Predicting cardiac arrest on the wards: a nested case-control study

Enhancing discharge decision-making through continuous monitoring in an acute admission ward: a randomized controlled trial

In Acute Admission Wards, vital signs are commonly measured only intermittently. This may result in failure to detect early signs of patient deterioration and impede timely identification of patient stability, ultimately leading to prolonged stays and avoidable hospital admissions. Therefore, continuous vital sign monitoring may improve hospital efficacy. The objective of this randomized controlled trial was to evaluate the effect of continuous monitoring on the proportion of patients safely discharged home directly from an Acute Admission Ward. Patients were randomized to either the control group, which received usual care, or the sensor group, which additionally received continuous monitoring using a wearable sensor. The continuous measurements could be considered in discharge decision-making by physicians during the daily bedside rounds. Safe discharge was defined as no unplanned readmissions, emergency department revisits or deaths, within 30 days after discharge. Additionally, length of stay, the number of Intensive Care Unit admissions and Rapid Response Team calls were assessed. In total, 400 patients were randomized, of which 394 completed follow-up, with 196 assigned to the sensor group and 198 to the control group. The proportion of patients safely discharged home was 33.2% in the sensor group and 30.8% in the control group (p = 0.62). No significant differences were observed in secondary outcomes. The trial was terminated prematurely due to futility. In conclusion, continuous monitoring did not have an effect on the proportion of patients safely discharged from an Acute Admission Ward. Implementation challenges of continuous monitoring may have contributed to the lack of effect observed. Trial registration: https://clinicaltrials.gov/ct2/show/NCT05181111 . Registered: January 6, 2022.

Development of an enhanced scoring system to predict ICU readmission or in-hospital death within 24 hours using routine patient data from two NHS Foundation Trusts

RATIONALE

Intensive care units (ICUs) admit the most severely ill patients. Once these patients are discharged from the ICU to a step-down ward, they continue to have their vital signs monitored by nursing staff, with Early Warning Score (EWS) systems being used to identify those at risk of deterioration.

OBJECTIVES

We report the development and validation of an enhanced continuous scoring system for predicting adverse events, which combines vital signs measured routinely on acute care wards (as used by most EWS systems) with a risk score of a future adverse event calculated on discharge from the ICU.

DESIGN

A modified Delphi process identified candidate variables commonly available in electronic records as the basis for a 'static' score of the patient's condition immediately after discharge from the ICU. L1-regularised logistic regression was used to estimate the in-hospital risk of future adverse event. We then constructed a model of physiological normality using vital sign data from the day of hospital discharge. This is combined with the static score and used continuously to quantify and update the patient's risk of deterioration throughout their hospital stay.

SETTING

Data from two National Health Service Foundation Trusts (UK) were used to develop and (externally) validate the model.

PARTICIPANTS

A total of 12 394 vital sign measurements were acquired from 273 patients after ICU discharge for the development set, and 4831 from 136 patients in the validation cohort.

RESULTS

Outcome validation of our model yielded an area under the receiver operating characteristic curve of 0.724 for predicting ICU readmission or in-hospital death within 24 hours. It showed an improved performance with respect to other competitive risk scoring systems, including the National EWS (0.653).

CONCLUSIONS

We showed that a scoring system incorporating data from a patient's stay in the ICU has better performance than commonly used EWS systems based on vital signs alone.

TRIAL REGISTRATION NUMBER

ISRCTN32008295.

A novel structured debriefing program for consensus determinations of in-hospital cardiac arrest predictability and preventability

AIM

Hospital rapid response systems aim to stop preventable cardiac arrests, but defining preventability is a challenge. We developed a multidisciplinary consensus-based process to determine in-hospital cardiac arrest (IHCA) preventability based on objective measures.

METHODS

We developed an interdisciplinary ward IHCA debriefing program at an urban quaternary-care academic hospital. This group systematically reviewed all IHCAs weekly, reaching consensus determinations of the IHCA's cause and preventability across three mutually exclusive categories: 1) unpredictable (no evidence of physiologic instability < 1 h prior to and within 24 h of the arrest), 2) predictable but unpreventable (meeting physiologic instability criteria in the setting of either a poor baseline prognosis or a documented goals of care conversation) or 3) potentially preventable (remaining cases).

RESULTS

Of 544 arrests between 09/2015 and 11/2023, 339 (61%) were deemed predictable by consensus, with 235 (42% of all IHCAs) considered potentially preventable. Potentially preventable arrests disproportionately occurred on nights and weekends (70% vs 55%, p = 0.002) and were more frequently respiratory than cardiac in etiology (33% vs 15%, p < 0.001). Despite similar rates of ROSC across groups (67-70%), survival to discharge was highest in arrests deemed unpredictable (31%), followed by potentially preventable (21%), and then those deemed predictable but unpreventable which had the lowest survival rate (16%, p = 0.007).

CONCLUSIONS

Our IHCA debriefing procedures are a feasible and sustainable means of determining the predictability and potential preventability of ward cardiac arrests. This approach may be useful for improving quality benchmarks and care processes around pre-arrest clinical activities.

Real-time machine learning model to predict short-term mortality in critically ill patients: development and international validation

BACKGROUND

A real-time model for predicting short-term mortality in critically ill patients is needed to identify patients at imminent risk. However, the performance of the model needs to be validated in various clinical settings and ethnicities before its clinical application. In this study, we aim to develop an ensemble machine learning model using routinely measured clinical variables at a single academic institution in South Korea.

METHODS

We developed an ensemble model using deep learning and light gradient boosting machine models. Internal validation was performed using the last two years of the internal cohort dataset, collected from a single academic hospital in South Korea between 2007 and 2021. External validation was performed using the full Medical Information Mart for Intensive Care (MIMIC), eICU-Collaborative Research Database (eICU-CRD), and Amsterdam University Medical Center database (AmsterdamUMCdb) data. The area under the receiver operating characteristic curve (AUROC) was calculated and compared to that for the National Early Warning Score (NEWS).

RESULTS

The developed model (iMORS) demonstrated high predictive performance with an internal AUROC of 0.964 (95% confidence interval [CI] 0.963-0.965) and external AUROCs of 0.890 (95% CI 0.889-0.891) for MIMIC, 0.886 (95% CI 0.885-0.887) for eICU-CRD, and 0.870 (95% CI 0.868-0.873) for AmsterdamUMCdb. The model outperformed the NEWS with higher AUROCs in the internal and external validation (0.866 for the internal, 0.746 for MIMIC, 0.798 for eICU-CRD, and 0.819 for AmsterdamUMCdb; p < 0.001).

CONCLUSIONS

Our real-time machine learning model to predict short-term mortality in critically ill patients showed excellent performance in both internal and external validations. This model could be a useful decision-support tool in the intensive care units to assist clinicians.

Effect of a lower target oxygen saturation range on the risk of hypoxaemia and elevated NEWS2 scores at a university hospital: a retrospective study

BACKGROUND

The optimal target oxygen saturation (SpO2) range for hospital inpatients not at risk of hypercapnia is unknown. The objective of this study was to assess the impact on oxygen usage and National Early Warning Score 2 (NEWS2) of changing the standard SpO2 target range from 94-98% to 92-96%.

METHODS

In a metropolitan UK hospital, a database of electronic bedside SpO2 measurements, oxygen prescriptions and NEWS2 records was reviewed. Logistic regression was used to compare the proportion of hypoxaemic SpO2 values (<90%) and NEWS2 records ≥5 in 2019, when the target SpO2 range was 94-98%; with 2022, when the target range was 92-96%.

RESULTS

In 2019, 218 of 224 936 (0.10%) observations on room air and 162 of 11 328 (1.43%) on oxygen recorded an SpO2 <90%, and in 2022, 251 of 225 970 (0.11%) and 233 of 12 845 (1.81%), respectively (risk difference 0.04%, 95% CI 0.02% to 0.07%). NEWS2 ≥5 was observed in 3009 of 236 264 (1.27%) observations in 2019 and 4061 of 238 815 (1.70%) in 2022 (risk difference 0.43%, 0.36% to 0.50%; p<0.001). The proportion of patients using supplemental oxygen with hyperoxaemia (SpO2 100%) was 5.4% in 2019 and 3.9% in 2022 (OR 0.71, 0.63 to 0.81; p<0.001).

DISCUSSION

The proportion of observations with SpO2 <90% or NEWS2 ≥5 was greater with the 92-96% range; however, absolute differences were very small and of doubtful clinical relevance, in contrast to hyperoxaemia for which the proportion was markedly less in 2022. These findings support proposals that the British Thoracic Society oxygen guidelines could recommend a lower target SpO2 range.

Causes, Diagnostic Testing, and Treatments Related to Clinical Deterioration Events among High-Risk Ward Patients

OBJECTIVE

Timely intervention for clinically deteriorating ward patients requires that care teams accurately diagnose and treat their underlying medical conditions. However, the most common diagnoses leading to deterioration and the relevant therapies provided are poorly characterized. Therefore, we aimed to determine the diagnoses responsible for clinical deterioration, the relevant diagnostic tests ordered, and the treatments administered among high-risk ward patients using manual chart review.

DESIGN

Multicenter retrospective observational study.

