Improving ICU Risk Predictive Models Through Automation Designed for Resiliency Against Documentation Bias.

OBJECTIVES: Electronic health records enable automated data capture for risk models but may introduce bias. We present the Philips Critical Care Outcome Prediction Model (CCOPM) focused on addressing model features sensitive to data drift to improve benchmarking ICUs on mortality performance. DESIGN: Retrospective, multicenter study of ICU patients randomized in 3:2 fashion into development and validation cohorts. Generalized additive models (GAM) with features designed to mitigate biases introduced from documentation of admission diagnosis, Glasgow Coma Scale (GCS), and extreme vital signs were developed using clinical features representing the first 24 hours of ICU admission. SETTING: eICU Research Institute database derived from ICUs participating in the Philips eICU telecritical care program. PATIENTS: A total of 572,985 adult ICU stays discharged from the hospital between January 1, 2017, and December 31, 2018, were included, yielding 509,586 stays in the final cohort; 305,590 and 203,996 in development and validation cohorts, respectively. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Model discrimination was compared against Acute Physiology and Chronic Health Evaluation (APACHE) IVa/IVb models on the validation cohort using the area under the receiver operating characteristic (AUROC) curve. Calibration assessed by actual/predicted ratios, calibration-in-the-large statistics, and visual analysis. Performance metrics were further stratified by subgroups of admission diagnosis and ICU characteristics. Historic data from two health systems with abrupt changes in Glasgow Coma Scale (GCS) documentation were assessed in the year prior to and after data shift. CCOPM outperformed APACHE IVa/IVb for ICU mortality (AUROC, 0.925 vs 0.88) and hospital mortality (AUROC, 0.90 vs 0.86). Better calibration performance was also attained among subgroups of different admission diagnoses, ICU types, and over unique ICU-years. The CCOPM provided more stable predictions compared with APACHE IVa within an external cohort of greater than 120,000 patients from two health systems with known changes in GCS documentation. CONCLUSIONS: These mortality risk models demonstrated excellent performance compared with APACHE while appearing to mitigate bias introduced through major shifts in GCS documentation at two large health systems. This provides evidence to support using automated capture rather than trained personnel for capture of GCS data used in benchmarking ICUs on mortality performance.

Optimisation of clinical workflow and monitor settings safely reduces alarms in the NICU.

AIM: To address alarm fatigue, a new alarm management system which ensures a quicker delivery of alarms together with waveform information on nurses' handheld devices was implemented and settings optimised. The effects of this clinical implementation on alarm rates and nurses' responsiveness were measured in an 18-bed single family rooms neonatal intensive care unit (NICU). METHODS: The technical implementation of the alarm management system was followed by clinical workflow optimisation. Alarms and vital parameters from October 2017 to December 2019 were analysed. Measures included monitoring alarms, nurses' response to alarms and time spent by patients in different saturation ranges. A survey among nurses was performed to evaluate changes in alarm rate and use of protocols. RESULTS: A significant reduction of monitoring alarms per patient days was detected after the optimisation phase (in particular for SpO2 ≤ 80%, P < .001). More time was spent by infants within the optimal peripheral oxygen saturation range (88% < SpO2 < 95%, P < .001). Results from the surveys showed that false alarms are less likely to cause an inappropriate response after the optimisation phase. CONCLUSION: The implementation of an alarm management solution and an optimisation programme can safely reduce the alarm burden inside of the NICU environment.

Impact of a computerized decision support tool deployed in two intensive care units on acute kidney injury progression and guideline compliance: a prospective observational study.

