Access Emerg Med 2022; 14 :619–31. [Ref ID: 402] Devia Jaramillo, 2022 31 Devia Jaramillo G, Ibanez Pinilla M. Quick Sequential Organ Failure Assessment, Sequential Organ Failure Assessment, and procalcitonin for early diagnosis and prediction of death in elderly patients with suspicion of sepsis in the emergency department, based on sepsis-3 definition. Gerontology 2022; 68 :171–80. [Ref ID: 475] Dudaryk, 2021 55 Dudaryk R, Navas-Blanco JR, Ferreira TD, Epstein RH. Failure to clear intermediate lactate levels in ward patients with admission blood cultures did not increase the risk of intensive care unit transfer or in-hospital mortality: a retrospective cohort study. Cureus 2021; 13 :e13326. [Ref ID: 1308] Fang, 2015 54 Fang Y, Li C, Shao R, Yu H, Zhang Q, Zhao L. Prognostic significance of the angiopoietin-2/angiopoietin-1 and angiopoietin-1/ Tie-2 ratios for early sepsis in an emergency department. Crit Care 2015; 19 :367. [Ref ID: 1054] Gonzalez Del Castillo, 2019 53 Gonzalez Del Castillo J, Wilson DC, Clemente-Callejo C, Roman F, Bardes-Robles I, Jimenez I, et al . Biomarkers and clinical scores to identify patient populations at risk of delayed antibiotic administration or intensive care admission. Crit Care 2019; 23 :335. [Ref ID: 776] Guarino, 2023 25 Guarino M, Perna B, Cesaro AE, Spampinato MD, Previati R, Costanzini A, et al . Comparison between capillary and serum lactate levels in predicting short-term mortality of septic patients at the emergency department. Int J Mol Sci 2023; 24 :9121. [Ref ID: 232] Gunes Ozaydin, 2017 63 Gunes Ozaydin M, Guneysel O, Saridogan F, Ozaydin V. Are scoring systems sufficient for predicting mortality due to sepsis in the emergency department? Turk J Emerg Med 2017; 17 :25–8. [Ref ID: 950] Hargreaves, 2019 60 Hargreaves DS, de Carvalho JLJ, Smith L, Picton G, Venn R, Hodgson LE. Persistently elevated early warning scores and lactate identifies patients at high risk of mortality in suspected sepsis. Eur J Emerg Med 2019; 27 :125–31. [Ref ID: 369] Henning, 2017 24 Henning DJ, Puskarich MA, Self WH, Howell MD, Donnino MW, Yealy DM, et al . An emergency department validation of the SEP-3 sepsis and septic shock definitions and comparison with 1992 consensus definitions. Ann Emerg Med 2017; 70 :544–52.e5. [Ref ID: 920] Hong, 2016 43 Hong DY, Kim JW, Paik JH, Jung HM, Baek KJ, Park SO, et al . Value of plasma neutrophil gelatinase-associated lipocalin in predicting the mortality of patients with sepsis at the emergency department. Clin Chim Acta 2016; 452 :177–81. [Ref ID: 998] Hou, 2021 23 Hou SK, Lin HA, Chen SC, Lin CF, Lin SF. Monocyte distribution width, neutrophil-to-lymphocyte ratio, and platelet-to-lymphocyte ratio improves early prediction for sepsis at the emergency. J Pers Med 2021; 11 :732. [Ref ID: 333] Hunter, 2013 50 Hunter CL, Silvestri S, Dean M, Falk JL, Papa L. End-tidal carbon dioxide is associated with mortality and lactate in patients with suspected sepsis. Am J Emerg Med 2013; 31 :64–71. [Ref ID: 1160] Jekarl, 2019 46 Jekarl DW, Lee S, Kim M, Kim Y, Woo SH, Lee WJ. Procalcitonin as a prognostic marker for sepsis based on SEPSIS-3. J Clin Lab Anal 2019; 33 :e22996. [Ref ID: 771] Jiang, 2019 36 Jiang L, Caputo ND, Chang BP. Respiratory adjusted shock index for identifying occult shock and level of Care in Sepsis Patients. Am J Emerg Med 2019; 37 :506–9. [Ref ID: 763] Katsaros, 2022 35 * Katsaros K, Renieris G, Safarika A, Adami EM, Gkavogianni T, Giannikopoulos G, et al . Heparin binding protein for the early diagnosis and prognosis of sepsis in the emergency department: the PROMPT multicenter study. Shock 2022; 57 :518–25. [Ref ID: 477] Kostaki A, Wacker JW, Safarika A, Solomonidi N, Katsaros K, Giannikopoulos G, et al . A 29-MRNA host response whole-blood signature improves prediction of 28-day mortality and 7-day intensive care unit care in adults presenting to the emergency department with suspected acute infection and/or sepsis. Shock 2022; 58 :224–30. [Ref ID: 425] 75 Kece, 2016 42 Kece E, Yaka E, Yilmaz S, Dogan NT, Alyesil C, Pekdemir M. Comparison of diagnostic and prognostic utility of lactate and procalcitonin for sepsis in adult cancer patients presenting to emergency department with systemic inflammatory response syndrome. Turk J Emerg Med 2016; 16 :1–7. [Ref ID: 956] Kim, 2019 41 Kim SJ, Hwang SO, Kim YW, Lee JH, Cha KC. Procalcitonin as a diagnostic marker for sepsis/septic shock in the emergency department; a study based on Sepsis-3 definition. Am J Emerg Med 2019; 37 :272–6. [Ref ID: 766] Klimpel, 2019 49 Klimpel J, Weidhase L, Bernhard M, Gries A, Petros S. The impact of the Sepsis-3 definition on ICU admission of patients with infection. Scand J Trauma Resusc Emerg Med 2019; 27 :98. [Ref ID: 733] Lee, 2016 22 Lee WJ, Woo SH, Kim DH, Seol SH, Park SK, Choi SP, et al . Are prognostic scores and biomarkers such as procalcitonin the appropriate prognostic precursors for elderly patients with sepsis in the emergency department? Aging Clin Exp Res 2016; 28 :917–24. [Ref ID: 981] Lee, 2022 30 Lee JH, Kim SH, Jang JH, Park JH, Jo KM, No TH, et al . Clinical usefulness of biomarkers for diagnosis and prediction of prognosis in sepsis and septic shock. Medicine (United States) 2022; 101 : E31895 . [Ref ID: 385] Lucas, 2021 76 Lucas G, Bartolf A, Kroll N, De Thabrew AU, Murtaza Z, Kumar S, et al . Procalcitonin (PCT) level in the emergency department identifies a high-risk cohort for all patients treated for possible sSepsis. EJIFCC 2021; 32 :20–6. [Ref ID: 1307] Macdonald, 2017 37 Macdonald SPJ, Bosio E, Neil C, Arendts G, Burrows S, Smart L, et al . Resistin and NGAL are associated with inflammatory response, endothelial activation and clinical outcomes in sepsis. Inflamm Res 2017; 66 :611–9. [Ref ID: 910] Magrini, 2013 77 Magrini L, Travaglino F, Marino R, Ferri E, De Berardinis B, Cardelli P, et al . Procalcitonin variations after emergency department admission are highly predictive of hospital mortality in patients with acute infectious diseases. Eur Rev Med Pharmacol Sci 2013; 17 (Suppl. 1):133–42. [Ref ID: 1194] Magrini, 2014 28 Magrini L, Gagliano G, Travaglino F, Vetrone F, Marino R, Cardelli P, et al . Comparison between white blood cell count, procalcitonin and C reactive protein as diagnostic and prognostic biomarkers of infection or sepsis in patients presenting to the emergency department. Clin Chem Lab Med 2014; 52 :1465–72. [Ref ID: 1099] Malinovska, 2022 38 Malinovska A, Hinson JS, Badaki-Makun O, Hernried B, Smith A, Debraine A, et al . Monocyte distribution width as part of a broad pragmatic sepsis screen in the emergency department. J Am Coll Emerg Physicians Open 2022; 3 :e12679. [Ref ID: 270] Musikatavorn, 2015 47 Musikatavorn K, Thepnimitra S, Komindr A, Puttaphaisan P, Rojanasarntikul D. Venous lactate in predicting the need for intensive care unit and mortality among nonelderly sepsis patients with stable hemodynamic. Am J Emerg Med 2015; 33 :925–30. [Ref ID: 1062] Noparatkailas, 2023 33 Noparatkailas N, Inchai J, Deesomchok A. Blood lactate level and the predictor of death in non-shock septic patients. Indian J Crit Care Med 2023; 27 :93–100. [Ref ID: 308] Parke, 2023 67 Parke A, Unge C, Yu D, Sunden-Cullberg J, Stralin K. Plasma calprotectin as an indicator of need of transfer to intensive care in patients with suspected sepsis at the emergency department. BMC Emerg Med 2023; 23 :16. [Ref ID: 305] Puskarich, 2015 29 Puskarich MA, Nandi U, Shapiro NI, Trzeciak S, Kline JA, Jones AE. Detection of microRNAs in patients with sepsis. J Acute Dis 2015; 4 :101–6. [Ref ID: 1063] Reshmi, 2021 26 Reshmi K, Oommen M, Belgundi P, Paul T, Mehta A. Prognostic role of N-terminal prohormone of brain natriuretic peptide for patients in the medical intensive care unit with severe sepsis. Lung India 2021; 38 :438–41. [Ref ID: 556] Ruangsomboon, 2020 58 Ruangsomboon O, Panjaikaew P, Monsomboon A, Chakorn T, Permpikul C, Limsuwat C. Diagnostic and prognostic utility of presepsin for sepsis in very elderly patients in the emergency department. Clin Chim Acta 2020; 510: 723–32. [Ref ID: 682] Shetty, 2018 51 Shetty AL, Thompson K, Byth K, Macaskill P, Green M, Fullick M, et al . Serum lactate cut-offs as a risk stratification tool for in-hospital adverse outcomes in emergency department patients screened for suspected sepsis. BMJ Open 2018; 8 :015492. [Ref ID: 876] Shim, 2019 39 Shim BS, Yoon YH, Kim JY, Cho YD, Park SJ, Lee ES, et al . Clinical value of whole blood procalcitonin using point of care testing, quick sequential organ failure assessment score, c-reactive protein and lactate in emergency department patients with suspected infection. J Clin Med 2019; 8 :833. [Ref ID: 726] Singer, 2014 27 Singer AJ, Taylor M, Domingo A, Ghazipura S, Khorasonchi A, Thode HC Jr, et al . Diagnostic characteristics of a clinical screening tool in combination with measuring bedside lactate level in emergency department patients with suspected sepsis. Acad Emerg Med 2014; 21 :853–7. [Ref ID: 1124] Sohn, 2019 40 Sohn YW, Jang HY, Park S, Lee Y, Cho YS, Park J, et al . Validation of quick sequential organ failure assessment score for poor outcome prediction among emergency department patients with suspected infection. Clin Exp Emerg Med 2019; 6 :314–20. [Ref ID: 740] Suttapanit, 2021 32 Suttapanit K, Wisan M, Sanguanwit P, Prachanukool T. Prognostic accuracy of VqSOFA for predicting 28-day mortality in patients with suspected sepsis in the emergency department. Shock 2021; 56 :368–73. [Ref ID: 530] Ueno, 2021 48 Ueno R, Masubuchi T, Shiraishi A, Gando S, Abe T, Kushimoto S, et al . Quick sequential organ failure assessment score combined with other sepsis-related risk factors to predict in-hospital mortality: post-hoc analysis of prospective multicenter study data. PLOS ONE 2021; 16 :e0254343. [Ref ID: 544] Uusitalo-Seppala, 2013 52 Uusitalo-Seppala R, Huttunen R, Aittoniemi J, Koskinen P, Leino A, Vahlberg T, et al . Pentraxin 3 (PTX3) is associated with severe sepsis and fatal disease in emergency room patients with suspected infection: a prospective cohort study. PLOS ONE 2013; 8 :e53661. [Ref ID: 1963] Xie, 2021 44 Xie Y, Li B, Lin Y, Shi F, Chen W, Wu W, et al . Combining blood-based biomarkers to predict mortality of sepsis at arrival at the emergency department. Med Sci Monit 2021; 27 :e929527. [Ref ID: 593] Zhang, 2019 45 Zhang Q, Li CS. Risk stratification and prognostic evaluation of endothelial cell-specific molecule1, von Willebrand factor, and a disintegrin-like and metalloprotease with thrombospondin type 1 motif for sepsis in the emergency department: an observational study. Exp Ther Med 2019; 17 :4527–35. [Ref ID: 797] Appendix 4. Excluded studies and reason for exclusion TABLE 3 Characteristics of selected excluded studies and reason for exclusion View in own window Author Year Primary reason for exclusion from the review Reference Agnello 2020 Outcome not relevant Agnello L, Bivona G, Vidali M, Scazzone C, Giglio RV, Iacolino G, et al . Monocyte distribution width (MDW) as a screening tool for sepsis in the emergency department. Clin Chem Lab Med 2020; 58 :1951–7. [Ref ID: 352] Agrawal 2013 Wrong exposure. No focus on biomarker. Agrawal A, Matthay MA, Kangelaris KN, Stein J, Chu JC, Imp BM, et al . Plasma angiopoietin-2 predicts the onset of acute lung injury in critically ill patients. Am J Respir Crit Care Med 2013; 187 :736–42. [Ref ID: 1162] Aluisio 2016 Population not relevant. Study of confirmed sepsis only. Aluisio AR, Jain A, Baron BJ, Sarraf S, Sinert R, Legome E, et al . The prognostic role of non-critical lactate levels for in-hospital survival time among ED patients with sepsis. Am J Emerg Med 2016; 34 :170–3. [Ref ID: 1005] Andersen 2016 Population not relevant. Only patients with septic shock. Andersen LW, Liu X, Montissol S, Holmberg MJ, Sulmonte C, Balkema JL, et al . Cytochrome C in patients with septic shock. Shock 2016; 45 :512–7. [Ref ID: 187] April 2017 Population not relevant. Only patients with septic shock. April MD, Donaldson C, Tannenbaum LI, Moore T, Aguirre J, Pingree A, et al . Emergency department septic shock patient mortality with refractory hypotension vs. hyperlactatemia: a retrospective cohort study. Am J Emerg Med 2017; 35 :1474–9. [Ref ID: 936] Arnau-Barres 2019 Population not relevant. Only recruited patients with confirmed sepsis. Arnau-Barres I, Guerri-Fernandez R, Luque S, Sorli L, Vazquez O, Miralles R. Serum albumin is a strong predictor of sepsis outcome in elderly patients. Eur J Clin Microbiol Infect Dis 2019; 38 :743–6. [Ref ID: 762] Arnold 2013 Outcome not relevant Arnold RC, Sherwin R, Shapiro NI, O’Connor JL, Glaspey L, Singh S, et al . Multicenter observational study of the development of progressive organ dysfunction and therapeutic interventions in normotensive sepsis patients in the emergency department. Acad Emerg Med 2013; 20 :433–40. [Ref ID: 1189] Baldira 2020 Setting not relevant. Included ICU and ward as well as ED . Baldira J, Ruiz-Rodriguez JC, Wilson DC, Ruiz-Sanmartin A, Cortes A, Chiscano L, et al . Biomarkers and clinical scores to aid the identification of disease severity and intensive care requirement following activation of an in-hospital sepsis code. Ann Intensive Care 2020; 10 :7. [Ref ID: 718] Battista 2016 Population not relevant. Analysis is restricted to confirmed sepsis. Battista S, Audisio U, Galluzzo C, Maggiorotto M, Masoero M, Forno D, et al . Assessment of Diagnostic and Prognostic Role of Copeptin in the Clinical Setting of Sepsis. Biomed Res Int 2016; 2016 :3624730. [Ref ID: 988] Bewersdorf 2017 Prediction model development or validation study Bewersdorf JP, Hautmann O, Kofink D, Abdul Khalil A, Zainal Abidin I, Loch A. The SPEED (sepsis patient evaluation in the emergency department) score: a risk stratification and outcome prediction tool. Eur J Emerg Med 2017; 24 :170–5. [Ref ID: 917] Bhat 2015 Wrong exposure. Study of lactate clearance, not lactate at presentation. Bhat SR, Swenson KE, Francis MW, Wira CR. Lactate clearance predicts survival among patients in the emergency department with severe sepsis. West J Emerg Med 2015; 16 :1118–26. [Ref ID: 1036] Bhavani 2019 Prediction model development or validation study Bhavani SV, Carey KA, Gilbert ER, Afshar M, Verhoef PA, Churpek MM. Identifying Novel Sepsis Subphenotypes Using Temperature Trajectories. Am J Respir Crit Care Med 2019; 200 :327–35. [Ref ID: 371] Bildik 2023 Diagnostic study Bildik B, Kalafat UM, Dorter M, Can D, Cander B, Neijmann ST, et al . Diagnostic value of hepcidin in patients with sepsis and septic shock. Ann Clin Anal Med 2023; 14 :101–5. [Ref ID: 299] Biyikli 2018 Population not relevant. Only recruited patients with confirmed sepsis. Biyikli E, Kayipmaz AE, Kavalci C. Effect of platelet-lymphocyte ratio and lactate levels obtained on mortality with sepsis and septic shock. Am J Emerg Med 2018; 36 :647–50. [Ref ID: 853] Blum 2015 Population not relevant. Included only those who died. Blum A, Zoubi AA, Kuria S, Blum N. High serum lactate level may predict death within 24 hours. Open Med (Wars) 2015; 10 :318–22. [Ref ID: 1978] BouChebl 2020 Population not relevant. Only recruited patients with confirmed sepsis. Bou Chebl R, Jamali S, Sabra M, Safa R, Berbari I, Shami A, et al . Lactate/albumin ratio as a predictor of in-hospital mortality in septic patients presenting to the emergency department. Front Med 2020; 7 :550182. [Ref ID: 681] BouChebl 2020 Population not relevant. Not a suspected sepsis population. Bou Chebl R, Jamali S, Mikati N, Al Assaad R, Abdel Daem K, Kattouf N, et al . Relative hyperlactatemia in the emergency department. Front Med 2020; 7 :561. [Ref ID: 675] BouChebl 2021 Population not relevant. Only recruited patients with confirmed sepsis. Bou Chebl R, Geha M, Assaf M, Kattouf N, Haidar S, Abdeldaem K, et al . The prognostic value of the lactate/albumin ratio for predicting mortality in septic patients presenting to the emergency department: a prospective study. Ann Med 2021; 53 :2268–77. [Ref ID: 503] Boussina 2023 Prediction model development or validation study Boussina A, Wardi G, Shashikumar SP, Malhotra A, Zheng K, Nemati S. Representation learning and spectral clustering for the development and external validation of dynamic sepsis phenotypes: observational cohort study. J Med Internet Res 2023; 25 :e45614. [Ref ID: 223] Cao 2022 Outcome not relevant. Composite outcome only. Cao JD, Wang ZC, Wang YL, Li HC, Gu CM, Bai ZG, et al . Risk factors for progression of urolith associated with obstructive urosepsis to severe sepsis or septic shock. BMC Urol 2022; 22 :46. [Ref ID: 480] Capp 2015 Population not relevant. Study of confirmed sepsis only. Capp R, Horton CL, Takhar SS, Ginde AA, Peak DA, Zane R, et al . Predictors of patients who present to the emergency department with sepsis and progress to septic shock between 4 and 48 hours of emergency department arrival. Crit Care Med 2015; 43 :983–8. [Ref ID: 1032] Carvas 2016 Population not relevant. Only patients with severe sepsis and septic shock. Carvas JM, Montanha G, Canelas C, Silva C, Esteves F. Impact of compliance with a sepsis resuscitation bundle in a Portuguese emergency department. Acta Med Port 2016; 29 :88–94. [Ref ID: 972] Casagranda 2015 Population not relevant. Study of confirmed sepsis only. Casagranda I, Vendramin C, Callegari T, Vidali M, Calabresi A, Ferrandu G, et