SETTING

Inpatient medical-surgical wards at four health systems from 2006-2020 PATIENTS: Randomly selected patients (1,000 from each health system) with clinical deterioration, defined by reaching the 95th percentile of a validated early warning score, electronic Cardiac Arrest Risk Triage (eCART), were included.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Clinical deterioration was confirmed by a trained reviewer or marked as a false alarm if no deterioration occurred for each patient. For true deterioration events, the condition causing deterioration, relevant diagnostic tests ordered, and treatments provided were collected. Of the 4,000 included patients, 2,484 (62%) had clinical deterioration confirmed by chart review. Sepsis was the most common cause of deterioration (41%; n=1,021), followed by arrhythmia (19%; n=473), while liver failure had the highest in-hospital mortality (41%). The most common diagnostic tests ordered were complete blood counts (47% of events), followed by chest x-rays (42%), and cultures (40%), while the most common medication orders were antimicrobials (46%), followed by fluid boluses (34%), and antiarrhythmics (19%).

CONCLUSIONS

We found that sepsis was the most common cause of deterioration, while liver failure had the highest mortality. Complete blood counts and chest x-rays were the most common diagnostic tests ordered, and antimicrobials and fluid boluses were the most common medication interventions. These results provide important insights for clinical decision-making at the bedside, training of rapid response teams, and the development of institutional treatment pathways for clinical deterioration.

KEY POINTS

Question: What are the most common diagnoses, diagnostic test orders, and treatments for ward patients experiencing clinical deterioration? Findings: In manual chart review of 2,484 encounters with deterioration across four health systems, we found that sepsis was the most common cause of clinical deterioration, followed by arrythmias, while liver failure had the highest mortality. Complete blood counts and chest x-rays were the most common diagnostic test orders, while antimicrobials and fluid boluses were the most common treatments. Meaning: Our results provide new insights into clinical deterioration events, which can inform institutional treatment pathways, rapid response team training, and patient care.

Surveillance of high-risk early postsurgical patients for real-time detection of complications using wireless monitoring (SHEPHERD study): results of a randomized multicenter stepped wedge cluster trial

Background

Vital signs measurements on the ward are performed intermittently. This could lead to failure to rapidly detect patients with deteriorating vital signs and worsens long-term outcome. The aim of this study was to test the hypothesis that continuous wireless monitoring of vital signs on the postsurgical ward improves patient outcome.

Methods

In this prospective, multicenter, stepped-wedge cluster randomized study, patients in the control group received standard monitoring. The intervention group received continuous wireless monitoring of heart rate, respiratory rate and temperature on top of standard care. Automated alerts indicating vital signs deviation from baseline were sent to ward nurses, triggering the calculation of a full early warning score followed. The primary outcome was the occurrence of new disability three months after surgery.

Results

The study was terminated early (at 57% inclusion) due to COVID-19 restrictions. Therefore, only descriptive statistics are presented. A total of 747 patients were enrolled in this study and eligible for statistical analyses, 517 patients in the control group and 230 patients in the intervention group, the latter only from one hospital. New disability at three months after surgery occurred in 43.7% in the control group and in 39.1% in the intervention group (absolute difference 4.6%).

Conclusion

This is the largest randomized controlled trial investigating continuous wireless monitoring in postoperative patients. While patients in the intervention group seemed to experience less (new) disability than patients in the control group, results remain inconclusive with regard to postoperative patient outcome due to premature study termination.

Clinical trial registration

ClinicalTrials.gov, ID: NCT02957825.

Wearable heart rate variability and atrial fibrillation monitoring to improve clinically relevant endpoints in cardiac surgery-a systematic review

Background

This systematic review aims to highlight the untapped potential of heart rate variability (HRV) and atrial fibrillation (AF) monitoring by wearable health monitoring devices as a critical diagnostic tool in cardiac surgery (CS) patients. We reviewed established predictive capabilities of HRV and AF monitoring in specific cardiosurgical scenarios and provide a perspective on additional predictive properties of wearable health monitoring devices that need to be investigated.

Methods

After screening most relevant databases, we included 33 publications in this review. Perusing these publications on HRV's prognostic value, we could identify HRV as a predictor for sudden cardiac death, mortality after acute myocardial infarction (AMI), and post operative atrial fibrillation (POAF). With regards to standard AF assessment, which typically includes extensive periods of unrecorded cardiac activity, we demonstrated that continuous monitoring via wearables recorded significant cardiac events that would otherwise have been missed.

Results

Photoplethysmography and single-lead electrocardiogram (ECG) were identified as the most useful and convenient technical assessment modalities, and their advantages and disadvantages were described in detail. As a call to further action, we observed that the scientific community has relatively extensively explored wearable AF screening, whereas HRV assessment to improve relevant clinical outcomes in CS is rarely studied; it still has great potential to be leveraged.

Conclusions

Therefore, risk assessment in CS would benefit greatly from earlier preoperative and postoperative AF detection, comprehensive and accurate assessment of cardiac health through HRV metrics, and continuous long-term monitoring. These should be achievable via commercially available wearables.

Rapid response team activation prior to in-hospital cardiac arrest: Areas for improvements based on a national cohort study

INTRODUCTION

Rapid response teams (RRTs) are designed to improve the "chain of prevention" of in-hospital cardiac arrest (IHCA). We studied the 30-day survival of patients reviewed by RRTs within 24 hours prior to IHCA, as compared to patients not reviewed by RRTs.

METHODS

A nationwide cohort study based on the Swedish Registry of Cardiopulmonary Resuscitation, between January 1st, 2014, and December 31st, 2021. An explorative, hypothesis-generating additional in-depth data collection from medical records was performed in a small subgroup of general ward patients reviewed by RRTs.

RESULTS

In all, 12,915 IHCA patients were included. RRT-reviewed patients (n = 2,058) had a lower unadjusted 30-day survival (25% vs 33%, p < 0.001), a propensity score based Odds ratio for 30-day survival of 0.92 (95% Confidence interval 0.90-0.94, p < 0.001) and were more likely to have a respiratory cause of IHCA (22% vs 15%, p < 0.001). In the subgroup (n = 82), respiratory distress was the most common RRT trigger, and 24% of the RRT reviews were delayed. Patient transfer to a higher level of care was associated with a higher 30-day survival rate (20% vs 2%, p < 0.001).

CONCLUSION

IHCA preceded by RRT review is associated with a lower 30-day survival rate and a greater likelihood of a respiratory cause of cardiac arrest. In the small explorative subgroup, respiratory distress was the most common RRT trigger and delayed RRT activation was frequent. Early detection of respiratory abnormalities and timely interventions may have a potential to improve outcomes in RRT-reviewed patients and prevent further progress into IHCA.

Deep Learning-Based Early Warning Score for Predicting Clinical Deterioration in General Ward Cancer Patients

BACKGROUND

Cancer patients who are admitted to hospitals are at high risk of short-term deterioration due to treatment-related or cancer-specific complications. A rapid response system (RRS) is initiated when patients who are deteriorating or at risk of deteriorating are identified. This study was conducted to develop a deep learning-based early warning score (EWS) for cancer patients (Can-EWS) using delta values in vital signs.

METHODS

A retrospective cohort study was conducted on all oncology patients who were admitted to the general ward between 2016 and 2020. The data were divided into a training set (January 2016-December 2019) and a held-out test set (January 2020-December 2020). The primary outcome was clinical deterioration, defined as the composite of in-hospital cardiac arrest (IHCA) and unexpected intensive care unit (ICU) transfer.

RESULTS

During the study period, 19,739 cancer patients were admitted to the general wards and eligible for this study. Clinical deterioration occurred in 894 cases. IHCA and unexpected ICU transfer prevalence was 1.77 per 1000 admissions and 43.45 per 1000 admissions, respectively. We developed two models: Can-EWS V1, which used input vectors of the original five input variables, and Can-EWS V2, which used input vectors of 10 variables (including an additional five delta variables). The cross-validation performance of the clinical deterioration for Can-EWS V2 (AUROC, 0.946; 95% confidence interval [CI], 0.943-0.948) was higher than that for MEWS of 5 (AUROC, 0.589; 95% CI, 0.587-0.560; p < 0.001) and Can-EWS V1 (AUROC, 0.927; 95% CI, 0.924-0.931). As a virtual prognostic study, additional validation was performed on held-out test data. The AUROC and 95% CI were 0.588 (95% CI, 0.588-0.589), 0.890 (95% CI, 0.888-0.891), and 0.898 (95% CI, 0.897-0.899), for MEWS of 5, Can-EWS V1, and the deployed model Can-EWS V2, respectively. Can-EWS V2 outperformed other approaches for specificities, positive predictive values, negative predictive values, and the number of false alarms per day at the same sensitivity level on the held-out test data.

CONCLUSIONS

We have developed and validated a deep learning-based EWS for cancer patients using the original values and differences between consecutive measurements of basic vital signs. The Can-EWS has acceptable discriminatory power and sensitivity, with extremely decreased false alarms compared with MEWS.

Wearable Sensors for Respiration Monitoring: A Review

This paper provides an overview of flexible and wearable respiration sensors with emphasis on their significance in healthcare applications. The paper classifies these sensors based on their operating frequency distinguishing between high-frequency sensors, which operate above 10 MHz, and low-frequency sensors, which operate below this level. The operating principles of breathing sensors as well as the materials and fabrication techniques employed in their design are addressed. The existing research highlights the need for robust and flexible materials to enable the development of reliable and comfortable sensors. Finally, the paper presents potential research directions and proposes research challenges in the field of flexible and wearable respiration sensors. By identifying emerging trends and gaps in knowledge, this review can encourage further advancements and innovation in the rapidly evolving domain of flexible and wearable sensors.