BACKGROUND: Acute kidney injury (AKI) affects a large proportion of the critically ill and is associated with worse patient outcomes. Early identification of AKI can lead to earlier initiation of supportive therapy and better management. In this study, we evaluate the impact of computerized AKI decision support tool integrated with the critical care clinical information system (CCIS) on patient outcomes. Specifically, we hypothesize that integration of AKI guidelines into CCIS will decrease the proportion of patients with Stage 1 AKI deteriorating into higher stages of AKI. METHODS: The study was conducted in two intensive care units (ICUs) at University Hospitals Bristol, UK, in a before (control) and after (intervention) format. The intervention consisted of the AKIN guidelines and AKI care bundle which included guidance for medication usage, AKI advisory and dashboard with AKI score. Clinical data and patient outcomes were collected from all patients admitted to the units. AKI stage was calculated using the Acute Kidney Injury Network (AKIN) guidelines. Maximum AKI stage per admission, change in AKI stage and other metrics were calculated for the cohort. Adherence to eGFR-based enoxaparin dosing guidelines was evaluated as a proxy for clinician awareness of AKI. RESULTS: Each phase of the study lasted a year, and a total of 5044 admissions were included for analysis with equal numbers of patients for the control and intervention stages. The proportion of patients worsening from Stage 1 AKI decreased from 42% (control) to 33.5% (intervention), p = 0.002. The proportion of incorrect enoxaparin doses decreased from 1.72% (control) to 0.6% (intervention), p < 0.001. The prevalence of any AKI decreased from 43.1% (control) to 37.5% (intervention), p < 0.05. CONCLUSIONS: This observational study demonstrated a significant reduction in AKI progression from Stage 1 and a reduction in overall development of AKI. In addition, a reduction in incorrect enoxaparin dosing was also observed, indicating increased clinical awareness. This study demonstrates that AKI guidelines coupled with a newly designed AKI care bundle integrated into CCIS can impact patient outcomes positively.

Perioperative measurement of core body temperature using an unobtrusive passive heat flow sensor.

Clinicians strive to maintain normothermia, which requires measurement of core-body temperature and may necessitate active warming of patients. Monitoring temperature currently requires invasive probes. This work investigates a novel foam-based flexible sensor worn behind the ear for the measurement of core body temperature. This observational study uses the device prototype and clinical data to compare three methods for calculating the temperature from this sensor: a basic heat-flow model, a new dynamic model that addresses changing surrounding temperatures and one that combines the dynamic model with a correction for adhesive quality. Clinical validation was performed with 21 surgical patients (average length of surgery 4.4 h) using an esophageal temperature probe as reference. The operative period was divided into four segments: normal periods (with stable surrounding temperatures), surrounding temperatures increasing due to the use of the Bair Hugger™, stable periods during Bair Hugger™ use and surrounding temperatures decreasing due to its removal. The error bias and limits of agreement over these segments were on average of - 0.05 ± 0.28 °C (95% limits of agreement) overall. The dynamic model outperformed the simple heat-flow model for periods of surrounding temperature changes (12.7% of total time) while it had a similar, high, performance for the temperature-stable periods. The results suggest that our proposed topical sensor can replace invasive core temp sensors and provide a means of consistently measuring core body temperature despite surrounding temperature shifts.

A review of recent advances in data analytics for post-operative patient deterioration detection.

Most deaths occurring due to a surgical intervention happen postoperatively rather than during surgery. The current standard of care in many hospitals cannot fully cope with detecting and addressing post-surgical deterioration in time. For millions of patients, this deterioration is left unnoticed, leading to increased mortality and morbidity. Postoperative deterioration detection currently relies on general scores that are not fully able to cater for the complex post-operative physiology of surgical patients. In the last decade however, advanced risk and warning scoring techniques have started to show encouraging results in terms of using the large amount of data available peri-operatively to improve postoperative deterioration detection. Relevant literature has been carefully surveyed to provide a summary of the most promising approaches as well as how they have been deployed in the perioperative domain. This work also aims to highlight the opportunities that lie in personalizing the models developed for patient deterioration for these particular post-surgical patients and make the output more actionable. The integration of pre- and intra-operative data, e.g. comorbidities, vitals, lab data, and information about the procedure performed, in post-operative early warning algorithms would lead to more contextualized, personalized, and adaptive patient modelling. This, combined with careful integration in the clinical workflow, would result in improved clinical decision support and better post-surgical care outcomes.

Features of Heart Rate Variability Capture Regulatory Changes During Kangaroo Care in Preterm Infants.

