Comparison of qSOFA and Hospital Early Warning Scores for prognosis in suspected sepsis in emergency department patients: a systematic review

Development of an Electronic Clinical Surveillance Measure for Unnecessary Rapid Antibiotic Administration in Suspected Sepsis

BACKGROUND

Substantial efforts focus on monitoring and reducing delays in antibiotic treatment for sepsis, but little has been done to characterize the balancing measure of sepsis overtreatment. We aimed to establish preliminary validity and usefulness of electronic health record (EHR) data-derived criteria for sepsis overtreatment surveillance (SEP-OS).

METHODS

We evaluated adults with potential sepsis (≥2 Systemic Inflammatory Response Syndrome criteria within 6 hours of arrival) presenting to the emergency department of 12 hospitals, excluding patients with shock. We defined SEP-OS as the proportion of patients receiving rapid IV antibiotics (≤3 hours) who did not ultimately meet the Centers for Disease Control Adult Sepsis Event "true sepsis" definition. We evaluated the frequency and characteristics of patients meeting overtreatment criteria and outcomes associated with sepsis overtreatment.

RESULTS

Of 113 764 eligible patients, the prevalence of sepsis overtreatment was 22.5%. The measure met prespecified criteria for reliability, content, construct, and criterion validity. Patients classified by the SEP-OS overtreatment criteria had higher median antibiotic days (4 days [IQR, 2-5] vs 1 day [1-2]; P < .01), longer median length of stay (4 days [2-6] vs 3 days [2-5]; P < .01), higher hospital mortality (2.4% vs 2.1%; P = .01), and higher frequency of Clostridium difficile infection within 6 months of hospital discharge (P < .01) compared with "true negative" cases.

CONCLUSIONS

We developed a novel, valid EHR metric for clinical surveillance of sepsis overtreatment. Applying this metric to a large cohort of potential sepsis patients revealed a high rate of overtreatment and provides a useful tool to inform sepsis quality-improvement targets.

NEWS2, S/F-ratio and ROX-index at emergency department for the prediction of adverse outcomes in COVID-19 patients: An external validation study

BACKGROUND

In the context of the COVID-19 pandemic, the early and accurate identification of patients at risk of deterioration was crucial in overcrowded and resource-limited emergency departments. This study conducts an external validation for the evaluation of the performance of the National Early Warning Score 2 (NEWS2), the S/F ratio, and the ROX index at ED admission in a large cohort of COVID-19 patients from Colombia, South America, assessing the net clinical benefit with decision curve analysis.

METHODS

A prospective cohort study was conducted on 6907 adult patients with confirmed COVID-19 admitted to a tertiary care ED in Colombia. The study evaluated the diagnostic performance of NEWS2, S/F ratio, and ROX index scores at ED admission using the area under the receiver operating characteristic curve (AUROC) for discrimination, calibration, and decision curve analysis for the prediction of intensive care unit admission, invasive mechanical ventilation, and in-hospital mortality.

RESULTS

We included 6907 patients who presented to the ED with confirmed SARS-CoV-2 infection from March 2020 to November 2021. Mean age was 51 (35-65) years and 50.4% of patients were males. The rate of intensive care unit admission was 28%, and in-hospital death was 9.8%. All three scores have good discriminatory performance for the three outcomes based on the AUROC. S/F ratio showed miscalibration at low predicted probabilities and decision curve analysis indicated that the NEWS2 score provided a greater net benefit compared to other scores across at a 10% threshold to decide ED admission at a high-level of care facility.

CONCLUSIONS

The NEWS2, S/F ratio, and ROX index at ED admission have good discriminatory performances in COVID-19 patients for the prediction of adverse outcomes, but the NEWS2 score has a higher net benefit underscoring its clinical utility in optimizing patient management and resource allocation in emergency settings.

Comparison of modified quick Sequential Organ Failure Assessment models as triage tools for febrile patients

OBJECTIVE

Effective triage of febrile patients in the emergency department is crucial during times of overcrowding to prioritize care and allocate resources, especially during pandemics. However, available triage tools often require laboratory data and lack accuracy. We aimed to develop a simple and accurate triage tool for febrile patients by modifying the quick Sequential Organ Failure Assessment (qSOFA) score.

METHODS

We retrospectively analyzed data from 7,303 febrile patients and created modified versions of qSOFA using factors identified through multivariable analysis. The performance of these modified qSOFAs in predicting in-hospital mortality and intensive care unit (ICU) admission was compared using the area under the receiver operating characteristic curve (AUROC).

RESULTS

Through multivariable analysis, the identified factors were age ("A" factor), male sex ("M" factor), oxygen saturation measured by pulse oximetry (SpO2; "S" factor), and lactate level ("L" factor). The AUROCs of ASqSOFA (in-hospital mortality: 0.812 [95% confidence interval, 0.789-0.835]; ICU admission: 0.794 [95% confidence interval, 0.771-0.817]) were simple and not inferior to those of other more complex models (e.g., ASMqSOFA, ASLqSOFA, and ASMLqSOFA). ASqSOFA also displayed significantly higher AUROC than other triage scales, such as the Modified Early Warning Score and Korean Triage and Acuity Scale. The optimal cutoff score of ASqSOFA for the outcome was 2, and the score for redistribution to a lower level emergency department was 0.

CONCLUSION

We demonstrated that ASqSOFA can be employed as a simple and efficient triage tool for emergency febrile patients to aid in resource distribution during overcrowding. It also may be applicable in prehospital settings for febrile patient triage.

Oncology Intensive Care Units: Distinguishing Features and Clinical Considerations

The rapidly advancing field of cancer therapeutics has led to increased longevity among cancer patients as well as increasing complexity of cancer-related illness and associated comorbid conditions. As a result, institutions and organizations that specialize in the in-patient care of cancer patients have similarly evolved to meet the constantly changing needs of this unique patient population. Within these institutions, the intensive care units that specialize in the care of critically ill cancer patients represent an especially unique clinical resource. This article explores some of the defining and distinguishing characteristics associated with oncology ICUs.

Prehospital early warning scores for adults with suspected sepsis: the PHEWS observational cohort and decision-analytic modelling study

Background

Guidelines for sepsis recommend treating those at highest risk within 1 hour. The emergency care system can only achieve this if sepsis is recognised and prioritised. Ambulance services can use prehospital early warning scores alongside paramedic diagnostic impression to prioritise patients for treatment or early assessment in the emergency department.

Objectives

To determine the accuracy, impact and cost-effectiveness of using early warning scores alongside paramedic diagnostic impression to identify sepsis requiring urgent treatment.

Design

Retrospective diagnostic cohort study and decision-analytic modelling of operational consequences and cost-effectiveness.

Setting

Two ambulance services and four acute hospitals in England.

Participants

Adults transported to hospital by emergency ambulance, excluding episodes with injury, mental health problems, cardiac arrest, direct transfer to specialist services, or no vital signs recorded.

Interventions

Twenty-one early warning scores used alongside paramedic diagnostic impression, categorised as sepsis, infection, non-specific presentation, or other specific presentation.

Main outcome measures

Proportion of cases prioritised at the four hospitals; diagnostic accuracy for the sepsis-3 definition of sepsis and receiving urgent treatment (primary reference standard); daily number of cases with and without sepsis prioritised at a large and a small hospital; the minimum treatment effect associated with prioritisation at which each strategy would be cost-effective, compared to no prioritisation, assuming willingness to pay £20,000 per quality-adjusted life-year gained.

Results

Data from 95,022 episodes involving 71,204 patients across four hospitals showed that most early warning scores operating at their pre-specified thresholds would prioritise more than 10% of cases when applied to non-specific attendances or all attendances. Data from 12,870 episodes at one hospital identified 348 (2.7%) with the primary reference standard. The National Early Warning Score, version 2 (NEWS2), had the highest area under the receiver operating characteristic curve when applied only to patients with a paramedic diagnostic impression of sepsis or infection (0.756, 95% confidence interval 0.729 to 0.783) or sepsis alone (0.655, 95% confidence interval 0.63 to 0.68). None of the strategies provided high sensitivity (> 0.8) with acceptable positive predictive value (> 0.15). NEWS2 provided combinations of sensitivity and specificity that were similar or superior to all other early warning scores. Applying NEWS2 to paramedic diagnostic impression of sepsis or infection with thresholds of > 4, > 6 and > 8 respectively provided sensitivities and positive predictive values (95% confidence interval) of 0.522 (0.469 to 0.574) and 0.216 (0.189 to 0.245), 0.447 (0.395 to 0.499) and 0.274 (0.239 to 0.313), and 0.314 (0.268 to 0.365) and 0.333 (confidence interval 0.284 to 0.386). The mortality relative risk reduction from prioritisation at which each strategy would be cost-effective exceeded 0.975 for all strategies analysed.

Limitations

We estimated accuracy using a sample of older patients at one hospital. Reliable evidence was not available to estimate the effectiveness of prioritisation in the decision-analytic modelling.

Conclusions

No strategy is ideal but using NEWS2, in patients with a paramedic diagnostic impression of infection or sepsis could identify one-third to half of sepsis cases without prioritising unmanageable numbers. No other score provided clearly superior accuracy to NEWS2. Research is needed to develop better definition, diagnosis and treatments for sepsis.

Study registration

This study is registered as Research Registry (reference: researchregistry5268).

Funding

This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 17/136/10) and is published in full in Health Technology Assessment; Vol. 28, No. 16. See the NIHR Funding and Awards website for further award information.

Early sepsis care with the National Early Warning Score 2-guided Sepsis Hour-1 Bundle in the emergency department: hybrid type 1 effectiveness-implementation pilot stepped wedge randomised controlled trial (NEWS-1 TRIPS) protocol

INTRODUCTION

Early sepsis treatment in the emergency department (ED) is crucial to improve patient survival. Despite international promulgation, the uptake of the Surviving Sepsis Campaign (SSC) Hour-1 Bundle (lactate measurement, blood culture, broad-spectrum antibiotics, 30 mL/kg crystalloid for hypotension/lactate ≥4 mmol/L and vasopressors for hypotension during/after fluid resuscitation within 1 hour of sepsis recognition) is low across healthcare settings. Delays in sepsis recognition and a lack of high-quality evidence hinder its implementation. We propose a novel sepsis care model (National Early Warning Score, NEWS-1 care), in which the SSC Hour-1 Bundle is triggered objectively by a high NEWS-2 (≥5). This study aims to determine the feasibility of a full-scale type 1 hybrid effectiveness-implementation trial on the NEWS-1 care in multiple EDs.

METHODS AND ANALYSIS

We will conduct a pilot type 1 hybrid trial and prospectively recruit 200 patients from 4 public EDs in Hong Kong cluster randomised in a stepped wedge design over 10 months. All study sites will start with an initial period of standard care and switch in random order at 2-month intervals to the NEWS-1 care unidirectionally. The implementation evaluation will employ mixed methods guided by the Reach, Effectiveness, Adoption, Implementation and Maintenance framework, which includes qualitative and quantitative data from focus group interviews, staff survey and clinical record reviews. We will analyse the 14 feasibility outcomes as progression criteria to a full-scale trial, including trial acceptability to patients and staff, patient and staff recruitment rates, accuracy of sepsis screening, protocol adherence, accessibility to follow-up data, safety and preliminary clinical impacts of the NEWS1 care, using descriptive statistics.

ETHICS AND DISSEMINATION

The institutional review boards of all study sites approved this study. This study will establish the feasibility of a full-scale hybrid trial. We will disseminate the findings through peer-reviewed publications, conference presentations and educational activities.

TRIAL REGISTRATION NUMBER

NCT05731349.

Comparing 11 early warning scores and three shock indices in early sepsis prediction in the emergency department

BACKGROUND

This study aimed to evaluate the discriminatory performance of 11 vital sign-based early warning scores (EWSs) and three shock indices in early sepsis prediction in the emergency department (ED).

METHODS

We performed a retrospective study on consecutive adult patients with an infection over 3 months in a public ED in Hong Kong. The primary outcome was sepsis (Sepsis-3 definition) within 48 h of ED presentation. Using c-statistics and the DeLong test, we compared 11 EWSs, including the National Early Warning Score 2 (NEWS2), Modified Early Warning Score, and Worthing Physiological Scoring System (WPS), etc., and three shock indices (the shock index [SI], modified shock index [MSI], and diastolic shock index [DSI]), with Systemic Inflammatory Response Syndrome (SIRS) and quick Sequential Organ Failure Assessment (qSOFA) in predicting the primary outcome, intensive care unit admission, and mortality at different time points.

RESULTS

We analyzed 601 patients, of whom 166 (27.6%) developed sepsis. NEWS2 had the highest point estimate (area under the receiver operating characteristic curve [AUROC] 0.75, 95%CI 0.70-0.79) and was significantly better than SIRS, qSOFA, other EWSs and shock indices, except WPS, at predicting the primary outcome. However, the pooled sensitivity and specificity of NEWS2 ≥ 5 for the prediction of sepsis were 0.45 (95%CI 0.37-0.52) and 0.88 (95%CI 0.85-0.91), respectively. The discriminatory performance of all EWSs and shock indices declined when used to predict mortality at a more remote time point.

CONCLUSION

NEWS2 compared favorably with other EWSs and shock indices in early sepsis prediction but its low sensitivity at the usual cut-off point requires further modification for sepsis screening.

Prehospital early warning scores for adults with suspected sepsis: retrospective diagnostic cohort study

BACKGROUND

Ambulance services need to identify and prioritise patients with sepsis for early hospital assessment. We aimed to determine the accuracy of early warning scores alongside paramedic diagnostic impression to identify sepsis that required urgent treatment.

METHODS

We undertook a retrospective diagnostic cohort study involving adult emergency medical cases transported to Sheffield Teaching Hospitals ED by Yorkshire Ambulance Service in 2019. We used routine ambulance service data to calculate 21 early warning scores and categorise paramedic diagnostic impressions as sepsis, infection, non-specific presentation or other presentation. We linked cases to hospital records and identified those meeting the sepsis-3 definition who received urgent hospital treatment for sepsis (reference standard). Analysis determined the accuracy of strategies that combined early warning scores at varying thresholds for positivity with paramedic diagnostic impression.

RESULTS

We linked 12 870/24 955 (51.6%) cases and identified 348/12 870 (2.7%) with a positive reference standard. None of the strategies provided sensitivity greater than 0.80 with positive predictive value greater than 0.15. The area under the receiver operating characteristic curve for the National Early Warning Score, version 2 (NEWS2) applied to patients with a diagnostic impression of sepsis or infection was 0.756 (95% CI 0.729, 0.783). No other early warning score provided clearly superior accuracy to NEWS2. Paramedic impression of sepsis or infection had sensitivity of 0.572 (0.519, 0.623) and positive predictive value of 0.156 (0.137, 0.176). NEWS2 thresholds of >4, >6 and >8 applied to patients with a diagnostic impression of sepsis or infection, respectively, provided sensitivities and positive predictive values of 0.522 (0.469, 0.574) and 0.216 (0.189, 0.245), 0.447 (0.395, 0.499) and 0.274 (0.239, 0.313), and 0.314 (0.268, 0.365) and 0.333 (0.284, 0.386).

CONCLUSION

No strategy is ideal but using NEWS2 alongside paramedic diagnostic impression of infection or sepsis could identify one-third to half of sepsis cases without prioritising unmanageable numbers. No other score provided clearly superior accuracy to NEWS2.

TRIAL REGISTRATION NUMBER

researchregistry5268, https://www.researchregistry.com/browse-the-registry%23home/registrationdetails/5de7bbd97ca5b50015041c33/.

The reverse shock index multiplied by the Glasgow Coma Scale score can predict the need for initial resuscitation in patients suspected of sepsis

For patients suspected of sepsis, early recognition of the need for initial resuscitation is key in management. This study evaluated the ability of a modified shock index - the reverse shock index multiplied by the Glasgow Coma Scale score (rSIG) - to predict the need for initial resuscitation in patients with sepsis. This retrospective study involved adults with infection who were admitted to a Japanese tertiary care hospital from an emergency department between January and November 2020. The rSIG, modified Early Warning Score (MEWS), quick Sequential Organ Failure Assessment (qSOFA), and original shock index (SI) values were recorded using initial vital signs. The primary outcome was the area under the receiver-operating characteristic curve (AUROC) for the composite outcome consisting of vasopressor use, mechanical ventilation, and 72-h mortality. Secondary outcomes were the AUROCs for each component of the primary outcome and 28-day mortality. As a result, the primary outcome was met by 67 of the 724 patients (9%). The AUROC was significantly higher for the rSIG than for the other tools (rSIG 0.84 [0.78 - 0.88]; MEWS 0.78 [0.71 - 0.84]; qSOFA 0.72 [0.65 - 0.79]; SI 0.80 [0.74 - 0.85]). Compared with MEWS and qSOFA, the rSIG also had a higher AUROC for vasopressor use and mechanical ventilation, but not for 72-h mortality or in-hospital mortality. The rSIG could be a simple and reliable predictor of the need for initial resuscitation in patients suspected of sepsis.

Combining C-reactive protein and quick sequential organ failure assessment (qSOFA) to improve prognostic accuracy for sepsis and mortality in adult inpatients: A systematic review

Background and Aims

Infections are common in hospitals, and if mismanaged can develop into sepsis, a leading cause of death and disability worldwide. This study aimed to examine whether combining C-reactive protein (CRP) with the quick sequential organ failure assessment (qSOFA) improves its accuracy for predicting mortality and sepsis in adult inpatients.

Methods

PubMed, MEDLINE, EMBASE, Scopus, Web of Science, Science Direct, CINAHL, Open Grey, Grey Literature Report, and the Clinical Trials registry were searched using CRP and qSOFA search terms. Title, abstract, and full-text screening were performed by two independent reviewers using pre-determined eligibility criteria, followed by data extraction and a risk of bias assessment using the Quality Assessment tool for Diagnostic Accuracy Studies 2 (QUADAS-2). Disagreements were settled through discussion and consultation with a third reviewer.

Results

Four retrospective studies with a total of 2070 patients were included in this review. Adding CRP to qSOFA improved the Area Under the Receiver Operating Characteristic Curve up to 9.7% for predicting mortality and by 14.9% for identifying sepsis. The sensitivity and specificity of the combined score for mortality prediction were available in two studies. CRP improved the sensitivity of qSOFA by 43% and 71% while only decreasing the specificity by 12% and 7%, respectively. A meta-analysis was not performed due to study heterogeneity.

Conclusion

This comprehensive review provided initial evidence that combining CRP with qSOFA may improve the accuracy of qSOFA alone in identifying sepsis or patients at risk of dying in hospital. The combined tool demonstrated the potential to improve patient outcomes, with implications for low-resource settings given its simplicity and low-cost.

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

BACKGROUND

The aim of this study was to determine whether: (1) the quick Sequential (Sepsis-related) Organ Failure Assessment (qSOFA) and National Early Warning Score (NEWS) clinical prediction tools alone, (2) modified versions of these prediction tools that integrate lactate into their scores, or (3) use of the two tools in tandem with lactate better predicts in-hospital 28-day mortality among adult EDpatients with suspected infection.

METHODS

From 1 January through 31 December 2018, this retrospective cohort study enrolled consecutive adult patients with suspected infection evaluated at two EDs in France. Patients were included if blood cultures were obtained and non-prophylactic antibiotics were administered in the ED. qSOFA, NEWS criteria and lactate measurements were recorded when patients were clinically suspected of having an infection. Two composite scores (lactate qSOFA (LqSOFA) and lactate NEWS (LNEWS)) integrating lactate were created. Diagnostic test performances for predicting in-hospital mortality within 28days were assessed for qSOFA≥2, LqSOFA≥2, qSOFA≥2 or lactate≥2 mmol/L, and for NEWS≥7, LNEWS≥7, and NEWS≥7 or lactate≥2 mmol/L.

RESULTS

1003 patients were included, 130 (13%) of whom had died by day 28. Sensitivities for 28-day mortality were 50% (95%CI41% to 59%) for qSOFA≥2,69% (95% CI60% to 77%) for LqSOFA≥2,77% (95% CI69% to 84%) for qSOFA or lactate≥2 mmol/L; and 69% (95% CI60% to 77%) for NEWS≥7, 80% (95% CI72% to 86%) for LNEWS≥7, 87% (95% CI80% to 92%) for NEWS≥7 or lactate≥2 mmol/L.

CONCLUSION

Lactate used in tandem with qSOFA or NEWS yielded higher sensitivities in predicting in-hospital 28-day mortality, as compared with integration of lactate into these prediction tools or usage of the tools independently.

Aetiology of hospitalized fever and risk of death at Arua and Mubende tertiary care hospitals in Uganda from August 2019 to August 2020

BACKGROUND

Epidemiology of febrile illness in Uganda is shifting due to increased HIV treatment access, emerging viruses, and increased surveillance. We investigated the aetiology and outcomes of acute febrile illness in adults presenting to hospital using a standardized testing algorithm of available assays in at Arua and Mubende tertiary care hospitals in Uganda.

METHODS

We recruited adults with a ≥ 38.0 °C temperature or history of fever within 48 h of presentation from August 2019 to August 2020. Medical history, demographics, and vital signs were recorded. Testing performed included a complete blood count, renal and liver function, malaria smears, blood culture, and human immunodeficiency virus (HIV). When HIV positive, testing included cryptococcal antigen, CD4 count, and urine lateral flow lipoarabinomannan assay for tuberculosis. Participants were followed during hospitalization and at a 1-month visit. A Cox proportional hazard regression was performed to evaluate for baseline clinical features and risk of death.

RESULTS

Of 132 participants, the median age was 33.5 years (IQR 24 to 46) and 58.3% (n = 77) were female. Overall, 73 (55.3%) of 132 had a positive microbiologic result. Among those living with HIV, 31 (68.9%) of 45 had at least one positive assay; 16 (35.6%) had malaria, 14 (31.1%) tuberculosis, and 4 (8.9%) cryptococcal antigenemia. The majority (65.9%) were HIV-negative; 42 (48.3%) of 87 had at least one diagnostic assay positive; 24 (27.6%) had positive malaria smears and 1 was Xpert MTB/RIF Ultra positive. Overall, 16 (12.1%) of 132 died; 9 (56.3%) of 16 were HIV-negative, 6 died after discharge. High respiratory rate (≥ 22 breaths per minute) (hazard ratio [HR] 8.05; 95% CI 1.81 to 35.69) and low (i.e., < 92%) oxygen saturation (HR 4.33; 95% CI 1.38 to 13.61) were identified to be associated with increased risk of death.

CONCLUSION

In those with hospitalized fever, malaria and tuberculosis were common causes of febrile illness, but most deaths were non-malarial, and most HIV-negative participants did not have a positive diagnostic result. Those with respiratory failure had a high risk of death.

Building artificial intelligence and machine learning models : a primer for emergency physicians

There has been a rise in the number of studies relating to the role of artificial intelligence (AI) in healthcare. Its potential in Emergency Medicine (EM) has been explored in recent years with operational, predictive, diagnostic and prognostic emergency department (ED) implementations being developed. For EM researchers building models de novo, collaborative working with data scientists is invaluable throughout the process. Synergism and understanding between domain (EM) and data experts increases the likelihood of realising a successful real-world model. Our linked manuscript provided a conceptual framework (including a glossary of AI terms) to support clinicians in interpreting AI research. The aim of this paper is to supplement that framework by exploring the key issues for clinicians and researchers to consider in the process of developing an AI model.

Comparing complaint-based triage scales and early warning scores for emergency department triage

Background

Emergency triage systems are used globally to prioritise care based on patients’ needs. These systems are commonly based on patient complaints, while the need for timely interventions on regular hospital wards is usually assessed with early warning scores (EWS). We aim to directly compare the ability of currently used triage scales and EWS scores to recognise patients in need of urgent care in the ED.

Methods

We performed a retrospective, single-centre study on all patients who presented to the ED of a Dutch Level 1 trauma centre, between 1 September 2018 and 24 June 2020 and for whom a Netherlands Triage System (NTS) score as well as a Modified Early Warning Score (MEWS) was recorded. The performance of these scores was assessed using surrogate markers for true urgency and presented using bar charts, cross tables and a paired area under the curve (AUC).

Results

We identified 12 317 unique patient visits where NTS and MEWS scores were documented during triage. A paired comparison of the AUC of these scores showed that the MEWS score had a significantly better AUC than the NTS for predicting the need for hospital admission (0.65 vs 0.60; p<0.001) or 30-day all-cause mortality (0.70 vs 0.60; p<0.001). Furthermore, when non-urgent MEWS scores co-occur with urgent NTS scores, the MEWS score seems to more accurately capture the urgency level that is warranted.

Conclusions

The results of this study suggest that EWSs could potentially be used to replace the current emergency triage systems.

Critical Care Database Comprising Patients With Infection

Patients treated in the intensive care unit (ICU) are closely monitored and receive intensive treatment. Such aggressive monitoring and treatment will generate high-granularity data from both electronic healthcare records and nursing charts. These data not only provide infrastructure for daily clinical practice but also can help to inform clinical studies. It is technically challenging to integrate and cleanse medical data from a variety of sources. Although there are several open-access critical care databases from western countries, there is a lack of this kind of database for Chinese adult patients. We established a critical care database involving patients with infection. A large proportion of these patients have sepsis and/or septic shock. High-granularity data comprising laboratory findings, baseline characteristics, medications, international statistical classification of diseases (ICD) code, nursing charts, and follow-up results were integrated to generate a comprehensive database. The database can be utilized for a variety of clinical studies. The dataset is fully accessible at PhysioNet(https://physionet.org/content/icu-infection-zigong-fourth/1.0/).