Continuous monitoring of patients in and after the acute admission ward to improve clinical pathways: study protocol for a randomized controlled trial (Optimal-AAW)

BACKGROUND

Because of high demand on hospital beds, hospitals seek to reduce patients' length of stay (LOS) while preserving the quality of care. In addition to usual intermittent vital sign monitoring, continuous monitoring might help to assess the patient's risk of deterioration, in order to improve the discharge process and reduce LOS. The primary aim of this monocenter randomized controlled trial is to assess the effect of continuous monitoring in an acute admission ward (AAW) on the percentage of patients who are discharged safely.

METHODS

A total of 800 patients admitted to the AAW, for whom it is equivocal whether they can be discharged directly after their AAW stay, will be randomized to either receive usual care without (control group) or with additional continuous monitoring of heart rate, respiratory rate, posture, and activity, using a wearable sensor (sensor group). Continuous monitoring data are provided to healthcare professionals and used in the discharge decision. The wearable sensor keeps collecting data for 14 days. After 14 days, all patients fill in a questionnaire to assess healthcare use after discharge and, if applicable, their experience with the wearable sensor. The primary outcome is the difference in the percentage of patients who are safely discharged home directly from the AAW between the control and sensor group. Secondary outcomes include hospital LOS, AAW LOS, intensive care unit (ICU) admissions, Rapid Response Team calls, and unplanned readmissions within 30 days. Furthermore, facilitators and barriers for implementing continuous monitoring in the AAW and at home will be investigated.

DISCUSSION

Clinical effects of continuous monitoring have already been investigated in specific patient populations for multiple purposes, e.g., in reducing the number of ICU admissions. However, to our knowledge, this is the first Randomized Controlled Trial to investigate effects of continuous monitoring in a broad patient population in the AAW.

TRIAL REGISTRATION

https://clinicaltrials.gov/ct2/show/NCT05181111 . Registered on 6 January 2022. Start of recruitment: 7 December 2021.

The United Kingdom's National Early Warning Score: should everyone use it? A narrative review

This review critiques the benefits and drawbacks of the United Kingdom's National Early Warning Score (NEWS). Potential developments for the future are considered, as well as the role for NEWS in an emergency department (ED). The ability of NEWS to predict death within 24 h has been well validated in multiple clinical settings. It provides a common language for the assessment of clinical severity and can be used to trigger clinical interventions. However, it should not be used as the only metric for risk stratification as its ability to predict mortality beyond 24 h is not reliable and greatly influenced by other factors. The main drawbacks of NEWS are that measuring it requires trained professionals, it is time consuming and prone to calculation error. NEWS is recommended for use in acute UK hospitals, where it is linked to an escalation policy that reflects postgraduate experience; patients with lower NEWS are first assessed by a junior clinician and those with higher scores by more senior staff. This policy was based on expert opinion that did not consider workload implications. Nevertheless, its implementation has been shown to improve the efficient recording of vital signs. How and who should respond to different NEWS levels is uncertain and may vary according to the clinical setting and resources available. In the ED, simple triage scores which are quicker and easier to use may be more appropriate determinants of acuity. However, any alternative to NEWS should be easier and cheaper to use and provide evidence of outcome improvement.

Machine learning to predict in-hospital cardiac arrest from patients presenting to the emergency department

In-hospital cardiac arrest (IHCA) in the emergency department (ED) is not uncommon but often fatal. Using the machine learning (ML) approach, we sought to predict ED-based IHCA (EDCA) in patients presenting to the ED based on triage data. We retrieved 733,398 ED records from a tertiary teaching hospital over a 7 year period (Jan. 1, 2009-Dec. 31, 2015). We included only adult patients (≥ 18 y) and excluded cases presenting as out-of-hospital cardiac arrest. Primary outcome (EDCA) was identified via a resuscitation code. Patient demographics, triage data, and structured chief complaints (CCs), were extracted. Stratified split was used to divide the dataset into the training and testing cohort at a 3-to-1 ratio. Three supervised ML models were trained and performances were evaluated and compared to the National Early Warning Score 2 (NEWS2) and logistic regression (LR) model by the area under the receiver operating characteristic curve (AUC). We included 316,465 adult ED records for analysis. Of them, 636 (0.2%) developed EDCA. Of the constructed ML models, Random Forest outperformed the others with the best AUC result (0.931, 95% CI 0.911-0.949), followed by Gradient Boosting (0.930, 95% CI 0.909-0.948) and Extra Trees classifier (0.915, 95% CI 0.892-0.936). Although the differences between each of ML models and LR (AUC: 0.905, 95% CI 0.882-0.926) were not significant, all constructed ML models performed significantly better than using the NEWS2 scoring system (AUC 0.678, 95% CI 0.635-0.722). Our ML models showed excellent discriminatory performance to identify EDCA based only on the triage information. This ML approach has the potential to reduce unexpected resuscitation events if successfully implemented in the ED information system.

Clinician Perspectives on Barriers and Enablers to Implementing an Inpatient Oncology Early Warning System: A Mixed-Methods Study

PURPOSE

To elicit end-user and stakeholder perceptions regarding design and implementation of an inpatient clinical deterioration early warning system (EWS) for oncology patients to better fit routine clinical practices and enhance clinical impact.

METHODS

In an explanatory-sequential mixed-methods study, we evaluated a stakeholder-informed oncology early warning system (OncEWS) using surveys and semistructured interviews. Stakeholders were physicians, advanced practice providers (APPs), and nurses. For qualitative data, we used grounded theory and thematic content analysis via the constant comparative method to identify determinants of OncEWS implementation.

RESULTS

Survey respondents generally agreed that an oncology-focused EWS could add value beyond clinical judgment, with nurses endorsing this notion significantly more strongly than other clinicians (nurse: median 5 on a 6-point scale [6 = strongly agree], interquartile range 4-5; doctors/advanced practice providers: 4 [4-5]; P = .005). However, some respondents would not trust an EWS to identify risk accurately (n = 36 [42%] somewhat or very concerned), while others were concerned that institutional culture would not embrace such an EWS (n = 17 [28%]).Interviews highlighted important aspects of the EWS and the local context that might facilitate implementation, including (1) a model tailored to the subtleties of oncology patients, (2) transparent model information, and (3) nursing-centric workflows. Interviewees raised the importance of sepsis as a common and high-risk deterioration syndrome.

CONCLUSION

Stakeholders prioritized maximizing the degree to which the OncEWS is understandable, informative, actionable, and workflow-complementary, and perceived these factors to be key for translation into clinical benefit.

Development and Validation of the Medical Emergency Team-Risk Prediction Model for Clinical Deterioration in Acute Hospital Patients, at Time of an Emergency Admission

OBJECTIVES

To develop and validate a prediction model to estimate the risk of Medical Emergency Team (MET) review, within 48 hours of an emergency admission, using information routinely available at the time of hospital admission.

DESIGN

Development and validation of a multivariable risk model using prospectively collected data. Transparent Reporting of a multivariable model for Individual Prognosis Or Diagnosis recommendations were followed to develop and report the prediction model.

SETTING

A 560-bed teaching hospital, with a 22-bed ICU and 24-hour Emergency Department in Melbourne, Australia.

PATIENTS

A total of 45,170 emergency admissions of 30,064 adult patients (≥18 yr), with an inpatient length of stay greater than 24 hours, admitted under acute medical or surgical hospital services between 2015 and 2017.

MEASUREMENTS AND MAIN RESULTS

The outcome was MET review within 48 hours of emergency admission. Thirty candidate variables were selected from a routinely collected hospital dataset based on their availability to clinicians at the time of admission. The final model included nine variables: age; comorbid alcohol-related behavioral diagnosis; history of heart failure, chronic obstructive pulmonary disease (COPD), or renal disease; admitted from residential care; Charlson Comorbidity Index score 1 or 2, or 3+; at least one planned and one emergency admission in the last year; and admission diagnosis and one interaction (past history of COPD × admission diagnosis). The discrimination of the model was comparable in the training (C-statistics 0.82; 95% CI, 0.81-0.83) and the validation set (0.81; 0.80-0.83). Calibration was reasonable for training and validation sets.

CONCLUSIONS

Using only nine predictor variables available to clinicians at the time of admission, the MET-risk model can predict the risk of MET review during the first 48 hours of an emergency admission. Model utility in improving patient outcomes requires further investigation.

Predicting Bacteremia among Septic Patients Based on ED Information by Machine Learning Methods: A Comparative Study

Introduction: Bacteremia is a common but life-threatening infectious disease. However, a well-defined rule to assess patient risk of bacteremia and the urgency of blood culture is lacking. The aim of this study is to establish a predictive model for bacteremia in septic patients using available big data in the emergency department (ED) through logistic regression and other machine learning (ML) methods. Material and Methods: We conducted a retrospective cohort study at the ED of National Cheng Kung University Hospital in Taiwan from January 2015 to December 2019. ED adults (≥18 years old) with systemic inflammatory response syndrome and receiving blood cultures during the ED stay were included. Models I and II were established based on logistic regression, both of which were derived from support vector machine (SVM) and random forest (RF). Net reclassification index was used to determine which model was superior. Results: During the study period, 437,969 patients visited the study ED, and 40,395 patients were enrolled. Patients diagnosed with bacteremia accounted for 7.7% of the cohort. The area under the receiver operating curve (AUROC) in models I and II was 0.729 (95% CI, 0.718–0.740) and 0.731 (95% CI, 0.721–0.742), with Akaike information criterion (AIC) of 16,840 and 16,803, respectively. The performance of model II was superior to that of model I. The AUROC values of models III and IV in the validation dataset were 0.730 (95% CI, 0.713–0.747) and 0.705 (0.688–0.722), respectively. There is no statistical evidence to support that the performance of the model created with logistic regression is superior to those created by SVM and RF. Discussion: The advantage of the SVM or RF model is that the prediction model is more elastic and not limited to a linear relationship. The advantage of the LR model is that it is easy to explain the influence of the independent variable on the response variable. These models could help medical staff identify high-risk patients and prevent unnecessary antibiotic use. The performance of SVM and RF was not inferior to that of logistic regression. Conclusions: We established models that provide discrimination in predicting bacteremia among patients with sepsis. The reported results could inspire researchers to adopt ML in their development of prediction algorithms.