OBJECTIVE: To determine whether heart rate variability (HRV) can serve as a surrogate measure to track regulatory changes during kangaroo care, a period of parental coregulation distinct from regulation within the incubator. STUDY DESIGN: Nurses annotated the starting and ending times of kangaroo care for 3 months. The pre-kangaroo care, during-kangaroo care, and post-kangaroo care data were retrieved in infants with at least 10 accurately annotated kangaroo care sessions. Eight HRV features (5 in the time domain and 3 in the frequency domain) were used to visually and statistically compare the pre-kangaroo care and during-kangaroo care periods. Two of these features, capturing the percentage of heart rate decelerations and the extent of heart rate decelerations, were newly developed for preterm infants. RESULTS: A total of 191 kangaroo care sessions were investigated in 11 preterm infants. Despite clinically irrelevant changes in vital signs, 6 of the 8 HRV features (SD of normal-to-normal intervals, root mean square of the SD, percentage of consecutive normal-to-normal intervals that differ by >50 ms, SD of heart rate decelerations, high-frequency power, and low-frequency/high-frequency ratio) showed a visible and statistically significant difference (P <.01) between stable periods of kangaroo care and pre-kangaroo care. HRV was reduced during kangaroo care owing to a decrease in the extent of transient heart rate decelerations. CONCLUSION: HRV-based features may be clinically useful for capturing the dynamic changes in autonomic regulation in response to kangaroo care and other changes in environment and state.

Machine learning modelling for predicting the utilization of invasive and non-invasive ventilation throughout the ICU duration.

The goal of this work is to develop a Machine Learning model to predict the need for both invasive and non-invasive mechanical ventilation in intensive care unit (ICU) patients. Using the Philips eICU Research Institute (ERI) database, 2.6 million ICU patient data from 2010 to 2019 were analyzed. This data was randomly split into training (63%), validation (27%), and test (10%) sets. Additionally, an external test set from a single hospital from the ERI database was employed to assess the model's generalizability. Model performance was determined by comparing the model probability predictions with the actual incidence of ventilation use, either invasive or non-invasive. The model demonstrated a prediction performance with an AUC of 0.921 for overall ventilation, 0.937 for invasive, and 0.827 for non-invasive. Factors such as high Glasgow Coma Scores, younger age, lower BMI, and lower PaCO2 were highlighted as indicators of a lower likelihood for the need for ventilation. The model can serve as a retrospective benchmarking tool for hospitals to assess ICU performance concerning mechanical ventilation necessity. It also enables analysis of ventilation strategy trends and risk-adjusted comparisons, with potential for future testing as a clinical decision tool for optimizing ICU ventilation management.

An ear-worn sensor for the detection of gait impairment after abdominal surgery.

Surgery to the trunk often results in a change of gait, most pronounced during walking. This change is usually transient, often as a result of wound pain, and returns to normal as the patient recovers. Quantifying and monitoring gait impairment therefore represents a novel means of functional postoperative home recovery follow-up. Until now, this type of assessment could only be made in a gait lab, which is both expensive and labor intensive to administer on a large scale. The objective of this work is to validate the use of an ear-worn activity recognition (e-AR) sensor for quantification of gait impairment after abdominal wall and perianal surgery. The e-AR sensor was used on 2 comparative simulated data sets (N = 32) of truncal impairment to observe walking patterns. The sensor was also used to observe the walking patterns of preoperative and postoperative surgical patients who had undergone abdominal wall (n = 5) and perianal surgery (n = 5). Methods for multiresolution feature extraction, selection, and classification are investigated using the raw ear-sensor data. Results show that the method demonstrates a good separation between impaired and nonimpaired classes for both simulated and real patient data sets. This indicates that the e-AR sensor may be used as a tool for the pervasive assessment of postoperative gait impairment, as part of functional recovery monitoring, in patients at their own homes.

Machine Learning for Benchmarking Critical Care Outcomes.

