Comparison of risk prediction scoring systems for ward patients: a retrospective nested case-control study

SOFA score performs worse than age for predicting mortality in patients with COVID-19

The use of the Sequential Organ Failure Assessment (SOFA) score, originally developed to describe disease morbidity, is commonly used to predict in-hospital mortality. During the COVID-19 pandemic, many protocols for crisis standards of care used the SOFA score to select patients to be deprioritized due to a low likelihood of survival. A prior study found that age outperformed the SOFA score for mortality prediction in patients with COVID-19, but was limited to a small cohort of intensive care unit (ICU) patients and did not address whether their findings were unique to patients with COVID-19. Moreover, it is not known how well these measures perform across races. In this retrospective study, we compare the performance of age and SOFA score in predicting in-hospital mortality across two cohorts: a cohort of 2,648 consecutive adult patients diagnosed with COVID-19 who were admitted to a large academic health system in the northeastern United States over a 4-month period in 2020 and a cohort of 75,601 patients admitted to one of 335 ICUs in the eICU database between 2014 and 2015. We used age and the maximum SOFA score as predictor variables in separate univariate logistic regression models for in-hospital mortality and calculated area under the receiver operator characteristic curves (AU-ROCs) and area under precision-recall curves (AU-PRCs) for each predictor in both cohorts. Among the COVID-19 cohort, age (AU-ROC 0.795, 95% CI 0.762, 0.828) had a significantly better discrimination than SOFA score (AU-ROC 0.679, 95% CI 0.638, 0.721) for mortality prediction. Conversely, age (AU-ROC 0.628 95% CI 0.608, 0.628) underperformed compared to SOFA score (AU-ROC 0.735, 95% CI 0.726, 0.745) in non-COVID-19 ICU patients in the eICU database. There was no difference between Black and White COVID-19 patients in performance of either age or SOFA Score. Our findings bring into question the utility of SOFA score-based resource allocation in COVID-19 crisis standards of care.

A nomogram for predicting mortality risk within 30 days in sepsis patients admitted in the emergency department: A retrospective analysis

OBJECTIVE

To establish and validate an individualized nomogram to predict mortality risk within 30 days in patients with sepsis from the emergency department.

METHODS

Data of 1205 sepsis patients who were admitted to the emergency department in a tertiary hospital between Jun 2013 and Sep 2021 were collected and divided into a training group and a validation group at a ratio of 7:3. The independent risk factors related to 30-day mortality were identified by univariate and multivariate analysis in the training group and used to construct the nomogram. The model was evaluated by receiver operating characteristic (ROC) curve, calibration chart and decision curve analysis. The model was validated in patients of the validation group and its performance was confirmed by comparing to other models based on SOFA score and machine learning methods.

RESULTS

The independent risk factors of 30-day mortality of sepsis patients included pro-brain natriuretic peptide, lactic acid, oxygenation index (PaO2/FiO2), mean arterial pressure, and hematocrit. The AUCs of the nomogram in the training and verification groups were 0.820 (95% CI: 0.780-0.860) and 0.849 (95% CI: 0.783-0.915), respectively, and the respective P-values of the calibration chart were 0.996 and 0.955. The DCA curves of both groups were above the two extreme curves, indicating high clinical efficacy. The AUC values were 0.847 for the model established by the random forest method and 0.835 for the model established by the stacking method. The AUCs of SOFA model in the model and validation groups were 0.761 and 0.753, respectively.

CONCLUSION

The sepsis nomogram can predict the risk of death within 30 days in sepsis patients with high accuracy, which will be helpful for clinical decision-making.

Illness severity characteristics and outcomes of patients remaining on an acute ward following medical emergency team review: a latent profile analysis

BACKGROUND

Patients requiring medical emergency team (MET) review have complex clinical needs, and most remain on the ward after review. Current detection instruments cannot identify post-MET patient requirements, meaning patients remain undistinguished, potentially resulting in missed management opportunities. We propose that deteriorating patients will cluster along dimensions of illness severity and that these clusters may be used to strengthen patient risk management practices.

OBJECTIVE

To identify and define the number of illness severity clusters and report outcomes among ward patients following MET review.

STUDY DESIGN AND SETTING

This retrospective cohort study examined the clinical records of 1500 adult ward patients following MET review at an Australian quaternary hospital. Three-step latent profile analysis methods were used to determine clusters using Sequential Organ Failure Assessment (SOFA) and Nursing Activities Score (NAS) as illness severity indicators. Study outcomes were (1) hospital mortality, (2) unplanned intensive care unit (ICU) admission and (3) subsequent MET review.

RESULTS

Patients were unplanned (73.9%) and medical (57.5%) admissions with at least one comorbidity (51.4%), and complex combinations of acuity (SOFA range 1-17) and dependency (NAS range 22.4%-148.5%). Five clusters are reported. Patients in cluster 1 were equivalent to clinically stable general ward patients. Organ failure and complexity increased with cluster progression-clusters 2 and 3 were equivalent to subspecialty/higher-dependency wards, and clusters 4 and 5 were equivalent to ICUs. Patients in cluster 5 had the greatest odds for death (OR 26.2, 95% CI 23.3 to 31.3), unplanned ICU admission (OR 3.1, 95% CI 3.0 to 3.1) and subsequent MET review (OR 2.4, 95% CI 2.4 to 2.6).

CONCLUSION

The five illness severity clusters may be used to define patients at risk of poorer outcomes who may benefit from enhanced levels of monitoring and targeted care.

The Accuracy of Rapid Emergency Medicine Score in Predicting Mortality in Non-Surgical Patients: A Systematic Review and Meta-Analysis

BACKGROUND

Emergency department (ED) physicians often need to quickly assess patients and determine vital signs to prioritize them by the severity of their condition and make optimal treatment decisions. Effective triage requires optimal scoring systems to accelerate and positively influence the treatment of trauma cases. To this end, a variety of scoring systems have been developed to enable rapid assessment of ED patients. The present systematic review and meta-analysis aimed to investigate the accuracy of the rapid emergency medicine score (REMS) system in predicting the mortality rate in non-surgical ED patients.

METHODS

A systematic search of articles published between 1990 and 2020 was conducted using various scientific databases (Medline, Embase, Scopus, Web of Science, ProQuest, Cochrane Library, IranDOC, Magiran, and Scientific Information Database). Both cross-sectional and cohort studies assessing the REMS system to predict mortality in ED settings were considered. Two reviewers appraised the selected articles independently using the National Institutes of Health (NIH) quality assessment tool. The random-effects model was used for meta-analysis. I² index and Q statistic were used to examine heterogeneity between the articles.

RESULTS

The search resulted in 1,310 hits from which, 29 articles were eventually selected. Out of these, for 25 articles, the area under the curve value of REMS ranged from 0.52 to 0.986. The predictive power of REMS for the in-hospital mortality rate was high in 19 articles (67.85%) and low in nine articles (32.15%).

CONCLUSION

The results showed that the REMS system is an effective tool to predict mortality in non-surgical patients presented to the ED. However, further evidence using high-quality design studies is required to substantiate our findings.