Performance of Contactless Respiratory Rate Monitoring by Albus HomeTM, an Automated System for Nocturnal Monitoring at Home: A Validation Study

Respiratory rate (RR) is a clinically important predictor of cardio-respiratory deteriorations. The mainstay of clinical measurement comprises the manual counting of chest movements, which is variable between clinicians and limited to sporadic readings. Emerging solutions are limited by poor adherence and acceptability or are not clinically validated. Albus Home^TM is a contactless and automated bedside system for nocturnal respiratory monitoring that overcomes these limitations. This study aimed to validate the accuracy of Albus Home compared to gold standards in real-world sleeping environments. Participants undertook overnight monitoring simultaneously using Albus Home and gold-standard polygraphy with thoraco-abdominal respiratory effort belts (SomnomedicsEU). Reference RR readings were obtained by clinician-count of polygraphy data. For both the Albus system and reference, RRs were measured in 30-s segments, reported as breaths/minute, and compared. Accuracy was defined as the percentage of RRs from the Albus system within ±2 breaths/minute of reference counts. Across a diverse validation set of 32 participants, the mean accuracy exceeded 98% and was maintained across different participant characteristics. In a Bland-Altman analysis, Albus RRs had strong agreement with reference mean differences and the limits of agreement of -0.4 and ±1.2 breaths/minute, respectively. Albus Home is a contactless yet accurate system for automated respiratory monitoring. Validated against gold -standard methods, it enables long-term, reliable nocturnal monitoring without patient burden.

Noncontact Longitudinal Respiratory Rate Measurements in Healthy Adults Using Radar-Based Sleep Monitor (Somnofy): Validation Study

BACKGROUND

Respiratory rate (RR) is arguably the most important vital sign to detect clinical deterioration. Change in RR can also, for example, be associated with the onset of different diseases, opioid overdoses, intense workouts, or mood. However, unlike for most other vital parameters, an easy and accurate measuring method is lacking.

OBJECTIVE

This study aims to validate the radar-based sleep monitor, Somnofy, for measuring RRs and investigate whether events affecting RR can be detected from personalized baselines calculated from nightly averages.

METHODS

First, RRs from Somnofy for 37 healthy adults during full nights of sleep were extensively validated against respiratory inductance plethysmography. Then, the night-to-night consistency of a proposed filtered average RR was analyzed for 6 healthy participants in a pilot study in which they used Somnofy at home for 3 months.

RESULTS

Somnofy measured RR 84% of the time, with mean absolute error of 0.18 (SD 0.05) respirations per minute, and Bland-Altman 95% limits of agreement adjusted for repeated measurements ranged from -0.99 to 0.85. The accuracy and coverage were substantially higher in deep and light sleep than in rapid eye movement sleep and wake. The results were independent of age, sex, and BMI, but dependent on supine sleeping position for some radar orientations. For nightly filtered averages, the 95% limits of agreement ranged from -0.07 to -0.04 respirations per minute. In the longitudinal part of the study, the nightly average was consistent from night to night, and all substantial deviations coincided with self-reported illnesses.

CONCLUSIONS

RRs from Somnofy were more accurate than those from any other alternative method suitable for longitudinal measurements. Moreover, the nightly averages were consistent from night to night. Thus, several factors affecting RR should be detectable as anomalies from personalized baselines, enabling a range of applications. More studies are necessary to investigate its potential in children and older adults or in a clinical setting.

Prognostic Value of the Intensive Care Respiratory Distress Observation Scale on ICU Admission

BACKGROUND

The association between dyspnea and mortality has not been demonstrated in the ICU setting. We tested the hypothesis that dyspnea (self-reported respiratory discomfort) or its observational correlates (5-item intensive care Respiratory Distress Observation Scale [IC-RDOS]) assessed on ICU admission would be associated with ICU mortality.

METHODS

Ancillary analysis of single-center data prospectively collected from 220 communicative ICU subjects allocated to a derivation cohort of 120 subjects and a separate validation cohort of 100 subjects. Dyspnea was assessed dichotomously (yes/no), with a dyspnea visual analog scale (measured in mm), and IC-RDOS was calculated. Multivariate logistic regression was used to identify factors associated with ICU and hospital mortality.

RESULTS

Dyspnea was reported by 69 (58%; median 45 [interquartile range [IQR] 32-60] mm) and 47 (47%; 38 [IQR 26-48] mm) subjects in the derivation and validation cohorts, respectively. IC-RDOS was 2.3 (1.2-3.1) and 2.4 (1.3-2.8), respectively. IC-RDOS values were higher in subjects with dyspnea than in subjects without dyspnea in both the derivation cohort (2.6 [2.2-4.6] vs 1.4 [0.9-2.4], P < .001) and the validation cohort (2.6 [2.3-4.4] vs 2.2 [1.0-2.8], P < .001). On multivariate analysis of the derivation cohort, admission for hemorrhagic shock (odds ratio 13.98), IC-RDOS (odds ratio 1.77), and Simplified Acute Physiology Score II (odds ratio 1.10) was associated with ICU mortality. Areas under the receiving operating characteristic curve of IC-RDOS to predict ICU mortality were 0.785 and 0.794 in the derivation and validation cohorts, respectively.

CONCLUSIONS

IC-RDOS, an observational correlate of dyspnea, but not dyspnea itself, was associated with higher mortality in ICU subjects.

Machine Learning-Based Cardiac Arrest Prediction for Early Warning System

The early warning system detects early and responds quickly to emergencies in high-risk patients, such as cardiac arrest in hospitalized patients. However, traditional early warning systems have the problem of frequent false alarms due to low positive predictive value and sensitivity. We conducted early prediction research on cardiac arrest using time-series data such as biosignal and laboratory data. To derive the data attributes that affect the occurrence of cardiac arrest, we performed a correlation analysis between the occurrence of cardiac arrest and the biosignal data and laboratory data. To improve the positive predictive value and sensitivity of early cardiac arrest prediction, we evaluated the performance according to the length of the time series of measured biosignal data, laboratory data, and patient data range. We propose a machine learning and deep learning algorithm: the decision tree, random forest, logistic regression, long short-term memory (LSTM), gated recurrent unit (GRU) model, and the LSTM–GRU hybrid model. We evaluated cardiac arrest prediction models. In the case of our proposed LSTM model, the positive predictive value was 85.92% and the sensitivity was 89.70%.

Nursing surveillance of respiratory adverse events among hospitalized adults: A systematic review to guide evidence-based practice

BACKGROUND

Nursing surveillance (NS) involves the purposeful, ongoing acquisition, interpretation, and synthesis of patient data for clinical decision-making. Surveillance is used to identify patients with early signs of distress and prevent adverse events. The processes that support and measure the outcomes of nursing surveillance are not clearly specified.

AIM

The aim of this systematic review was to describe the impact of NS on respiratory adverse events for adult hospitalized patients.

METHODS

The PRISMA model guided this systematic search of Academic Search Complete (EBSCOhost), CINAHL Complete (EBSCOhost), Nursing & Allied Health (ProQuest), and PubMed databases for articles published between 1990 and 2019. Search terms included nursing surveillance, data points typically attributed to nursing surveillance, adult hospitalized patients, and adverse respiratory events. The protocol for this review was registered as PROSPERO: CRD42020147557.

RESULTS

Of the 2907 references screened, 67 full-text articles were reviewed and 10 were eligible for inclusion. Research on nursing surveillance in the presence of respiratory deterioration is limited. Six studies used assessment tools that were generated from early warning scores, and four used research or institutionally designed trigger criteria. Surveillance, like other types of nursing care, was difficult to isolate and measure. Although components of surveillance were described in the selected studies, the nurse's role was not explicitly identified. Further research is required to highlight the role nursing surveillance plays in clinical decision-making to keep patients safe.

LINKING EVIDENCE TO ACTION

The attributes of NS provide a useful intervention guide for the hospitalized patient at risk of deterioration. Early warning score techniques provide empirical evidence for identifying patients at risk of deterioration. The findings of this study provide evidence of the significance for research focused on the attributes of NS relative to responding to patients at risk of deterioration.