OBJECTIVES: Enhancing critical care efficacy involves evaluating and improving system functioning. Benchmarking, a retrospective comparison of results against standards, aids risk-adjusted assessment and helps healthcare providers identify areas for improvement based on observed and predicted outcomes. The last two decades have seen the development of several models using machine learning (ML) for clinical outcome prediction. ML is a field of artificial intelligence focused on creating algorithms that enable computers to learn from and make predictions or decisions based on data. This narrative review centers on key discoveries and outcomes to aid clinicians and researchers in selecting the optimal methodology for critical care benchmarking using ML. METHODS: We used PubMed to search the literature from 2003 to 2023 regarding predictive models utilizing ML for mortality (592 articles), length of stay (143 articles), or mechanical ventilation (195 articles). We supplemented the PubMed search with Google Scholar, making sure relevant articles were included. Given the narrative style, papers in the cohort were manually curated for a comprehensive reader perspective. RESULTS: Our report presents comparative results for benchmarked outcomes and emphasizes advancements in feature types, preprocessing, model selection, and validation. It showcases instances where ML effectively tackled critical care outcome-prediction challenges, including nonlinear relationships, class imbalances, missing data, and documentation variability, leading to enhanced results. CONCLUSIONS: Although ML has provided novel tools to improve the benchmarking of critical care outcomes, areas that require further research include class imbalance, fairness, improved calibration, generalizability, and long-term validation of published models.

Machine learning prediction of the total duration of invasive and non-invasive ventilation During ICU Stay.

Predicting the duration of ventilation in the ICU helps in assessing the risk of ventilator-induced lung injury, ensuring sufficient oxygenation, and optimizing resource allocation. Prior models provided a prediction of total duration without distinguishing between invasive and non-invasive ventilation. This work proposes two independent gradient boosting regression models for predicting the duration of invasive and non-invasive ventilation based on commonly available ICU features. These models are trained on 2.6 million patient stays across 350 US hospitals between 2010 to 2019. The mean absolute error (MAE) for the prediction of duration was 2.08 days for invasive ventilation and 0.36 days for non-invasive ventilation. The total ventilation duration predicted by our model had MAE of 2.38 days, which outperformed the gold standard (APACHE) with MAE of 3.02 days. The feature importance analysis of the trained models showed that, for invasive ventilation, high average heart rate, diagnosis of respiratory infection and admissions from locations other than the operating room were associated with longer ventilation durations. For non-invasive ventilation, higher respiratory rates and having any GCS measurement were associated with longer durations.

GenHPF: General Healthcare Predictive Framework for Multi-task Multi-source Learning.

Despite the remarkable progress in the development of predictive models for healthcare, applying these algorithms on a large scale has been challenging. Algorithms trained on a particular task, based on specific data formats available in a set of medical records, tend to not generalize well to other tasks or databases in which the data fields may differ. To address this challenge, we propose General Healthcare Predictive Framework (GenHPF), which is applicable to any EHR with minimal preprocessing for multiple prediction tasks. GenHPF resolves heterogeneity in medical codes and schemas by converting EHRs into a hierarchical textual representation while incorporating as many features as possible. To evaluate the efficacy of GenHPF, we conduct multi-task learning experiments with single-source and multi-source settings, on three publicly available EHR datasets with different schemas for 12 clinically meaningful prediction tasks. Our framework significantly outperforms baseline models that utilize domain knowledge in multi-source learning, improving average AUROC by 1.2%P in pooled learning and 2.6%P in transfer learning while also showing comparable results when trained on a single EHR dataset. Furthermore, we demonstrate that self-supervised pretraining using multi-source datasets is effective when combined with GenHPF, resulting in a 0.6 pretraining. By eliminating the need for preprocessing and feature engineering, we believe that this work offers a solid framework for multi-task and multi-source learning that can be leveraged to speed up the scaling and usage of predictive algorithms in healthcare.1.

Incidence and outcomes of AKI in postoperative patients admitted to ICU using full KDIGO criteria - a cohort study.