al . Usefulness of suPAR in the risk stratification of patients with sepsis admitted to the emergency department. Intern Emerg Med 2015; 10 :725–30. [Ref ID: 1078] Cetin 2021 Population not relevant. Only patients with severe sepsis and septic shock. Cetin M, Oray NC, Bayram B, Calan OG. The prognostic value of ischemia-modified albumin in patients with sepsis. Niger J Clin Pract 2021; 24 :680–4. [Ref ID: 622] Cha 2015 Outcome not relevant Cha YS, Yoon JM, Jung WJ, Kim YW, Kim TH, Kim OH, et al . Evaluation of usefulness of myeloperoxidase index (MPXI) for differential diagnosis of systemic inflammatory response syndrome (SIRS) in the emergency department. Emerg Med J 2015; 32 :304–7. [Ref ID: 1876] Chae 2020 Population not relevant. Analysis of confirmed sepsis only Chae BR, Kim YJ, Lee YS. Prognostic accuracy of the sequential organ failure assessment (SOFA) and quick SOFA for mortality in cancer patients with sepsis defined by systemic inflammatory response syndrome (SIRS). Support Care Cancer 2020; 28 :653–9. [Ref ID: 720] Chae 2022 Prediction model development or validation study Chae B, Kim S, Lee YS. Development and validation of a new risk scoring system for solid tumor patients with suspected infection. Sci Rep 2022; 12 :3442. [Ref ID: 414] Chaftari 2021 Prediction model development or validation study Chaftari P, Qdaisat A, Chaftari AM, Maamari J, Li Z, Lupu F, et al . Prognostic value of procalcitonin, C-reactive protein, and lactate levels in emergency evaluation of cancer patients with suspected infection. Cancers (Basel) 2021; 13 :4087. [Ref ID: 578] Charoentanyarak 2021 Setting not relevant. Community hospital. Charoentanyarak S, Sawunyavisuth B, Deepai S, Sawanyawisuth K. A point-of-care serum lactate level and mortality in adult sepsis patients: a community hospital setting. J Prim Care Community Health 2021; 12 :21501327211000233. [Ref ID: 1461] Chaudhari 2022 Population not relevant. Only recruited patients with confirmed sepsis. Chaudhari M, Agarwal N. Study of significance of serum lactate kinetics in sepsis as mortality predictor. Indian J Crit Care Med 2022; 26 :589–93. [Ref ID: 452] Chen 2013 Outcome not relevant Chen YX, Li CS. Prognostic value of adrenomedullin in septic patients in the ED . Am J Emerg Med 2013; 31 :1017–21. [Ref ID: 1167] Chen 2014 Population not relevant. Confirmed sepsis only analysed Chen YX, Li CS. Arterial lactate improves the prognostic performance of severity score systems in septic patients in the ED . Am J Emerg Med 2014; 32 :982–6. [Ref ID: 1079] Chen 2022 Population not relevant. Confirmed sepsis only analysed. Chen L, Chen K, Hong Y, Xing L, Zhang J, Zhang K, et al . The landscape of isoform switches in sepsis: a multicenter cohort study. Sci Rep 2022; 12 :10276. [Ref ID: 490] Chen 2017 Prediction model development or validation study Chen KF, Liu SH, Li CH, Wu CC, Chaou CH, Tzeng IS, et al . Development and validation of a parsimonious and pragmatic CHARM score to predict mortality in patients with suspected sepsis. Am J Emerg Med 2017; 35 (4):640‐6. [Ref ID: 84] Choe 2016 Population not relevant. Septic shock patients only. Choe EA, Shin TG, Jo IJ, Hwang SY, Lee TR, Cha WC, et al . The prevalence and clinical significance of low procalcitonin levels among patients with severe sepsis or septic shock in the emergency department. Shock 2016; 46 (1):37–43. [Ref ID: 962] Christensen 2022 Outcome not relevant Christensen EE, Binde C, Leegaard M, Tonby K, Dyrhol-Riise AM, Kvale D, et al . Diagnostic accuracy and added value of infection biomarkers in patients with possible sepsis in the emergency department. Shock 2022; 58 :251–9. [Ref ID: 410] Chu 2021 Population not relevant. Only recruited patients with confirmed sepsis. Chu CC, Su CM, Chen FC, Cheng CY, Cheng HH, Te Kung C. The timing of last hemodialysis influences the prognostic value of serum lactate levels in predicting mortality of end-stage renal disease patients with sepsis in the emergency department. Medicine (United States) 2021; 100 :E24474. [Ref ID: 245] Chukwulebe 2021 Outcome not relevant Chukwulebe SB, Gaieski DF, Bhardwaj A, Mulugeta-Gordon L, Shofer FS, Dean AJ. Early hemodynamic assessment using NICOM in patients at risk of developing sepsis immediately after emergency department triage. Scand J Trauma Resusc Emerg Med 2021; 29 :23. [Ref ID: 535] Clar 2021 Population not relevant. Vague description but assumed that study population had confirmed sepsis. Clar J, Oltra MR, Benavent R, Pinto C, Ruiz A, Sanchez MT, et al . Prognostic value of diagnostic scales in community-acquired sepsis mortality at an emergency service. Prognosis in community-acquired sepsis. BMC Emerg Med 2021; 21 :161. [Ref ID: 502] Covino 2020 Population not relevant. Not a suspected sepsis population. Covino M, Piccioni A, Bonadia N, Onder G, Sabia L, Carbone L, et al . Early procalcitonin determination in the emergency department and clinical outcome of community-acquired pneumonia in old and oldest old patients. Eur J Intern Med 2020; 79 :51–7. [Ref ID: 690] Covino 2021 Outcome not relevant Covino M, Gallo A, Montalto M, De Matteis G, Burzo ML, Simeoni B, et al . The role of early procalcitonin determination in the emergency department in adults hospitalized with fever. Medicina (Kaunas) 2021; 57 :179. [Ref ID: 536] Dadeh 2016 Population not relevant. Patients with SIRS only. Dadeh AA, Wuthisuthimethawee P. Serum lactate levels as a prognostic predictor of septic shock in emergency department patients with systemic inflammatory response syndrome (SIRS) at Songklanagarind hospital. J Med Assoc Thai 2016; 99 :913–8. [Ref ID: 958] Datta 2015 Population not relevant. Population is ‘clinically unwell’ and not suspected sepsis. Datta D, Walker C, Gray AJ, Graham C. Arterial lactate levels in an emergency department are associated with mortality: a prospective observational cohort study. Emerg Med J 2015; 32 :673–7. [Ref ID: 1052] Davoudian 2022 Population not relevant. Analysis of confirmed sepsis only Davoudian S, Piovani D, Desai A, Mapelli SN, Leone R, Sironi M, et al . A cytokine/ PTX3 prognostic index as a predictor of mortality in sepsis. Front Immunol 2022; 13 :979232. [Ref ID: 434] DeCoux 2015 Population not relevant De Coux A, Tian Y, Deleon-Pennell KY, Nguyen NT, De Castro Bras LE, Flynn ER, et al . Plasma glycoproteomics reveals sepsis outcomes linked to distinct proteins in common pathways. Crit Care Med 2015; 43 :2049–58. [Ref ID: 1057] Dettmer 2015 Outcome not relevant Dettmer MR, Mohr NM, Fuller BM. Sepsis-associated pulmonary complications in emergency department patients monitored with serial lactate: an observational cohort study. J Crit Care 2015; 30 :1163–8. [Ref ID: 1038] Dharaniyadewi 2013 Outcome not relevant Dharaniyadewi D, Lie KC, Sukmana N, Rumende CM. Effect of semi-quantitative procalcitonin assay on the adequacy of empirical antibiotics and mortality in septic patients. Crit Care 2013; 17 :P15. [Ref ID: 99] Di 2020 Population not relevant. Only recruited patients with confirmed sepsis. Di S, Wang Y, Sun L, Zhao H, Guo L. The value of TLR-4, CRP , PCT and WBC levels in assessing the diagnosis and prognosis of sepsis patients. Int J Clin Exp Med 2020; 13 :9419–28. [Ref ID: 645] Doenyas-Barak 2016 Wrong exposure. No focus on biomarker. Doenyas-Barak K, Beberashvili I, Marcus R, Efrati S. Lactic acidosis and severe septic shock in metformin users: a cohort study. Crit Care 2016; 20 :10. [Ref ID: 1002] Dolatabadi 2015 Population not relevant. Only analysed patients with confirmed sepsis. Dolatabadi AA, Memary E, Amini A, Shojaee M, Abdalvand A, Hatamabadi HR. Efficacy of measuring procalcitonin levels in determination of prognosis and early diagnosis of bacterial resistance in sepsis. Niger Med J 2015; 56 :17–22. [Ref ID: 1982] Dorsett 2017 Wrong exposure. No focus on biomarker. Dorsett M, Kroll M, Smith CS, Asaro P, Liang SY, Moy HP. qSOFA has poor sensitivity for prehospital identification of severe sepsis and septic shock. Prehosp Emerg Care 2017; 21 :489–97. [Ref ID: 952] Drumheller 2016 Population not relevant. Only patients with severe sepsis and septic shock. Drumheller BC, Agarwal A, Mikkelsen ME, Sante SC, Weber AL, Goyal M, et al . Risk factors for mortality despite early protocolized resuscitation for severe sepsis and septic shock in the emergency department. J Crit Care 2016; 31 :13–20. [Ref ID: 1014] Duplessis 2018 Population not relevant. Only analysed sepsis patients as a secondary analysis of larger study. Duplessis C, Gregory M, Frey K, Bell M, Truong L, Schully K, et al . Evaluating the discriminating capacity of cell death (apoptotic) biomarkers in sepsis. J Intensive Care 2018; 6 :72. [Ref ID: 873] Endo 2014 Population not relevant Endo S, Suzuki Y, Takahashi G, Shozushima T, Ishikura H, Murai A, et al . Presepsin as a powerful monitoring tool for the prognosis and treatment of sepsis: a multicenter prospective study. J Infect Chemother 2014; 20 :30–4. [Ref ID: 1108] Fang 2016 Population not relevant. Appears to only analyse patients with confirmed sepsis. Fang YY, Shao R, Yu H, Zhang Q, Wang MM, Li CS. Prognostic significance of C5a2 on polymorphonuclear neutrophil and C5a2intra/C5a2 ratio level for early sepsis in an ED . Am J Emerg Med 2016; 34 :2084–9. [Ref ID: 1007] Fernando 2018 Outcome not relevant. Composite outcome only. Fernando SM, Barnaby DP, Herry CL, Gallagher EJ, Shapiro NI, Seely AJE. Helpful only when elevated: initial serum lactate in stable emergency department patients with sepsis is specific, but not sensitive for future deterioration. J Emerg Med 2018; 54 :766–73. [Ref ID: 850] Filho 2016 Population not relevant. Only patients with severe sepsis and septic shock. Filho RR, Rocha LL, Correa TD, Souza Pessoa CM, Colombo G, Cesar Assuncao MS. Blood lactate levels cutoff and mortality prediction in sepsis – time for a reappraisal? A retrospective cohort study. Shock 2016; 46 :480–5. [Ref ID: 959] Franchini 2015 Outcome not relevant Franchini S, Marciano T, Sorlini C, Campochiaro C, Tresoldi M, Sabbadini MG, et al . Serum CXCL12 levels on hospital admission predict mortality in patients with severe sepsis/septic shock. Am J Emerg Med 2015; 33 :1802–4. [Ref ID: 1034] Galtung 2022 Prediction model development or validation study Galtung N, Diehl-Wiesenecker E, Lehmann D, Markmann N, Bergstrom WH, Wacker J, et al . Prospective validation of a transcriptomic severity classifier among patients with suspected acute infection and sepsis in the emergency department. Eur J Emerg Med 2022; 29 :357–65. [Ref ID: 430] Gao 2022 Population not relevant. Septic shock patients only. Gao J, Chen S, Kong T, Wen D, Yang Q. The relationship between soluble CD73 and the incidence of septic shock in severe sepsis patients: a cross-sectional analysis of data from a prospective FINNAKI study. Ann Transl Med 2022; 10 :302. [Ref ID: 468] Genga 2018 Prediction model development or validation study Genga KR, Trinder M, Kong HJ, Li X, Leung AKK, Shimada T, et al . CETP genetic variant rs1800777 (allele A) is associated with abnormally low HDL-C levels and increased risk of AKI during sepsis. Sci Rep 2018; 8 :16764. [Ref ID: 1624] Gicheru 2023 Population not relevant. Only included patients with confirmed sepsis. Gicheru B, Shah J, Wachira B, Omuse G, Maina D. The diagnostic accuracy of an initial point-of-care lactate at the emergency department as a predictor of in-hospital mortality among adult patients with sepsis and septic shock. Front Med 2023; 10 :1173286. [Ref ID: 239] Ginde 2014 Population not relevant. Only analysed confirmed sepsis patients. Ginde AA, Blatchford PJ, Trzeciak S, Hollander JE, Birkhahn R, Otero R, et al . Age-related differences in biomarkers of acute inflammation during hospitalization for sepsis. Shock 2014; 42 :99–107. [Ref ID: 1127] Gomez-Ramos 2018 Outcome not relevant. Composite outcome only. Gomez-Ramos JJ, Marin-Medina A, Prieto-Miranda SE, Davalos-Rodriguez IP, Alatorre-Jimenez MA, Esteban-Zubero E. Determination of plasma lactate in the emergency department for the early detection of tissue hypoperfusion in septic patients. Am J Emerg Med 2018; 36 :1418–22. [Ref ID: 822] Gornet 2021 Population not relevant. ED patients suspected of infection. (Related to ID 573 Leroux 2021.) Gornet M, Leroux P, Ramont L, De Ruffi S, Giordano Orsini G, Losset X, et al . Lack of admission biomarkers’ clinical utility in outcomes prediction in patients suspected with infection in the emergency department. Am J Emerg Med 2021; 47 :109–14. [Ref ID: 576] Gotmaker 2017 Population not relevant. Patients with early septic shock in ED . Gotmaker R, Peake SL, Forbes A, Bellomo R. Mortality is greater in septic patients with hyperlactatemia than with refractory hypotension. Shock 2017; 48 :294–300. [Ref ID: 914] Guarino 2022 Population not relevant. Only recruited patients with confirmed sepsis. Guarino M, Perna B, De Giorgi A, Gambuti E, Alfano F, Catanese EM, et al . A 2-year retrospective analysis of the prognostic value of MqSOFA compared to lactate, NEWS and qSOFA in patients with sepsis. Infection 2022; 50 :941–8. [Ref ID: 1386] Guerin 2014 Setting not relevant. ICU and ED . Guerin E, Orabona M, Raquil MA, Giraudeau B, Bellier R, Gibot S, et al . Circulating immature granulocytes with T-cell killing functions predict sepsis deterioration*. Crit Care Med 2014; 42 :2007–18. [Ref ID: 1894] Guirgis 2021 Prediction model development or validation study Guirgis FW, Black LP, Henson M, Labilloy G, Smotherman C, Hopson C, et al . A hypolipoprotein sepsis phenotype indicates reduced lipoprotein antioxidant capacity, increased endothelial dysfunction and organ failure, and worse clinical outcomes. Crit Care 2021; 25 :341. [Ref ID: 518] Guirgis 2023 Prediction model development or validation study Guirgis FW, Jacob V, Wu D, Henson M, Daly-Crews K, Hopson C, et al . DHCR7 expression predicts porsusor outcomes and mortality from sepsis. Crit Care Explor 2023; 5 :e0929. [Ref ID: 1212] Ha 2015 Population not relevant. Mortality is analysed for patients with complicated sepsis only. Ha SO, Park SH, Park JS, Huh JW, Lim CM, Koh Y, et al . Fraction of immature granulocytes reflects severity but not mortality in sepsis. Scand J Clin Lab Invest 2015; 75 :36–43. [Ref ID: 1050] Henning 2019 Population not relevant. Only patients with signs of critical illness. Henning DJ, Bhatraju PK, Johnson NJ, Kosamo S, Shapiro NI, Zelnick LR, et al . Physician judgment and circulating biomarkers predict 28-day mortality in emergency department patients. Crit Care Med 2019; 47 :1513–21. [Ref ID: 262] Hicks 2014 Outcome not relevant Hicks CW, Engineer RS, Benoit JL, Dasarathy S, Christenson RH, Peacock WF. Procalcitonin as a biomarker for early sepsis in the emergency department. Eur J Emerg Med 2014; 21 :112–7. [Ref ID: 1131] Holder 2016 Outcome not relevant. Composite outcome only. Holder AL, Gupta N, Lulaj E, Furgiuele M, Hidalgo I, Jones MP, et al . Predictors of early progression to severe sepsis or shock among emergency department patients with nonsevere sepsis. Int J Emerg Med 2016; 9 :1–11. [Ref ID: 1000] Hou 2017 Prediction model development or validation study Hou PC, Filbin MR, Ngo L, Aird WC, Shapiro NI, Wang H, et al . Endothelial permeability and hemostasis in septic shock: results from the ProCESS trial. Chest 2017; 15 2:22–31. [Ref ID: 894] Hou 2017 Prediction model development or validation study Hou PC, Filbin MR, Wang H, Ngo L, Aird WC, Shapiro NI, et al . Endothelial permeability and hemostasis in septic shock: results from the ProCESS trial. Chest 2017; 152 :22‐31. [Ref ID: 172] Hou 2020 Population not relevant. Only recruited patients with confirmed sepsis. Hou X, Liu C, Lian H, Xu Z, Ma L, Zang X, et al . The value of neutrophil gelatinase-associated lipocalin and citrullinated alpha enolase peptide-1 antibody in diagnosis, classification, and prognosis for patients with sepsis. Medicine (United States) 2020; 99 :E21893. [Ref ID: 250] Hsiao 2019 Population not relevant. Only recruited patients with confirmed sepsis. Hsiao SY, Lai YR, Kung CT, Tsai NW, Su CM, Huang CC, et al . alpha-1-acid glycoprotein concentration as an outcome predictor in adult patients with sepsis. Biomed Res Int 2019; 2019 :3174896. [Ref ID: 750] Hsu 2019 Population not relevant. Appear to be patients with confirmed sepsis. No relevant outcome. Hsu YC, Hsu CW. Septic acute kidney injury patients in emergency department: the risk factors and its correlation to serum lactate. Am J Emerg Med 2019; 37 :204–8. [Ref ID: 765] Hu 2022 Population not relevant. Only recruited patients with confirmed sepsis. Hu H, Jiang JY, Yao N. Comparison of different versions of the quick sequential organ failure assessment for predicting in-hospital mortality of sepsis patients: a retrospective observational study. World J Emerg Med 2022; 13 :114–9. [Ref ID: 1279] Huang 2018 Population not relevant. Only recruited patients with confirmed sepsis. Huang Z, Jiang H, Cui X, Liang G, Chen Y, Wang T, et al . Elevated serum levels of lipoprotein-associated phospholipase A2 predict mortality rates in patients with sepsis. Mol Med Report 2018; 17 :1791–8. [Ref ID: 862] Hwang 2014 Population not relevant. Only patients with septic shock or severe sepsis. Hwang SY, Shin TG, Jo IJ, Jeon K, Suh GY, Lee TR, et al . Association between hemodynamic presentation and outcome in sepsis patients. Shock 2014; 42 :205–10. [Ref ID: 