External validation of a triage tool for predicting cardiac arrest in the emergency department

Early recognition and prevention comprise the first ring of the Chain of Survival for in-hospital cardiac arrest (IHCA). We previously developed and internally validated an emergency department (ED) triage tool, Emergency Department In-hospital Cardiac Arrest Score (EDICAS), for predicting ED-based IHCA. We aimed to externally validate this novel tool in another ED population. This retrospective cohort study used electronic clinical warehouse data from a tertiary medical center with approximately 130,000 ED visits per year. We retrieved data from 268,208 ED visits over a 2-year period. We selected one ED visit per person and excluded out-of-hospital cardiac arrest or children. Patient demographics and computerized triage information were retrieved, and the EDICAS was calculated to predict the ED-based IHCA. A total of 145,557 adult ED patients were included. Of them, 240 (0.16%) developed IHCA. The EDICAS showed excellent discrimination with an area under the receiver operating characteristic (AUROC) of 0.88. The AUROC of the EDICAS outperformed those of other early warning scores (0.80 for Modified Early Warning Score [MEWS] and 0.83 for Rapid Emergency Medicine Score [REMS]) in the same ED population. An EDICAS of 6 or above (i.e., high-risk patients) corresponded to a sensitivity of 33%, a specificity of 97%, and a positive likelihood ratio of 12.2. In conclusion, we externally validated a tool for predicting imminent IHCA in the ED and demonstrated its superior performance over other early warning scores. The real-world impact of the EDICAS warning system with appropriate interventions would require a future prospective study.

Quick Sequential Organ Failure Assessment Score and the Modified Early Warning Score for Predicting Clinical Deterioration in General Ward Patients Regardless of Suspected Infection

BACKGROUND

The quick sequential organ failure assessment (qSOFA) score is suggested to use for screening patients with a high risk of clinical deterioration in the general wards, which could simply be regarded as a general early warning score. However, comparison of unselected admissions to highlight the benefits of introducing qSOFA in hospitals already using Modified Early Warning Score (MEWS) remains unclear. We sought to compare qSOFA with MEWS for predicting clinical deterioration in general ward patients regardless of suspected infection.

METHODS

The predictive performance of qSOFA and MEWS for in-hospital cardiac arrest (IHCA) or unexpected intensive care unit (ICU) transfer was compared with the areas under the receiver operating characteristic curve (AUC) analysis using the databases of vital signs collected from consecutive hospitalized adult patients over 12 months in five participating hospitals in Korea.

RESULTS

Of 173,057 hospitalized patients included for analysis, 668 (0.39%) experienced the composite outcome. The discrimination for the composite outcome for MEWS (AUC, 0.777; 95% confidence interval [CI], 0.770-0.781) was higher than that for qSOFA (AUC, 0.684; 95% CI, 0.676-0.686; P < 0.001). In addition, MEWS was better for prediction of IHCA (AUC, 0.792; 95% CI, 0.781-0.795 vs. AUC, 0.640; 95% CI, 0.625-0.645; P < 0.001) and unexpected ICU transfer (AUC, 0.767; 95% CI, 0.760-0.773 vs. AUC, 0.716; 95% CI, 0.707-0.718; P < 0.001) than qSOFA. Using the MEWS at a cutoff of ≥ 5 would correctly reclassify 3.7% of patients from qSOFA score ≥ 2. Most patients met MEWS ≥ 5 criteria 13 hours before the composite outcome compared with 11 hours for qSOFA score ≥ 2.

CONCLUSION

MEWS is more accurate that qSOFA score for predicting IHCA or unexpected ICU transfer in patients outside the ICU. Our study suggests that qSOFA should not replace MEWS for identifying patients in the general wards at risk of poor outcome.

Usefulness of a mobile phone application for respiratory rate measurement in adult patients

AIMS

Respiratory rate measurement is one of the core nursing skills for early detection of deterioration of a patient's condition. Nevertheless, it is sometimes bothersome to visually measure respiratory rate over 1 min. Respiratory rate measurement using a mobile phone application "RRate" has been reported to be accurate and completed in a short time. However, it has only been investigated in children. The aim of this study was to validate the "RRate" compared with the 1-min method in adult patients.

METHODS

This was a cross-sectional study in the setting of a nursing school. Videos of the movement of the thorax during respiration of adult patients were made. Nursing students watched these videos and measured respiratory rate with each method. Bland-Altman analysis was used to calculate bias and limits of agreement. The times taken for the measurements were compared using a t test.

RESULTS

A total of 59 nursing students participated. When compared to the reference measurement, the one measured using "RRate" and the one measured over 1 min showed a bias of 0.40 breaths per minute and 0.65 breaths per minute, limits of agreement of -2.86 to 3.67 breaths per minute and -2.11 to 3.41 breaths per minute, respectively. The mean measurement time for "RRate" was 22.8 s (95% CI 13.9-36.6), which was significantly shorter than the 65.8 s (95% CI 61.0-73.2) for the measurement over 1 min (p < .001).

CONCLUSIONS

Respiratory rate can be measured accurately in a shorter time using a mobile phone application in adult patients.

Variation Across Hospitals in In-Hospital Cardiac Arrest Incidence Among Medicare Beneficiaries

IMPORTANCE

Although survival for in-hospital cardiac arrest (IHCA) has improved substantially over the last 2 decades, survival rates have plateaued in recent years. A better understanding of hospital differences in IHCA incidence may provide important insights regarding best practices for prevention of IHCA.

OBJECTIVE

To determine the incidence of IHCA among Medicare beneficiaries, and evaluate hospital variation in incidence of IHCA.

DESIGN, SETTING, AND PARTICIPANTS

This observational cohort study analyzes 2014 to 2017 data from 170 hospitals participating in the Get With The Guidelines-Resuscitation registry, linked to Medicare files. Participants were adults aged 65 years and older. Statistical analysis was performed from January to December 2021.

EXPOSURES

Case-mix index, teaching status, and nurse-staffing.

MAIN OUTCOMES AND MEASURES

Hospital incidence of IHCA among Medicare beneficiaries was estimated as the number of IHCA patients divided by the total number of hospital admissions. Multivariable hierarchical regression models were used to calculate hospital incidence rates adjusted for differences in patient case-mix and evaluate the association of hospital variables with IHCA incidence.

RESULTS

Among a total of 4.5 million admissions at 170 hospitals, 38 630 patients experienced an IHCA during 2014 to 2017. Among the 38 630 patients with IHCAs, 7571 (19.6%) were non-Hispanic Black, 26 715 (69.2%) were non-Hispanic White, and 16 732 (43.3%) were female; the mean (SD) age at admission was 76.3 (7.8) years. The median risk-adjusted IHCA incidence was 8.5 per 1000 admissions (95% CI, 8.2-9.0 per 1000 admissions). After adjusting for differences in case-mix index, IHCA incidence varied markedly across hospitals ranging from 2.4 per 1000 admissions to 25.5 per 1000 admissions (IQR, 6.6-11.4; median odds ratio, 1.51 [95% CI, 1.44-1.58]). Among hospital variables, a higher case-mix index, higher nurse staffing, and teaching status were associated with a lower hospital incidence of IHCA.

CONCLUSIONS AND RELEVANCE

This cohort study found that the incidence of IHCA varies markedly across hospitals, and hospitals with higher nurse staffing and teaching status had lower IHCA incidence rates. Future studies are needed to better understand processes of care at hospitals with exceptionally low IHCA incidence to identify best practices for cardiac arrest prevention.

Continuous Monitoring of Vital Signs With Wearable Sensors During Daily Life Activities: Validation Study

Background

Continuous telemonitoring of vital signs in a clinical or home setting may lead to improved knowledge of patients’ baseline vital signs and earlier detection of patient deterioration, and it may also facilitate the migration of care toward home. Little is known about the performance of available wearable sensors, especially during daily life activities, although accurate technology is critical for clinical decision-making.

Objective

The aim of this study is to assess the data availability, accuracy, and concurrent validity of vital sign data measured with wearable sensors in volunteers during various daily life activities in a simulated free-living environment.

Methods

Volunteers were equipped with 4 wearable sensors (Everion placed on the left and right arms, VitalPatch, and Fitbit Charge 3) and 2 reference devices (Oxycon Mobile and iButton) to obtain continuous measurements of heart rate (HR), respiratory rate (RR), oxygen saturation (SpO₂), and temperature. Participants performed standardized activities, including resting, walking, metronome breathing, chores, stationary cycling, and recovery afterward. Data availability was measured as the percentage of missing data. Accuracy was evaluated by the median absolute percentage error (MAPE) and concurrent validity using the Bland-Altman plot with mean difference and 95% limits of agreement (LoA).

Results

A total of 20 volunteers (median age 64 years, range 20-74 years) were included. Data availability was high for all vital signs measured by VitalPatch and for HR and temperature measured by Everion. Data availability for HR was the lowest for Fitbit (4807/13,680, 35.14% missing data points). For SpO₂ measured by Everion, median percentages of missing data of up to 100% were noted. The overall accuracy of HR was high for all wearable sensors, except during walking. For RR, an overall MAPE of 8.6% was noted for VitalPatch and that of 18.9% for Everion, with a higher MAPE noted during physical activity (up to 27.1%) for both sensors. The accuracy of temperature was high for VitalPatch (MAPE up to 1.7%), and it decreased for Everion (MAPE from 6.3% to 9%). Bland-Altman analyses showed small mean differences of VitalPatch for HR (0.1 beats/min [bpm]), RR (−0.1 breaths/min), and temperature (0.5 °C). Everion and Fitbit underestimated HR up to 5.3 (LoA of −39.0 to 28.3) bpm and 11.4 (LoA of −53.8 to 30.9) bpm, respectively. Everion had a small mean difference with large LoA (−10.8 to 10.4 breaths/min) for RR, underestimated SpO₂ (>1%), and overestimated temperature up to 2.9 °C.

Conclusions

Data availability, accuracy, and concurrent validity of the studied wearable sensors varied and differed according to activity. In this study, the accuracy of all sensors decreased with physical activity. Of the tested sensors, VitalPatch was found to be the most accurate and valid for vital signs monitoring.

In-hospital Cardiac Arrest in Patients With Sepsis: A National Cohort Study

Background: Little is known about the risk of in-hospital cardiac arrest (IHCA) among patients with sepsis. We aimed to characterize the incidence and outcome of IHCA among patients with sepsis in a national database. We then determined the major risk factors associated with IHCA among sepsis patients.

Methods: We used data from a population-based cohort study based on the National Health Insurance Research Database of Taiwan (NHRID) between 2000 and 2013. We used Martin's implementation that combined the explicit ICD-9 CM codes for sepsis and six major organ dysfunction categories. IHCA among sepsis patients was identified by the presence of cardiopulmonary resuscitation procedures. The survival impact was analyzed with the Cox proportional-hazards model using inverse probability of treatment weighting (IPTW). The risk factors were identified by logistic regression models with 10-fold cross-validation, adjusting for competing risks.

Results: We identified a total of 20,022 patients with sepsis, among whom 2,168 developed in-hospital cardiac arrest. Sepsis patients with a higher burden of comorbidities and organ dysfunction were more likely to develop in-hospital cardiac arrest. Acute respiratory failure, hematological dysfunction, renal dysfunction, and hepatic dysfunction were associated with increased risk of IHCA. Regarding the source of infection, patients with respiratory tract infections were at the highest risk, whereas patients with urinary tract infections and primary bacteremia were less likely to develop IHCA. The risk of IHCA correlated well with age and revised cardiac risk index (RCRI). The final competing risk model concluded that acute respiratory failure, male gender, and diabetes are the three strongest predictors for IHCA. The effect of IHCA on survival can last 1 year after hospital discharge, with an IPTW-weighted hazard ratio of 5.19 (95% CI: 5.06, 5.35) compared to patients who did not develop IHCA.

Conclusion: IHCA in sepsis patients had a negative effect on both short- and long-term survival. The risk of IHCA among hospitalized sepsis patients was strongly correlated with age and cardiac risk index. The three identified risk factors can help clinicians to identify patients at higher risk for IHCA.

Intelligent patient monitoring for proactive alerting of key personnel in intensive care: A single-center study

A code blue event is an emergency code to indicate when a patient goes into cardiac arrest and needs resuscitation. In this paper, we model the binary response of a intensive care unit (ICU) patients experiencing a code-blue event, starting with vital time-series data of patients in 12 ICU beds. Our study introduces day-of and day-ahead risk scoring models trained against ground truth information on per-patient-per-day code-blue events, starting with multi-variate vital-time-series-sequences of varying durations with a plurality of engineered features capturing temporal variations of these signals. Actionable events, including code-blue events, aggregated by patient by day were predicted on the day-of or day-ahead with an overall accuracy of over 80% in our best models. Such models have potential to improve healthcare delivery by providing just-in-time alerting, enabling proactive and preventative clinical interventions, through continuous patient monitoring.

Clinical evaluation of a wearable sensor for mobile monitoring of respiratory rate on hospital wards

A wireless and wearable system was recently developed for mobile monitoring of respiratory rate (RR). The present study was designed to compare RR mobile measurements with reference capnographic measurements on a medical-surgical ward. The wearable sensor measures impedance variations of the chest from two thoracic and one abdominal electrode. Simultaneous measurements of RR from the wearable sensor and from the capnographic sensor (1 measure/minute) were compared in 36 ward patients. Patients were monitored for a period of 182 ± 56 min (range 68–331). Artifact-free RR measurements were available 81% of the monitoring time for capnography and 92% for the wearable monitoring system (p < 0.001). A total of 4836 pairs of simultaneous measurements were available for analysis. The average reference RR was 19 ± 5 breaths/min (range 6–36). The average difference between the wearable and capnography RR measurements was − 0.6 ± 2.5 breaths/min. Error grid analysis showed that the proportions of RR measurements done with the wearable system were 89.7% in zone A (no risk), 9.6% in zone B (low risk) and < 1% in zones C, D and E (moderate, significant and dangerous risk). The wearable method detected RR values > 20 (tachypnea) with a sensitivity of 81% and a specificity of 93%. In ward patients, the wearable sensor enabled accurate and precise measurements of RR within a relatively broad range (6–36 b/min) and the detection of tachypnea with high sensitivity and specificity.

Prediction of In-Hospital Cardiac Arrest Using Shallow and Deep Learning

Sudden cardiac arrest can leave serious brain damage or lead to death, so it is very important to predict before a cardiac arrest occurs. However, early warning score systems including the National Early Warning Score, are associated with low sensitivity and false positives. We applied shallow and deep learning to predict cardiac arrest to overcome these limitations. We evaluated the performance of the Synthetic Minority Oversampling Technique Ratio. We evaluated the performance using a Decision Tree, a Random Forest, Logistic Regression, Long Short-Term Memory model, Gated Recurrent Unit model, and LSTM–GRU hybrid models. Our proposed Logistic Regression demonstrated a higher positive predictive value and sensitivity than traditional early warning systems.

Technical considerations for evaluating clinical prediction indices: a case study for predicting code blue events with MEWS

Objective.There have been many efforts to develop tools predictive of health deterioration in hospitalized patients, but comprehensive evaluation of their predictive ability is often lacking to guide implementation in clinical practice. In this work, we propose new techniques and metrics for evaluating the performance of predictive alert algorithms and illustrate the advantage of capturing the timeliness and the clinical burden of alerts through the example of the modified early warning score (MEWS) applied to the prediction of in-hospital code blue events.Approach. Different implementations of MEWS were calculated from available physiological parameter measurements collected from the electronic health records of ICU adult patients. The performance of MEWS was evaluated using conventional and a set of non-conventional metrics and approaches that take into account the timeliness and practicality of alarms as well as the false alarm burden.Main results. MEWS calculated using the worst-case measurement (i.e. values scoring 3 points in the MEWS definition) over 2 h intervals significantly reduced the false alarm rate by over 50% (from 0.19/h to 0.08/h) while maintaining similar sensitivity levels as MEWS calculated from raw measurements (∼80%). By considering a prediction horizon of 12 h preceding a code blue event, a significant improvement in the specificity (∼60%), the precision (∼155%), and the work-up to detection ratio (∼50%) could be achieved, at the cost of a relatively marginal decrease in sensitivity (∼10%).Significance. Performance aspects pertaining to the timeliness and burden of alarms can aid in understanding the potential utility of a predictive alarm algorithm in clinical settings.

Validation of a New Contactless and Continuous Respiratory Rate Monitoring Device Based on Ultra-Wideband Radar Technology

Respiratory rate (RR) is typically the first vital sign to change when a patient decompensates. Despite this, RR is often monitored infrequently and inaccurately. The Circadia Contactless Breathing Monitor™ (model C100) is a novel device that uses ultra-wideband radar to monitor RR continuously and un-obtrusively. Performance of the Circadia Monitor was assessed by direct comparison to manually scored reference data. Data were collected across a range of clinical and non-clinical settings, considering a broad range of user characteristics and use cases, in a total of 50 subjects. Bland-Altman analysis showed high agreement with the gold standard reference for all study data, and agreement fell within the predefined acceptance criteria of ±5 breaths per minute (BrPM). The 95% limits of agreement were -3.0 to 1.3 BrPM for a nonprobability sample of subjects while awake, -2.3 to 1.7 BrPM for a clinical sample of subjects while asleep, and -1.2 to 0.7 BrPM for a sample of healthy subjects while asleep. Accuracy rate, using an error margin of ±2 BrPM, was found to be 90% or higher. Results demonstrate that the Circadia Monitor can effectively and efficiently be used for accurate spot measurements and continuous bedside monitoring of RR in low acuity settings, such as the nursing home or hospital ward, or for remote patient monitoring.

In-hospital cardiac arrest and preceding National Early Warning Score (NEWS): A retrospective case-control study

We aimed to describe and evaluate the National Early Warning Score (NEWS) in the 24 hours preceding an in-hospital cardiac arrest among general somatic ward patients.The 24 hours preceding the in-hospital cardiac arrest were divided into four timespans and analysed by a medical record review of 127:254 matched case-control patients. The median NEWS ranged from 3 (2-6) to 6 (3-9) points for cases vs 1 (0-3) to 1 (0-3) point for controls. The proportion of cases ranged from 23-45% at high risk vs 3-6% for controls. The NEWS high-risk category was associated with an increase of 3.17 (95% confidence interval (CI) 1.66-6.04) to 4.43 (95% CI 2.56-7.67) in odds of in-hospital cardiac arrest compared to the low-risk category.NEWS, with its intuitive and for healthcare staff easy to interpret risk classification, is suitable for discriminating deteriorating patients with major deviating vital signs scoring high risk on NEWS.

The EMS Modified Early Warning Score (EMEWS): A Simple Count of Vital Signs as a Predictor of Out-of-Hospital Cardiac Arrests

Objective: For patients at risk for out-of-hospital cardiac arrest (OHCA) after Emergency Medical Services (EMS) arrival, outcomes may be mitigated by identifying impending arrests and intervening before they occur. Tools such as the Modified Early Warning Score (MEWS) have been developed to determine the risk of arrest, but involve relatively complicated algorithms that can be impractical to compute in the prehospital environment. A simple count of abnormal vital signs, the "EMS Modified Early Warning Score" (EMEWS), may represent a more practical alternative. We sought to compare to the ability of MEWS and EMEWS to identify patients at risk for EMS-witnessed OHCA.Methods: We conducted a retrospect analysis of the 2018 ESO Data Collaborative database of EMS encounters. Patients without cardiac arrest before EMS arrival were categorized into those who did or did not have an EMS-witnessed arrest. MEWS was evaluated without its temperature component (MEWS-T). The performance of MEWS-T and EMEWS in predicting EMS witnessed arrest was evaluated by comparing receiver-operating characteristic curves.Results: Of 369,064 included encounters, 4,651 were EMS witnessed arrests. MEWS-T demonstrated an area under the curve (AUC) of 0.79 (95% CI: 0.79 - 0.80), with 86.8% sensitivity and 51.0% specificity for MEWS-T ≥ 3. EMEWS demonstrated an AUC of 0.74 (95% CI: 0.73 - 0.75), with 81.3% sensitivity and 53.9% specificity for EMEWS ≥ 2.Conclusions: EMEWS showed a similar ability to predict EMS-witnessed cardiac arrest compared to MEWS-T, despite being significantly simpler to compute. Further study is needed to evaluate whether the implementation of EMEWS can aid EMS clinicians in anticipating and preventing OHCA.

Machine learning for early prediction of in-hospital cardiac arrest in patients with acute coronary syndromes

Background

Previous studies have used machine leaning to predict clinical deterioration to improve outcome prediction. However, no study has used machine learning to predict cardiac arrest in patients with acute coronary syndrome (ACS). Algorithms are required to generate high‐performance models for predicting cardiac arrest in ACS patients with multivariate features.

Hypothesis

Machine learning algorithms will significantly improve outcome prediction of cardiac arrest in ACS patients.

Methods

This retrospective cohort study reviewed 166 ACS patients who had in‐hospital cardiac arrest. Eight machine learning algorithms were trained using multivariate clinical features obtained 24 h prior to the onset of cardiac arrest. All machine learning models were compared to each other and to existing risk prediction scores (Global Registry of Acute Coronary Events, National Early Warning Score, and Modified Early Warning Score) using the area under the receiver operating characteristic curve (AUROC).

Results

The XGBoost model provided the best performance with regard to AUC (0.958 [95%CI: 0.938–0.978]), accuracy (88.9%), sensitivity (73%), negative predictive value (89%), and F1 score (80%) compared with other machine learning models. The K‐nearest neighbor model generated the best specificity (99.3%) and positive predictive value (93.8%) metrics, but had low and unacceptable values for sensitivity and AUC. Most, but not all, machine learning models outperformed the existing risk prediction scores.

Conclusions

The XGBoost model, which was generated based on a machine learning algorithm, has high potential to be used to predict cardiac arrest in ACS patients. This proposed model significantly improves outcome prediction compared to existing risk prediction scores.

Research protocol for the validation of a new portable technology for real-time continuous monitoring of Early Warning Score (EWS) in hospital practice and for an early-stage multistakeholder assessment

INTRODUCTION

The real-time continuous monitoring of vital parameters in patients affected by multiple chronic conditions and/or COVID-19 can lead to several benefits to the Italian National Healthcare System (IT-NHS). The UBiquitous Integrated CARE (UBICARE) technology is a novel health digital platform at the validation stage in hospital setting. UBICARE might support the urgent need for digitalisation and early intervention, as well as minimise the face-to-face delivery of care in both hospital and community-based care settings. This research protocol aims to design an early-stage assessment of the multidimensional impact induced by UBICARE within the IT-NHS alongside technology validation in a hospital ward.

METHODS AND ANALYSIS

The targeted patients will be medium/high-risk hypertensive individuals as an illustrative first example of how UBICARE might bring benefits to susceptible patients. A mixed-method study will be applied to incorporate to the validation study a multistakeholder perspective, including perceived patient experiences and preferences, and facilitate technology adoption. First, semistructured interviews will be undertaken with a variety of stakeholders including clinicians, health managers and policy-makers to capture views on the likely technology utility, economic sustainability, impact of adoption in hospital practice and alternative adoption scenarios. Second, a monocentric, non-randomised and non-comparative clinical study, supplemented by the administration of standardised usability questionnaires to patients and health professionals, will validate the use of UBICARE in hospital practice. Finally, the results of the previous stages will be discussed in a multidisciplinary-facilitated workshop with IT-NHS relevant stakeholders to reconcile stakeholders' perspectives. Limitations include a non-random recruitment strategy in the clinical study, small sample size of the key stakeholders and potential stakeholder recruitment bias introduced by the research technique.

ETHICS AND DISSEMINATION

The Ethics Committee for Clinical Experimentation of Tuscany Region approved the protocol. Participation in this study is voluntary. Study results will be disseminated through peer-reviewed publications and academic conferences.

The Importance of Respiratory Rate Monitoring: From Healthcare to Sport and Exercise

Respiratory rate is a fundamental vital sign that is sensitive to different pathological conditions (e.g., adverse cardiac events, pneumonia, and clinical deterioration) and stressors, including emotional stress, cognitive load, heat, cold, physical effort, and exercise-induced fatigue. The sensitivity of respiratory rate to these conditions is superior compared to that of most of the other vital signs, and the abundance of suitable technological solutions measuring respiratory rate has important implications for healthcare, occupational settings, and sport. However, respiratory rate is still too often not routinely monitored in these fields of use. This review presents a multidisciplinary approach to respiratory monitoring, with the aim to improve the development and efficacy of respiratory monitoring services. We have identified thirteen monitoring goals where the use of the respiratory rate is invaluable, and for each of them we have described suitable sensors and techniques to monitor respiratory rate in specific measurement scenarios. We have also provided a physiological rationale corroborating the importance of respiratory rate monitoring and an original multidisciplinary framework for the development of respiratory monitoring services. This review is expected to advance the field of respiratory monitoring and favor synergies between different disciplines to accomplish this goal.

Early warning scores for detecting deterioration in adult hospital patients: systematic review and critical appraisal of methodology

OBJECTIVE

To provide an overview and critical appraisal of early warning scores for adult hospital patients.

DESIGN

Systematic review.

DATA SOURCES

Medline, CINAHL, PsycInfo, and Embase until June 2019.

ELIGIBILITY CRITERIA FOR STUDY SELECTION

Studies describing the development or external validation of an early warning score for adult hospital inpatients.

RESULTS

13 171 references were screened and 95 articles were included in the review. 11 studies were development only, 23 were development and external validation, and 61 were external validation only. Most early warning scores were developed for use in the United States (n=13/34, 38%) and the United Kingdom (n=10/34, 29%). Death was the most frequent prediction outcome for development studies (n=10/23, 44%) and validation studies (n=66/84, 79%), with different time horizons (the most frequent was 24 hours). The most common predictors were respiratory rate (n=30/34, 88%), heart rate (n=28/34, 83%), oxygen saturation, temperature, and systolic blood pressure (all n=24/34, 71%). Age (n=13/34, 38%) and sex (n=3/34, 9%) were less frequently included. Key details of the analysis populations were often not reported in development studies (n=12/29, 41%) or validation studies (n=33/84, 39%). Small sample sizes and insufficient numbers of event patients were common in model development and external validation studies. Missing data were often discarded, with just one study using multiple imputation. Only nine of the early warning scores that were developed were presented in sufficient detail to allow individualised risk prediction. Internal validation was carried out in 19 studies, but recommended approaches such as bootstrapping or cross validation were rarely used (n=4/19, 22%). Model performance was frequently assessed using discrimination (development n=18/22, 82%; validation n=69/84, 82%), while calibration was seldom assessed (validation n=13/84, 15%). All included studies were rated at high risk of bias.

CONCLUSIONS

Early warning scores are widely used prediction models that are often mandated in daily clinical practice to identify early clinical deterioration in hospital patients. However, many early warning scores in clinical use were found to have methodological weaknesses. Early warning scores might not perform as well as expected and therefore they could have a detrimental effect on patient care. Future work should focus on following recommended approaches for developing and evaluating early warning scores, and investigating the impact and safety of using these scores in clinical practice.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42017053324.

Predictive powers of the Modified Early Warning Score and the National Early Warning Score in general ward patients who activated the medical emergency team

Background

The current early warning scores may be insufficient for medical emergency teams (METs) to use in assessing the severity and the prognosis of activated patients. We evaluated the predictive powers of the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS) for 28-day mortality and to analyze predictors of 28-day mortality in general ward patients who activate the MET.

Methods

Adult general ward inpatients who activated the MET in a tertiary referral teaching hospital between March 2009 and December 2016 were included. The demographic and clinical characteristics and physiologic parameters at the time of MET activation were collected, and MEWS and NEWS were calculated.

Results

A total of 6,729 MET activation events were analyzed. Patients who died within 28 days were younger (mean age 60 vs 62 years), were more likely to have malignancy (72% vs 53%), were more likely to be admitted to the medical department rather than the surgical department (93% vs 80%), had longer intervals from admission to MET activation (median, 7 vs 5 days), and were less likely to activate the MET during nighttime hours (5 PM to 8 AM) (61% vs 66%) compared with those who did not die within 28 days (P < 0.001 for all comparisons). The areas under the receiver operating characteristic curves of MEWS and NEWS for 28-day mortality were 0.58 (95% CI, 0.56–0.59) and 0.60 (95% CI, 0.59–0.62), which were inferior to that of the logistics regression model (0.73; 95% CI, 0.72–0.74; P < 0.001 for both comparisons).

Conclusions

Both the MEWS and NEWS had poor predictive powers for 28-day mortality in patients who activated the MET. A new scoring system is needed to stratify the severity and prognosis of patients who activated the MET.

Characteristics of Rapid Response Calls in the United States: An Analysis of the First 402,023 Adult Cases From the Get With the Guidelines Resuscitation-Medical Emergency Team Registry

OBJECTIVES

To characterize the rapid response team activations, and the patients receiving them, in the American Heart Association-sponsored Get With The Guidelines Resuscitation-Medical Emergency Team cohort between 2005 and 2015.

DESIGN

Retrospective multicenter cohort study.

SETTING

Three hundred sixty U.S. hospitals.

PATIENTS

Consecutive adult patients experiencing rapid response team activation.

INTERVENTIONS

Rapid response team activation.

MEASUREMENTS AND MAIN RESULTS

The cohort included 402,023 rapid response team activations from 347,401 unique healthcare encounters. Respiratory triggers (38.0%) and cardiac triggers (37.4%) were most common. The most frequent interventions-pulse oximetry (66.5%), other monitoring (59.6%), and supplemental oxygen (62.0%)-were noninvasive. Fluids were the most common medication ordered (19.3%), but new antibiotic orders were rare (1.2%). More than 10% of rapid response teams resulted in code status changes. Hospital mortality was over 14% and increased with subsequent rapid response activations.

CONCLUSIONS

Although patients requiring rapid response team activation have high inpatient mortality, most rapid response team activations involve relatively few interventions, which may limit these teams' ability to improve patient outcomes.

Are current wireless monitoring systems capable of detecting adverse events in high-risk surgical patients? A descriptive study

BACKGROUND

Adverse events are common in high-risk surgical patients, but early detection is difficult. Recent innovations have resulted in wireless and 'wearable' sensors, which may capture patient deterioration at an early stage, but little is known regarding their ability to timely detect events. The objective of this study is to describe the ability of currently available wireless sensors to detect adverse events in high-risk patients.

METHODS

A descriptive analysis was performed of all vital signs trend data obtained during an observational comparison study of wearable sensors for vital signs monitoring in high-risk surgical patients during the initial days of recovery at a surgical step-down unit (SDU) and subsequent traumatology or surgical oncology ward. Heart rate (HR), respiratory rate (RR) and oxygen saturation (SpO₂) were continuously recorded. Vital sign trend patterns of patients that developed adverse events were described and compared to vital sign recordings of patients without occurrence of adverse events. Two wearable patch sensors were used (SensiumVitals and HealthPatch), a bed-based mattress sensor (EarlySense) and a patient-worn monitor (Masimo Radius-7).

RESULTS

Twenty adverse events occurred in 11 of the 31 patients included. Atrial fibrillation (AF) was most common (20%). The onset of AF was recognizable as a sudden increase in HR in all recordings, and all patients with new-onset AF after esophagectomy developed other postoperative complications. Patients who developed respiratory insufficiency showed an increase in RR and a decrease in SpO₂, but an increase in HR was not always visible. In patients without adverse events, temporary periods of high HR and RR are observed as well, but these were transient and less frequent.

CONCLUSIONS

Current systems for remote wireless patient monitoring on the ward are capable of detecting abnormalities in vital sign patterns in patients who develop adverse events. Remote patient monitoring may have potential to improve patient safety by generating early warnings for deterioration to nursing staff.

Risk Factors for Mortality in Severely Ill Children Admitted to a Tertiary Referral Hospital in Malawi

In low-resource settings, many children are severely ill at arrival to hospital. The risk factors for mortality among such ill children are not well-known. Understanding which of these patients are at the highest risk could assist in the allocation of limited resources to where they are most needed. A cohort study of severely ill children treated in the resuscitation room of the pediatric emergency department at Queen Elizabeth Central Hospital in Malawi was conducted over a 6-month period in 2017. Data on signs and symptoms, vital signs, blood glucose levels, and nutritional status were collected and linked with in-hospital mortality data. The factors associated with in-hospital mortality were analyzed using multivariable logistic regression. Data for 1,359 patients were analyzed and 118 (8.7%) patients died. The following factors were associated with mortality: presence of any severely deranged vital sign, unadjusted odds ratio (UOR) 2.6 (95% CI 1.7–4.0) and adjusted odds ratio (AOR) 3.2 (95% CI 2.0–5.0); severe dehydration, UOR 2.6 (1.4–5.1) and AOR 2.8 (1.3–6.0); hypoglycemia glycemia (< 5 mmol/L), UOR 3.6 (2.2–5.8) and AOR 2.7 (1.6–4.7); and severe acute malnutrition, UOR 5.8 (3.5–9.6) and AOR 5.7 (3.3–10.0). This study suggests that among severely sick children, increased attention should be given to those with hypo/low glycemia, deranged vital signs, malnutrition, and severe dehydration to avert mortality among these high-risk patients.

Evaluation of a new respiratory monitoring tool "Early Warning ScoreO2" for patients admitted at the emergency department with dyspnea

BACKGROUND

Many scores derived from Early Warning Scores have been developed to detect patients at risk of poor outcome. Few of these scores incorporate the oxygen flow rate while this is a major marker in patients with respiratory complaint. We developed and evaluated a new automatable monitoring tool (Early Warning Score O₂: EWS.O₂) that incorporates cardio-respiratory parameters (Respiratory rate, Heart rate, SpO₂, and FiO₂ derived from oxygen flow rate), aiming to achieve early detection of poor outcome among patients with dyspnea.

METHODS

All patients presenting at an emergency department for dyspnea from June 2011 to June 2018 with available initial value (nurse triage) of respiratory parameters were included. Our primary endpoint was a composite criterion including the use of non-invasive ventilation, ICU admission and death. The Area under the Receiver Operating Characteristic curve (AUROC) of the SpO₂/FiO₂ index, NEWS, NEWS2, and the EWS.O₂ were compared, including in subgroup analysis by final diagnosis or oxygen supplementation.

RESULTS

Among the 1729 patients retrieved, the composite outcome was observed in 288 (16.7%). The EWS.O₂ displayed better or comparable predictive accuracy at triage (AUROC: 0.704, 95% CI 0.672-0.736) compared to NEWS (0.662, p < 0.01), NEWS2 (0.672, p = 0.02) and SpO₂/FiO₂ (0.695, p = 0.46).

CONCLUSIONS

This new ScoreO₂ is equivalent or superior to common early warning scores and index to predict poor outcome at first medical contact. This score may be automatically and continuously recorded with new closed-loop devices to titrate oxygen flow. Further prospective studies will allow to verify its accuracy at multiple time points of the patient's journey.

Remote wireless vital signs monitoring on the ward for early detection of deteriorating patients: A case series

INTRODUCTION

Remote wireless monitoring is a new technology that allows the continuous recording of ward patients' vital signs, supporting nurses by measuring vital signs frequently and accurately. A case series is presented to illustrate how these systems might contribute to improved patient surveillance.

METHODS AND RESULTS

Five hospitals in three European countries installed a remote wireless vital signs monitoring system on medical or surgical wards. Heart rate, respiratory rate and temperature were measured by the system every 2 min. Four cases of (paroxysmal) atrial fibrillation are presented, two cases of sepsis and one case each of pyrexia, cardiogenic pulmonary edema and pulmonary embolisms. All cases show that the remote monitoring system revealed the first signs of ventilatory and circulatory deterioration before a change in the trends of the respective values became obvious by manual vital signs measurement.

DISCUSSION

This case series illustrates that a wireless remote vital signs monitoring system on medical and surgical wards has the potential to reduce time to detect deteriorating patients.

Performance of patient acuity rating by rapid response team nurses for predicting short-term prognosis

Background

Although scoring and machine learning methods have been developed to predict patient deterioration, bedside assessment by nurses should not be overlooked. This study aimed to evaluate the performance of subjective bedside assessment of the patient by the rapid response team (RRT) nurses in predicting short-term patient deterioration.

Methods

Patients noticed by RRT nurses based on the vital sign instability, abnormal laboratory results, and direct contact via phone between November 1, 2016, and December 12, 2017, were included. Five RRT nurses visited the patients according to their shifts and assessed the possibility of patient deterioration. Patient acuity rating (PAR), a scale of 1–7, was used as the tool of bedside assessment. Other scores, including the modified early warning score, VitalPAC early warning score, standardised early warning score, and cardiac arrest risk triage, were calculated afterwards. The performance of these scores in predicting mortality and/or intensive care unit admission within 1 day was compared by calculating the area under the receiver operating curve.

Results

A total of 1,426 patients were included in the study, of which 258 (18.1%) died or were admitted to the intensive care unit within 1 day. The area under the receiver operating curve of PAR was 0.87 (95% confidence interval [CI] 0.84–0.89), which was higher than those of modified early warning score (0.66, 95% CI 0.62–0.70), VitalPAC early warning score (0.69, 95% CI 0.66–0.73), standardised early warning score (0.67, 95% CI 0.63–0.70) and cardiac arrest risk triage (0.63, 95% CI 0.59–0.66) (P<0.001).

Conclusions

PAR assessed by RRT nurses can be a useful tool for assessing short-term patient prognosis in the RRT setting.