STUDY OBJECTIVE: Acute kidney injury (AKI) is a serious complication in postoperative ICU patients. The incidence of AKI varies substantially based on the type of surgery and definition used. This study focuses on the incidence of AKI in postoperative ICU patients using full KDIGO criteria and related outcomes regarding to different types of surgery. DESIGN: Retrospective cohort study. SETTING: Tertiary level university hospital, eight anaesthesiological/surgical ICUs, between 2016 and 2018. PATIENTS: 6261 adult patients. MEASUREMENTS: Primary outcome was 28-day all-cause mortality in different stages of AKI according to complete KDIGO criteria. MAIN RESULTS: We found 3497 (55.9%) postoperative ICU patients with AKI. The severity distribution of AKI stage 1 to 3 was 19.7%, 28.4% and 7.8%, respectively, and 235 (4%) patients received RRT. The 28-day mortality was 3% (n = 205). Increasing AKI severity was associated with increased 28-day mortality when adjusted for other variables (AKI 2°: OR 2.81; 95% CI 1.55 to 5.24; p < 0.001 and AKI 3°: OR 11.37.; 95% CI 5.91 to 22.55; p < 0.001). Besides AKI stages 2 and 3, age (OR 1.02; 95% CI 1.01 to 1.04, p < 0.001), NYHA IV (OR 2.23; 95% CI 1.03 to 4.43, p = 0.042), need for surgical reintervention within 48 h (OR 2.92; 95% CI 1.76 to 4.72, p = 0.001), urgent surgery (OR 1.78; 95% CI 1.15 to 2.71, p = 0.01), emergency surgery (OR 2.63; 95% CI 1.58 to 4.31, p = 0.001), vascular surgery (OR 2.01; 95% CI 1.06 to 3.98, p = 0.033), and orthopedic and trauma surgery (OR 3.79; 95% CI 1.98 to 7.09, p < 0.001) versus cardiac surgery was significantly associated with increased risk for 28-days mortality in multivariate analysis. CONCLUSION: AKI based on full KDIGO criteria is very common in postoperative ICU patients and it is associated with stepwise increase in 28-days mortality.

Implementation of an automated early warning scoring system in a surgical ward: Practical use and effects on patient outcomes.

INTRODUCTION: Early warning scores (EWS) are being increasingly embedded in hospitals over the world due to their promise to reduce adverse events and improve the outcomes of clinical patients. The aim of this study was to evaluate the clinical use of an automated modified EWS (MEWS) for patients after surgery. METHODS: This study conducted retrospective before-and-after comparative analysis of non-automated and automated MEWS for patients admitted to the surgical high-dependency unit in a tertiary hospital. Operational outcomes included number of recorded assessments of the individual MEWS elements, number of complete MEWS assessments, as well as adherence rate to related protocols. Clinical outcomes included hospital length of stay, in-hospital and 28-day mortality, and ICU readmission rate. RESULTS: Recordings in the electronic medical record from the control period contained 7929 assessments of MEWS elements and were performed in 320 patients. Recordings from the intervention period contained 8781 assessments of MEWS elements in 273 patients, of which 3418 were performed with the automated EWS system. During the control period, 199 (2.5%) complete MEWS were recorded versus 3991 (45.5%) during intervention period. With the automated MEWS systems, the percentage of missing assessments and the time until the next assessment for patients with a MEWS of ≥2 decreased significantly. The protocol adherence improved from 1.1% during the control period to 25.4% when the automated MEWS system was involved. There were no significant differences in clinical outcomes. CONCLUSION: Implementation of an automated EWS system on a surgical high dependency unit improves the number of complete MEWS assessments, registered vital signs, and adherence to the EWS hospital protocol. However, this positive effect did not translate into a significant decrease in mortality, hospital length of stay, or ICU readmissions. Future research and development on automated EWS systems should focus on data management and technology interoperability.

Evidence-based Clinical Decision Support Systems for the prediction and detection of three disease states in critical care: A systematic literature review.