1090] Hwang 2020 Population not relevant. Only recruited patients with confirmed sepsis. Hwang TS, Park HW, Park HY, Park YS. Prognostic value of severity score change for septic shock in the emergency room. Diagnostics 2020; 10 :743. [Ref ID: 229] Innocenti 2014 Population not relevant. Only patients with severe sepsis and septic shock. Innocenti F, Bianchi S, Guerrini E, Vicidomini S, Conti A, Zanobetti M, et al . Prognostic scores for early stratification of septic patients admitted to an emergency department-high dependency unit. Eur J Emerg Med 2014; 21 :254–9. [Ref ID: 1110] Innocenti 2019 Setting not relevant. Emergency Department High Dependency Unit (ED-HDU) Innocenti F, Gori AM, Giusti B, Tozzi C, Donnini C, Meo F, et al . Prognostic value of sepsis-induced coagulation abnormalities: an early assessment in the emergency department. Intern Emerg Med 2019; 14 :459–66. [Ref ID: 809] Ishikawa 2021 Population not relevant. Study of SIRS not sepsis Ishikawa S, Teshima Y, Otsubo H, Shimazui T, Nakada TA, Takasu O, et al . Risk prediction of biomarkers for early multiple organ dysfunction in critically ill patients. BMC Emerg Med 2021; 21 :132. [Ref ID: 522] Javed 2017 Population not relevant. Only patients with severe sepsis. Javed A, Guirgis FW, Sterling SA, Puskarich MA, Bowman J, Robinson T, et al . Clinical predictors of early death from sepsis. J Crit Care 2017; 42 :30–4. [Ref ID: 79] Jekarl 2019 Population not relevant. Only recruited patients with confirmed sepsis. Jekarl DW, Kim JY, Ha JH, Lee S, Yoo J, Kim M, et al . Diagnosis and prognosis of sepsis based on use of cytokines, chemokines, and growth factors. Dis Markers 2019; 2019 :1089107. [Ref ID: 746] Jeon 2021 Population not relevant. Only included severe sepsis patients. Jeon SY, Ryu S, Oh SK, Park JS, You YH, Jeong WJ, et al . Lactate dehydrogenase to albumin ratio as a prognostic factor for patients with severe infection requiring intensive care. Medicine (United States) 2021; 100 : E27538 . [Ref ID: 322] Jeong 2018 Outcome not relevant Jeong HS, Lee TH, Bang CH, Kim JH, Hong SJ. Risk factors and outcomes of sepsis-induced myocardial dysfunction and stress-induced cardiomyopathy in sepsis or septic shock: a comparative retrospective study. Medicine 2018; 97 :e0263. [Ref ID: 1651] Jessen 2017 Prediction model development or validation study Jessen MK, Skibsted S, Shapiro NI. Number of organ dysfunctions predicts mortality in emergency department patients with suspected infection: a multicenter validation study. Eur J Emerg Med 2017; 24 :176–82. [Ref ID: 911] Julienne 2022 Population not relevant. Patients with suspected infection (population too broad). Julienne J, Douillet D, Mozziconacci MS, Callahan JC. Prognostic accuracy of using lactate in addition to the quick Sequential Organ Failure Assessment score and the National Early Warning Score for emergency department patients with suspected infection. Emerg Med J 2022; 40 :28–35. [Ref ID: 380] Junhasavasdikul 2016 Population not relevant. Only analysed confirmed sepsis patients. Junhasavasdikul D, Theerawit P, Ingsathit A, Kiatboonsri S. Lactate and combined parameters for triaging sepsis patients into intensive care facilities. J Crit Care 2016; 33 :71–7. [Ref ID: 971] Kahveci 2021 Population not relevant. Confirmed sepsis recruited. Kahveci U, Ozkan S, Melekoglu A, Usul E, Ozturk G, Cetin E, et al . The role of plasma presepsin levels in determining the incidence of septic shock and mortality in patients with sepsis. J Infect Dev Ctries 2021; 15 :123–30. [Ref ID: 601] Kandori 2019 Population not relevant. Not a suspected sepsis population. Kandori K, Okada Y, Matsuyama T, Morita S, Ehara N, Miyamae N, et al . Prognostic ability of the sequential organ failure assessment score in accidental hypothermia: a multi-institutional retrospective cohort study. Scand J Trauma Resusc Emerg Med 2019; 27 :103. [Ref ID: 732] Kangelaris 2015 Outcome not relevant Kangelaris KN, Prakash A, Liu KD, Aouizerat B, Woodruff PG, Erle DJ, et al . Increased expression of neutrophil-related genes in patients with early sepsis-induced ARDS. Am J Physiol Lung Cell Mol Physiol 2015; 308 :L1102–13. [Ref ID: 1046] Kangelaris 2021 Population not relevant. Only analysed confirmed sepsis patients. Kangelaris KN, Clemens R, Fang X, Jauregui A, Liu T, Vessel K, et al . A neutrophil subset defined by intracellular olfactomedin 4 is associated with mortality in sepsis. Am J Physiol Lung Cell Mol Physiol 2021; 320 :L892–L902. [Ref ID: 1477] KilincToker 2021 Population not relevant. Confirmed sepsis only recruited. Kilinc Toker A, Kose S, Turken M. Comparison of SOFA score, SIRS , qSOFA , and qSOFA + L criteria in the diagnosis and prognosis of sepsis. Eurasian J Med 2021; 53 :40–7. [Ref ID: 1309] Kim 2013 Wrong exposure. Focus on an increase in red blood cell distribution width (RDW) from baseline. Kim CH, Park JT, Kim EJ, Han JH, Han JS, Choi JY, et al . An increase in red blood cell distribution width from baseline predicts mortality in patients with severe sepsis or septic shock. Crit Care 2013; 17 :R282. [Ref ID: 1186] Kim 2015 Population not relevant. Only patients with severe sepsis and septic shock. Kim MH, Ahn JY, Song JE, Choi H, Ann HW, Kim JK, et al . The C-reactive protein/albumin ratio as an independent predictor of mortality in patients with severe sepsis or septic shock treated with early goal-directed therapy. PLOS ONE 2015; 10 :e0132109. [Ref ID: 1029] Kim 2022 Population not relevant. Only patients with septic shock. Kim SM, Ryoo SM, Shin TG, Park YS, Jo YH, Lim TH, et al . Prognostic factors for late death in septic shock survivors: a multi-center, prospective, registry-based observational study. Intern Emerg Med 2022; 17 :865–71. [Ref ID: 492] Kim 2020 Prediction model development or validation study Kim YJ, Kang J, Kim MJ, Ryoo SM, Kang GH, Shin TG, et al . Development and validation of the VitaL CLASS score to predict mortality in stage IV solid cancer patients with septic shock in the emergency department: a multi-center, prospective cohort study. BMC Med 2020; 18 :390. [Ref ID: 661] Knox 2015 Population not relevant. Only patients with severe sepsis and septic shock. Knox DB, Lanspa MJ, Kuttler KG, Brewer SC, Brown SM. Phenotypic clusters within sepsis-associated multiple organ dysfunction syndrome. Intensive Care Med 2015. [Ref ID: 377] Ko 2016 Population not relevant. Only recruited patients with confirmed sepsis. Ko YH, Ji YS, Park SY, Kim SJ, Song PH. Procalcitonin determined at emergency department as na early indicator of progression to septic shock in patient with sepsis associated with ureteral calculi. Int Braz J Urol 2016; 42 :270–6. [Ref ID: 1022] Kumar 2019 Population not relevant. Only patients with severe sepsis and septic shock. Kumar S, Jangpangi G, Bhalla A, Sharma N. Role of central venous oxygen saturation in prognostication of patients with severe sepsis and septic shock in emergency medical services. Int J Crit Illn Inj Sci 2019; 9 :164–71. [Ref ID: 1740] Kung 2014 Population not relevant. Only recruited severe sepsis and septic shock patients. Kung CT, Su CM, Chang HW, Cheng HH, Hsiao SY, Tsai TC, et al . Serum adhesion molecules as outcome predictors in adult severe sepsis patients requiring mechanical ventilation in the emergency department. Clin Biochem 2014; 47 :38–43. [Ref ID: 378] Kung 2015 Population not relevant. Only patients with severe sepsis and septic shock. Kung CT, Su CM, Chang HW, Cheng HH, Hsiao SY, Tsai TC, et al . The prognostic value of leukocyte apoptosis in patients with severe sepsis at the emergency department. Clin Chim Acta 2015; 438 :364–9. [Ref ID: 1068] Lafon 2020 Outcome not relevant. Composite outcome only. Lafon T, Cazalis MA, Vallejo C, Tazarourte K, Blein S, Pachot A, et al . Prognostic performance of endothelial biomarkers to early predict clinical deterioration of patients with suspected bacterial infection and sepsis admitted to the emergency department. Ann Intensive Care 2020; 10 :113. [Ref ID: 692] Langley 2013 Prediction model development or validation study Langley RJ, Tsalik EL, van Velkinburgh JC, Glickman SW, Rice BJ, Wang C, et al . An integrated clinico-metabolomic model improves prediction of death in sepsis. Sci Transl Med 2013; 5 :195ra95. [Ref ID: 1195] Lavoignet 2019 Population not relevant. Not a suspected sepsis population. Lavoignet CE, Le Borgne P, Chabrier S, Bidoire J, Slimani H, Chevrolet-Lavoignet J, et al . White blood cell count and eosinopenia as valuable tools for the diagnosis of bacterial infections in the ED . Eur J Clin Microbiol Infect Dis 2019; 38 :1523–32. [Ref ID: 83] Lee 2016 Population not relevant. Only analysed confirmed sepsis patients. Lee YK, Hwang SY, Shin TG, Jo IJ, Suh GY, Jeon K. Prognostic value of lactate and central venous oxygen saturation after early resuscitation in sepsis patients. PLOS ONE 2016; 11 :e0153305. [Ref ID: 996] Lee 2021 Population not relevant. Only recruited patients with confirmed sepsis. Lee SG, Song J, Park DW, Moon S, Cho HJ, Kim JY, et al . Prognostic value of lactate levels and lactate clearance in sepsis and septic shock with initial hyperlactatemia: a retrospective cohort study according to the Sepsis-3 definitions. Medicine (United States) 2021; 100 :E24835. [Ref ID: 244] Lee 2021 Population not relevant. Only recruited severe sepsis and septic shock patients. Lee GT, Hwang SY, Park JE, Jo IJ, Kim WY, Chung SP, et al . Diagnostic accuracy of lactate levels after initial fluid resuscitation as a predictor for 28 day mortality in septic shock. Am J Emerg Med 2021; 46 :392–7. [Ref ID: 528] Lee 2021 Population not relevant. Septic shock patients only. Lee MS, Shin TG, Kim WY, Jo YH, Hwang YJ, Choi SH, et al . Hypochloraemia is associated with 28-day mortality in patients with septic shock: a retrospective analysis of a multicentre prospective registry. Emerg Med J 2021; 38 :423–9. [Ref ID: 1487] Lee 2022 Prediction model development or validation study Lee HJ, Ko BS, Ryoo SM, Han E, Suh GJ, Choi SH, et al . Modified cardiovascular SOFA score in sepsis: development and internal and external validation. BMC Med 2022; 20 :263. [Ref ID: 390] Leisman 2016 Population not relevant. Only patients with severe sepsis and septic shock. Leisman D, Wie B, Doerfler M, Bianculli A, Ward MF, Akerman M, et al . Association of fluid resuscitation initiation within 30 minutes of severe sepsis and septic shock recognition with reduced mortality and length of stay. Ann Emerg Med 2016; 68 :298–311. [Ref ID: 977] Leroux 2021 Population not relevant. ED patients suspected of infection. Leroux P, De Ruffi S, Ramont L, Gornet M, Giordano Orsini G, Losset X, et al . Clinical outcome predictive value of procalcitonin in patients suspected with infection in the emergency department. Emerg Med Int 2021; 2021 :2344212. [Ref ID: 573] Li 2018 Wrong exposure. Prognostic score, not biomarker. Li D, Zhou Y, Yu J, Yu H, Xia Y, Zhang L, et al . Evaluation of a novel prognostic score based on thrombosis and inflammation in patients with sepsis: a retrospective cohort study. Clin Chem Lab Med 2018; 56 :1182–92. [Ref ID: 826] Liang 2022 Population not relevant. Only patients with septic shock. Liang CY, Yang YY, Hung CC, Wang TH, Hsu YC. Prognostic values of the timing of antibiotic administration and the sepsis bundle component in elderly patients with septic shock: a retrospective study. Shock 2022; 57 :181–8. [Ref ID: 485] Lin 2023 Outcome not relevant Lin SF, Lin HA, Pan YH, Hou SK. A novel scoring system combining Modified Early Warning Score with biomarkers of monocyte distribution width, white blood cell counts, and neutrophil-to-lymphocyte ratio to improve early sepsis prediction in older adults. Clin Chem Lab Med 2023; 61 :162–72. [Ref ID: 393] Ling 2023 Population not relevant. Only analysed confirmed sepsis patients. Ling H, Chen M, Dai J, Zhong H, Chen R, Shi F. Evaluation of qSOFA combined with inflammatory mediators for diagnosing sepsis and predicting mortality among emergency department. Clin Chim Acta 2023; 544 :117352. [Ref ID: 273] Liu 2013 Population not relevant. Analysis of confirmed sepsis only Liu B, Chen YX, Yin Q, Zhao YZ, Li CS. Diagnostic value and prognostic evaluation of presepsin for sepsis in an emergency department. Crit Care 2013; 17 :R244. [Ref ID: 1163] Liu 2018 Population not relevant. Confirmed sepsis only analysed; no relevant outcomes. Liu XW, Ma T, Liu W, Cai Q, Wang L, Song HW, et al . Sustained increase in angiopoietin-2, heparin-binding protein, and procalcitonin is associated with severe sepsis. J Crit Care 2018; 45 :14–9. [Ref ID: 827] Liu 2019 Outcome not relevant Liu XW, Ma T, Cai Q, Wang L, Song HW, Liu Z. Elevation of serum PARK7 and IL-8 levels is associated with acute lung injury in patients with severe sepsis/septic shock. J Intensive Care Med 2019; 34 :662–8. [Ref ID: 754] Liu 2021 Population not relevant. Only recruited patients with confirmed sepsis. Liu H, Wang SY, Li XD, Yang S, Liu XY, Yang CW, et al . The expression of TRPM7 in serum of patients with sepsis, its influences on inflammatory factors and prognosis, and its diagnostic value. Eur Rev Med Pharmacol Sci 2021; 23 :3926–32. [Ref ID: 590] Liu 2022 Outcome not relevant Liu Q, Gao Y, Yang T, Zhou Z, Lin K, Li T, et al . nCD64 index as a novel inflammatory indicator for the early prediction of prognosis in infectious and non-infectious inflammatory diseases: an observational study of febrile patients. Front Immunol 2022; 13 :905060. [Ref ID: 437] Lokhandwala 2017 Wrong exposure. Focus on lactate reduction, not lactate at presentation. Lokhandwala S, Andersen LW, Nair S, Patel P, Cocchi MN, Donnino MW. Absolute lactate value vs. relative reduction as a predictor of mortality in severe sepsis and septic shock. J Crit Care 2017; 37 :179–84. [Ref ID: 924] Lopez-Izquierdo 2020 Population not relevant. Not a suspected sepsis population. López -Izquierdo R, Martin-Rodríguez F, Santos Pastor J, Garcia Criado J, Fadrique Millan L, Carbajosa Rodriguez V, et al . Can capillary lactate improve early warning scores in emergency department? an observational, prospective, multicentre study. Int J Clin Pract 2020. [Ref ID: 349] Lowe 2018 Population not relevant. Septic shock patients only. Lowe KM, Heffner AC, Karvetski CH. Clinical factors and outcomes of dialysis-dependent end-stage renal disease patients with emergency department septic shock. J Emerg Med 2018; 54 :16–24. [Ref ID: 900] Lu 2020 Outcome not relevant Lu KL, Hsiao CY, Wu CY, Yen CL, Tsai CY, Jenq CC, et al . Delayed fever and acute kidney injury in patients with urinary tract infection. J Clin Med 2020; 9 :1–11. [Ref ID: 634] Macdonald 2014 Outcome not relevant Macdonald SPJ, Stone SF, Neil CL, Van Eeden PE, Fatovich DM, Arendts G, et al . Sustained elevation of resistin, NGAL and IL-8 are associated with severe sepsis/septic shock in the emergency department. PLOS ONE 2014; 9 :e110678. [Ref ID: 1135] Machado 2020 Setting not relevant. Included ward and ED . Machado FR, Cavalcanti AB, Monteiro MB, Sousa JL, Bossa A, Bafi AT, et al . Predictive accuracy of the quick sepsis-related organ failure assessment score in Brazil: a prospective multicenter study. Am J Respir Crit Care Med 2020. [Ref ID: 363] Maeda 2021 Outcome not relevant Maeda T, Paralkar J, Kuno T, Patrawalla P. Inhaled albuterol use and impaired lactate clearance in patients with sepsis: a retrospective cohort study. J Intensive Care Med 2021; 36 :284–9. [Ref ID: 606] Manohar 2018 Outcome not relevant Manohar V, Prasad BS, Raj S, Sreekrishnan TP, Gireesh Kumar KP. The eminence of neutrophil-lymphocyte count ratio in predicting bacteremia for community-acquired infections at an emergency medicine department in a tertiary care setting. J Emerg Trauma Shock 2018; 11 :271–5. [Ref ID: 871] Manzon 2015 Population not relevant. Patients with SIRS only. Manzon C, Barrot L, Besch G, Barbot O, Desmettre T, Capellier G, et al . Capillary lactate as a tool for the triage nurse among patients with SIRS at emergency department presentation: a preliminary report. Ann Intensive Care 2015; 5 . [Ref ID: 1064] Mearelli 2018 Prediction model development or validation study Mearelli F, Fiotti N, Giansante C, Casarsa C, Orso D, De Helmersen M, et al . Derivation and validation of a biomarker-based clinical algorithm to rule out sepsis from noninfectious systemic inflammatory response syndrome at emergency department admission: a multicenter prospective study. Crit Care Med 2018; 46 :1421–9. [Ref ID: 816] Mearelli 2020 Prediction model development or validation study Mearelli F, Barbati G, Casarsa C, Giansante C, Breglia A, Spica A, et al . The integration of qSOFA with clinical variables and serum biomarkers improves the prognostic value of qSOFA alone in patients with suspected or confirmed sepsis at ed admission. J Clin Med 2020; 9 :1205. [Ref ID: 633] Melero-Guijarro 2023 Outcome not relevant Melero-Guijarro L, Sanz-Garcia A, Martin-Rodriguez F, Lipari V, Mazas Perez Oleaga C, Carvajal Altamiranda S, et al . Prehospital qSOFA , mSOFA, and NEWS2 performance for sepsis prediction: a prospective, multi-center, cohort study. Front Med 2023; 10 :1149736. [Ref ID: 255] Mendoza 2022 Outcome not relevant Mendoza D, Ascuntar J, Rosero O, Jaimes F. Improving the diagnosis and prognosis of sepsis according to the sources of infection. Emerg Med J 2022; 39 : 279–83. [Ref ID: 404] Mikkelsen 2013 Population