A Review of Mortality Risk Prediction Models in Smartphone Applications

Healthcare professionals in healthcare systems need access to freely available, real-time, evidence-based mortality risk prediction smartphone applications to facilitate resource allocation. The objective of this study is to evaluate the quality of smartphone mobile health applications that include mortality prediction models, and corresponding information quality. We conducted a systematic review of commercially available smartphone applications in Google Play for Android, and iTunes for iOS smartphone applications. We performed initial screening, data extraction, and rated smartphone application quality using the Mobile Application Rating Scale: user version (uMARS). The information quality of smartphone applications was evaluated using two patient vignettes, representing low and high risk of mortality, based on critical care data from the Medical Information Mart for Intensive Care (MIMIC) III database. Out of 3051 evaluated smartphone applications, 33 met our final inclusion criteria. We identified 21 discrete mortality risk prediction models in smartphone applications. The most common mortality predicting models were Sequential Organ Failure Assessment (SOFA) (n = 15) and Acute Physiology and Clinical Health Assessment II (n = 13). The smartphone applications with the highest quality uMARS scores were Observation-NEWS 2 (4.64) for iOS smartphones, and MDCalc Medical Calculator (4.75) for Android smartphones. All SOFA-based smartphone applications provided consistent information quality with the original SOFA model for both the low and high-risk patient vignettes. We identified freely available, high-quality mortality risk prediction smartphone applications that can be used by healthcare professionals to make evidence-based decisions in critical care environments.

Predicting Patient Deterioration: A Review of Tools in the Digital Hospital Setting

BACKGROUND

Early warning tools identify patients at risk of deterioration in hospitals. Electronic medical records in hospitals offer real-time data and the opportunity to automate early warning tools and provide real-time, dynamic risk estimates.

OBJECTIVE

This review describes published studies on the development, validation, and implementation of tools for predicting patient deterioration in general wards in hospitals.

METHODS

An electronic database search of peer reviewed journal papers from 2008-2020 identified studies reporting the use of tools and algorithms for predicting patient deterioration, defined by unplanned transfer to the intensive care unit, cardiac arrest, or death. Studies conducted solely in intensive care units, emergency departments, or single diagnosis patient groups were excluded.

RESULTS

A total of 46 publications were eligible for inclusion. These publications were heterogeneous in design, setting, and outcome measures. Most studies were retrospective studies using cohort data to develop, validate, or statistically evaluate prediction tools. The tools consisted of early warning, screening, or scoring systems based on physiologic data, as well as more complex algorithms developed to better represent real-time data, deal with complexities of longitudinal data, and warn of deterioration risk earlier. Only a few studies detailed the results of the implementation of deterioration warning tools.

CONCLUSIONS

Despite relative progress in the development of algorithms to predict patient deterioration, the literature has not shown that the deployment or implementation of such algorithms is reproducibly associated with improvements in patient outcomes. Further work is needed to realize the potential of automated predictions and update dynamic risk estimates as part of an operational early warning system for inpatient deterioration.

The Value of Physiological Scoring Criteria in Predicting the In-Hospital Mortality of Acute Patients; a Systematic Review and Meta-Analysis

INTRODUCTION

There is no comprehensive meta-analysis on the value of physiological scoring systems in predicting the mortality of critically ill patients. Therefore, the present study intended to conduct a systematic review and meta-analysis to collect the available clinical evidence on the value of physiological scoring systems in predicting the in-hospital mortality of acute patients.

METHOD

An extensive search was performed on Medline, Embase, Scopus, and Web of Science databases until the end of year 2020. Physiological models included Rapid Acute Physiology Score (RAPS), Rapid Emergency Medicine Score (REMS), modified REMS (mREMS), and Worthing Physiological Score (WPS). Finally, the data were summarized and the findings were presented as summary receiver operating characteristics (SROC), sensitivity, specificity and diagnostic odds ratio (DOR).

RESULTS

Data from 25 articles were included. The overall analysis showed that the area under the SROC curve of REMS, RAPS, mREMS, and WPS criteria were 0.83 (95% CI: 0.79-0.86), 0.89 (95% CI: 0.86-0.92), 0.64 (95% CI: 0.60-0.68) and 0.86 (95% CI: 0.83-0.89), respectively. DOR for REMS, RAPS, mREMS and WPS models were 11 (95% CI: 8-16), 13 (95% CI: 4-41), 2 (95% CI: 2-4) and 17 (95% CI: 5-59) respectively. When analyses were limited to trauma patients, the DOR of the REMS and RAPS models were 112 and 431, respectively. Due to the lack of sufficient studies, it was not possible to limit the analyses for mREMS and WPS.

CONCLUSION

The findings of the present study showed that three models of RAPS, REMS and WPS have a high predictive value for in-hospital mortality. In addition, the value of these models in trauma patients is much higher than other patient settings.

A catalogue of tools and variables from crisis and routine care to support decision-making about allocation of intensive care beds and ventilator treatment during pandemics: Scoping review

PURPOSE

This scoping review sought to identify objective factors to assist clinicians and policy-makers in making consistent, objective and ethically sound decisions about resource allocation when healthcare rationing is inevitable.

MATERIALS AND METHODS

Review of guidelines and tools used in ICUs, hospital wards and emergency departments on how to best allocate intensive care beds and ventilators either during routine care or developed during previous epidemics, and association with patient outcomes during and after hospitalisation.

RESULTS

Eighty publications from 20 countries reporting accuracy or validity of prognostic tools/algorithms, or significant correlation between prognostic variables and clinical outcomes met our eligibility criteria: twelve pandemic guidelines/triage protocols/consensus statements, twenty-two pandemic algorithms, and 46 prognostic tools/variables from non-crisis situations. Prognostic indicators presented here can be combined to create locally-relevant triage algorithms for clinicians and policy makers deciding about allocation of ICU beds and ventilators during a pandemic. No consensus was found on the ethical issues to incorporate in the decision to admit or triage out of intensive care.

CONCLUSIONS

This review provides a unique reference intended as a discussion starter for clinicians and policy makers to consider formalising an objective a locally-relevant triage consensus document that enhances confidence in decision-making during healthcare rationing of critical care and ventilator resources.

A Simulated Prospective Evaluation of a Deep Learning Model for Real-Time Prediction of Clinical Deterioration Among Ward Patients

OBJECTIVES

The National Early Warning Score, Modified Early Warning Score, and quick Sepsis-related Organ Failure Assessment can predict clinical deterioration. These scores exhibit only moderate performance and are often evaluated using aggregated measures over time. A simulated prospective validation strategy that assesses multiple predictions per patient-day would provide the best pragmatic evaluation. We developed a deep recurrent neural network deterioration model and conducted a simulated prospective evaluation.

DESIGN

Retrospective cohort study.

SETTING

Four hospitals in Pennsylvania.

PATIENTS

Inpatient adults discharged between July 1, 2017, and June 30, 2019.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We trained a deep recurrent neural network and logistic regression model using data from electronic health records to predict hourly the 24-hour composite outcome of transfer to ICU or death. We analyzed 146,446 hospitalizations with 16.75 million patient-hours. The hourly event rate was 1.6% (12,842 transfers or deaths, corresponding to 260,295 patient-hours within the predictive horizon). On a hold-out dataset, the deep recurrent neural network achieved an area under the precision-recall curve of 0.042 (95% CI, 0.04-0.043), comparable with logistic regression model (0.043; 95% CI 0.041 to 0.045), and outperformed National Early Warning Score (0.034; 95% CI, 0.032-0.035), Modified Early Warning Score (0.028; 95% CI, 0.027- 0.03), and quick Sepsis-related Organ Failure Assessment (0.021; 95% CI, 0.021-0.022). For a fixed sensitivity of 50%, the deep recurrent neural network achieved a positive predictive value of 3.4% (95% CI, 3.4-3.5) and outperformed logistic regression model (3.1%; 95% CI 3.1-3.2), National Early Warning Score (2.0%; 95% CI, 2.0-2.0), Modified Early Warning Score (1.5%; 95% CI, 1.5-1.5), and quick Sepsis-related Organ Failure Assessment (1.5%; 95% CI, 1.5-1.5).