Background: Clinical decision support (CDS) systems have emerged as tools providing intelligent decision making to address challenges of critical care. CDS systems can be based on existing guidelines or best practices; and can also utilize machine learning to provide a diagnosis, recommendation, or therapy course. Methods: This research aimed to identify evidence-based study designs and outcome measures to determine the clinical effectiveness of clinical decision support systems in the detection and prediction of hemodynamic instability, respiratory distress, and infection within critical care settings. PubMed, ClinicalTrials.gov and Cochrane Database of Systematic Reviews were systematically searched to identify primary research published in English between 2013 and 2018. Studies conducted in the USA, Canada, UK, Germany and France with more than 10 participants per arm were included. Results: In studies on hemodynamic instability, the prediction and management of septic shock were the most researched topics followed by the early prediction of heart failure. For respiratory distress, the most popular topics were pneumonia detection and prediction followed by pulmonary embolisms. Given the importance of imaging and clinical notes, this area combined Machine Learning with image analysis and natural language processing. In studies on infection, the most researched areas were the detection, prediction, and management of sepsis, surgical site infections, as well as acute kidney injury. Overall, a variety of Machine Learning algorithms were utilized frequently, particularly support vector machines, boosting techniques, random forest classifiers and neural networks. Sensitivity, specificity, and ROC AUC were the most frequently reported performance measures. Conclusion: This review showed an increasing use of Machine Learning for CDS in all three areas. Large datasets are required for training these algorithms; making it imperative to appropriately address, challenges such as class imbalance, correct labelling of data and missing data. Recommendations are formulated for the development and successful adoption of CDS systems.

HMM assessment of quality of movement trajectory in laparoscopic surgery.

Laparoscopic surgery poses many different constraints for the operating surgeon, resulting in a slow uptake of advanced laparoscopic procedures. Traditional approaches to the assessment of surgical performance rely on prior classification of a cohort of surgeons' technical skills for validation, which may introduce subjective bias to the outcome. In this study, Hidden Markov Models (HMMs) are used to learn surgical maneuvers from 11 subjects with mixed abilities. By using the leave-one-out method, the HMMs are trained without prior clustering of subjects into different skill levels, and the output likelihood indicates the similarity of a particular subject's motion trajectories to those of the group. The results show that after a short period of training, the novices become more similar to the group when compared to the initial pre-training assessment. The study demonstrates the strength of the proposed method in ranking the quality of trajectories of the subjects, highlighting its value in minimizing the subjective bias in skills assessment for minimally invasive surgery.

Unobtrusive sleep state measurements in preterm infants - A review.

Sleep is important for the development of preterm infants. During sleep, neural connections are formed and the development of brain regions is triggered. In general, various rudimentary sleep states can be identified in the preterm infant, namely active sleep (AS), quiet sleep (QS) and intermediate sleep (IS). As the infant develops, sleep states change in length and organization, with these changes as important indicators of brain development. As a result, several methods have been deployed to distinguish between the different preterm infant sleep states, among which polysomnography (PSG) is the most frequently used. However, this method is limited by the use of adhesive electrodes or patches that are attached to the body by numerous cables that can disturb sleep. Given the importance of sleep, this review explores more unobtrusive methods that can identify sleep states without disturbing the infant. To this end, after a brief introduction to preterm sleep states, an analysis of the physiological characteristics associated with the different sleep states is provided and various methods of measuring these physiological characteristics are explored. Finally, the advantages and disadvantages of each of these methods are evaluated and recommendations for neonatal sleep monitoring proposed.

Unobtrusive Monitoring of Neonatal Brain Temperature Using a Zero-Heat-Flux Sensor Matrix.

The temperature of preterm neonates must be maintained within a narrow window to ensure their survival. Continuously measuring their core temperature provides an optimal means of monitoring their thermoregulation and their response to environmental changes. However, existing methods of measuring core temperature can be very obtrusive, such as rectal probes, or inaccurate/lagging, such as skin temperature sensors and spot-checks using tympanic temperature sensors. This study investigates an unobtrusive method of measuring brain temperature continuously using an embedded zero-heat-flux (ZHF) sensor matrix placed under the head of the neonate. The measured temperature profile is used to segment areas of motion and incorrect positioning, where the neonate's head is not above the sensors. We compare our measurements during low motion/stable periods to esophageal temperatures for 12 preterm neonates, measured for an average of 5 h per neonate. The method we propose shows good correlation with the reference temperature for most of the neonates. The unobtrusive embedding of the matrix in the neonate's environment poses no harm or disturbance to the care work-flow, while measuring core temperature. To address the effect of motion on the ZHF measurements in the current embodiment, we recommend a more ergonomic embedding ensuring the sensors are continuously placed under the neonate's head.

Pattern discovery in critical alarms originating from neonates under intensive care.

Patient monitoring generates a large number of alarms, the vast majority of which are false. Excessive non-actionable medical alarms lead to alarm fatigue, a well-recognized patient safety issue. While multiple approaches to reduce alarm fatigue have been explored, patterns in alarming and inter-alarm relationships, as they manifest in the clinical workspace, are largely a black-box and hamper research efforts towards reducing alarms. The aim of this study is to detect opportunities to safely reduce alarm pressure, by developing techniques to identify, capture and visualize patterns in alarms. Nearly 500 000 critical medical alarms were acquired from a neonatal intensive care unit over a 20 month period. Heuristic techniques were developed to extract the inter-alarm relationships. These included identifying the presence of alarm clusters, patterns of transition from one alarm category to another, temporal associations amongst alarms and determination of prevalent sequences in which alarms manifest. Desaturation, bradycardia and apnea constituted 86% of all alarms and demonstrated distinctive periodic increases in the number of alarms that were synchronized with nursing care and enteral feeding. By inhibiting further alarms of a category for a short duration of time (30 s/60 s), non-actionable physiological alarms could be reduced by 20%. The patterns of transition from one alarm category to another and the time duration between such transitions revealed the presence of close temporal associations and multiparametric derangement. Examination of the prevalent alarm sequences reveals that while many sequences comprised of multiple alarms, nearly 65% of the sequences were isolated instances of alarms and are potentially irreducible. Patterns in alarming, as they manifest in the clinical workspace were identified and visualized. This information can be exploited to investigate strategies for reducing alarms.

A wearable mobility assessment device for total knee replacement: A longitudinal feasibility study.

BACKGROUND: Total knee replacement currently lacks robust indications and objective follow-up metrics. Patients and healthcare staff are under-equipped to optimise outcomes. This study aims to investigate the feasibility of using an ear-worn motion sensor (e-AR, Imperial College London) to conduct objective, home-based mobility assessments in the peri-operative setting. METHODS: Fourteen patients on the waiting list for knee replacement, and 15 healthy subjects, were recruited. Pre-operatively, and at 1, 3, 6, 12 and 24 weeks post-operatively, patients underwent functional mobility testing (Timed Up and Go), knee examination (including range of motion), and an activity protocol whilst wearing the e-AR sensor. Features extracted from sensor motion data were used to assess patient performance and predict patients' recovery phase. RESULTS: Sensor-derived peri-operative mobility trends correlated with clinical measures in several activities, allowing functional recovery of individual subjects to be profiled and compared, including the detection of a complication. Sensor data features enabled classification of subjects into normal, pre-operative and 24-week post-operative groups with 89% (median) accuracy. Classification accuracy was reduced to 69% when including all time intervals. DISCUSSION: This study demonstrates a novel, objective method of assessing peri-operative mobility, which could be used to supplement surgical decision-making and facilitate community-based follow-up.

Predicting free-living energy expenditure using a miniaturized ear-worn sensor: an evaluation against doubly labeled water.

Accurate estimation of daily total energy expenditure (EE)is a prerequisite for assisted weight management and assessing certain health conditions. The use of wearable sensors for predicting free-living EE is challenged by consistent sensor placement, user compliance, and estimation methods used. This paper examines whether a single ear-worn accelerometer can be used for EE estimation under free-living conditions.An EE prediction model as first derived and validated in a controlled setting using healthy subjects involving different physical activities. Ten different activities were assessed showing a tenfold cross validation error of 0.24. Furthermore, the EE prediction model shows a mean absolute deviation(MAD) below 1.2 metabolic equivalent of tasks. The same model was applied to a free-living setting with a different population for further validation. The results were compared against those derived from doubly labeled water. In free-living settings, the predicted daily EE has a correlation of 0.74, p 0.008, and a MAD of 272 kcal day. These results demonstrate that laboratory-derived prediction models can be used to predict EE under free-living conditions [corrected].