not relevant. Patients with severe sepsis only Mikkelsen ME, Shah CV, Meyer NJ, Gaieski DF, Lyon S, Miltiades AN, et al . The epidemiology of acute respiratory distress syndrome in patients presenting to the emergency department with severe sepsis. Shock 2013; 40 :375–81. [Ref ID: 379] Mitra 2020 Population not relevant. Only analysed confirmed sepsis patients. Mitra B, Roman C, Charters KE, O’Reilly G, Gantner D, Cameron PA. Lactate, bicarbonate and anion gap for evaluation of patients presenting with sepsis to the emergency department: a prospective cohort study. Emerg Med Australas 2020; 32 :20–4. [Ref ID: 715] Mohr 2018 Prediction model development or validation study Mohr NM, Vakkalanka JP, Faine BA, Skow B, Harland KK, Dick-Perez R, et al . Serum anion gap predicts lactate poorly, but may be used to identify sepsis patients at risk for death: a cohort study. J Crit Care 2018; 44 :223–8. [Ref ID: 831] Nandi 2021 Population not relevant. Only recruited patients with confirmed sepsis. Nandi U, Jones AE, Puskarich MA. Group IIA secretory phospholipase 2 independently predicts mortality and positive blood culture in emergency department sepsis patients. J Am Coll Emerg Physicians Open 2021; 2 :e12460. [Ref ID: 265] Narendra 2022 Population not relevant. Of the 92 patients with suspected sepsis recruited, 19 who died before day 3 were excluded Narendra S, Wyawahare M, Adole PS. Presepsin vs procalcitonin as predictors of sepsis outcome. J Assoc Physicians India 2022; 70 :11–2. [Ref ID: 224] Ng 2022 Prediction model development or validation study Ng ML, Kuan WS, Pakkiri LS, Goh ECH, Wu LH, Drum CL. Deep phenotyping of oxidative stress in emergency room patients reveals homoarginine as a novel predictor of sepsis severity, length of hospital stay, and length of intensive care unit stay. Front Med 2022; 9 :1033083. [Ref ID: 301] Nino 2017 Setting not relevant. ICU and ED . Nino ME, Serrano SE, Nino DC, McCosham DM, Cardenas ME, Villareal VP, et al . TIMP1 and MMP9 are predictors of mortality in septic patients in the emergency department and intensive care unit unlike MMP9/TIMP1 ratio: Multivariate model. PLOS ONE 2017; 12 :0171191. [Ref ID: 941] Nowak 2016 Population not relevant. Only recruited patients with confirmed sepsis. Nowak RM, Reed BP, Nanayakkara P, DiSomma S, Moyer ML, Millis S, et al . Presenting hemodynamic phenotypes in ED patients with confirmed sepsis. Am J Emerg Med 2016; 34 :2291–7. [Ref ID: 1009] Okonkwo 2020 Population not relevant. Only patients with septic shock. Okonkwo E, Rozario N, Heffner AC. Presentation and outcomes of end stage liver disease patients presenting with septic shock to the emergency department. Am J Emerg Med 2020; 38 :1408–13. [Ref ID: 691] Ozkan 2021 Population not relevant. Only patients with community-acquired pneumonia, or pneumonia focal sepsis. Ozkan S, Kahveci U, Hur I, Halici A. Prognostic importance of serum presepsin level in pneumonia focal sepsis and its relationship with other biomarkers and clinical severity scores. Saudi Med J 2021; 42 :994–1001. [Ref ID: 562] Park 2013 Population not relevant Park JH, Wee JH, Choi SP, Park KN. Serum procalcitonin level for the prediction of severity in women with acute pyelonephritis in the ED : value of procalcitonin in acute pyelonephritis. Am J Emerg Med 2013; 31 :1092–7. [Ref ID: 1950] Park 2019 Outcome not relevant Park HS, Kim JW, Lee KR, Hong DY, Park SO, Kim SY, et al . Urinary neutrophil gelatinase-associated lipocalin as a biomarker of acute kidney injury in sepsis patients in the emergency department. Clin Chim Acta 2019; 495 :552–5. [Ref ID: 755] Pei 2022 Outcome not relevant Pei Y, Zhou G, Wang P, Shi F, Ma X, Zhu J. Serum cystatin C, kidney injury molecule-1, neutrophil gelatinase-associated lipocalin, klotho and fibroblast growth factor-23 in the early prediction of acute kidney injury associated with sepsis in a Chinese emergency cohort study. Eur J Med Res 2022; 27 :39. [Ref ID: 481] Perman 2020 Population not relevant. Only patients with severe sepsis. Perman SM, Mikkelsen ME, Goyal M, Ginde A, Bhardwaj A, Drumheller B, et al . The sensitivity of qSOFA calculated at triage and during emergency department treatment to rapidly identify sepsis patients. Sci Rep 2020; 10 :20395. [Ref ID: 630] Peschanski 2016 Population not relevant. Only included confirmed sepsis patients. Peschanski N, Chenevier-Gobeaux C, Mzabi L, Lucas R, Ouahabi S, Aquilina V, et al . Prognostic value of PCT in septic emergency patients. Ann Intensive Care 2016; 6 :47. [Ref ID: 989] Puskarich 2013 Population not relevant. Only included septic shock patients. Puskarich MA, Trzeciak S, Shapiro NI, Albers AB, Heffner AC, Kline JA, et al . Whole blood lactate kinetics in patients undergoing quantitative resuscitation for severe sepsis and septic shock. Chest 2013; 143 :1548–53. [Ref ID: 1149] Puskarich 2018 Population not relevant. Focus on severe sepsis. Puskarich MA, Cornelius DC, Bandyopadhyay S, McCalmon M, Tramel R, Dale WD, et al . Phosphatidylserine expressing platelet microparticle levels at hospital presentation are decreased in sepsis non-survivors and correlate with thrombocytopenia. Thromb Res 2018; 168 :138–44. [Ref ID: 95] Quinten 2016 Outcome not relevant Quinten VM, Van Meurs M, Ter Maaten JC, Ligtenberg JJM. Trends in vital signs and routine biomarkers in patients with sepsis during resuscitation in the emergency department: a prospective observational pilot study. BMJ Open 2016; 6 :e009718. [Ref ID: 965] Reaven 2022 Population not relevant. Only patients with septic shock. Reaven MS, Rozario NL, McCarter MSJ, Heffner AC. Incidence and risk factors associated with early death in patients with emergency department septic shock. Acute Crit Care 2022; 37 :193–201. [Ref ID: 1281] Ren 2023 Population not relevant. Only recruited patients with confirmed sepsis. Ren E, Xiao H, Li J, Yu H, Liu B, Wang G, et al . Clinical characteristics and predictors of mortality differ between pulmonary and abdominal sepsis. Shock 2023; 60 :42–50.. [Ref ID: 240] Ryoo 2018 Population not relevant. Septic shock patients only Ryoo SM, Ahn R, Lee J, Sohn CH, Seo DW, Huh JW, et al . Timing of repeated lactate measurement in patients with septic shock at the emergency department. Am J Med Sci 2018; 356 :97–102. [Ref ID: 820] Saeed 2019 Prediction model development or validation study Saeed K, Wilson DC, Bloos F, Schuetz P, Van Der Does Y, Melander O, et al . The early identification of disease progression in patients with suspected infection presenting to the emergency department: a multi-centre derivation and validation study. Crit Care 2019; 23 :40. [Ref ID: 753] Seo 2016 Prediction model development or validation study Seo MH, Choa M, You JS, Lee HS, Hong JH, Park YS, et al . Hypoalbuminemia, low base excess values, and tachypnea predict 28-day mortality in severe sepsis and septic shock patients in the emergency department. Yonsei Med J 2016; 57 :1361–9. [Ref ID: 73] Serano 2019 Population not relevant. Only patients with septic shock. Serano AMN, Alonso JV, Pinero GR, Camacho AR, Benet JS, Vaquero M. Biomarkers in shock patients and their value as a prognostic tool; a prospective multi-center cohort study. Bull Emerg Trauma 2019; 7 :232–9. [Ref ID: 786] Serrano 2022 Prediction model development or validation study Serrano L, Ruiz LA, Perez S, Espana PP, Gomez A, Cilloniz C, et al . Estimating the risk of bacteraemia in hospitalised patients with pneumococcal pneumonia. J Infect 2022; 85 :644–51. [Ref ID: 282] Seymour 2013 Prediction model development or validation study Seymour CW, Yende S, Scott MJ, Pribis J, Mohney RP, Bell LN, et al . Metabolomics in pneumonia and sepsis: an analysis of the GenIMS cohort study. Intensive Care Med 2013; 39 :1423–34. [Ref ID: 167] Shankar-Hari 2018 Outcome not relevant. Composite outcome only. Shankar-Hari M, Datta D, Wilson J, Assi V, Stephen J, Weir CJ, et al . Early PREdiction of sepsis using leukocyte surface biomarkers: the ExPRES-sepsis cohort study. Intensive Care Med 2018; 44 :1836–48. [Ref ID: 835] Shao 2015 Population not relevant. Only analysed patients with confirmed sepsis. Shao R, Li CS, Fang Y, Zhao L, Hang C. Low B and T lymphocyte attenuator expression on CD4+ T cells in the early stage of sepsis is associated with the severity and mortality of septic patients: a prospective cohort study. Crit Care 2015; 19 :308. [Ref ID: 1056] Shetty 2016 Outcome not relevant Shetty AL, Brown T, Booth T, Van KL, Dor-Shiffer DE, Vaghasiya MR, et al . Systemic inflammatory response syndrome-based severe sepsis screening algorithms in emergency department patients with suspected sepsis. Emerg Med Australas 2016; 28 :287–94. [Ref ID: 991] Shetty 2017 Outcome not relevant. Composite outcome only. Shetty A, MacDonald SPJ, Williams JM, van Bockxmeer J, de Groot B, Esteve Cuevas LM, et al . Lactate >=2 mmol/L plus qSOFA improves utility over qSOFA alone in emergency department patients presenting with suspected sepsis. Emerg Med Australas 2017; 29 :626–34. [Ref ID: 903] Shimazui 2021 Population not relevant. Not a suspected sepsis population. Shimazui T, Nakada TA, Yazaki M, Mayumi T, Takasu O, Matsuda K, et al . Blood interleukin-6 levels predict multiple organ dysfunction in critically ill patients. Shock 2021; 55 :790–5. [Ref ID: 532] Shin 2016 Outcome not relevant Shin TG, Jo IJ, Hwang SY, Jeon K, Suh GY, Choe E, et al . Comprehensive interpretation of central venous oxygen saturation and blood lactate levels during resuscitation of patients with severe sepsis and septic shock in the emergency department. Shock 2016; 45 :4–9. [Ref ID: 1003] Shin 2018 Population not relevant. Septic shock patients only. Shin J, Hwang SY, Jo IJ, Kim WY, Ryoo SM, Kang GH, et al . Prognostic value of the lactate/albumin ratio for predicting 28-day mortality in critically ill sepsis patients. Shock 2018; 50 :545–50. [Ref ID: 836] Shinde 2023 Outcome not relevant. Unclear outcome. Shinde VV, Jha A, Natarajan MSS, Vijayakumari V, Govindaswamy G, Sivaasubramani S, et al . Serum procalcitonin vs. SOFA score in predicting outcome in sepsis patients in medical intensive care unit. Indian J Crit Care Med 2023; 27 :348–51. [Ref ID: 247] Sinto 2020 Population not relevant. Patients with suspected infection (population too broad). Sinto R, Suwarto S, Lie KC, Harimurti K, Widodo D, Pohan HT. Prognostic accuracy of the quick Sequential Organ Failure Assessment (qSOFA)-lactate criteria for mortality in adults with suspected bacterial infection in the emergency department of a hospital with limited resources. Emerg Med J 2020; 37 :363–9. [Ref ID: 684] Sivayoham 2019 Prediction model development or validation study Sivayoham N, Blake LA, Tharimoopantavida SE, Chughtai S, Hussain AN, Cecconi M, et al . The REDS score: a new scoring system to risk-stratify emergency department suspected sepsis: a derivation and validation study. BMJ Open 2019; 9 :e030922. [Ref ID: 787] Song 2016 Population not relevant. Only patients with septic shock. Song JE, Kim MH, Jeong WY, Jung IY, Oh DH, Kim YC, et al . Mortality risk factors for patients with septic shock after implementation of the surviving sepsis campaign bundles. Infect Chemother 2016; 48 :199–208. [Ref ID: 1018] Song 2019 Population not relevant. Only analysed patients with confirmed sepsis. Song J, Park DW, Moon S, Cho HJ, Park JH, Seok H, et al . Diagnostic and prognostic value of interleukin-6, pentraxin 3, and procalcitonin levels among sepsis and septic shock patients: a prospective controlled study according to the sepsis-3 definitions. BMC Infect Dis 2019; 19 :968. [Ref ID: 736] Song 2019 Population not relevant. Study of community acquired pneumonia (CAP). Song H, Moon HG, Kim SH. Efficacy of quick Sequential Organ Failure Assessment with lactate concentration for predicting mortality in patients with community-acquired pneumonia in the emergency department. Clin Exp Emerg Med 2019; 6 :1–8. [Ref ID: 1751] Sonmez 2020 Population not relevant. Only recruited patients with confirmed sepsis. Sonmez BM, Celikbas AK. Sepsis-related mortality with SOFA and qSOFA in emergency department patients. Ann Clin Anal Med 2020; 11 :359–64. [Ref ID: 668] Su 2018 Population not relevant. Patients with Parkinson’s disease and suspected serious infection. Su CM, Kung CT, Chen FC, Cheng HH, Hsiao SY, Lai YR, et al . Manifestations and outcomes of patients with Parkinson’s disease and serious infection in the emergency department. Biomed Res Int 2018; 2018 :6014896. [Ref ID: 845] Sugimoto 2021 Population not relevant. Only recruited patients with confirmed sepsis. Sugimoto M, Takayama W, Murata K, Otomo Y. The impact of lactate clearance on outcomes according to infection sites in patients with sepsis: a retrospective observational study. Sci Rep 2021; 11 :22394. [Ref ID: 497] Swenson 2018 Population not relevant. Only patients with severe sepsis. Swenson KE, Dziura JD, Aydin A, Reynolds J, Wira CR. Evaluation of a novel 5-group classification system of sepsis by vasopressor use and initial serum lactate in the emergency department. Intern Emerg Med 2018; 13 :257–68. [Ref ID: 892] Takada 2020 Population not relevant Takada T, Hoogland J, Yano T, Fujii K, Fujiishi R, Miyashita J, et al . Added value of inflammatory markers to vital signs to predict mortality in patients suspected of severe infection. Am J Emerg Med 2020; 38 :1389–95. [Ref ID: 1543] Teng 2019 Population not relevant. Study of influenza patients. Teng F, Wan TT, Guo SB, Liu X, Cai JF, Qi X, et al . Outcome prediction using the Mortality in Emergency Department Sepsis score combined with procalcitonin for influenza patients. Med Clin (Barc) 2019; 153 :411–7. [Ref ID: 769] Tsalik 2021 Prediction model development or validation study Tsalik EL, Henao R, Montgomery JL, Nawrocki JW, Aydin M, Lydon EC, et al . Discriminating bacterial and viral infection using a rapid host gene expression test. Crit Care Med 2021; 49 :1651–63. [Ref ID: 269] Ulla 2013 Population not relevant. Only analysed confirmed sepsis patients. Ulla M, Pizzolato E, Lucchiari M, Loiacono M, Soardo F, Forno D, et al . Diagnostic and prognostic value of presepsin in the management of sepsis in the emergency department: a multicenter prospective study. Crit Care 2013; 17 :R168. [Ref ID: 1166] Upadhyaya 2023 Outcome not relevant. Composite outcome only. Upadhyaya DP, Tarabichi Y, Prantzalos K, Ayub S, Kaelber DC, Sahoo SS. Characterizing the importance of hematologic biomarkers in screening for severe sepsis using machine learning interpretability methods. medRxiv 2023. [Ref ID: 1196] van Tienhoven 2020 Population not relevant. Only patients with normo- and hyper-lactaemia. van Tienhoven AJ, van Beers CAJ, Siegert CEH, Nanayakkara PWB. The utility of peripheral venous lactate in emergency department patients with normal and higher lactate levels: a prospective observational study. Acute Med 2020; 19 :125–30. [Ref ID: 640] Velasquez 2013 Population not relevant. Patients with suspected infection (population too broad). Velasquez S, Matute JD, Gamez LY, Enriquez LE, Gomez ID, Toro F, et al . Characterization of nCD64 expression in neutrophils and levels of s-TREM-1 and HMGB-1 in patients with suspected infection admitted in an emergency department. Biomedica 2013; 33 :643–52. [Ref ID: 1192] Wang 2014 Population not relevant. Analysis of confirmed sepsis only. Wang M, Zhang Q, Zhao X, Dong G, Li C. Diagnostic and prognostic value of neutrophil gelatinase-associated lipocalin, matrix metalloproteinase-9, and tissue inhibitor of matrix metalloproteinases-1 for sepsis in the emergency department: an observational study. Crit Care 2014; 18 :634. [Ref ID: 1087] Wang 2020 Population not relevant. Only recruited patients with confirmed sepsis. Wang L, Zhou W, Wang K, He S, Chen Y. Predictive value of circulating plasma mitochondrial DNA for Sepsis in the emergency department: observational study based on the sepsis-3 definition. BMC Emerg Med 2020; 20 :25. [Ref ID: 709] Wang 2021 Population not relevant. Only analysed patients with confirmed sepsis. Wang TH, Hsu YC. Red cell distribution width as a prognostic factor and its comparison with lactate in patients with sepsis. Diagnostics 2021; 11 :1474. [Ref ID: 227] Wang 2022 Population not relevant. Only recruited patients with confirmed sepsis. Wang L, Tang C, He S, Chen Y, Xie C. Combined suPAR and qSOFA for the prediction of 28-day mortality in sepsis patients. Signa Vitae 2022; 18 :119–27. [Ref ID: 470] Wardi 2017 Population not relevant. Only patients with severe sepsis and septic shock. Wardi G, Wali AR, Villar J, Tolia V, Tomaszewski C, Sloane C, et al . Unexpected intensive care transfer of admitted patients with severe sepsis. J Intensive Care 2017; 5 :43. [Ref ID: 943] Wasserman 2019 Population not relevant. Only recruited patients with confirmed sepsis. Wasserman A, Karov R, Shenhar-Tsarfaty S, Paran Y, Zeltzer D, Shapira I, et al . Septic patients presenting with apparently normal C-reactive protein: a point of caution for the ER physician. Medicine 2019; 98 :e13989. [Ref ID: 1603] Webb 2020 Outcome not relevant Webb AL, Kramer N, Rosario J, Dub L, Lebowitz D, Amico K, et al . Delta lactate (three-hour lactate minus initial lactate) prediction of in-hospital death in sepsis patients. Cureus 2020; 12 :e7863. [Ref ID: 1326] Webb 2020 Population not relevant. Only patients with severe sepsis. Webb AL, Kramer N, Stead