CONCLUSIONS

Commonly used early warning scores for clinical decompensation, along with a logistic regression model and a deep recurrent neural network model, show very poor performance characteristics when assessed using a simulated prospective validation. None of these models may be suitable for real-time deployment.

Outcomes prediction in longitudinal data: Study designs evaluation, use case in ICU acquired sepsis

Outcomes' prediction in Electronic Health Records (EHR) and specifically in Critical Care is increasingly attracting more exploration and research. In this study, we used clinical data from the Intensive Care Unit (ICU), focusing on ICU acquired sepsis. Looking at the current literature, several evaluation approaches are reported, inspired by epidemiological designs, in which some do not always reflect real-life application's conditions. This problem seems relevant generally to outcomes' prediction in longitudinal EHR data, or generally longitudinal data, while in this study we focused on ICU data. Unlike in most previous studies that investigated all sepsis admissions, we focused specifically on ICU-Acquired Sepsis. Due to the sparse nature of the longitudinal data, we employed the use of Temporal Abstraction and Time Interval-Related Patterns discovery, which are further used as classification features. Two experiments were designed using three different outcomes prediction study designs from the literature, implementing various levels of real-life conditions to evaluate the prediction models. The first experiment focused on predicting whether a patient would suffer from ICU-acquired sepsis and when during her admission, given a sliding observation time window, and the comparison of the three study designs behavior. The second experiment focused only on predicting whether the patient will suffer from ICU-acquired sepsis, based on data taken relatively to his admission start time. Our results show that using Temporal Discretization for Classification (TD4C) led to better performance than using the Equal-Width Discretization, Knowledge-Based, or SAX. Also, using two states abstraction was better than three or four. Using the default Binary TIRP representation method performed better than Mean Duration, Horizontal Support, and horizontally normalized horizontal support. Using XGBoost as a classifier performed better than Logistic Regression, Neural Net, or Random Forest. Additionally, it is demonstrated why the use of case-crossover-control is most appropriate for real life application conditions evaluation, unlike other incomplete designs that may even result in "better performance".

Early Warning Scores to Predict Noncritical Events Overnight in Hospitalized Medical Patients: A Prospective Case Cohort Study

BACKGROUND

Physicians are often called to evaluate patients overnight with varying levels of clinical deterioration. Early warning scores predict critical clinical deterioration in patients; however, it is unknown whether they are able to reliably predict which patients will need to be seen overnight and whether these patients will require further resource use.

METHODS

A prospective case cohort study of 522 patient nights in a single tertiary care hospital in Vancouver, British Columbia, Canada, was conducted to assess the ability of Modified Early Warning Score (MEWS) and National Early Warning Score (NEWS) to predict patients who will need to be seen overnight by physicians and will require other healthcare resources. Prediction ability was assessed using area under the receiver operating characteristic curve and logistic regression models.

RESULTS

The MEWS and NEWS both significantly predicted which patients needed to be seen overnight, and area under the receiver operating characteristic curves (95% confidence interval) for MEWS and NEWS were 0.72 (0.66-0.78) and 0.69 (0.63-0.76), respectively. Odds ratios (95% confidence interval) for MEWS and NEWS predicting need to be seen overnight were 1.52 (1.34-1.73) and 1.22 (1.14-1.31), respectively.

CONCLUSIONS

Both MEWS and NEWS have fair ability to predict patients who will need to be seen overnight. This may be useful for improving handover and resource allocation for overnight care.

A Machine Learning Prediction Model of Respiratory Failure Within 48 Hours of Patient Admission for COVID-19: Model Development and Validation

BACKGROUND

Predicting early respiratory failure due to COVID-19 can help triage patients to higher levels of care, allocate scarce resources, and reduce morbidity and mortality by appropriately monitoring and treating the patients at greatest risk for deterioration. Given the complexity of COVID-19, machine learning approaches may support clinical decision making for patients with this disease.

OBJECTIVE

Our objective is to derive a machine learning model that predicts respiratory failure within 48 hours of admission based on data from the emergency department.

METHODS

Data were collected from patients with COVID-19 who were admitted to Northwell Health acute care hospitals and were discharged, died, or spent a minimum of 48 hours in the hospital between March 1 and May 11, 2020. Of 11,525 patients, 933 (8.1%) were placed on invasive mechanical ventilation within 48 hours of admission. Variables used by the models included clinical and laboratory data commonly collected in the emergency department. We trained and validated three predictive models (two based on XGBoost and one that used logistic regression) using cross-hospital validation. We compared model performance among all three models as well as an established early warning score (Modified Early Warning Score) using receiver operating characteristic curves, precision-recall curves, and other metrics.

RESULTS

The XGBoost model had the highest mean accuracy (0.919; area under the curve=0.77), outperforming the other two models as well as the Modified Early Warning Score. Important predictor variables included the type of oxygen delivery used in the emergency department, patient age, Emergency Severity Index level, respiratory rate, serum lactate, and demographic characteristics.

CONCLUSIONS

The XGBoost model had high predictive accuracy, outperforming other early warning scores. The clinical plausibility and predictive ability of XGBoost suggest that the model could be used to predict 48-hour respiratory failure in admitted patients with COVID-19.

The illness severity of patients reviewed by the medical emergency team: A scoping review

BACKGROUND

Medical emergency teams (METs) are internationally used to manage hospitalised deteriorating patients. Although triggers for MET review and hospital outcomes have previously been widely reported, the illness severity at the point of MET review has not been reported. As such, levels of clinical acuity and patient dependency representing the risk of exposure to short-term adverse clinical outcomes remain largely unknown.

OBJECTIVE

This scoping review sought to understand the illness severity of MET review recipients in terms of acuity and dependency.

METHODS

This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews. The published and grey literature since 2009 was searched to identify relevant articles reporting illness severity scores associated with hospitalised adult inpatients reviewed by a MET. After applying the inclusion and exclusion criteria, 17 articles (16 quantitative studies, one mixed-methods study) were reviewed, summarised, collated, and reported.

RESULTS

A total of 17 studies reported clinical acuity metrics for patients reviewed by a MET. No studies described an integrated risk score encompassing acuity, patient dependency, or wider parameters that might be associated with increased patient risk or the need for intervention. Multi-MET review, the use of specialist interventions, and delayed/transfer to the intensive care unit were associated with a greater risk of clinical deterioration, higher clinical acuity score, and predicted mortality risk. A single dependency metric was not reported although organisational levels of care, the duration of MET review, MET interventions, chronic illness, and frailty were inferred proxy measures.

CONCLUSION

Of the 17 studies reviewed, no single study provided an integrated assessment of illness severity from which to stratify risk or support patient management processes. Patients reviewed by a MET have variable and rapidly changing health needs that make them particularly vulnerable. The lack of high-quality data reporting acuity and dependency limits our understanding of true clinical risk and subsequent opportunities for pathway development.