TG, Mangal R, Lebowitz D, Dub L, et al . Serum procalcitonin level is associated with positive blood cultures, in-hospital mortality, and septic shock in emergency department sepsis patients. Cureus 2020; 12 :e7812. [Ref ID: 1327] Wittayachamnankul 2020 Outcome not relevant Wittayachamnankul B, Apaijai N, Sutham K, Chenthanakij B, Liwsrisakun C, Jaiwongkam T, et al . High central venous oxygen saturation is associated with mitochondrial dysfunction in septic shock: a prospective observational study. J Cell Mol Med 2020; 24 :6485–94. [Ref ID: 701] Wiwatcharagoses 2016 Population not relevant. All either confirmed sepsis or bacteraemia. Wiwatcharagoses K, Kingnakom A. Procalcitonin under investigation as a means of detecting severe sepsis, septic shock and bacteremia at emergency department, Rajavithi hospital. J Med Assoc Thai 2016; 99 :S63–8. [Ref ID: 990] Yamamoto 2015 Prediction model development or validation study Yamamoto S, Yamazaki S, Shimizu T, Takeshima T, Fukuma S, Yamamoto Y, et al . Prognostic utility of serum CRP levels in combination with CURB-65 in patients with clinically suspected sepsis: a decision curve analysis. BMJ Open 2015; 5 :e007049. [Ref ID: 376] Yan 2021 Population not relevant. Study of confirmed sepsis only. Yan S, Zhang G. Predictive performance of critical illness scores and procalcitonin in sepsis caused by different gram-stain bacteria. Clinics 2021; 76 :e2610. [Ref ID: 504] Yang 2020 Population not relevant. Septic shock patients only. Yang WS, Kang HD, Jung SK, Lee YJ, Oh SH, Kim YJ, et al . A mortality analysis of septic shock, vasoplegic shock, and cryptic shock classified by the Third International Consensus Definitions (Sepsis-3). Clin Respir J 2020; 14 :857–863. [Ref ID: 354] Yang 2021 Outcome not relevant Yang WS, Kim YJ, Ryoo SM, Kim WY. Independent risk factors for sepsis-associated cardiac arrest in patients with septic shock. Int J Environ Res Public Health 2021; 18 :4971. [Ref ID: 567] Yang 2022 Population not relevant. Study of confirmed sepsis only. Yang L, Lin Y, Zhang X, Wei B, Wang J, Liu B. Predictive value of combination of procalcitonin and predisposition, infection, response, and organ dysfunction (PIRO) system in septic patients with positive blood cultures in the emergency department. Infect Drug Resist 2022; 15 :6189–202. [Ref ID: 309] Yin 2013 Population not relevant. Only analysed confirmed sepsis patients. Yin Q, Liu B, Chen Y, Zhao Y, Li C. The role of soluble thrombomodulin in the risk stratification and prognosis evaluation of septic patients in the emergency department. Thromb Res 2013; 132 :471–6. [Ref ID: 1190] Yin 2014 Population not relevant. Only analysed confirmed sepsis patients. Yin Q, Liu B, Chen Y, Zhao Y, Li C. Prognostic value of the International Society on Thrombosis and Haemostasis scoring system for overt disseminated intravascular coagulation in emergency department sepsis. Infection 2014; 42 :629–37. [Ref ID: 1130] Yin 2020 Population not relevant. Only recruited patients with confirmed sepsis. Yin WP, Li JB, Zheng XF, An L, Shao H, Li CS. Effect of neutrophil CD64 for diagnosing sepsis in emergency department. World J Emerg Med 2020; 11 :79–86. [Ref ID: 287] Yu 2019 Population not relevant. Only patients with septic shock. Yu G, Yoo SJ, Lee SH, Kim JS, Jung S, Kim YJ, et al . Utility of the early lactate area score as a prognostic marker for septic shock patients in the emergency department. Acute Crit Care 2019; 34 :126–32. [Ref ID: 1743] Yu 2022 Outcome not relevant Yu S, Song SA, June KR, Park HY, Lee JN. Clinical Performance of monocyte distribution width for early detection of sepsis in emergency department patients: a prospective study. Ann Lab Med 2022; 42 :286–9. [Ref ID: 1401] Yu 2022 Population not relevant. Only analysed patients with confirmed sepsis. Yu B, Chen M, Zhang Y, Cao Y, Yang J, Wei B, et al . Diagnostic and prognostic value of interleukin-6 in emergency department sepsis patients. Infect Drug Resist 2022; 15 :5557–66. [Ref ID: 442] Yu 2019 Prediction model development or validation study Yu H, Nie L, Liu A, Wu K, Hsein YC, Yen DW, et al . Combining procalcitonin with the qSOFA and sepsis mortality prediction. Medicine (United States) 2019; 98 :e15981. [Ref ID: 789] Zhang 2014 Population not relevant. Only confirmed sepsis in analysis. Zhang Q, Dong G, Zhao X, Wang M, Li CS. Prognostic significance of hypothalamic-pituitary-adrenal axis hormones in early sepsis: a study performed in the emergency department. Intensive Care Med 2014; 40 :1499–508. [Ref ID: 1082] Zhang 2021 Population not relevant. Only recruited patients with confirmed sepsis. Zhang J, He XH, Yang J, Guo SB. Role of red blood cell distribution width in predicting the prognosis of patients with sepsis. Hong Kong J Emerg Med 2021; 28 :199–204. [Ref ID: 500] Zhang 2022 Population not relevant. Only analysed patients with confirmed sepsis. Zhang F, Wan T, Liu X, Guo S. Prediction of short-term mortality in elderly patients with sepsis using immunoglobulin G2: an observational study. Heliyon 2022; 8 :e12642. [Ref ID: 1230] Zhang 2013 Population not relevant. Confirmed sepsis only recruited. Zhang XH, Dong Y, Chen YD, Zhou P, Wang JD, Wen FQ. Serum N-terminal pro-brain natriuretic peptide level is a significant prognostic factor in patients with severe sepsis among Southwest Chinese Population. Eur Rev Med Pharmacol Sci 2013; 17 :517–21. [Ref ID: 1155] Zhao 2013 Population not relevant. Only analysed confirmed sepsis patients. Zhao Y, Li C, Jia Y. Evaluation of the mortality in emergency department sepsis score combined with procalcitonin in septic patients. Am J Emerg Med 2013; 31 :1086–91. [Ref ID: 1156] Zhao 2018 Population not relevant. Only analysed confirmed sepsis patients. Zhao Y, Jia Y, Li C, Fang Y, Shao R. The risk stratification and prognostic evaluation of soluble programmed death-1 on patients with sepsis in emergency department. Am J Emerg Med 2018; 36 :43–8. [Ref ID: 868] Zhao 2020 Population not relevant. Only recruited patients with confirmed sepsis. Zhao J, He Y, Xu P, Liu J, Ye S, Cao Y. Serum ammonia levels on admission for predicting sepsis patient mortality at D28 in the emergency department: A 2-center retrospective study. Medicine (United States) 2020; 99 :E19477. [Ref ID: 246] Zhou 2020 Population not relevant. Only recruited patients with confirmed sepsis. Zhou H, Lan T, Guo S. Prognostic prediction value of qSOFA , SOFA , and admission lactate in septic patients with community-acquired pneumonia in emergency department. Emerg Med Int 2020; 2020 :7979353. [Ref ID: 710] Zhou 2020 Population not relevant. Only recruited patients with confirmed sepsis. Zhou HJ, Lan TF, Guo SB. Outcome prediction value of National Early Warning Score in septic patients with community-acquired pneumonia in emergency department: a single-center retrospective cohort study. World J Emerg Med 2020; 11 :206–15. [Ref ID: 290] Appendix 5. Included studies TABLE 4 Characteristics of the included studies View in own window Author, year, country Study size, n Age Mean (SD) Sex/gender, % male Race (%) Pre-morbid functional status/frailty score Comorbidities (%) Clinical score(s), median (IQR) a % who are in care (e.g. nursing homes, DNACPR) Confirmed sepsis (%) Biomarker(s) Athan, 2023, 69 Australia 1716 60 (45–75) b 53.2 White 86.6 NR NR NR 0 NR CRP Aboriginal, Torres Strait Islander, Pacific Islander 10.0 Terminal illness, and documented limits on goals of care excluded Asian, South Asian/Indian 1.6 Other 1.7 Baumann, 2020, 70 USA 2584 60 (46–73) b 51.9 White 39.3 NR NR NR 0 67.8 sepsis Lactate African American 32.6 Hospice or DNACPR patients excluded 34.4 severe sepsis Hispanic 15.8 Asian 4.4 Native American 0.5 Other 7.5 Bolanaki, 2021, 61 Germany 742 68 (56–78) b 58 NR NR Previous MI 15.9 Heart failure 37.9 PAD 12.9 Cerebrovascular disease 12.8 Dementia 5.4 Chronic respiratory disease 36.5 Collagenosis 2.8 Ulcers 6.2 Mild liver disease 2.8 Severe liver disease 2.1 Hemiplegia 2.6 DM (no end organ damage) 20.2 DM (end organ damage) 5.4 Medium- severe renal disease 6.4 Tumour 10.9 Metastatic solid tumour 7.2 Leukaemia 1.4 Lymphoma 3.2 AIDS 0.5 CCI (median, IQR) 2 (1–3) qSOFA Score 1 : 77.1% Score 2 : 20.9% Score 3 : 2.0% Patients admitted for palliative care with a life expectancy of < 1 month were excluded 27.2 PCT Boland, 2016, 72 USA 112 24 (22–100) c 45 NR NR NR NR NR 24 Lactate Caramello, 2020, 62 Italy 469 69.7 (16.4) 57 NR NR Neoplasm 28% DM 23% History of heart failure, CAD or stroke 24% Immunosuppressed 13% qSOFA 1 (0–1) MEWS 3 (1–4) NR Defined by ED physician: 96.1 sepsis 3.9 septic shock Sepsis-3 definition: 62 sepsis Lactate PCT Castello, 2019, 59 Italy 101 80 (73–88) b 56.4 NR NR Heart failure 24.8 Previous stroke 18.8 Dementia 27.7 COPD 16.8 DM 32.7 Neoplasia 23.8 Arterial hypertension 70.3 CKD 30.7 qSOFA Score 2 : 68.3%; Score 3 : 31.7% NR 91.1 OPN Cheng, 2018, 64 Taiwan (Province of China) 7087 67.3 (15.8) 57.6 NR NR Liver cirrhosis 8.7 DM 34.6 Chronic renal insufficiency 21.4 CHF 7.9 Cerebrovascular disease 15.7 Malignancy 24 NR NR NR Lactate Choo, 2020, 57 Republic of Korea 300 75 (66–81) b 49.3 NR NR NR qSOFA 1 (0–2) NR 44.7 sepsis 28.0 septic shock IMA PCT Lactate hsCRP Contenti, 2015, 73 France 103 70 (20) 46.4 NR NR None 47.6 HT 29.1 DM 10.7 Cancer 12.6 Dementia 14.6 NR NR 38.8 sepsis 61.2 severe sepsis Lactate Covino, 2021, 34 Italy 6595 [including 422 suspected sepsis and 6173 non-suspected sepsis (i.e. non-sepsis)] 71 (58–81) b 55.6 NR NR Severe obesity 2.0 Hypertension 28.7 CAD 16.7 Heart failure 19.0 PVD 22.0 Dementia 6.4 COPD 14.2 DM 22.2 CKD 22.4 Leukaemia/lymphoma 8.3 Malignancy 27.1 CCI [median (IQR)] 5 (3–7) NR NR 6.4 PCT Dadeh, 2022, 74 Thailand 92 68 (–) b 52.2 NR NR DM 26.1 Chronic renal failure 8.7 Immunosuppressive use 9.8 HIV infection 2.2 Chemotherapy session 31.5 Other 37 NEWS 24-hour non-survivor, 12 (10.5–12.5) 24-hour survivor 8 (6–9) Palliative care patients excluded Patients who refused resuscitation excluded NR (haemoculture positive 35.9%) Lactate NEWS-L Devia Jarmillo, 2022, 31 Colombia 179 Total group 77 (6.7) NR NR NR NR NR NR NR PCT Lactate 45 non-survivors NR 51.1 non-survivors NR NR Non-survivors: Chronic renal failure on dialysis 22.2 Severe liver disease 15.6 CCI > 5 64.4 qSOFA Reported for non-survivors as ‘ n = 7, 38.9%’ (unclear) NR 71.1 Survivors 134 NR Survivors 50 NR NR Survivors: Chronic renal failure on dialysis 31.3 Severe liver disease 29.9 CCI > 5 66.9 Reported for survivors as ‘ n = 11. 61.1%’ (unclear) NR 47.8 D’Onofrio, 2021, 56 Belgium 1690 70 (55–80) b 57.8 NR NR Cardiac comorbidities 17.8 HT 22.1 Chronic pulmonary disease 15.6 Cerebrovascular disease 8.2 Renal insufficiency 15.1 Liver disease 3.1 DM 15.1 Solid malignancies 10.4 Solid metastatic malignancies 10.5 Haematological malignancies 3.2 CCI [median (IQR)] 1 (0–3) NR NR NR Lactate Dudaryk, 2021, 55 USA 10,716 63 (51–75) b 50 NR NR NR NR Hospice patients excluded NR Lactate Fang, 2015, 54 China 440 total group NR NR NR NR NR NR NR 40 sepsis 19.1 severe sepsis 16.6 septic shock Ang-1 Ang-2 Tie-2 PCT 107 SIRS 68 (56–76) b 52.6 NR NR NR MEDS 5 (3–6) APACHE II 6 (3–8) NR 0 176 Sepsis 71 (59–78) b 61 NR NR NR MEDS 7 (6–8) APACHE II 7 (6.07–9) NR 100 84 severe sepsis 74 (61–78) b 60.3 NR NR NR MEDS 19 (18–22) APACHE II 15.5 (10–22) NR 100 73 septic shock 75 (65–78) 58.1 NR NR NR MEDS 25 (24–26) APACHE II 21 (16.1–29) NR 100 Gonzalez del Castillo, 2019, 53 Spain 684 65.1 (19.6) 53.5 NR NR CVD 37.3 DM 22.2 Immunodeficiency 15.4 Liver disease 7.9 Malignancy 26.8 Neurological disorders 19.3 Respiratory disease 26.3 Renal disease 18.7 NEWS 2 (1–5) qSOFA 0 (0–1) NR CRP PCT Lactate MR-pro ADM Guarino, 2023, 25 Italy 203 total group NR NR NR NR NR NR NR 92.4 Capillary lactate Serum lactate 175 survived 85 (79–90) b 43.4 NR NR NR NEWS 8 (5–10) NR NR 28 deceased 84 (71–88) b 39.3 NR NR NR NEWS 10 (9–13) NR NR Gunes Ozaydin, 2017, 63 Turkey 200 74 (15) 55 NR NR 24 chronic pulmonary disease 22 DM 21malignancy 19 CHF 19 cerebrovascular disease 14 chronic renal disease, 5 chronic liver disease MEDS [mean (SD)] 9.3 (–) NR NR Lactate Hargreaves, 2023, 60 UK 1233 79 (68–86) b 54 NR NR Congestive cardiac failure 17 DM 26 Vascular disease 26 Liver disease 34 NEWS 6 (4–8) NR 23 Lactate Henning, 2017, 24 USA 7637 (total group) (4526 with available lactate level) NR NR NR NR NR NR NR SEP-3 criteria qSOFA ≥ 2 15.9% (95% CI: 15.1% to 16.7%) Lactate or qSOFA f 2132 cohort 1 56.9 (20.8) 52.2 NR NR DM 26.5 COPD 12.6 Malignancy 2.0 Liver disease 6.5 qSOFA > 2 n = 314 (14.7%) NR 14.7 ( qSOFA ≥2) 4618 cohort 2 59.9 (19.9) 51.6 NR NR DM 22.5 COPD 7.7 Malignancy 2.7 Liver disease 6.2 qSOFA > 2 n = 689 (14.9%) NR 14.9 ( qSOFA ≥2) 50.3 (1992 consensus guideline) 1004 cohort 3 54.4 (18.9) 50.9 NR NR DM 23.8 COPD 16.6 Malignancy 5.6 Liver disease 5.4 qSOFA > 2 n = 211 (23.8%) NR 21.0 ( qSOFA ≥ 2) NR (1992 consensus guideline) Hong, 2016, 43 Republic of Korea 470 74 (61–81) b 47.7 NR NR HT 40.2 DM 26.2 Cerebrovascular disease 17.7 Chronic liver disease 11.1 Chronic heart disease 8.1 COPD 4.9 NR Palliative care admissions excluded NR CRP PCT NGAL Hou, 2021, 23 Taiwan (Province of China) 8698 before propensity score matching (total group) NR NR NR NR NR NR NR 3.5 CRP NLR PLR MDW 308 before propensity matching (sepsis) 68.7 (17.4) 47.4 NR NR HT 48.4 DM 35.9 CAD 27.9 Cerebrovascular disease 8.4 ESRD 5.2 Pulmonary disease 1.0 Malignant disease 4.6 GCS [mean (SD)] 13.5 (2.9) NR 100 8390 before propensity score matching (non-sepsis) 58.5 (21.1) 46.7 NR NR HT 34.4 DM 20.8 CAD 21.0 Cerebrovascular disease 3.8 ESRD 2.5 Pulmonary disease 0.7 Malignant disease 3.5 GCS [mean (SD)] 14.5 (1.8) NR 0 1480 after propensity score matching (total group) NR NR NR NR NR NR NR 20 296 after propensity score matching (sepsis) 68.6 (17.7) 45.6 NR NR HT 48.3 DM 35.8 CAD 27.4 Cerebrovascular disease 8.5 ESRD 4.4 Pulmonary disease 1.0 Malignant disease 4.7 GCS [mean (SD)] 13.5 (2.9) NR 100 1184 after propensity score matching (non-sepsis) 69.4 (18) 45.6 NR NR HT 51.4 DM 35.3 CAD 27.5 Cerebrovascular disease 6.4 ESRD 5.0 pulmonary disease 1.4 malignant disease 4.5 GCS [mean (SD)] 13.7 (2.7) NR 0 Hunter, 2013, 50 USA 201 NR 53 NR NR NR NR NR 66.7 suspected sepsis 19.9 severe sepsis 13.4 septic shock Lactate Jekarl, 2019, 46 Republic of Korea 248 70.1 (14.9) 50 NR NR NR GCS [mean (SD)] 13.4 (3.1) NR 74.6 PCT CRP Jiang, 2019, 36 USA 408 61 (18.7) 49 NR NR NR RASI NR NR NR Lactate RASI Katsaros, 2022, 35 Greece 371 total group NR NR NR NR NR NR NR 44.7 HBP 29 28-day non-survivors 83.1 (7.9) 58.6 NR NR DM (type 2) 27.6 Chronic heart failure 20.7 Coronary heart disease 17.2 COPD 31.0 Chronic renal disease 6.9 Corticosteroid intake 3.4 Chemotherapy 13.8 Non-metastatic solid tumour 6.9 Metastatic solid tumour 6.9 Ischaemic stroke 17.2 Atrial fibrillation 13.8 Dementia 27.6 Parkinson’s disease 3.4 Nephrolithiasis 0 Gallstones 6.9 Depression 6.9 CCI [mean (SD)] 5.72 (2.96) qSOFA [mean (SD)] 1.48 (1.02) APACHE II [mean (SD)] 17.46 (7.62) NR NR 342 28-day survivors 60.7 (21.2) 50.9 NR NR DM (type 2) 17.0 Chronic heart failure 8.2 Coronary heart disease 10.2 COPD 11.4 Chronic renal disease 6.1 Corticosteroid intake 3.5 Chemotherapy 4.1 Non-metastatic solid tumour 3.8 Metastatic solid tumour 5.0 Ischaemic stroke 5.0 Atrial fibrillation 9.9 Dementia 11.1 Parkinson’s disease 2.6 Nephrolithiasis 2.8 Gallstones 6.4 Depression 8.8 CCI [mean (SD)] 2.81 (2.93) qSOFA [mean (SD)] 0.53 (0.64) APACHE II [mean (SD)] 7.91 (5.09) NR NR Kece, 2016, 42 Turkey 86 61 (54–69) b 68.6 NR NR DM 16.2 HT 31.3 COPD 11.6 CHF 9.3 Chronic renal failure 3.4 None 45.4 Other 2.3 NR NR 25.6 Lactate PCT Kim, 2019, 41 Republic of Korea 866 73 (61–80) b 56 NR NR NR NR NR 66.9 (sepsis and septic shock) PCT DNI CRP Klimpel, 2019, 49 Germany 916 74 (62–82) b 56.3 NR NR At least one chronic disease 60.7 qSOFA score ≥ 2 23.6% Decision for end-of-life care on admission to the ED exclusion criterion NR (23.6 of those with qSOFA ≥2) Lactate Lee, 2016, 22 Republic of Korea 36 76.5 (70.5–81.5) b 55.6 NR NR NR Abbreviated MEDS ICU admission group 13.5 (12.0–15.0) General ward admission group 8.0 (6.0–8.0) NR NR PCT IL10 IL6 IL5 CRP Lee, 2022, 30 Republic of Korea 249 69.31 (16.56) 44.2 NR NR CVD 24.1 Cerebrovascular 10.4 Chronic pulmonary 12.4 CKD 22.1 Chronic liver 7.6 Malignancy 17.3 Bacteraemia 20.9 AKI 35.3 APACHE II Mean (SD) 10.25 (12.14) SAPS III Mean (SD) 30 (31.38) NR 39.4 sepsis 14.1 septic shock PSEP PCT CRP Lactate Lucas, 2021, 76 UK 1242 62% > 65 year 49.1 NR NR NR NR NR NR PCT MacDonald, 2017, 37 Australia 186 (total group) NR NR NR NR NR NR NR 37.1 sepsis 47.8 septic shock Resistan NGAL VCAM-1 ICAM-1 IL-6 IL-10 Lactate 28 Infection 59 (44–72) b 57 NR NR CCI [median (IQR)] 1 (0–3) MEDS 4 (3–6) SOFA 1 (0–1) NR 0 69 Sepsis 71 (55–78) b 68 NR NR CCI [median (IQR)] 3 (2–4) MEDS 7 (5–10) SOFA 3 (2–5) NR 100 Septic shock 67 (51–76) b 63 NR NR CCI [median (IQR)] 2 (1–4) MEDS 11 (8–13) SOFA 6 (5–9) NR Magrini, 2013, 77 Italy 261 72.7 (15.1) 46 NR NR NR NR NR 36.7 PCT CRP Magrini, 2014, 28 Italy 513 72.7 (15.1) 51.3 NR NR DM 23.8 CVD 59.1 Neoplasms 5.8 COPD 25.1 CKD 12.9 NR NR 43.1 PCT WBC CRP Malinovska, 2022, 38 USA 7952 50 (34–63) 47.2 Black 57.4 White 31.4 Other 11.1 NR CAD 7.7 Cancer 14.5 Cerebrovascular disease 4.9 DM 13.3 Heart failure 6.0 HT 25.5 Immunosuppression 10.7 Kidney disease 10.5 Liver disease 11.3 Prior respiratory failure 0.5 qSOFA ≥ 2 1.2% NR 1.7 sepsis without shock 0.5 septic shock MDW Musikatavorn, 2015, 47 Thailand 392 44 (14.2) 53.1 NR NR DM 18.1 Chronic renal failure 9.2 Chronic liver diseases 8.7 Active malignancy 31.6 Receiving chemotherapy 23.2 Neutropenia 12.0 Immunosuppressive drugs 11.5 REMS ≥ 6 13.5% NR NR ( SIRS criteria ≥ 65.6%) Lactate Noparatkailas, 2023, 33 Thailand 448 total group 71 (59–87) b 44.6 NR Most prevalent comorbidities: Cancer 29.9 Cerebrovascular accident 19.6 DM 19.0 NR NR Lactate 200 blood lactate < 2 mmol/l 71.5 (59–81) b 39 NR NR Blood lactate < 2 mmol/l: CAD 8.0 CHF 13.0 COPD 11.5 Cerebrovascular accident 19.5 DM 20.5 ESRD 14.5 Cirrhosis 2.5 Cancer 23.5 Immunocompromised 10.0 NEWS 7 (–) qSOFA 1 (1–1) NR NR 248 blood lactate ≥ 2 mmol/l 70 (59–82) b 49.2 NR NR Blood lactate > 2 mmol/l: CAD 7.3 CHF 12.9 COPD 13.7 Cerebrovascular accident 19.8 DM 17.7 ESRD 8.9 Cirrhosis 6.9 Cancer 35.1 Immunocompromised 5.6 NEWS 5 (–) qSOFA 1 (1–2) NR NR Parke, 2023, 67 Sweden 351 total group NR NR NR NR NR NR NR NR Calprotectin PCT CRP 66 Infection: transfer to ICU/ HDU 71.5 (18–94) b 65 NR NR MI 7.5 Heart failure 21 PAD 10.5 Stroke 16.5 Dementia 12 COPD 10.5 Connective tissue disease 4.5 Peptic ulcer disease 6 Liver disease 7.5 DM 24 Hemiplegia 1.5 Moderate to severe kidney failure 19.5 Solid tumour 15 Lymphoma 6 Leukaemia 7.5 HIV/AIDS 0 CCI [median (IQR)] 5 (0–15) NR NR 86 sepsis 24 septic shock 253 infection: transfer to ward 73 (19–96) b 60 NR NR MI 14 (%) Heart failure 17 PAD 9 Stroke 17 Dementia 12 COPD 18 Connective tissue disease 7 Peptic ulcer disease 3 Liver disease 5.5 DM 23 Hemiplegia 4 Moderate to severe kidney failure 21 Solid tumour 21 Lymphoma 3 Leukaemia 5 HIV/AIDS 0 CCI [median (IQR)] 5 (0–18) NR NR 81 sepsis 0 septic shock 32 no infection NR NR NR NR NR NR NR NR Puskarich, 2015, 29 USA 69 NR NR NR NR NR NR DNR patients excluded NR miR-146a miR-223 miR-150 29 sepsis 55.1 (16.5) 65.5 White 72.4 African American 27.6 Asian 0 Ethnicity: Non-Hispanic 96.6 Hispanic 3.4 NR DM 31 CHF 17 COPD 14 Malignancy 14 NR DNR patients excluded NR 40 septic shock 62 (18.6) 62.5 White 75 African American 22.5 Asian 2.5 Ethnicity: Non-Hispanic 95 Hispanic 0.05 NR DM 35 CHF 25 COPD 15 Malignancy 15 NR DNR patients excluded NR Reshmi, 2021, 26 India 215 66.9 (12.7) 69.3 NR NR NR DM, chronic liver disease, CKD measured and in model NR NR NR NT-pro BNP Ruangsomboon, 2020, 58 Thailand 250 83.4 (5.4) b 39.2 NR NR DNR status 29.6% DM 41.6 HT 72 Dyslipidaemia 40.8 CKD or ESRD 27.6 CAD 23.2 Cerebrovascular disease 20.4 Cancer 20.4 NEWS [mean (SD)] 6.7 (3.4) qSOFA [mean (SD)] 1.4 (0.7) REMS [mean (SD)] 10 (2.8) Patients with DNR status 29.6 36.4 sepsis 19.2 septic shock PSEP PCT CRP Shetty, 2018, 51 Australia 12,349 72.6 (58.1–82.6) b NR NR NR NR NR NR NR Lactate Shim, 2019, 39 Republic of Korea 199 67.1 (17.1) 50.3 NR NR NR qSOFA score > 2 37.2 Patients with limited life expectancy due to chronic diseases were excluded NR qSOFA score > 2 in 37.2% of patients Whole blood PCT Serum PCT Lactate CRP Singer, 2014, 27 USA 258 64 (19) 54 White 82 NR Comorbidities measured (no % given): DM COPD CHF CAD HIV ESRD Malignancy Organ transplant Indwelling vascular line Residence of a nursing home NR Measured but NR 81.6 Lactate Sohn, 2019, 40 Republic of Korea 2698 52.9 (21.4) 46.8 NR NR NR qSOFA score 0 61.7 1 30.7 2 6.3 3 1.2 NR NR Lactate Suttapanit, 2021, 32 Thailand 1139 total group NR NR NR NR NR NR Patients with DNACPR order excluded NR Lactate 118 deceased 70 (67–73) d 49 NR NR HT 50.8 DM 33.9 CKD 25.4 Coronary heart disease 25.4 Liver cirrhosis 12.7 Immunocompromised 40.7 Cerebrovascular disease 16.9 qSOFA 4 (4–5) NR NR 1021 survived 70 (69–71) d 46 NR NR HT 55.5 DM 36.4 CKD 25.7 Coronary heart disease 22.4 Liver cirrhosis 9.8 Immunocompromised 34.8 Cerebrovascular disease 16.2 qSOFA 1 (1–2) NR NR Ueno, 2021, 48 Japan 951 79 (68.5–85) b 59.3 NR Clinical frailty score [median (IQR)] 4 (3–6) NR qSOFA 1 (1–2) NR NR Pulmonary sepsis: 59.6% in non-survivors 45.9% in survivors Urinary sepsis 12.6% in non-survivors 45.9% in survivors Lactate Uusitalo-Seppala, 2013, 52 Finland 537 64 (–) b (Range 18–100) 57.7 NR NR DM 15.1 Malignancy (solid or haematological) 17.7 Rheumatic disease 9.3 Chronic renal insufficiency 3.4 CVD 53.8 COPD/asthma 20.1 NR NR 66.5 CRP PCT PTX3 Xie, 2021, 44 China 90 74 (26–97) b 64.4 NR NR Hemiplegic stroke 28.9 DM 34.4 HT 53.3 CVD 26.7 Chronic renal failure 37.8 Chronic lung disease 17.8 NR Patients with DNACPR status were excluded NR NWR IL-6 PCT Lactate Zhang, 2019, 45 China 301 total group NR NR NR NR NR NR NR 25.6 sepsis 43.9 severe sepsis 17.2 septic shock vWF ADAMTS‑13 40 SIRS 65 (14–85) b 63.6 NR NR COPD 30.0 CVD 32.5 Cerebrovascular diseases 22.5 DM 27.5 Other 10.0 NR NR 0 77 sepsis 65 (20–93) b 55.8 NR NR COPD 12 35.1 CVD 37.7 Cerebrovascular diseases 24.7 DM 35.1 Other 15.6 MEDS 8 (5.5–10.5) APACHE II Mean (SD) 9.1 (3.6) NR 100 132 severe sepsis 69 (26–90) 61.4 NR NR COPD 38.6 CVD 40.2 Cerebrovascular diseases 29.6 DM 37.1 Other 18.9 MEDS 9 (7–11) APACHE II Mean (SD) 12 (5.2) NR 100 52 septic shock 67 (67–94) b 65.4 NR NR COPD 30.8 CVD 32.7 Cerebrovascular diseases 38.4 DM 40.4 Others 19.2 MEDS 11 (8–14) APACHE II Mean (SD) 16.2 (6.6) NR 100 AIDS, acquired immunodeficiency syndrome; AKI, acute kidney injury; APACHE II, Acute Physiology And Chronic Health Evaluation II; CAD, coronary artery disease; CCI, Charlson comorbidity index; CHF, congestive heart failure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CVD, cardiovascular disease; DM, diabetes mellitus; DNR, do not resuscitate; ESRD, end-stage renal disease; GCS, Glasgow Coma Scale; HT, hypertension; HIV, human immunodeficiency virus; IQR, interquartile range; MEWS, modified early warning score; MI, myocardial infarction; miR , micro ribonucleic acid; NEWS-L, sum of the NEWS + the serum lactate level; NR, not reported; PAD, peripheral arterial disease; PVD, peripheral vascular disease; REMS , Rapid Emergency Medical Score; SAPS, simplified acute physiology score III; SEP-3, Sepsis-3; WBC, White blood count. a Clinical scores presented as median (IQR) unless otherwise stated. b Median (IQR). c Range. d Age presented as median (IQR) but labelled in paper as mean (standard deviation). e Lactate > 2.0 mmol/l and 4.0 mmol/l added to the SEP-3 sepsis definition. Appendix 6. Studies’ selection criteria and definitions of suspected sepsis TABLE 5 Inclusion and exclusion criteria and definition of suspected sepsis used in the included studies View in own window Author, year, country Recruitment period Study inclusion criteria (additional definition of suspected sepsis, if reported) Study exclusion criteria Athan, 2023, 69 Australia 2021 January–2021 June > 18 years old with suspected sepsis (emergency clinicians’ decision to perform a blood culture) ED re-presentations, terminal illness, and documented limits on goals of care Baumann, 2020, 70 USA 2016 January–2016 April Admitted to an observation unit, inpatient ward, or ICU from the ED with a presumed infectious disease-related illness Transferred from an outside hospital, admitted primarily for reasons other than their infectious disease illness, hospice or DNACPR patients, no lactate measured Bolanaki, 2021, 61 Germany 2017 January–2018 March ≥ 18 years old with a qSOFA score of at least 1 (patients with one qSOFA point to detect patients likely to develop sepsis in the ED) Acute trauma, acute STEMI, suspected stroke, admission for palliative care with a life expectancy of < 1 month, pregnancy, referrals from other hospitals following prior in-hospital treatment Boland, 2016, 72 USA 2011 July–2013 August A history of recent infection or recent suspected infection, ≥ 18 years old If met initial criteria, body temperature, heart rate and respiratory rate measured to further assess eligibility. Patients enrolled if ≥ 2 of the SIRS criteria confirmed (patients meeting conventional definition of sepsis, i.e. ≥ 2 SIRS criteria plus evidence of infection) Pregnancy Caramello, 2020, 62 Italy 2018 May–2019 March ≥18 years old, diffused infection of any severity, defined both by sepsis definition scores and by clinical criteria: Refused to participate, low acuity infection (localised infection without general symptoms and normal vital parameters), those in whom the positivity of the SIRS or qSOFA criteria attributable in first hypothesis to a non-infectious event (i.e. trauma, CAD, stroke, acute pancreatitis) 1. ED arrival with suspicion of infection on a clinical or instrumental basis, associated with signs of SIRS 2. Identified as septic according to clinical judgement (suspicion of infection on a clinical or instrumental basis, associated with signs of SIRS) Castello, 2019, 59 Italy 2016 October–2018 March Presenting to ED with suspected sepsis according to the Sepsis-3 criteria (suspected infection and qSOFA ≥ 2) consecutively enrolled (suspected infection and qSOFA ≥ 2; Sepsis-3 criteria) Pregnancy, < 18 years old, lack of signed informed consent Cheng, 2018, 64 Taiwan (Province of China) 2007 January–2013 December Patients ≥ 18 years old who visited ED with SIRS , received parenteral antibiotics, and had blood culture collected Only patients with available serum lactate levels checked at the ED were included in analysis (ACCP/SCCM definitions used, and sepsis was defined as infections consisting of ≥ 2 SIRS criteria) NR Choo, 2020, 57 Republic of Korea 2018 July–2018 December Adult aged ≥ 18 years old with potential infection, patients who had tests within 1 hour of admission to ED (≥ 2 signs of SIRS considered to have a potential infection) Non-infectious conditions (such as trauma, thermal injury, sterile inflammation, pancreatitis, heart failure, drug and gas intoxication, gastrointestinal bleeding, seizure, cardiac arrhythmia and shock due to non-infectious causes) based on clinical and laboratory information, patients diagnosed with a disease known to elevate IMA levels, (such as acute coronary syndrome, stroke, PE, PVD and cardiac arrest) patients who had undergone chemotherapy, patients who required surgical intervention Contenti, 2015, 73 France 2013 December–2014 March > 18 years old, ≥ 2 SIRS criteria, suspected infection Patients who did not have arterial blood gas done Covino, 2021, 34 Italy 2014 January–2019 December Admitted to the ED with fever and then hospitalised, > 18 years, fever (temperature ≥ 38 °C) or chills within 24 hours from presentation to the ED as the main symptom, availability of a PCT determination obtained < 24 hours since ED access ( qSOFA score and clinical presentation) < 18 years old, pregnancy Dadeh, 2022, 74 Thailand 2021 March–2021 November ≥ 18 years old who visited ED with an initial NEWS ≥ 5 and diagnosed with sepsis (patients clinically suspected of infection with NEWS ≥ 5 and serum lactate > 2 mmol/l) Pregnancy, lactation, referred patients, traumatic patients, incomplete information, palliative patients, patients who refused resuscitation Devia Jarmillo, 2022, 31 Colombia 2016 January–2017 January > 65 years old, assisted in the hospital resuscitation service with suspected sepsis, underwent a PCT test, immunoassay for the in vitro quantitative determination of PCT during first 12 hours of admission to the ED , arterial lactate measurement was performed upon admission to the resuscitation unit. Chronic liver disease, CKD were included Referred to other institutions, incomplete data in the medical records regarding diagnostic tests evaluated D’Onofrio, 2021, 56 Belgium 2019 February–2020 March Adult patients, presenting with suspected sepsis (patients for whom blood cultures were drawn) NR Dudaryk, 2021, 55 USA 2017 October–2019 November Non-hospice patients, ≥ 18 years old, admitted through ED in whom blood cultures were ordered in the ED (practitioner entertained a diagnosis of possible sepsis from a blood culture having been obtained in the ED because the entered test order indication was ‘the patient has a known or suspected infection’ ) Admitted to the hospice service Fang, 2015, 54 China 2014 August–2015 January Suspected infection with ≥ 2 SIRS criteria < 18 years old, > 80 years old, pregnancy, breastfeeding, neutropenia (defined as < 1000 neutrophils/mm 3 ), HIV infection, HT or constantly taking ACE inhibitors, angiotensin II receptor antagonists, or renin inhibitors, chronic intake of corticosteroids (any daily oral intake of 1 mg/kg or more of equivalent prednisone for more than 1 month), recent use of doses of unfractionated or low-molecular-weight heparin, lack of informed consent by the patients or their relatives Gonzalez del Castillo, 53 2019, Spain 2018 May–2018 July ≥ 18 years old, presenting to ED with a clinical suspicion of infection (clinical suspicion of infection, which could be made based on vital signs, main presenting symptoms, the request for a blood culture, or overall laboratory findings during standard ED assessment) < 18 years old, pregnancy, refusal to participate, patients without obvious clinical signs or symptoms of infection Guarino, 2023, 25 Italy 2021 October–2022 May Clinical suspicion of infectious disease, qSOFA ≥ 2, ≥ 18 years old, a signed informed consent was obtained from each involved patient (or their relatives in case of overall severe clinical conditions) NR GunesOzaydin, 2017, 63 Turkey 2014 March–2014 August ≥18 years old, ≥ 2 SIRS features, clinical infection Non-sepsis diagnosis (e.g. pulmonary embolism, trauma), those who did not have at least two SIRS features, pregnancy Hargreaves, 2023, 60 UK 2015–2017 ≥ 18 years old, only if brought in by ambulance and identified as ‘suspicion of sepsis’ at ED triage NR Henning, 2017, 24 USA Cohort 1 2003 December–2004 September Cohort 2 2005 September–2006 September Cohort 3 2004 July–2005 June Cohorts 1 and 2: Patients with an ED admission diagnosis consistent with infection (i.e. pneumonia) or possibly infection-related (i.e. shortness of breath) Cohort 3: ED patients ≥18 years, hospital admission and suspected infection defined by antibiotic administration in the ED [Meeting any of the following criteria: (1) a documented source of infection; (2) documentation of a clinical diagnosis of infection by the ED clinician; or (3) administration of antibiotics in the ED] NR Hong, 2016, 43 South Korea 2012 November–2014 September > 18 years old, fulfilled diagnostic criteria of the 2001 International Sepsis Definitions Conference < 18 years old, admitted for haemodialysis or peritoneal dialysis, transferred from other hospitals, admission for palliative care Hou, 2021, 23 Taiwan (Province of China) 2020 January–2020 November Patients presenting to ED with infectious diseases (medical record confirmed by study authors), examination of patients by ED physicians, and completion of laboratory tests within 2 hours after arrival to the ED < 20 years old, no definite diagnosis of infectious disease after study author review, and no laboratory tests Hunter, 2013, 50 USA 2009 January–2010 March ≥ 18 years old, presented to ED with suspected infection and ≥ 2 of the SIRS criteria Refused standard therapy, cranial facial abnormalities that would prevent measurement of ETCO 2 , known history of acute asthma exacerbation or COPD, were hyperthermic from environmental causes, were intubated before ED arrival, had been intubated and ventilated in the ED before enrolment Jekarl, 2019, 46 Republic of Korea 2016 June–2017 February ≥ 18 years old, admitted to ED and diagnosed with suspected bacterial infection by an ED physician. (Clinical assessment of signs and symptoms, laboratory and radiological results, concomitant administration of oral or parenteral antibiotics, and sampling of body fluid cultures including blood, urine, cerebrospinal fluid and peritoneal fluid.) Evidence of an immunocompromised state (e.g. malignancy), evidence of a viral infection including respiratory virus or hepatitis virus Jiang, 2019, 36 USA 2017 July–2018 January Deemed to have a suspicion for sepsis by the treating provider, had an initial lactate drawn within the first hour of care Elevated lactate for reasons other than sepsis (e.g. bleeding, cardiogenic shock) Katsaros, 2022, 35 Greece 2017 September–2018 September Adults of either gender with suspected infection and presence of at least one of the following: None 1. core temperature > 38 °C or < 36 °C 2. heart rate > 90 beats per minute 3. respiratory rate > 20/minute 4. self- reported fever/chills Kece, 2016, 42 Turkey 2014 February–2014 August Adult cancer patients who met ≥ 2 of the SIRS criteria. Patients whose probable cause of SIRS were assumed to be infection and had been verified clinically infected were accepted as sepsis Haematological and thyroid malignancies, liver dysfunction, trauma patients, presenting to ED because of seizure, missing data for follow-up, treated with intravenous fluid therapy of more than 500cc when determined as candidates for study, blood was obtained for lactate and procalcitonin levels, using antibiotics at presentation, receiving parenteral fluid therapy at home Kim, 2019, 41 Republic of Korea 2011 March–2016 June > 18 years old, directly visited ED with infection-related disease, underwent laboratory tests including serum PCT and lactate within one hour of ED arrival. [Sepsis defined as SOFA score of 2 or more consequent to the infection (Sepsis-3 definition). Patients with SOFA