Incidence and Risk Factors of Delirium in the Intensive Care Unit: A Prospective Cohort

Purpose

The purpose of this study was to determine the incidence, risk factors, and impact of delirium on outcomes in ICU patients. In addition, the scoring systems were measured consecutively to characterize how these scores changed with time in patients with and without delirium. Material and Methods. A prospective cohort study enrolling 400 consecutive patients admitted to the ICU between 2018 and 2019 due to trauma or surgery. Patients were followed up for the development of delirium over ICU days using the Confusion Assessment Method (CAM) for the ICU and Intensive Care Delirium Screening Checklist (ICDSC). Cox model logistic regression analysis was used to explore delirium risk factors.

Results

Delirium occurred in 108 (27%) patients during their ICU stay, and the median onset of delirium was 4 (IQR 3–4) days after admission. According to multivariate cox regression, the expected hazard for delirium was 1.523 times higher in patients who used mechanical ventilator as compared to those who did not (HR: 1.523, 95% CI: 1.197-2.388, P < 0.001).

Conclusion

Our findings suggest that an important opportunity for improving the care of critically ill patients may be the determination of modifiable risk factors for delirium in the ICU. In addition, the scoring systems (APACHE IV, SOFA, and RASS) are useful for the prediction of delirium in critically ill patients.

Urosepsis in Older Adults: Epidemiologic Trends in Florida

The incidence and geographic distribution of urosepsis, a life-threatening condition in older adults, is not well understood. The Florida State Inpatient Databases (2012-2014) showed an increase in the incidence of community-acquired urosepsis (5.37 to 6.16 per 1000), particularly among Hispanic older adults residing in low socioeconomic, urban areas with large numbers of nursing homes. These findings suggest a state policy is needed to address community-based preventative care and education for early detection of urosepsis in low-income urban areas. It is important for local health departments to partner with nursing homes to address disparities in care that disproportionally impact Hispanics.

Demonstrating the consequences of learning missingness patterns in early warning systems for preventative health care: A novel simulation and solution

When using tree-based methods to develop predictive analytics and early warning systems for preventive healthcare, it is important to use an appropriate imputation method to prevent learning the missingness pattern. To demonstrate this, we developed a novel simulation that generated synthetic electronic health record data using a variational autoencoder with a custom loss function, which took into account the high missing rate of electronic health data. We showed that when tree-based methods learn missingness patterns (correlated with adverse events) in electronic health record data, this leads to decreased performance if the system is used in a new setting that has different missingness patterns. Performance is worst in this scenario when the missing rate between those with and without an adverse event is the greatest. We found that randomized and Bayesian regression imputation methods mitigate the issue of learning the missingness pattern for tree-based methods. We used this information to build a novel early warning system for predicting patient deterioration in general wards and telemetry units: PICTURE (Predicting Intensive Care Transfers and other UnfoReseen Events). To develop, tune, and test PICTURE, we used labs and vital signs from electronic health records of adult patients over four years (n = 133,089 encounters). We analyzed primary outcomes of unplanned intensive care unit transfer, emergency vasoactive medication administration, cardiac arrest, and death. We compared PICTURE with existing early warning systems and logistic regression at multiple levels of granularity. When analyzing PICTURE on the testing set using all observations within a hospital encounter (event rate = 3.4%), PICTURE had an area under the receiver operating characteristic curve (AUROC) of 0.83 and an adjusted (event rate = 4%) area under the precision-recall curve (AUPR) of 0.27, while the next best tested method-regularized logistic regression-had an AUROC of 0.80 and an adjusted AUPR of 0.22. To ensure system interpretability, we applied a state-of-the-art prediction explainer that provided a ranked list of features contributing most to the prediction. Though it is currently difficult to compare machine learning-based early warning systems, a rudimentary comparison with published scores demonstrated that PICTURE is on par with state-of-the-art machine learning systems. To facilitate more robust comparisons and development of early warning systems in the future, we have released our variational autoencoder's code and weights so researchers can (a) test their models on data similar to our institution and (b) make their own synthetic datasets.

Implementing a system for the real-time risk assessment of patients considered for intensive care

BACKGROUND

Delay in identifying deterioration in hospitalised patients is associated with delayed admission to an intensive care unit (ICU) and poor outcomes. For the HAVEN project (HICF ref.: HICF-R9-524), we have developed a mathematical model that identifies deterioration in hospitalised patients in real time and facilitates the intervention of an ICU outreach team. This paper describes the system that has been designed to implement the model. We have used innovative technologies such as Portable Format for Analytics (PFA) and Open Services Gateway initiative (OSGi) to define the predictive statistical model and implement the system respectively for greater configurability, reliability, and availability.

RESULTS

The HAVEN system has been deployed as part of a research project in the Oxford University Hospitals NHS Foundation Trust. The system has so far processed > 164,000 vital signs observations and > 68,000 laboratory results for > 12,500 patients and the algorithm generated score is being evaluated to review patients who are under consideration for transfer to ICU. No clinical decisions are being made based on output from the system. The HAVEN score has been computed using a PFA model for all these patients. The intent is that this score will be displayed on a graphical user interface for clinician review and response.

CONCLUSIONS

The system uses a configurable PFA model to compute the HAVEN score which makes the system easily upgradable in terms of enhancing systems' predictive capability. Further system enhancements are planned to handle new data sources and additional management screens.

Early warning score validation methodologies and performance metrics: a systematic review

BACKGROUND

Early warning scores (EWS) have been developed as clinical prognostication tools to identify acutely deteriorating patients. In the past few years, there has been a proliferation of studies that describe the development and validation of novel machine learning-based EWS. Systematic reviews of published studies which focus on evaluating performance of both well-established and novel EWS have shown conflicting conclusions. A possible reason is the heterogeneity in validation methods applied. In this review, we aim to examine the methodologies and metrics used in studies which perform EWS validation.

METHODS

A systematic review of all eligible studies from the MEDLINE database and other sources, was performed. Studies were eligible if they performed validation on at least one EWS and reported associations between EWS scores and inpatient mortality, intensive care unit (ICU) transfers, or cardiac arrest (CA) of adults. Two reviewers independently did a full-text review and performed data abstraction by using standardized data-worksheet based on the TRIPOD (Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis) checklist. Meta-analysis was not performed due to heterogeneity.

RESULTS

The key differences in validation methodologies identified were (1) validation dataset used, (2) outcomes of interest, (3) case definition, time of EWS use and aggregation methods, and (4) handling of missing values. In terms of case definition, among the 48 eligible studies, 34 used the patient episode case definition while 12 used the observation set case definition, and 2 did the validation using both case definitions. Of those that used the patient episode case definition, 18 studies validated the EWS at a single point of time, mostly using the first recorded observation. The review also found more than 10 different performance metrics reported among the studies.

CONCLUSIONS

Methodologies and performance metrics used in studies performing validation on EWS were heterogeneous hence making it difficult to interpret and compare EWS performance. Standardizing EWS validation methodology and reporting can potentially address this issue.

Heart Rate Variability, Clinical and Laboratory Measures to Predict Future Deterioration in Patients Presenting With Sepsis

BACKGROUND

Risk stratification of patients presenting to the emergency department (ED) with sepsis can be challenging. We derived and evaluated performance of a predictive model containing clinical, laboratory, and heart rate variability (HRV) measures to quantify risk of deterioration in this population.

METHODS

ED patients aged 21 and older satisfying the 1992 consensus conference criteria for sepsis and able to consent (directly or through a surrogate) were enrolled (n = 1,247). Patients had clinical, laboratory, and HRV data recorded within 1 h of ED presentation, and were followed to identify deterioration within 72 h.

RESULTS

Eight hundred thirty-two patients had complete data, of whom 68 (8%) reached at least one endpoint. Optimal predictive performance was derived from a combination of laboratory values and HRV metrics with an area under the receiver-operating curve (AUROC) of 0.80 (95% CI, 0.65-0.92). This combination of variables was superior to clinical (AUROC = 0.69, 95% CI, 0.54-0.83), laboratory (AUROC = 0.77, 95% CI, 0.63-0.90), and HRV measures (AUROC = 0.76, 95% CI, 0.61-0.90) alone. The HRV+LAB model identified a high-risk cohort of patients (14% of all patients) with a 4.3-fold (95% CI, 3.2-5.4) increased risk of deterioration (incidence of deterioration: 35%), as well as a low-risk group (61% of all patients) with 0.2-fold (95% CI 0.1-0.4) risk of deterioration (incidence of deterioration: 2%).

CONCLUSIONS

A model that combines HRV and laboratory values may help ED physicians evaluate risk of deterioration in patients with sepsis and merits validation and further evaluation.

Prognostic Performance of SOFA, qSOFA, and SIRS in Kidney Transplant Recipients Suffering from Infection: A Retrospective Observational Study

INTRODUCTION

The prognostic performance of scoring systems for illness severity in infectious kidney transplant recipients (KTRs) is rarely reported. We investigated the ability of the scores for the quick Sequential Organ Failure Assessment (qSOFA), Sequential Organ Failure Assessment (SOFA) and Systemic Inflammatory Response Syndrome (SIRS) to predict in-hospital mortality, intensive care unit (ICU) admission and mechanical ventilation (MV) requirement.

METHODS

This was a second analysis of a retrospective observational study. Scores for SIRS, SOFA and qSOFA were calculated upon hospitalization (infection onset was before hospitalization) or on the day of infection onset (infection episodes were during hospitalization). The primary outcome was in-hospital mortality. The secondary outcomes were ICU admission and MV requirement. Binary logistic regression and area under the receiver operating characteristic curve (AUC) were employed to assess prognostic performance.

RESULTS

A total of 161 infectious episodes occurred in 97 KTRs. Forty patients (41%) experienced more than one episode. The SOFA score was available in 161 infections, and scores for qSOFA and SIRS were available in 160 infections. The SIRS score was not different between KTRs with opposite outcomes. The qSOFA score was higher in infections necessitating MV. The SOFA score was significantly higher in the deceased, those needing ICU admission, MV, and for those with positive etiology results. The SOFA score was the only independent predictor of in-hospital mortality, ICU admission, and MV requirement, and the AUCs were 0.879, 0.815, and 0.784, respectively. The optimum cutoff value of predicting the three outcomes was SOFA score ≥ 3.

CONCLUSIONS

The SOFA score (but not those for SIRS and qSOFA) independently predicted in-hospital mortality, ICU admission, and MV requirement in infectious KTRs.

Developing a framework for evidence-based grading and assessment of predictive tools for clinical decision support

BACKGROUND

Clinical predictive tools quantify contributions of relevant patient characteristics to derive likelihood of diseases or predict clinical outcomes. When selecting predictive tools for implementation at clinical practice or for recommendation in clinical guidelines, clinicians are challenged with an overwhelming and ever-growing number of tools, most of which have never been implemented or assessed for comparative effectiveness. To overcome this challenge, we have developed a conceptual framework to Grade and Assess Predictive tools (GRASP) that can provide clinicians with a standardised, evidence-based system to support their search for and selection of efficient tools.

METHODS

A focused review of the literature was conducted to extract criteria along which tools should be evaluated. An initial framework was designed and applied to assess and grade five tools: LACE Index, Centor Score, Well's Criteria, Modified Early Warning Score, and Ottawa knee rule. After peer review, by six expert clinicians and healthcare researchers, the framework and the grading of the tools were updated.

RESULTS

GRASP framework grades predictive tools based on published evidence across three dimensions: 1) Phase of evaluation; 2) Level of evidence; and 3) Direction of evidence. The final grade of a tool is based on the highest phase of evaluation, supported by the highest level of positive evidence, or mixed evidence that supports a positive conclusion. Ottawa knee rule had the highest grade since it has demonstrated positive post-implementation impact on healthcare. LACE Index had the lowest grade, having demonstrated only pre-implementation positive predictive performance.

CONCLUSION

GRASP framework builds on widely accepted concepts to provide standardised assessment and evidence-based grading of predictive tools. Unlike other methods, GRASP is based on the critical appraisal of published evidence reporting the tools' predictive performance before implementation, potential effect and usability during implementation, and their post-implementation impact. Implementing the GRASP framework as an online platform can enable clinicians and guideline developers to access standardised and structured reported evidence of existing predictive tools. However, keeping GRASP reports up-to-date would require updating tools' assessments and grades when new evidence becomes available, which can only be done efficiently by employing semi-automated methods for searching and processing the incoming information.

Neutrophil to lymphocyte ratio as a predictor of treatment response and mortality in septic shock patients in the intensive care unit

Background/aim

While C-reactive protein (CRP) is a well-studied marker for predicting treatment response and mortality in sepsis, it was aimed to assess the efficacy of the neutrophil lymphocyte ratio (NLR) as a predictor of mortality and treatment response in sepsis patients in the intensive care unit (ICU).

Materials and methods

In this retrospective cross-sectional study, sepsis patients were divided according to the presence of septic shock on the 1st day of ICU stay, and then subgrouped according to mortality. Patient demographics, acute physiologic and chronic health evaluation II and sequential organ failure assessment scores, NLR and CRP (on the 1st, 3rd, and last day in the ICU), microbiology data, antibiotic responses, ICU data, and mortality were recorded. Receiver operating characteristic (ROC) curves for the area under curve (AUC) were calculated for the inflammatory markers and ICU severity scores for mortality.

Results

Of the 591 (65% male) enrolled patients, 111 (18.8%) were nonsurvivors with shock, 117 (19.8%) were survivors with shock, 330 (55.8%) were survivors without shock, and 33 (5.6%) were nonsurvivors without shock. On the 1st day of ICU stay, the NLR and CRP were similar in all of the groups. On the 3rd day of antibiotic response, the NLR was increased (11.8) in the nonresponsive patients when compared with the partially responsive (11.0) and responsive (8.5) patients. If the NLR was ≥15 on the 3rd day, the mortality odds ratio was 6.96 (CI: 1.4–34.1, P < 0.017). The NLR and CRP on the 1st, 3rd, and last day of ICU stay (0.52, 0.58, 0.78 and 0.56, 0.70, 0.78, respectively) showed a similar increasing trend for mortality.

Conclusion

The NLR can predict mortality and antibiotic responsiveness in ICU patients with sepsis and septic shock. If the NLR is >15 on the 3rd day of postantibiotic initiation, the risk of mortality is high and treatment should be reviewed carefully.

The Rapid Emergency Medicine Score: A Critical Appraisal of Its Measurement Properties and Applicability to the Air Retrieval Environment

OBJECTIVE

The Rapid Emergency Medicine Score (REMS) was designed to predict in-hospital mortality using variables that are available in the prehospital setting. The objective of this article is to critically appraise the development and summarize the evidence regarding the measurement properties (sensitivity, reliability and validity) of the REMS.

METHODS

A literature search was performed identifying all studies describing the REMS. The original validation study was critically appraised for its development. All other studies that reported any measurement properties of the REMS were also appraised for evidence of calibration, reliability, and validity.

RESULTS

In total, 26 studies reported on the measurement properties of the REMS. Overall, the REMS was developed with robust methodology and has good sensibility with adequate content and face validity. It is easy to understand and feasible to be calculated within minutes of patient assessment. The REMS has the necessary measurement properties to be both a predictive and evaluative clinical index to measure prehospital severity of illness; however, no studies have adequately addressed the intra or inter-rater reliability of the score.

CONCLUSIONS

There is evidence to support the use of the REMS as a predictive or evaluative instrument. In most studies, it performed as well or better than other illness severity scores in predicting mortality.

Multicenter derivation and validation of an early warning score for acute respiratory failure or death in the hospital

BACKGROUND

Acute respiratory failure occurs frequently in hospitalized patients and often starts before ICU admission. A risk stratification tool to predict mortality and risk for mechanical ventilation (MV) may allow for earlier evaluation and intervention. We developed and validated an automated electronic health record (EHR)-based model-Accurate Prediction of Prolonged Ventilation (APPROVE)-to identify patients at risk of death or respiratory failure requiring >= 48 h of MV.

METHODS

This was an observational study of adults admitted to four hospitals in 2013 or a fifth hospital in 2017. Clinical data were extracted from the EHRs. The 2013 patients were randomly split 50:50 into a derivation/validation cohort. The qualifying event was death or intubation leading to MV >= 48 h. Random forest method was used in model derivation. APPROVE was calculated retrospectively whenever data were available in 2013, and prospectively every 4 h after hospital admission in 2017. The Modified Early Warning Score (MEWS) and National Early Warning Score (NEWS) were calculated at the same times as APPROVE. Clinicians were not alerted except for APPROVE in 2017cohort.

RESULTS

There were 68,775 admissions in 2013 and 2258 in 2017. APPROVE had an area under the receiver operator curve of 0.87 (95% CI 0.85-0.88) in 2013 and 0.90 (95% CI 0.84-0.95) in 2017, which is significantly better than the MEWS and NEWS in 2013 but similar to the MEWS and NEWS in 2017. At a threshold of > 0.25, APPROVE had similar sensitivity and positive predictive value (PPV) (sensitivity 63% and PPV 21% in 2013 vs 64% and 16%, respectively, in 2017). Compared to APPROVE in 2013, at a threshold to achieve comparable PPV (19% at MEWS > 4 and 22% at NEWS > 6), the MEWS and NEWS had lower sensitivity (16% for MEWS and NEWS). Similarly in 2017, at a comparable sensitivity threshold (64% for APPROVE > 0.25 and 67% for MEWS and NEWS > 4), more patients who triggered an alert developed the event with APPROVE (PPV 16%) while achieving a lower false positive rate (FPR 5%) compared to the MEWS (PPV 7%, FPR 14%) and NEWS (PPV 4%, FPR 25%).

CONCLUSIONS

An automated EHR model to identify patients at high risk of MV or death was validated retrospectively and prospectively, and was determined to be feasible for real-time risk identification.

TRIAL REGISTRATION

ClinicalTrials.gov, NCT02488174 . Registered on 18 March 2015.

Factors associated with recurrence and mortality in central line-associated bloodstream infections: a retrospective cohort study

Background

Central line-associated bloodstream infections (CLABSIs) are associated with increased mortality, hospital length of stay, and cost. Antimicrobial treatment guidelines for CLABSIs are primarily based on expert opinion. We hypothesized that shorter antimicrobial treatment duration is associated with decreased 60-day recurrence-free survival.

Methods

A retrospective cohort study of all adults with hospital-acquired CLABSIs (HA-CLABSIs) over 5 years at a single tertiary care academic hospital was performed. The time from the end of effective antimicrobial treatment until recurrence of infection or mortality, censored at 60 days after the end of antimicrobial treatment, represented the primary outcome. Effective antimicrobial treatment was defined as the administration of at least one antimicrobial to which the causative organism was sensitive.

Results

A total of 366 cases met eligibility criteria. The median Sequential Organ Failure Assessment (SOFA) score was 6 (interquartile range (IQR) 4–8). Patients were treated for a median of 15 (IQR 10–20) days with effective antimicrobials. The incidence of 60-day mortality or recurrence after completion of the antimicrobial course was 22.1% (81 patients). In a Cox proportional-hazards model, antimicrobial treatment duration (hazard ratio (HR) = 0.35; 95% confidence interval (CI) 0.26–0.48), SOFA score (HR = 1.16; 95% CI 1.09–1.22), and age (HR = 1.021; 95% CI = 1.004–1.037) were associated with mortality or recurrence. The effect of antimicrobial treatment duration appeared to plateau after 15 days.

Conclusions

Longer antimicrobial treatment duration in patients with HA-CLABSIs is associated with improved recurrence-free survival during the first 60 days after infection. This effect appears to plateau after 15 days of treatment. Prospective studies are needed to definitively determine the optimal antimicrobial treatment duration for CLABSIs.

Electronic supplementary material

The online version of this article (10.1186/s13054-018-2206-7) contains supplementary material, which is available to authorized users.

Evaluation of the efficacy of the National Early Warning Score in predicting in-hospital mortality via the risk stratification

PURPOSE

To investigate the efficacy of the National Early Warning Score (NEWS) in predicting in-hospital mortality.

MATERIALS AND METHODS

This was a retrospective observational study and the electronic medical records of the patients were reviewed based on NEWS at the time of admission.

RESULTS

The performance of NEWS was effective in predicting hospital mortality (area under the curve: 0.765; 95% confidence interval: 0.659-0.846). Based on the Kaplan Meier survival curves, the survival time of patients who are at high risk according to NEWS was significantly shorter than that of patients who are at low risk (p < 0.001). Results of the multivariate Cox proportional hazards regression analysis showed that the hazard ratios of patients who are at medium and high risk based on NEWS were 2.6 and 4.7, respectively (p < 0.001). In addition, our study showed that the combination model that used other factors, such as age and diagnosis, was more effective than NEWS alone in predicting hospital mortality (NEWS: 0.765; combination model: 0.861; p < 0.005).

CONCLUSIONS

NEWS is a simple and useful bedside tool for predicting in-hospital mortality. In addition, the rapid response team must consider other clinical factors as well as screening tools to improve clinical outcomes.

SOFA score in septic patients: incremental prognostic value over age, comorbidities, and parameters of sepsis severity

Several widely used scoring systems for septic patients have been validated in an ICU setting, and may not be appropriate for other settings like Emergency Departments (ED) or High-Dependency Units (HDU), where a relevant number of these patients are managed. The purpose of this study is to find reliable tools for prognostic assessment of septic patients managed in an ED-HDU. In 742 patients diagnosed with sepsis/severe sepsis/septic shock, not-intubated, admitted in ED between June 2008 and April 2016, SOFA, qSOFA, PIRO, MEWS, Charlson Comorbidity Index, MEDS, and APACHE II were calculated at ED admission (T0); SOFA and MEWS were also calculated after 24 h of ED-High-Dependency Unit stay (T1). Discrimination and incremental prognostic value of SOFA score over demographic data and parameters of sepsis severity were tested. Primary outcome is 28-day mortality. Twenty-eight day mortality rate is 31%. The different scores show a modest-to-moderate discrimination (T0 SOFA 0.695; T1 SOFA 0.741; qSOFA 0.625; T0 MEWS 0.662; T1 MEWS 0.729; PIRO: 0.646; APACHE II 0.756; Charlson Comorbidity Index 0.596; MEDS 0.674, all p < 0.001). At a multivariate stepwise Cox analysis, including age, Charlson Comorbidity Index, MEWS, and lactates, SOFA shows an incremental prognostic ability both at T0 (RR 1.165, IC 95% 1.009-1.224, p < 0.0001) and T1 (RR 1.168, IC 95% 1.104-1.234, p < 0.0001). SOFA score shows a moderate prognostic stratification ability, and demonstrates an incremental prognostic value over the previous medical conditions and clinical parameters in septic patients.

SOFA score and left ventricular systolic function as predictors of short-term outcome in patients with sepsis

In a group of septic patients, we assess the short-term prognostic value of LV systolic performance, evaluated through conventional left ventricular ejection fraction (LVEF) and left ventricular global longitudinal strain (LV GLS). One hundred forty-seven patients with sepsis were recruited; LVEF by planimetry and peak GLS by 2D speckle tracking could be assessed within 24 h. The study population was stratified according to SOFA tertiles assessed at the time of the echocardiogram (G1: SOFA score <5; G2: SOFA score 5-7; G3: SOFA score >7). Day-7 follow-up data were used as reference. Patients in G2 and G3 show a significant hemodynamic derangement, paralleling the more pronounced organ damage by definition; nevertheless, LVEF and GLS are comparable among the three groups (both p > 0.1). All-cause mortality at day-7 follow-up is slightly lower in G1 (9%) versus G2 and G3 (14 and 26%, respectively, p = NS). Analyses through ROC curves focusing on day-7 mortality show that the SOFA score fairly correlates with events (AUC 0.635, p = 0.037), while low LVEF (AUC 0.35, p = 0.022) and less negative GLS (AUC 0.73, p = 0.001) do so. In multivariate analyses, mortality by day-7 follow-up is more likely per higher GLS (i.e., indicative of worst systolic dysfunction, HR 1.22/%, p = 0.005) and per increasing SOFA score (HR 1.22/unit, p = 0.010), whereas LVEF, adjusted for age and SOFA score, does not enter the prognostic model. In the very short term in patients with severe sepsis, LV systolic function assessment by means of GLS predicts the short-term prognosis, independent of SOFA.

Personalized Risk Scoring for Critical Care Prognosis Using Mixtures of Gaussian Processes

OBJECTIVE

In this paper, we develop a personalized real-time risk scoring algorithm that provides timely and granular assessments for the clinical acuity of ward patients based on their (temporal) lab tests and vital signs; the proposed risk scoring system ensures timely intensive care unit admissions for clinically deteriorating patients.

METHODS

The risk scoring system is based on the idea of sequential hypothesis testing under an uncertain time horizon. The system learns a set of latent patient subtypes from the offline electronic health record data, and trains a mixture of Gaussian Process experts, where each expert models the physiological data streams associated with a specific patient subtype. Transfer learning techniques are used to learn the relationship between a patient's latent subtype and her static admission information (e.g., age, gender, transfer status, ICD-9 codes, etc).

RESULTS

Experiments conducted on data from a heterogeneous cohort of 6321 patients admitted to Ronald Reagan UCLA medical center show that our score significantly outperforms the currently deployed risk scores, such as the Rothman index, MEWS, APACHE, and SOFA scores, in terms of timeliness, true positive rate, and positive predictive value.

CONCLUSION

Our results reflect the importance of adopting the concepts of personalized medicine in critical care settings; significant accuracy and timeliness gains can be achieved by accounting for the patients' heterogeneity.

SIGNIFICANCE

The proposed risk scoring methodology can confer huge clinical and social benefits on a massive number of critically ill inpatients who exhibit adverse outcomes including, but not limited to, cardiac arrests, respiratory arrests, and septic shocks.

Use of the low-frequency/high-frequency ratio of heart rate variability to predict short-term deterioration in emergency department patients with sepsis

OBJECTIVE

To examine the ability of the low-frequency/high-frequency (LF/HF) ratio of heart rate variability (HRV) analysis to identify patients with sepsis at risk of early deterioration.

METHODS

This is a prospective observational cohort study of patients with sepsis presenting to the Montefiore Medical Center ED from December 2014 through September 2015. On presentation, a single ECG Holter recording was obtained and analysed to obtain the LF/HF ratio of HRV. Initial Sequential Organ Failure Assessment (SOFA) scores were computed. Patients were followed for 72 hours to identify those with early deterioration.

RESULTS

466 patients presenting to the ED with sepsis were analysed. Thirty-two (7%) reached at least one endpoint within 72 hours. An LF/HF ratio <1 had a sensitivity and specificity of 34% (95% CI (19% to 53%)) and 82% (95% CI (78% to 85%)), respectively, with positive and negative likelihood ratios of 1.9 (95% CI (1.1 to 3.2)) and 0.8 (95% CI (0.6 to 1.0)). An initial SOFA score ≥3 had a sensitivity and specificity of 38% (95% CI (22% to 56%)) and 92% (95% CI (89% to 95%)), with positive and negative likelihood ratios of 4.9 (95% CI (2.8 to 8.6)) and 0.7 (95% CI (0.5 to 0.9)). The composite measure of HRV+SOFA had improved sensitivity (56%, 95% CI (38% to 73%)) but at the expense of specificity (77%, 95% CI (72% to 80%)), with positive and negative likelihood ratios of 2.4 (95% CI (1.7 to 3.4)) and 0.6 (95% CI (0.4 to 0.9)). Receiver operating characteristic analysis did not identify a superior alternate threshold for the LF/HF ratio. Kaplan-Meier survival functions differed significantly (p=0.02) between low (<1) and high (≥1) LF/HF groups.

CONCLUSIONS

While we found a statistically significant relationship between HRV, SOFA and HRV+SOFA, and early deterioration, none reliably functioned as a clinical predictive tool. More complex multivariable models will likely be required to construct models with clinical utility.

Sepsis Prevalence and Outcome on the General Wards and Emergency Departments in Wales: Results of a Multi-Centre, Observational, Point Prevalence Study

Data on sepsis prevalence on the general wards is lacking on the UK and in the developed world. We conducted a multicentre, prospective, observational study of the prevalence of patients with sepsis or severe sepsis on the general wards and Emergency Departments (ED) in Wales. During the 24-hour study period all patients with NEWS≥3 were screened for presence of 2 or more SIRS criteria. To be eligible for inclusion, patients had to have a high clinical suspicion of an infection, together with a systemic inflammatory response (sepsis) and evidence of acute organ dysfunction and/or shock (severe sepsis). There were 5317 in-patients in the 24-hour study period. Data were returned on 1198 digital data collection forms on patients with NEWS≥3 of which 87 were removed, leaving 1111 for analysis. 146 patients had sepsis and 144 patients had severe sepsis. Combined prevalence of sepsis and severe sepsis was 5.5% amongst all in-patients. Patients with sepsis had significantly higher NEWS scores (3 IQR 3-4 for non-sepsis and 4 IQR 3-6 for sepsis patients, respectively). Common organ dysfunctions in severe sepsis were hypoxia (47%), hypoperfusion (40%) and acute kidney injury (25%). Mortality at 90 days was 31% with a median (IQR) hospital free stay of 78 (36-85) days. Screening for sepsis, referral to Critical Care and completion of Sepsis 6 bundle was low: 26%, 16% and 12% in the sepsis group. Multivariable logistic regression analysis identified higher National Early Warning Score, diabetes, COPD, heart failure, malignancy and current or previous smoking habits as independent variables suggesting the diagnosis of sepsis. We observed that sepsis is more prevalent in the general ward and ED than previously suggested before and that screening and effective treatment for sepsis and severe sepsis is far from being operationalized in this environment, leading to high 90 days mortality.

Use of the APACHE II score to assess impact of therapeutic plasma exchange for critically ill patients with hypertriglyceride-induced pancreatitis

OBJECTIVES

Hypertriglyceridemic (HTG) pancreatitis carries significant morbidity and mortality and often requires intensive care unit (ICU) admission. Therapeutic plasma exchange (TPE) rapidly lowers serum triglyceride (TG) levels. However, evidence supporting TPE for HTG pancreatitis is lacking.

METHODS

Ten patients admitted to the ICU for HTG pancreatitis underwent TPE at our institution from 2005-2015. We retrospectively calculated the Acute Physiology and Chronic Health Examination II (APACHE II) score at the time of initial TPE and again after the final TPE session to assess the impact of triglyceride apheresis on morbidity and mortality associated with HTG pancreatitis.

RESULTS

All 10 patients had rapid reduction in TG level after TPE, but only 5 had improvement in their APACHE II score. The median APACHE II score decreased from 19% to 17% after TPE, correlating with an 8% and 9% decrease in median predicted non-operative and post-operative mortality, respectively. The APACHE II score did not differ statistically before and after TPE implementation in our patient group (p=0.39).

CONCLUSION

TPE is a clinically useful tool to rapidly lower TG levels, but its impact on mortality of HTG pancreatitis as assessed by the APACHE II score remains uncertain.

Who to handover: a case-control study of a novel scoring system to prioritise handover of internal medicine inpatients

BACKGROUND

Handover of patients between care providers is a critical event in patient care. There is, however, little evidence to guide the handover process, including determining which patients to handover.

AIM

Compare the ability of gestalt-based handover with two structured scores, the modified early warning score (MEWS) and our novel iHAND clinical decision support system, to predict which patients will be assessed by a physician overnight.

METHODS

This case-control study included 90 inpatients, comprising 32 patients assessed overnight (cases) and 58 patients not assessed overnight (controls) at a teaching hospital in British Columbia, Canada (May 2012). Gestalt, MEWS and iHAND scores were analysed against patients seen overnight using logistic regression and receiver-operating characteristic (ROC) curves.

RESULTS

Neither current gestalt-based handover practice (odds ratio (OR) 1.50, 95% CI 0.89 to 3.83) nor MEWS (OR 0.96, 95% CI 0.75 to 1.24, area under the ROC curve (AUC) 0.61, 95% CI 0.49 to 0.73) were significantly associated with need to be seen overnight. The iHAND score was associated with need to be seen (OR 1.93, 95% CI 1.24 to 3.02, AUC 0.70, 95% CI 0.60 to 0.81).

CONCLUSIONS

The iHAND score had moderate ability to predict which patients required assessment overnight, while MEWS score and current gestalt approach correlated poorly, suggesting the iHAND score may help prioritisation of patients likely to be seen overnight for handover.

Early warning score: a dynamic marker of severity and prognosis in patients with Gram-negative bacteraemia and sepsis

Background

Early Warning Score (EWS) is a physiological composite score of six bedside vital parameters, routinely used in UK hospitals. We evaluated the prognostic ability of EWS in Gram-negative bacteraemia causing sepsis.

Methods

We prospectively evaluated EWS as a marker of severity and prognosis in adult patients with Gram-negative bacteraemia. All adult patients with Gram-negative bacteraemia admitted to our tertiary Teaching hospital of the National Health Service in England were enrolled over 1 year period. The highest daily EWS score was recorded from 7 days before to 14 days after the date of onset of bacteraemia. The primary outcome was 28-day mortality.

Main results

A total of 245 consecutive adult patients with Gram-negative bacteraemia with sepsis were enrolled. On multivariate analysis, following variables were associated with death for every single unit change (odds ratio in the brackets): higher age (1.05), lower mean arterial pressure (1.03), lower serum bicarbonate (1.08), higher EWS (1.27), higher SOFA score (1.36), hospital-onset of infection (5.43) and need for vasopressor agents (16.4). EWS on day 0, 1, 2, and average 14-day score were significantly higher in patients who died by 28 days from the onset of bacteraemia [95 % CI 0.4–0.6] p < 0.001. A stepwise rise in EWS and failure of improvement in EWS by 2 points 48 h after the onset of bacteraemia were associated with poor outcome.

Conclusion

EWS is a simple and cost-effective bedside tool for the assessment of severity and prognosis of sepsis caused by Gram-negative bacteraemia.

Electronic supplementary material

The online version of this article (doi:10.1186/s12941-016-0139-z) contains supplementary material, which is available to authorized users.

Association between trends in clinical variables and outcome in intensive care patients with faecal peritonitis: analysis of the GenOSept cohort

Introduction

Patients admitted to intensive care following surgery for faecal peritonitis present particular challenges in terms of clinical management and risk assessment. Collaborating surgical and intensive care teams need shared perspectives on prognosis. We aimed to determine the relationship between dynamic assessment of trends in selected variables and outcomes.

Methods

We analysed trends in physiological and laboratory variables during the first week of intensive care unit (ICU) stay in 977 patients at 102 centres across 16 European countries. The primary outcome was 6-month mortality. Secondary endpoints were ICU, hospital and 28-day mortality. For each trend, Cox proportional hazards (PH) regression analyses, adjusted for age and sex, were performed for each endpoint.

Results

Trends over the first 7 days of the ICU stay independently associated with 6-month mortality were worsening thrombocytopaenia (mortality: hazard ratio (HR) = 1.02; 95% confidence interval (CI), 1.01 to 1.03; P <0.001) and renal function (total daily urine output: HR =1.02; 95% CI, 1.01 to 1.03; P <0.001; Sequential Organ Failure Assessment (SOFA) renal subscore: HR = 0.87; 95% CI, 0.75 to 0.99; P = 0.047), maximum bilirubin level (HR = 0.99; 95% CI, 0.99 to 0.99; P = 0.02) and Glasgow Coma Scale (GCS) SOFA subscore (HR = 0.81; 95% CI, 0.68 to 0.98; P = 0.028). Changes in renal function (total daily urine output and renal component of the SOFA score), GCS component of the SOFA score, total SOFA score and worsening thrombocytopaenia were also independently associated with secondary outcomes (ICU, hospital and 28-day mortality). We detected the same pattern when we analysed trends on days 2, 3 and 5. Dynamic trends in all other measured laboratory and physiological variables, and in radiological findings, changes inrespiratory support, renal replacement therapy and inotrope and/or vasopressor requirements failed to be retained as independently associated with outcome in multivariate analysis.

Conclusions

Only deterioration in renal function, thrombocytopaenia and SOFA score over the first 2, 3, 5 and 7 days of the ICU stay were consistently associated with mortality at all endpoints. These findings may help to inform clinical decision making in patients with this common cause of critical illness.

Electronic supplementary material

The online version of this article (doi:10.1186/s13054-015-0931-8) contains supplementary material, which is available to authorized users.