scores of 0 or 1 were categorised as having an infection] Transferred to other hospitals during their hospital stay, incomplete data, haematological malignancies, neutropenic fever, undergoing chemotherapy, received drugs inducing the suppression of bone marrow (such as steroids), no infection-related diagnosis at discharge summary Klimpel, 2019, 49 Germany 2017 May–2018 April Diagnosis of infection based on clinical, laboratory, radiological and/or microbiological evidence, non-surgical infections Surgical procedure considered a possibility during ward admission, pregnancy, < 18 years old, refusal to participate in the study, readmission during the study period, and decision for end-of-life care on admission to the ED Lee, 2016, 22 Republic of Korea 2013 January–2013 December ≥ 65 years old, hospitalised with sepsis at the ED [defined by microbiological tests (including the culture of body fluids) radiological analyses (including X-rays, ultrasonography, and computed tomography), and serology] (Clinically suspected sepsis was also diagnosed as sepsis) Trauma, myocardial infarction, cerebral infarction, SIRS due to nonbacterial origin, vague diagnosis, evidence of an immunocompromised state (e.g. malignancy), history of administration of antibiotics before visiting the ED within the previous 14 days Lee, 2022, 30 Republic of Korea 2019 May–2020 May Suspected sepsis, had all biomarkers including PCT , PSEP , CRP , lactate at ED NR Lucas, 2021, 76 UK 2019 August–2020 February Presenting to the ED fulfilling the NICE 2016 criteria for moderate to severe sepsis, ≥18 years old with at least one PCT test in ED NR MacDonald, 2017, 37 Australia 2010 March–2013 July Met criteria for sepsis according to the SSC definitions when in the ED . [Uncomplicated sepsis ( SOFA score < 2 and no requirement for organ support)] Transferred from other hospitals, inadequate research blood samples for analysis Magrini, 2013, 77 Italy 2008 October–2009 September Arrived in ED with signs of infections on the basis of anamnestic data, physical examination, vital parameters, instrumental and laboratory tests NR Magrini, 2014, 28 Italy NR Referred to ED with symptoms of infection, and in which a diagnosis of infection or sepsis was formulated as per International Guidelines for Management of Severe Sepsis and Septic Shock 2012 criteria [Signs/symptoms of local infections or sepsis (stated in abstract)] NR Malinovska, 2022, 38 USA 2020 January–2021 July Missing valid MDW (e.g. low sample volume or poor sample quality), MDW sample analyses performed more than 2 hours after blood collection, missing other CBC parameters (WBC count, neutrophils, lymphocytes) within 6 hours of arrival, repeat ED visits by the same patient during the study period Musikatavorn, 2015, 47 Thailand 2013 April–2014 October 18–65 years old, presented to the ED , received a confirmed or suspected diagnosis of acute (onset within the past 7 days) major infection during the study period. Major infection included: acute pyelonephritis, acute bronchitis/pneumonia, acute hepatobiliary tract infections, intra-abdominal abscesses, meningitis and other central nervous system infections, soft-tissue infections involving tissue below the dermis or covering more than 10 cm 2 of the skin surface, significant tropical infection, dengue fever, leptospirosis, typhus fever, or high fever (> 38.5 °C) from any infectious source. Eligible patients must have had stable haemodynamic (≥ 2 repeated measurements of SBP higher than 100 mmHg and MAP higher than 70 mmHg at initial presentation) Overt organ hypoperfusion (e.g. cold, clammy or mottling skin, altered mental status), GCS ≤ 12 or a decrease of larger than 1 compared with baseline, peripheral pulse oximetry ≤ 92% in ambient air, received more than 10 ml/kg intravenous fluid before lactate blood sampling, received an intravenous antibiotic for more than 1 hour before blood lactate sampling, SBP < 100 mmHg or required mechanical ventilator support within 1 hour after ED presentation, minor infection (e.g. uncomplicated upper respiratory tract infection) Noparatkailas, 2023, 33 Thailand 2018 January–2018 December ≥ 18 years old, admitted to a non-critical medical ward with a diagnosis of sepsis at ED were enrolled in the analysis. (Sepsis defined as the patient with suspected or confirmed infection who either met the SIRS criteria or qSOFA score of two or greater.) Initial blood lactate was not obtained at ED , any type of shock or cardiac arrest, seizure at presentation, concurrent use of metformin. Parke, 2023, 67 Sweden 2017 September–2018 December Triggered sepsis alert (patients show signs of organ dysfunction combined with symptoms of infection, namely fever, history of fever, or clinical suspicion of infection) Signs of organ dysfunction either one of A or B: (A) At least one of the following: oxygen saturation below 90% despite supplemental oxygen administration, respiratory rate > 30 per minute, heart rate > 130 beats per minute, SBP < 90 mmHg, or GCS < 8. (B) Blood lactate > 3.2 mmol/l combined with at least one of the following: oxygen saturation below 95% on room air, respiratory rate > 25 per minute, heart rate > 110 beats per minute, altered mental status, and temperature above 38.5 °C or below 35 °C. NR Puskarich, 2015, 29 USA NR Suspected or confirmed infection, ≥2 SIRS criteria, no hypotension after adequate fluid resuscitation. < 18 years old, pregnancy, had an established DNACPR order prior to enrolment, requirement for immediate surgery, unable to obtain written informed consent, with the following primary diagnoses: acute traumatic or burn injury, acute cerebrovascular event, acute coronary syndrome, acute pulmonary oedema, cardiac dysrhythmia, acute and active gastrointestinal bleeding, acute drug overdose Reshmi, 2021, 26 India NR Presented to ED with suspected sepsis (defined in accordance with the criteria of the consensus conference of the ACCP/SCCM) NR Ruangsomboon, 2020, 58 Thailand 2019 September–2020 January ≥ 75 years old, presenting to the ED with suspected sepsis. (Haemoculture taken and intravenous antibiotics prescribed. The physicians suspected sepsis based on SIRS or qSOFA , together with clinical judgement.) Transferred to the ED , had haemoculture taken before ED arrival Shetty, 2018, 51 Australia 2010 June–2013 December ≥ 18 years old, presenting to ED and entered into the sepsis register as part of the SEPSIS KILLS programme None Shim, 2019, 39 Republic of Korea 2018 June–2018 December Febrile sense, suspected infection, underwent serum PCT testing based on the ED physician’s decision < 18 years old, pregnancy, limited life expectancy due to chronic diseases Singer, 2014, 27 USA NR Adults presenting to ED with suspected infection based on presence of ≥ 1 of the following: temperature > 38 °C or < 36 °C, heart rate > 90 beats per minute, respiratory rate > 20 breaths per minute, or altered mental status. NR Sohn, 2019, 40 Republic of Korea 2016 January–2016 December > 18 years old, presenting to the ED with suspicion of infection, and for whom qSOFA score could be calculated, patients who were administered intravenous or oral antibiotics during the ED visit (including those who were prescribed antibiotics as discharge medication) were considered to have infection and were included in the study. ≤ 18 years old, presenting to ED with injury, presenting to ED for medical record copies or medical certification, qSOFA score could not be calculated Suttapanit, 2021, 32 Thailand 2019 February–2020 February ≥18 years old, visited the ED with suspected sepsis and underwent the sepsis protocol. [Defined by physicians using the (local) sepsis protocol ( qSOFA screening in Sepsis-3 criteria or physician judgement), measurement venous blood lactate and was indicated by the receipt of antibiotics and the absence of viral infection as identified by the laboratory.] Infection by microbes other than bacteria such as viruses, submission of a do-not-attempt resuscitation order, transfer to other hospitals, receipt of medical treatment at another hospital before arrival to our ED , and missing data Ueno, 2021, 48 Japan 2017 December–2018 February ≥ 16 years old, suspected to have an infection during stay in the ED (administration of antibiotics, order of microbiological investigations, or imaging request to identify infection focus) Not hospitalised, transferred to a non-participating hospital Uusitalo-Seppala, 2013, 52 Finland 2004–2005 Adults, admitted to ED with suspected infection, from whom a clinician had decided to take samples for blood cultures. NR Xie, 2021, 44 China 2017 December–2019 April > 18 years old and were examined by a physician for assessment and diagnosis of sepsis based on Sepsis-3.0 criteria when they presented to the ED . Pregnancy, < 18 years old, traumatic injury, cancer, required immediate surgical intervention, had been resuscitated from cardiopulmonary arrest, had DNACPR status, major surgery in the previous 30 days, antibacterial therapy in the 5 days prior to ED arrival Zhang, 2019, 45 China 2014 October–2015 October SIRS or sepsis according to ACCP/SCCM criteria, treated in ED < 18 years old, refusal to participate, psychiatric disorders, tumours, organ transplantation, long-term use of immunosuppressive agents, allergic reactions, dysfunction of ≥ 2 organs before the onset of the disease ACCP, American College of Chest Physicians; ACE, angiotensin-converting enzyme; CAD, coronary artery disease; CBC, complete blood count; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; ETCO 2 , end-tidal carbon dioxide; GCS, Glasgow Coma Scale; HIV, human immunodeficiency virus; HT, hypertension; NICE, National Institute for Health and Care Excellence; NR, not reported; PE, pulmonary embolism; PVD, peripheral vascular disease; SCCM, Society of Critical Care Medicine; SSC, surviving sepsis campaign; STEMI, ST-elevation myocardial infarction; WBC, white blood count. Appendix 7. List of prognostic biomarkers evaluated in the included studies TABLE 6 List of individual prognostic biomarkers evaluated in the included studies View in own window Novel biomarker ADAMTS-13 A disintegrin-like metalloproteinase with thrombospondin motif type 1 member 13 Ang-1 Angiopoietin-1 Ang-2 Angiopoietin-2 Calprotectin Calprotectin DNI Delta neutrophil index Endocan Endothelial cell-specific molecule-1 HBP Heparin-binding protein (aka CAP37) ICAM-1 Inter-cellular adhesion molecule-1 IL-5 Interleukin 5 IL-6 Interleukin 6 IL-10 Interleukin 10 IMA Ischaemia-modified albumin IMX-SEV-2 Inflammatix Severity 2 (29 host mRNAs) MDW Monocyte distribution width miR146a micro-RNA 146a miR150 micro-RNA 150 miR223 micro-RNA 223 MR-proADM Mid-regional proadrenomedullin NGAL Neutrophil gelatinase-associated lipocalin NLR Neutrophil-to-lymphocyte ratio NT-pro BNP N-terminal prohormone of brain natriuretic peptide NWR Neutrophil-to-white blood cell ratio OPN Osteopontin PLR Platelet-to-lymphocyte ratio PSEP Presepsin PTX3 High pentraxin 3 Resistan Resistan Tie-2 Tyrosine kinase with immunoglobulin-like and epidermal growth factor-like domains 2 VCAM-1 Vascular cell adhesion molecule-1 vWF von Willebrand factor vWF/ADAMTS-13 ratio vWF/ADAMTS-13 ratio Common biomarker Arterial pH Arterial pH CRP C-reactive protein ESR Erythrocyte sedimentation rate ETCO 2 End-tidal carbon dioxide Ferritin Ferritin hsCRP High-sensitivity CRP LAC Lactate PCT Procalcitonin WBC White blood cell TABLE 7 List of prognostic biomarker combinations evaluated in the included studies View in own window Combined biomarker with another biomarker Combined biomarker with clinical score Ang-1/ Tie-2 Ang-1/ Tie-2 + MEDS Ang-2/ Ang-1 Ang-1/ Tie-2 + PCT + MEDS HBP + PCT Ang-2/ Ang-1 + Ang-1/TIE-2 + PCT + MEDS IL-6 + LAC Ang-2/ Ang-1 + MEDS IL-6 + NWR Ang-2/ Ang-1 + PCT + MEDS IL-6 + PCT CRP + NEWS IL-6 + PCT + LAC IMA + qSOFA IMA + hsCRP IMX-SEV-2 + qSOFA IMA + LAC LAC + NEWS IMA + PCT LAC + qSOFA MDW + NLR LAC + qSOFA + age + CCI + CFS MDW + WBC LAC or qSOFA MDW + WBC + NLR ‘Model 1’ MDW and NLR MR-proADM + CRB-65 MDW and WBC MR-proADM + NEWS MDW and WBC and NLR MR-proADM + qSOFA MDW or NLR MR-proADM + SIRS MDW or WBC PCT + MEDS MDW or WBC or NLR PCT + qSOFA MR-proADM + CRP PSEP + PCT + REMS MR-proADM + LAC PSEP + REMS MR-proADM + PCT qSOFA score model [ MDW + NLR + PLR + qSOFA] with CRP NWR + IL-6 + LAC qSOFA score model [ MDW + NLR + PLR + qSOFA] without CRP NWR + IL-6 + PCT SIRS Score model [ MDW + NLR + PLR + SIRS] with CRP NWR + IL-6 + PCT + LAC SIRS Score model [ MDW + NLR + PLR + SIRS] without CRP NWR + LAC SOFA score model [ MDW + NLR + PLR + SOFA] with CRP NWR + PCT + LAC SOFA score model [ MDW + NLR + PLR + SOFA] without CRP PCT + CRP vWF/ADAMTS-13 ratio + MEDS PCT + LAC PCT + NWR PSEP + CRP PSEP + PCT WBC + CRP WBC + NLR WBC + PCT WBC + PCT + CRP WBC and NLR WBC or NLR CCI, Charlson comorbidity index; CFS, Clinical Frailty Score; LAC, lactate. Appendix 8. Overview of studies investigating novel biomarkers presented in the forest plots TABLE 8 Summary, objectives and results for studies investigating novel biomarkers presented in forest plots (excluding lactate, CRP and PCT) View in own window Author, year, country Number of participants, n (recruitment period) Biomarker Primary/secondary objective Biomarker Outcome Unit Estimate p -value Choo, 2020, 57 Republic of Korea 300 (2018 July–2018 December) IMA Primary IMA Septic shock OR 1.07 (1.02–1.11) 0.002 IMA Septic shock AUROC 0.681 (0.613–0.748) IMA + lactate Septic shock AUROC 0.806 (0.754–0.858) IMA + hsCRP Septic shock AUROC 0.667 (0.602–0.732) IMA + PCT Septic shock AUROC 0.771 (0.712–0.830) IMA + qSOFA Septic shock AUROC 0.882 (0.842–0.923) Fang, 2015, 54 China 440 + 55 control subjects (2014 August–2015 January) Ang-1 Ang-2 Plasma Tie-2 Primary Ang-1 Mortality OR 0.939 (0.802–1.1) 0.157 Ang-1 Mortality AUROC 0.778 (0.732–0.824) 0.024 Ang-2 Mortality OR 1.069 (1.023–0.117) 0.003 Ang-2 Mortality AUROC 0.794 (0.75–0.837) 0.018 Tie-2 Mortality OR 1.002 (0.998–1.072) 0.269 Tie-2 Mortality AUROC 0.803 (0.756–0.849), 0.024 Ang-2/ Ang-1 Mortality OR 0.987 (0.909–1.072) 0.026 Ang-2/ Ang-1 Mortality AUROC 0.845 (0.810–0.880) 0.018 Ang-1/ Tie-2 Mortality OR 0.011 (0.009–0.013) 0.024 Ang-1/ Tie-2 Mortality AUROC 0.808 (0.763–0.853) 0.023 Ang-2/ Ang-1 + MEDS score Mortality AUROC 0.857 (0.821–0.893) 0.018 Ang-1/ Tie-2 + MEDS score Mortality AUROC 0.844 (0.805–0.883) 0.02 Ang-2/ Ang-1 + PCT + MEDS Mortality AUROC 0.9 (0.866–0.934) 0.017 Ang-1/ Tie-2 + PCT + MEDS Mortality AUROC 0.894 (0.862–0.926) 0.016 Ang-2/ Ang-1 + Ang-1/TIE-2 + PCT + MEDS Mortality AUROC 0.925 (0.898–0.951) 0.013 Gonzalez del Castillo, 2019, 53 Spain 684 (2018 May–2018 July) MR-proADM Primary MR-proADM Mortality Univariate HR 4.1 (2.6–6.5) < 0.0001 MR-proADM Mortality Unadjusted OR 15.4 (6.2–38.4) MR-proADM Mortality Multivariate HR 3.8 (2.2–6.5) < 0.0001 MR-proADM Mortality AUROC 0.84 (0.79–0.89) MR-proADM + PCT Mortality Bivariate HR 5.7 (2.8–11.6) MR-proADM + lactate Mortality Bivariate HR 3.3 (2.0–5.5) MR-proADM + CRP Mortality Bivariate HR 4.1 (2.5–6.6) MR-proADM + SOFA Mortality Bivariate HR 2.4 (1.4–4.2) MR-proADM + qSOFA Mortality Bivariate HR 3 (1.8–5.1) MR-proADM + NEWS Mortality Bivariate HR 3.4 (2.0–5.6) MR-proADM + CRB-65 Mortality Bivariate HR 3 (1.8–5.2) MR-proADM + SIRS Mortality Bivariate HR 3.9 (2.4–6.3) MR-proADM Critical care Univariate OR 4.1 (2.3–7.1) < 0.0001 MR-proADM Critical care Multivariate OR 5.8 (3.1–10.8) < 0.0001 MR-proADM Critical care AUROC 0.79 (0.7–0.88) MR-proADM + PCT Critical care Bivariate OR 3.8 (1.7–8.4) < 0.0001 MR-proADM + lactate Critical care Bivariate OR 3.6 (2.0–6.8) < 0.0001 MR-proADM + CRP Critical care Bivariate OR 3.7 (2.1–6.4) < 0.0001 MR-proADM + SOFA Critical care Bivariate OR 4.5 (2.2–9.3) < 0.0001 MR-proADM + qSOFA Critical care Bivariate OR 3.7 (2.0–6.9) < 0.0001 MR-proADM + NEWS Critical care Bivariate OR 3.7 (2.0–6.7) < 0.0001 MR-proADM + CRB-65 Critical care Bivariate OR 4.1 (2.2–7.7) < 0.0001 MR-proADM + SIRS Critical care Bivariate OR 4.1 (2.4–7.8) < 0.0001 Hong, 2016, 43 Republic of Korea 470 (2012 November–2014 September) NGAL Primary NGAL Mortality Univariate HR 1.18 (1.102–1.228) < 0.0001 NAGL Mortality Multivariate HR 1.271 (1.071–1.561) < 0.0001 NGAL Mortality AUROC 0.797 (0.757–0.832) < 0.0001 Hou, 2021, 23 Taiwan (Province of China) 8698 (2020 January–2020 November) NLR PLR MDW Primary NLR > 9U Mortality Adjusted OR 041 (0.05–2.01) 0.3276 PLR > 210 Mortality Adjusted OR 0.24 (0.02–2.4) 0.0413 MDW > 20U Mortality Adjusted OR 4 (0.83–19.32) < 0.0001 NLR/PLR/MDW/ SOFA (Model) Mortality AUROC (Model excludes CRP) 0.912 (0.865–0.96) Katsaros, 2022, 35 Greece 371 (2017 September–2018 September) HPB Primary HPB > 19.8 mg/ml Mortality (early death < 72 hour) Univariate OR 7.44 (1.01–58.72) 0.048 HPB > 19.8 mg/ml Mortality (early death < 72 hour) Multivariate OR 48.59 (1.41–167.9) 0.032 HPB Mortality Unadjusted HR 2.2 (0.92, 5.3) HPB + PCT Mortality Unadjusted OR 3.53 (1.62, 7.67) Kim, 2019, 41 Republic of Korea 866 (2011 March–2016 June) DNI Secondary DNI Mortality AUROC 0.636 (0.577–0.694) < 0.0001 DNI Septic shock AUROC 0.742 (0.685–0.799) < 0.0001 Lee, 2022, 30 Republic of Korea 249 (2019 May–2020 May) PSEP Primary PSEP Mortality Unadjusted HR 1 (1–1.001) 0.077 PSEP Septic shock Univariate OR 1.000 (1.000–1.001) 0.009 MacDonald, 2017, 37 Australia 186 + 29 control subjects (2010 March–2013 July) Resistin NGAL VCAM-1 ICAM-1 IL-6 IL-10 Primary Resistan Mortality Adjusted OR 1.97 (1.31–2.99) 0.001 Resistan Mortality AUROC 0.69 (0.57–0.72) NGAL Mortality Adjusted OR 1.79 (1.18–2.71) 0.006 NGAL Mortality AUROC 0.65 (0.57–0.72) VCAM-1 Mortality Adjusted OR 1.06 (0.68–1.64) 0.79 VCAM-1 Mortality AUROC 0.49 (0.42–0.56) ICAM-1 Mortality Adjusted OR 0.85 (0.58–1.23) 0.38 ICAM-1 Mortality AUROC 0.45 (0.37–0.52) IL-6 Mortality Adjusted OR 1.12 (0.97–1.31) 0.12 IL-6 Mortality AUROC 0.56 (0.48–0.63) IL-10 Mortality Adjusted OR 1.15 (0.98–1.36) 0.1 IL-10 Mortality AUROC 0.55 (0.48–0.63) Resistan Septic shock Adjusted OR 2.54 (1.76–3.67) < 0.0001 Resistan Septic shock AUROC 0.72 (0.65–0.78) NGAL Septic shock Adjusted OR 2.84 (1.92–4.2) < 0.0001 NGAL Septic shock AUROC 0.74 (0.67–0.80) VCAM-1 Septic shock Adjusted OR 1.87 (1.28–2.73) 0.0001 VCAM-1 Septic shock AUROC 0.63 (0.56–0.7) ICAM-1 Septic shock Adjusted OR 1.41 (0.96–1.77) 0.028 ICAM-1 Septic shock AUROC 0.58 (0.5–0.65) IL-6 Septic shock Adjusted OR 1.28 (1.12–1.47) < 0.0001 IL-6 Septic shock AUROC 0.65 (0.58–0.72) IL-10 Septic shock Adjusted OR 1.25 (1.07–1.47) 0.005 IL-10 Septic shock AUROC 0.61 (0.54–0.68) Malinovska, 2022, 38 USA 7952 (2020 January–2021 July) MDW NLR Primary MDW Septic shock AUROC 0.85 (0.80–0.91) NLR Septic shock AUROC 0.81 (0.73–0.88) Ruangsomboon, 2020, 58 Thailand 250 (2019 September–2020 January) PSEP Primary PSEP Mortality OR 4 (2.2–7.4) PSEP Mortality Univariate HR 2.4 (1.6–3.6) PSEP Mortality Unadjusted multivariate HR 1.5 (0.9–2.4) PSEP Mortality Adjusted multivariate HR 0.5 (0.2–1.2) PSEP Mortality AUROC 0.683 (0.613–0.753) < 0.0001 PSEP + PCT Mortality AUROC 0.702 (0.634–0.77) < 0.0001 PSEP + CRP Mortality AUROC 0.700 (0.633–0.766) < 0.0001 PSEP + REMS Mortality AUROC 0.775 (95% CI, 0.713–0.84) < 0.0001 PSEP + PCT + REMS Mortality AUROC 0.78 (0.719–0.84) < 0.0001 PSEP Septic shock OR 3.4 (1.8–6.5) PSEP Septic shock AUROC 0.699 (0.62–0.778) < 0.0001 PSEP + PCT Septic shock AUROC 0.785 (0.714–0.855) < 0.0001 PSEP + CRP Septic shock AUROC 0.737 (0.657–0.817) < 0.0001 PSEP + REMS Septic shock AUROC 0.748 (0.678–0.819) < 0.0001 PSEP + PCT + REMS Septic shock AUROC 0.819 (0.757–0.881) < 0.0001 Uusitalo-Seppala, 2013, 52 Finalnd 537 (2004–2005) PTX3 Primary PTX3 cut off < 7.7 ng/ml Mortality Adjusted OR 2.37 (1.04–5.38) 0.04 PTX3 Mortality AUROC 0.69 (0.58–0.79) < 0.0001 PTX3 cut off < 14.1 ng/ml Mortality Unadjusted OR 3.68 (1.8–7.51) < 0.0001 PTX3 cut off < 14.1 ng/ml Mortality Unadjusted OR 2.48 (1.41–4.37) 0.002 PTX3 cut off < 14.1 ng/ml Mortality Unadjusted OR 1.76 (1.12–2.78) 0.015 PTX3 cut off < 14.1 ng/ml Septic shock Unadjusted OR 3.69 (1.46–9.28) 0.006 PTX3 cut off < 14.1 ng/ml Critical care Unadjusted OR 3.54 (1.87–6.73) < 0.0001 PTX3 cut off < 14.1 ng/ml Organ failure Unadjusted OR 3.25 (1.12–9.45) 0.03 Xie, 2021, 44 China 90 (2017 December–2019 April) NWR IL-6 Primary IL-6 Mortality Adjusted OR 1 (1–1.001) 0.009 NWR Mortality Adjusted OR 0.009 (0–1.052) 0.052 IL-6 Mortality AUROC 0.675 (0.534–0.816) 0.013 IL-6 + PCT Mortality AUROC 0.753 (0.639–0.868) < 0.0001 IL-6 + LAC Mortality AUROC 0.718 (0.578–0.858) 0.002 IL-6 + NWR Mortality AUROC 0.742 (0.614–0.871) 0.001 IL-6 + PCT + lactate Mortality AUROC 0.78 (0.662–0.898) < 0.0001 NWR Mortality AUROC 0.383 (0.24–0.525) 0.094 NWR + lactate Mortality AUROC 0.710 (0.529–0.870) 0.003 NWR + PCT Mortality AUROC 0.617 (0.475–0.76) 0.094 NWR + PCT + LAC Mortality AUROC 0.807 (0.706–0.908) < 0.0001 NWR + IL-6 + LAC Mortality AUROC 0.718 (0.578–0.858) 0.002 NWR + IL-6 + PCT Mortality AUROC 0.796 (0.69–0.901) < 0.0001 NWR + IL-6 + PCT + LAC Mortality AUROC 0.823 (0.723–0.924) < 0.0001 Zhang, 2019, 45 China 301 + 40 control subjects (2014 October–2015 October) ADAMTS‑13 Endocan vWF Primary ADAMTS-13 Mortality AUROC 0.761 (0.706–0.815) < 0.0001 Endocan Mortality AUROC 0.656 (0.595–0.717) < 0.0001 vWF Mortality AUROC 0.751 (0.693–0.81) < 0.0001 vWF/ADAMTS-13 ratio Mortality AUROC 0.79 (0.737–0.844) < 0.0001 vWF/ADAMTS-13 ratio + MEDS Mortality AUROC 0.856 (0.808–0.903) < 0.0001 LAC, lactate. Appendix 9. Additional outcome data TABLE 9 Additional outcome data for novel biomarkers (not presented in forest plots) View in own window Author Year Prognostic factor Outcome FU time (days) Events n Total N Unit Estimate Lower 95% CI Upper 95% CI Reported p -Value Adjustment Zhang 45 2019 ADAMTS-13 (a disintegrin-like and metalloprotease with thrombospondin type 1 member 13) Mortality 28 – – OR – – – 0.000 None Castello 59 2019 Arterial pH Mortality 30 35 101 HR 0.008 0.001 0.541 0.026 Age, sex, respiratory rate, GCS, CRP , arterial pH, plasma lactate, PaO 2 /FiO 2 (ratio between partial pressure of oxygen and fractional inspired oxygen), qSOFA= 3, SOFA Parke 67 2023 Calprotectin Critical care admission (Direct transfer to ICU/HDU) – – 351 AUROC 0.65 – – – None Lee 22 2016 IL-5 Critical care admission – 15 36 AUROC 0.81 – – – None Lee 22 2016 IL-6 Critical care admission – 15 36 AUROC 0.714 – – – None Lee 22 2016 IL-10 Critical care admission – 15 36 AUROC 0.724 – – – None Katsaros (Kastaki) 35 2022 IMX-SEV-2 ( IMX-SEV-2 ; 29 host mRNAs) Critical care admission (Need for ICU care) 7 27 397 AUROC 0.85 0.79 0.92 – None Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 OR 7.16 1.49 34.51 0.0141 Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , PLR)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 OR 4.23 0.75 23.69 0.1011 Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level, and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 OR 2.71 0.58 19.66 0.2450 Model 2 [ SOFA score ≥ 3 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level, and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 OR 5.63 1.16 27.4 0.0325 Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , PLR)] without CR (PS matching by age, sex, GCS score, triage level, and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 OR 3.69 0.65 21.07 0.1417 Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 AUROC 0.678 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , PLR)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 AUROC 0.678 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 AUROC 0.678 – – – Model 2 [ SOFA score ≥ 3 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 AUROC 0.678 – – – Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , PLR)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 MDW > 20U (Ref: ≤ 20U) Mortality 3 – 1480 AUROC 0.678 – – – Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , PLR)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Puskarich 29 2015 miR146a (micro-RNA) Mortality In -hospital 11 69 OR – – – NS None Puskarich 29 2015 miR146a (micro-RNA) Mortality In -hospital 11 69 OR – – – NS Adjusted (no detail) Puskarich 29 2015 miR150 (micro-RNA) Mortality In -hospital 11 69 OR – – – 0.000 None Puskarich 29 2015 miR150 (micro-RNA) Mortality In -hospital 11 69 OR – – – 0.003 Adjusted (including SOFA); no further detail Puskarich 29 2015 miR150 (micro-RNA) Mortality In -hospital 11 69 OR – – – 0.010 Adjusted (including initial LAC); no further detail Puskarich 29 2015 miR223 (micro-RNA) Mortality In -hospital 11 69 OR – – – NS None Puskarich 29 2015 miR223 (micro-RNA) Mortality In -hospital 11 69 OR – – – NS Adjusted (no detail) Gonzalez del Castillo 53 2019 MR-proADM Critical care admission ( ICU admission) 28 – – OR 10.7 3.1 36.3 – None Gonzalez del Castillo 53 2019 MR-proADM + CRB-65 score Critical care admission ( ICU admission) 28 23 684 AUROC 0.79 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + CRP Critical care admission ( ICU admission) 28 23 646 AUROC 0.81 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + LAC Critical care admission ( ICU admission) 28 22 533 AUROC 0.76 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + NEWS Critical care admission ( ICU admission) 28 23 684 AUROC 0.78 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + PCT Critical care admission ( ICU admission) 28 23 684 AUROC 0.79 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + qSOFA Critical care admission ( ICU admission) 28 23 684 AUROC 0.79 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + SIRS Critical care admission ( ICU admission) 28 23 684 AUROC 0.8 – – – Bivariate Gonzalez del Castillo 53 2019 MR-proADM + SOFA Critical care admission ( ICU admission) 28 23 684 AUROC 0.79 – – – Bivariate Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 OR 0.94 0.17 5.23 0.9446 Model 1 [ SIRS score ≥ 3 + biomarkers ( PLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 OR 0.75 0.13 4.17 0.7402 Model 1 [ SIRS score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level, and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 OR 0.41 0.05 2.01 0.3276 Model 2 [ SOFA score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 -- 1480 OR 0.58 0.11 3.19 0.5339 Model 3 [ qSOFA score ≥ 3 + biomarkers ( PLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 OR 0.5 0.09 2.71 0.4192 Model 3 [ qSOFA score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level, and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 AUROC 0.583 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( PLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 AUROC 0.583 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 AUROC 0.583 – – – Model 2 [ SOFA score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 AUROC 0.583 – – – Model 3 [ qSOFA score ≥ 3 + biomarkers ( PLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 NLR > 9U (Ref: ≤ 9U) Mortality 3 – 1480 AUROC 0.583 – – – Model 3 [ qSOFA score ≥ 3 + biomarkers ( PLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Parke 67 2023 NLR Critical care admission (Direct transfer to ICU/HDU) – – 351 AUROC 0.47 – – – None Reshmi 26 2021 NT-pro BNP > 2546 ng/ml (Ref: ≤ 2546 ng/ml) Mortality – 19 215 OR 0.343 – – 0.040 None Reshmi 26 2021 NT-pro BNP > 2546 ng/ml (Ref: ≤ 2546 ng/ml) Mortality – 19 215 OR 0.343 – – 0.031 DM, chronic liver disease (CLD), CKD, NT-pro BNP , blood culture, PCT , mechanical ventilation Castello 59 2019 OPN Mortality 30 34 92 HR – – – 0.482 None Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 OR 0.15 0.02 1.42 0.0975 Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 OR 0.15 0.02 1.37 0.0925 Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , MDW)] with CRP , SIRS , NLR , MDW (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 OR 0.22 0.01 1.62 0.1996 Model 2 [ SOFA score ≥ 3 + biomarkers ( NLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 OR 0.18 0.02 1.64 0.1265 Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 OR 0.17 0.02 1.57 0.1186 Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 AUROC 0.65 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 AUROC 0.65 – – – Model 1 [ SIRS score ≥ 3 + biomarkers ( NLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 AUROC 0.65 – – -- Model 2 [ SOFA score ≥ 3 + biomarkers ( NLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 AUROC 0.65 – – – Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , MDW)] without CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Hou 23 2021 Platelet-to-lymphocyte ratio (PLR) > 210 (Ref: ≤ 210U) Mortality 3 – 1480 AUROC 0.65 – – – Model 3 [ qSOFA score ≥ 2 + biomarkers ( NLR , MDW)] with CRP (PS matching by age, sex, GCS score, triage level and medical comorbidities) Lee 30 2022 PSEP Mortality 30 22 249 HR – – – NS Age, sex, underlying diseases (cardiovascular, cerebrovascular, chronic pulmonary, chronic kidney, chronic liver, malignancy), bacteraemia, AKI, laboratory [WBC, total bilirubin, blood urea nitrogen (BUN), creatinine, lactate dehydrogenase (LDH), LAC, partial pressure of carbon dioxide (PaO 2 ), CRP , PCT], APACHE II, SAPS III Lee 30 2022 PSEP Septic shock 0 35 249 HR – – – NS Age, sex, underlying diseases (cardiovascular, cerebrovascular, chronic pulmonary, chronic kidney, chronic liver, malignancy), bacteraemia, AKI, laboratory [WBC, total bilirubin, blood urea nitrogen (BUN), creatinine, lactate dehydrogenase (LDH), LAC, partial pressure of carbon dioxide (PaO 2 ), CRP , PCT], APACHE II, SAPS III Ruangsomboon 58 2020 Log PSEP Mortality 30 67 250 HR 1.5 0.9 2.4 – Adjusted (unclear detail) Uusitalo-Seppala 52 2013 PTX3 ≥ 7.7 ng/ml (Ref: < 7.7 ng/ml) Mortality 28 33 537 OR 2.55 1.13 5.74 0.024 PCT Zhang 45 2019 vWF/ADAMTS-13 ratio (vWF/a disintegrin-like metalloproteinase with thrombospondin motif type 1 member 13) Mortality 28 – – OR – – – 0.000 Endocan, MEDS AKI, acute kidney injury; APACHE II, Acute Physiology And Chronic Health Evaluation II; CCI, Charlson Comorbidity Index; CKD, chronic kidney disease; DM, diabetes mellitus; ESI, Emergency Severity Index; FiO 2 , fraction of inspired oxygen; GCS, Glasgow Coma Scale; LAC, lactate; NS, not significant; NR, not reported; PS matching, propensity score matching; SAPS III, simplified acute physiology score III; WBC, white blood cell. About the Series Health Technology Assessment ISSN (Electronic): 2046-4924 Article history The contractual start date for this research was in August 2023. This article began editorial review in December 2024 and was accepted for publication in August 2025. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Health Technology Assessment editors and publisher have tried to ensure the accuracy of the authors’ article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article. Last reviewed: December 2024; Accepted: August 2025. Copyright © 2026 Imamura et al . This work was produced by Imamura et al . under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/ . For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited. Bookshelf ID: NBK621512 DOI: 10.3310/GJMB1730 Share Views PubReader Print View Cite this Page Imamura M, Duggan SN, Vadiveloo T, et al. Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department: a synthesis of current evidence [Internet]. Southampton (UK): National Institute for Health and Care Research; 2026 Mar 25. doi: 10.3310/GJMB1730 PDF version of this title (1.7M) In this Page Background Aims and objectives Methods Results Discussion Patient and public involvement Equality, diversity and inclusion Conclusions Additional information List of supplementary material List of abbreviations References Clinical criteria for sepsis Literature search strategies List of included studies Excluded studies and reason for exclusion Included studies Studies’ selection criteria and definitions of suspected sepsis List of prognostic biomarkers evaluated in the included studies Overview of studies investigating novel biomarkers presented in the forest plots Additional outcome data Other titles in this collection Health Technology Assessment Related information NLM Catalog Related NLM Catalog Entries PMC PubMed Central citations PubMed Links to PubMed Similar articles in PubMed Folic acid supplementation and malaria susceptibility and severity among people taking antifolate antimalarial drugs in endemic areas. [Cochrane Database Syst Rev. 2022] Folic acid supplementation and malaria susceptibility and severity among people taking antifolate antimalarial drugs in endemic areas. Crider K, Williams J, Qi YP, Gutman J, Yeung L, Mai C, Finkelstain J, Mehta S, Pons-Duran C, Menéndez C, et al. Cochrane Database Syst Rev. 2022 Feb 1; 2(2022). Epub 2022 Feb 1. The future of Cochrane Neonatal. [Early Hum Dev. 2020] The future of Cochrane Neonatal. Soll RF, Ovelman C, McGuire W. Early Hum Dev. 2020 Nov; 150:105191. Epub 2020 Sep 12. Review Rapid tests to inform triage and antibiotic prescribing decisions for adults presenting with suspected acute respiratory infection: a rapid evidence synthesis of clinical effectiveness and cost-utility studies. [Health Technol Assess. 2025] Review Rapid tests to inform triage and antibiotic prescribing decisions for adults presenting with suspected acute respiratory infection: a rapid evidence synthesis of clinical effectiveness and cost-utility studies. Scandrett K, Colquitt J, Court R, Whiter F, Shinkins B, Takwoingi Y, Loveman E, Todkill D, Gill P, Lasserson D, et al. Health Technol Assess. 2025 May; 29(13):1-114. Biomarkers for assessing acute kidney injury for people who are being considered for admission to critical care: a systematic review and cost-effectiveness analysis. [Health Technol Assess. 2022] Biomarkers for assessing acute kidney injury for people who are being considered for admission to critical care: a systematic review and cost-effectiveness analysis. Brazzelli M, Aucott L, Aceves-Martins M, Robertson C, Jacobsen E, Imamura M, Poobalan A, Manson P, Scotland G, Kaye C, et al. Health Technol Assess. 2022 Jan; 26(7):1-286. Personality Theories. [StatPearls. 2026] Personality Theories. Gallios JM, Iyer V, Kaylor LE. StatPearls. 2026 Jan See reviews... See all... Recent Activity Clear Turn Off Turn On Clinical utility of biomarkers for outcomes prediction in adults with suspected ... Clinical utility of biomarkers for outcomes prediction in adults with suspected sepsis presenting to the emergency department: a synthesis of current evidence Your browsing activity is empty. Activity recording is turned off. Turn recording back on See more... Follow NCBI Twitter Facebook LinkedIn GitHub NCBI Insights Blog Connect with NLM Twitter Facebook Youtube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov