Showing papers by "Lionel Tarassenko published in 2020"

Early risk assessment for COVID-19 patients from emergency department data using machine learning

Abstract: Background Since its emergence in late 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused a pandemic, with more than 4.8 million reported cases and 310 000 deaths worldwide. While epidemiological and clinical characteristics of COVID-19 have been reported, risk factors underlying the transition from mild to severe disease among patients remain poorly understood. Methods In this retrospective study, we analysed data of 820 confirmed COVID-19 positive patients admitted to a two-site NHS Trust hospital in London, England, between January 1st and April 23rd, 2020, with a majority of cases occurring in March and April. We extracted anonymised demographic data, physiological clinical variables and laboratory results from electronic healthcare records (EHR) and applied multivariate logistic regression, random forest and extreme gradient boosted trees. To evaluate the potential for early risk assessment, we used data available during patients’ initial presentation at the emergency department (ED) to predict deterioration to one of three clinical endpoints in the remainder of the hospital stay: A) admission to intensive care, B) need for mechanical ventilation and C) mortality. Based on the trained models, we extracted the most informative clinical features in determining these patient trajectories. Results Considering our inclusion criteria, we have identified 126 of 820 (15%) patients that required intensive care, 62 of 808 (8%) patients needing mechanical ventilation, and 170 of 630 (27%) cases of in-hospital mortality. Our models learned successfully from early clinical data and predicted clinical endpoints with high accuracy, the best model achieving AUC-ROC scores of 0.75 to 0.83 (F1 scores of 0.41 to 0.56). Younger patient age was associated with an increased risk of receiving intensive care and ventilation, but lower risk of mortality. Clinical indicators of a patient’s oxygen supply and selected laboratory results were most predictive of COVID-19 patient trajectories. Conclusion Among COVID-19 patients machine learning can aid in the early identification of those with a poor prognosis, using EHR data collected during a patient’s first presentation at ED. Patient age and measures of oxygenation status during ED stay are primary indicators of poor patient outcomes.

The Mobile-Based 6-Minute Walk Test: Usability Study and Algorithm Development and Validation.

Abstract: Background: The 6-min walk test (6MWT) is a convenient method for assessing functional capacity in patients with cardiopulmonary conditions. It is usually performed in the context of a hospital clinic and thus requires the involvement of hospital staff and facilities, with their associated costs. Objective: This study aimed to develop a mobile phone–based system that allows patients to perform the 6MWT in the community. Methods: We developed 2 algorithms to compute the distance walked during a 6MWT using sensors embedded in a mobile phone. One algorithm makes use of the global positioning system to track the location of the phone when outdoors and hence computes the distance travelled. The other algorithm is meant to be used indoors and exploits the inertial sensors built into the phone to detect U-turns when patients walk back and forth along a corridor of fixed length. We included these algorithms in a mobile phone app, integrated with wireless pulse oximeters and a back-end server. We performed Bland-Altman analysis of the difference between the distances estimated by the phone and by a reference trundle wheel on 49 indoor tests and 30 outdoor tests, with 11 different mobile phones (both Apple iOS and Google Android operating systems). We also assessed usability aspects related to the app in a discussion group with patients and clinicians using a technology acceptance model to guide discussion. Results: The mean difference between the mobile phone-estimated distances and the reference values was −2.013 m (SD 7.84 m) for the indoor algorithm and −0.80 m (SD 18.56 m) for the outdoor algorithm. The absolute maximum difference was, in both cases, below the clinically significant threshold. A total of 2 pulmonary hypertension patients, 1 cardiologist, 2 physiologists, and 1 nurse took part in the discussion group, where issues arising from the use of the 6MWT in hospital were identified. The app was demonstrated to be usable, and the 2 patients were keen to use it in the long term. Conclusions: The system described in this paper allows patients to perform the 6MWT at a place of their convenience. In addition, the use of pulse oximetry allows more information to be generated about the patient’s health status and, possibly, be more relevant to the real-life impact of their condition. Preliminary assessment has shown that the developed 6MWT app is highly accurate and well accepted by its users. Further tests are needed to assess its clinical value.

Blood pressure monitoring in high-risk pregnancy to improve the detection and monitoring of hypertension (the BUMP 1 and 2 trials): protocol for two linked randomised controlled trials

Abstract: Introduction Self-monitoring of blood pressure (BP) in pregnancy could improve the detection and management of pregnancy hypertension, while also empowering and engaging women in their own care. Two linked trials aim to evaluate whether BP self-monitoring in pregnancy improves the detection of raised BP during higher risk pregnancies (BUMP 1) and whether self-monitoring reduces systolic BP during hypertensive pregnancy (BUMP 2). Methods and analyses Both are multicentre, non-masked, parallel group, randomised controlled trials. Participants will be randomised to self-monitoring with telemonitoring or usual care. BUMP 1 will recruit a minimum of 2262 pregnant women at higher risk of pregnancy hypertension and BUMP 2 will recruit a minimum of 512 pregnant women with either gestational or chronic hypertension. The BUMP 1 primary outcome is the time to the first recording of raised BP by a healthcare professional. The BUMP 2 primary outcome is mean systolic BP between baseline and delivery recorded by healthcare professionals. Other outcomes will include maternal and perinatal outcomes, quality of life and adverse events. An economic evaluation of BP self-monitoring in addition to usual care compared with usual care alone will be assessed across both study populations within trial and with modelling to estimate long-term cost-effectiveness. A linked process evaluation will combine quantitative and qualitative data to examine how BP self-monitoring in pregnancy is implemented and accepted in both daily life and routine clinical practice. Ethics and dissemination The trials have been approved by a Research Ethics Committee (17/WM/0241) and relevant research authorities. They will be published in peer-reviewed journals and presented at national and international conferences. If shown to be effective, BP self-monitoring would be applicable to a large population of pregnant women. Trial registration number NCT03334149

Identification of patients with atrial fibrillation: a big data exploratory analysis of the UK Biobank.

Abstract: Atrial Fibrillation (AF) is the most common cardiac arrhythmia, with an estimated prevalence of around 1.6% in the adult population. The analysis of the Electrocardiogram (ECG) data acquired in the UK Biobank represents an opportunity to screen for AF in a large sub-population in the UK. The main objective of this paper is to assess ten machine-learning methods for automated detection of subjects with AF in the UK Biobank dataset. Six classical machine-learning methods based on Support Vector Machines are proposed and compared with state-of-the-art techniques (including a deep-learning algorithm), and finally a combination of a classical machine-learning and deep learning approaches. Evaluation is carried out on a subset of the UK Biobank dataset, manually annotated by human experts. The combined classical machine-learning and deep learning method achieved an F1 score of 84.8% on the test subset, and a Cohen's Kappa coefficient of 0.83, which is similar to the inter-observer agreement of two human experts. The level of performance indicates that the automated detection of AF in patients whose data have been stored in a large database, such as the UK Biobank, is possible. Such automated identification of AF patients would enable further investigations aimed at identifying the different phenotypes associated with AF.

Home monitoring with technology-supported management in chronic heart failure: a randomised trial.

Abstract: Objectives We aimed to investigate whether digital home monitoring with centralised specialist support for remote management of heart failure (HF) is more effective in improving medical therapy and patients’ quality of life than digital home monitoring alone. Methods In a two-armed partially blinded parallel randomised controlled trial, seven sites in the UK recruited a total of 202 high-risk patients with HF (71.3 years SD 11.1; left ventricular ejection fraction 32.9% SD 15.4). Participants in both study arms were given a tablet computer, Bluetooth-enabled blood pressure monitor and weighing scales for health monitoring. Participants randomised to intervention received additional regular feedback to support self-management and their primary care doctors received instructions on blood investigations and pharmacological treatment. The primary outcome was the use of guideline-recommended medical therapy for chronic HF and major comorbidities, measured as a composite opportunity score (total number of recommended treatment given divided by the total number of opportunities the treatment should have been given, with a score 1 indicating 100% adherence to recommendations). Co-primary outcome was change in physical score of Minnesota Living with Heart Failure questionnaire. Results 101 patients were randomised to ‘enhanced self-management’ and 101 to ‘supported medical management’. At the end of follow-up, the opportunity score was 0.54 (95% CI 0.46 to 0.62) in the control arm and 0.61 (95% CI 0.52 to 0.70) in the intervention arm (p=0.25). Physical well-being of participants also did not differ significantly between the groups (17.4 (12.4) mean (SD) for control arm vs 16.5 (12.1) in treatment arm; p for change=0.84). Conclusions Central provision of tailored specialist management in a multi-morbid HF population was feasible. However, there was no strong evidence for improvement in use of evidence-based treatment nor health-related quality of life. Trial registration number ISRCTN86212709

Self-reported and objectively measured physical activity in people with and without chronic heart failure: UK Biobank analysis.

Abstract: Objective The impact of heart failure (HF) on perceived and objectively measured levels of physical activity (PA) can inform risk stratification and treatment recommendation. We aimed to compare self-reported and objectively measured PA levels in a large sample of participants with and without HF. Methods A validated PA questionnaire was used to estimate self-reported weekly PA among 1600 participants with HF and 387 580 participants without HF. Accelerometer data were studied in 596 participants with HF and 96 105 participants without HF for a period of 7 days. Using multivariable linear regression models, we compared the PA levels between participants with HF and without HF, focusing on both the average daily PA levels and the intensity of PAs throughout the day. Results PA levels were significantly lower in participants with HF using both self-report (excess metabolic equivalent of task hours per week of 26.5 (95% CI 24.7 to 28.4) vs 34.7 (95% CI 34.5 to 34.9), respectively (p Conclusion Perceived and objectively recorded PA levels of patients with chronic HF are significantly lower than those of individuals without HF. This difference is continuous throughout the different hours of the day, with individuals with HF being on average 16% less active than individuals without HF. In patients with HF, increases in everyday activity may be a potential alternative to structured exercise programmes.

Vital-sign circadian rhythms in patients prior to discharge from an ICU: a retrospective observational analysis of routinely recorded physiological data

Abstract: Background: Circadian deregulation in patients treated in an intensive care unit (ICU) is commonplace and is associated with complications such as immune system disruption and delirium. The presence and nature of circadian rhythms in the vital signs recorded in the ICU are not well documented, nor is their generalisability across different ICU populations. This paper investigates the presence of circadian rhythms in the 24 h prior to discharge from an ICU of patients who subsequently recovered. We hypothesise that vital-sign circadian rhythms will be observable in this cohort of patients, that these circadian rhythms will resemble known behaviour in healthy individuals, and that these circadian rhythms will be generalisable across different populations of ICU patients. Methods: Circadian rhythms are investigated across several commonly measured vital signs: systolic blood pressure, heart rate, respiratory rate, and temperature. The data employed in this paper are from patients in the MIMIC-III (2001–2012), eICU-CRD (2014–2015), and PICRAM (2009–2015) databases, spanning 198,205 patients across 211 hospitals in the USA and the UK. Evaluation of circadian rhythms encompasses a comparison between the observed rhythm profiles and peak-nadir excursions with those found in the literature, as well as the assessment of the correlation in rhythm profiles between databases. Results: Circadian patterns in all four vital signs were found to conform to those reported for non-ICU cohorts. Additionally, all vital-sign circadian profiles were correlated between databases at the p = 0.05 level. The peak-nadir excursion in the observed rhythms was suppressed by a factor of 2–5 relative to results found in the literature for cohorts of young, healthy individuals. Conclusions: Across three different ICU datasets, systolic blood pressure, heart rate, respiratory rate, and temperature showed circadian variation in the 24 h prior to discharge from an ICU. However, the amplitude of these variations was markedly reduced in comparison to cohorts of young, healthy adults. The observed circadian variation correlated strongly between databases, suggesting there is a generalisable state of circadian behaviour in ICU patients during (Continued on next page) *Correspondence: shaun.davidson@eng.ox.ac.uk 1Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, UK Full list of author information is available at the end of the article © The Author(s). 2020, corrected publication 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 InternationalLicense,which permits use, sharing, adaptation, distribution and reproduction in anymedium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Davidson et al. Critical Care (2020) 24:181 Page 2 of 13 (Continued from previous page) the 24 h prior to discharge from an ICU. This result has potential uses in monitoring patient recovery and early detection of complications such as delirium.

Non-contact vital-sign monitoring of patients undergoing haemodialysis treatment.

Abstract: A clinical study was designed to record a wide range of physiological values from patients undergoing haemodialysis treatment in the Renal Unit of the Churchill Hospital in Oxford. Video was recorded for a total of 84 dialysis sessions from 40 patients during the course of 1 year, comprising an overall video recording time of approximately 304.1 h. Reference values were provided by two devices in regular clinical use. The mean absolute error between the heart rate estimates from the camera and the average from two reference pulse oximeters (positioned at the finger and earlobe) was 2.8 beats/min for over 65% of the time the patient was stable. The mean absolute error between the respiratory rate estimates from the camera and the reference values (computed from the Electrocardiogram and a thoracic expansion sensor-chest belt) was 2.1 breaths/min for over 69% of the time for which the reference signals were valid. To increase the robustness of the algorithms, novel methods were devised for cancelling out aliased frequency components caused by the artificial light sources in the hospital, using auto-regressive modelling and pole cancellation. Maps of the spatial distribution of heart rate and respiratory rate information were developed from the coefficients of the auto-regressive models. Most of the periods for which the camera could not produce a reliable heart rate estimate lasted under 3 min, thus opening the possibility to monitor heart rate continuously in a clinical environment.

Wearability Testing of Ambulatory Vital Sign Monitoring Devices: Prospective Observational Cohort Study

Abstract: Background: Timely recognition of patient deterioration remains challenging. Ambulatory monitoring systems (AMSs) may provide support to current monitoring practices; however, they need to be thoroughly tested before implementation in the clinical environment for early detection of deterioration. Objective: The objective of this study was to assess the wearability of a selection of commercially available AMSs to inform a future prospective study of ambulatory vital sign monitors in an acute hospital ward. Methods: Five pulse oximeters (4 with finger probes and 1 wrist-worn only, collecting pulse rates and oxygen saturation) and 2 chest patches (collecting heart rates and respiratory rates) were selected to be part of this study: The 2 chest-worn patches were VitalPatch (VitalConnect) and Peerbridge Cor (Peerbridge); the 4 wrist-worn devices with finger probe were Nonin WristOx2 3150 (Nonin), Checkme O2+ (Viatom Technology), PC-68B, and AP-20 (both from Creative Medical); and the 1 solely wrist-worn device was Wavelet (Wavelet Health). Adult participants wore each device for up to 72 hours while performing usual “activities of daily living” and were asked to score the perceived exertion and perception of pain or discomfort by using the Borg CR-10 scale; thoughts and feelings caused by the AMS using the Comfort Rating Scale (CRS); and to provide general free text feedback. Median and IQRs were reported and nonparametric tests were used to assess differences between the devices’ CRS scores. Results: Quantitative scores and feedback were collected in 70 completed questionnaires from 20 healthy volunteers, with each device tested approximately 10 times. The Wavelet seemed to be the most wearable device (P<.001) with an overall median (IQR) CRS score of 1.00 (0.88). There were no statistically significant differences in wearability between the chest patches in the CRS total score; however, the VitalPatch was superior in the Attachment section (P=.04) with a median (IQR) score of 3.00 (1.00). General pain and discomfort scores and total percentage of time worn are also reflective of this. Conclusions: Our results suggest that adult participants prefer to wear wrist-worn pulse oximeters without a probe compressing the fingertip and they prefer to wear a smaller chest patch. A compromise between wearability, reliability, and accuracy should be made for successful and practical integration of AMSs within the hospital environment.

OxCOVID19 Database: a multimodal data repository for better understanding the global impact of COVID-19

Abstract: Oxford COVID-19 Database (OxCOVID19 Database) is a comprehensive source of information related to the COVID-19 pandemic. This relational database contains time-series data on epidemiology, government responses, mobility, weather and more across time and space for all countries at the national level, and for more than 50 countries at the regional level. It is curated from a variety of (wherever available) official sources. Its purpose is to facilitate the analysis of the spread of SARS-CoV-2 virus and to assess the effects of non-pharmaceutical interventions to reduce the impact of the pandemic. Our database is a freely available, daily updated tool that provides unified and granular information across geographical regions.

Protocol for a prospective, controlled, cross-sectional, diagnostic accuracy study to evaluate the specificity and sensitivity of ambulatory monitoring systems in the prompt detection of hypoxia and during movement.

Abstract: Introduction Automated continuous ambulatory monitoring may provide an alternative to intermittent manual vital signs monitoring. This has the potential to improve frequency of measurements, timely escalation of care and patient safety. However, a major barrier to the implementation of these wearable devices in the ward environment is their uncertain reliability, efficiency and data fidelity. The purpose of this study is to test performance of selected devices in a simulated clinical setting including during movement and low levels of peripheral oxygen saturation. Methods and analysis This is a single centre, prospective, controlled, cross-sectional, diagnostic accuracy study to determine the specificity and sensitivity of currently available ambulatory vital signs monitoring equipment in the detection of hypoxia and the effect of movement on data acquisition. We will recruit up to 45 healthy volunteers who will attend a single study visit; starting with a movement phase and followed by the hypoxia exposure phase where we will gradually decrease saturation levels down to 80%. We will simultaneously test one chest patch, one wrist worn only and three wrist worn with finger probe devices against ‘clinical standard ‘and ‘gold standard’ references. We will measure peripheral oxygen saturations, pulse rate, heart rate and respiratory rate continuously and arterial blood gases intermittently throughout the study. Ethics and dissemination This study has received ethical approval by the East of Scotland Research Ethics Service REC 2 (19/ES/0008). The results will be broadly distributed through conference presentations and peer-reviewed publications. Trial registration number ISRCTN61535692 registered on 10/06/2019.

Risk factors for clinical progression in patients with COVID-19: a retrospective study of electronic health record data in the United Kingdom

Abstract: Background: The novel coronavirus disease 2019 (COVID-19) outbreak presents a significant threat to global health. A better understanding of patient clinical profiles is essential to drive efficient and timely health service strategies. In this study, we aimed to identify risk factors for a higher susceptibility to symptomatic presentation with COVID-19 and a transition to severe disease. Methods: We analysed data on 2756 patients admitted to Chelsea & Westminster Hospital NHS Foundation Trust between 1st January and 23rd April 2020. We compared differences in characteristics between patients designated positive for COVID-19 and patients designated negative on hospitalisation and derived a multivariable logistic regression model to identify risk factors for predicting risk of symptomatic COVID-19. For patients with COVID-19, we used univariable and multivariable logistic regression to identify risk factors associated with progression to severe disease defined by: 1) admission to the hospital AICU, 2) the need for mechanical ventilation, 3) in-hospital mortality, and 4) at least one measurement of elevated D-dimer (equal or superior to 1,000 ug/L) indicative of increased risk of venous thromboembolism. Results: The patient population consisted of 1148 COVID-19 positive and 1608 COVID-19 negative patients. Age, sex, self-reported ethnicity, C-reactive protein, white blood cell count, respiratory rate, body temperature, and systolic blood pressure formed the most parsimonious model for predicting risk of symptomatic COVID-19 at hospital admission. Among 1148 patients with COVID-19, 116 (10.1%) were admitted to the AICU, 71 (6.2%) required mechanical ventilation, 368 (32.1%) had at least one record of D-dimer levels ≥1,000 μg/L, and 118 patients died. In the multivariable logistic regression, age (OR = 0.953 per 1 year, 95% CI: 0.937-0.968) C-reactive protein (OR = 1.004 per 1 mg/L, 95% CI: 1.002-1.007), and white blood cell counts (OR = 1.059 per 109/L, 95% CI: 1.010-1.111) were found to be associated with admission to the AICU. Age (OR = 0.973 per 1 year, 95% CI: 0.955-0.990), C-reactive protein (OR = 1.003 per 1 mg/L, 95% CI: 1.000-1.006) and sodium (OR = 0.915 per 1 mmol/L, 0.868-0.962) were associated with mechanical ventilation. Age (OR = 1.023 per 1 year, 95% CI: 1.004-1.043), CRP (OR = 1.004 per 1 mg/L, 95% CI: 1.002-1.006), and body temperature (OR = 0.723 per 1oC, 95% CI: 0.541-0.958) were associated with elevated D-dimer. For mortality, we observed associations with age (OR = 1.060 per 1 year, 95% CI: 1.040-1.082), female sex (OR = 0.442, 95% CI: 0.442, 95% CI: 0.245-0.777), Asian ethnic background (OR = 2.237 vs White ethnic background, 95% CI: 1.111-4.510), C-reactive protein (OR = 1.004 per 1 mg/L, 95% CI: 1.001-1.006), sodium (OR = 1.038 per 1 mmol/L, 95% CI: 1.001-1.006), and respiratory rate (OR = 1.054 per 1 breath/min, 95% CI: 1.024-1.087). Conclusion: Our analysis suggests there are several demographic, clinical and laboratory findings associated with a symptomatic presentation of COVID-19. Moreover, significant associations between patient deterioration were found with age, sex and specific blood markers, chiefly C-reactive protein, and could help early identification of patients at risk of poorer prognosis. Further work is required to clarify the extent to which our observations are relevant beyond current settings.

Monitoring activity of hip injury patients (MoHIP): a sub-study of the World Hip Trauma Evaluation observational cohort study.

Abstract: Hip fracture is common, affecting 20% of women and 10% of men during their lifetime. The trajectory of patients’ recovery as they transition from the acute hospital setting to their usual residence is poorly understood. Recently, the use of activity trackers to monitor physical activity during recovery has been investigated as a way to explore this trajectory. This prospective observational cohort study followed patients from hospital to home as they recovered from a hip fracture. Participants were recruited from a single centre and provided with a 3-axis logging accelerometer worn as a pendant, for 16 weeks from recruitment. Participants received monthly follow-up visits which included questions about wearing the monitor. Monthly activity monitor data were also downloaded. Participant activity was estimated from the monitor data using the calibrated “Euclidean Norm Minus One” (ENMO) metric. Polynomial mixed-effects modelling was used to evaluate the difference between the weekly activity trends of 2 groups of participants: those with and without independent mobility at 16 weeks (defined by whether aids or personal assistance were required to mobilise). Twenty-nine participants from 125 eligible patients were recruited. Of these, 19 (66%) reported being aware of wearing the monitor at least some of the time. Fourteen (48%) participants withdrew before study completion. Data for thirteen (45%) participants were of sufficient quantity to be included in the activity modelling procedure. Of these, 8 reported independent mobility at 16 weeks post-surgery, and 5 did not. By week 7, the weekly predicted mean ENMO ($$ {\overline{ENMO}}_W $$) values were significantly different between the two participant groups, demonstrating feasibility of the model’s ability to predict which patients will report independent mobility at 16 weeks. This is the first study to our knowledge to investigate acceptability and feasibility of a pendant-worn activity monitor in this patient cohort. Acceptability of wearing the monitor and feasibility of recruitment and retention of participants were limited. Future research into the use of activity monitors in this population should use minimally intrusive devices which are acceptable to this population. MoHIP is a sub-study of the World Hip Trauma Evaluation (WHiTE) Study (ISRCTN 63982700).

Early detection of physiological deterioration in post-surgical patients using wearable technology combined with an integrated monitoring system: a pre- and post-interventional study

Abstract: Objectives Late recognition of physiological deterioration is a frequent problem in hospital wards. We assessed whether ambulatory (wearable) physiological monitoring combined with a system that continuously merges physiological variables into a single “risk” score (VSI), changed care and outcome in patients after major surgery. Design Pre- and post-interventional study. Setting A single centre tertiary referral university hospital upper-gastrointestinal service. Participants Patients who underwent major upper-gastrointestinal surgery. Interventions Phase-I (pre-intervention phase): Patients received continuous wearable monitoring and standard care, but the VSI score was not available for clinical use. Phase-II (post-intervention phase): Patients received continuous wearable monitoring. In addition to standard care the VSI score was displayed for use in clinical practice. Measurements and Main Results 200 participants were monitored in phase-I. 207 participants were monitored in phase-II. Participants were monitored (median, interquartile range, IQR) for 30.2% (13.8-49.2) of available time in phase-I and 58.2% (33.1-75.2) of available time in phase-II. Clinical staff recorded observations more frequently in the 36 hours prior to a major adverse event (death, cardiac arrest or unplanned admission to intensive care) for phase-II participants (median, IQR, time between observations of 1.00, 0.50-2.08 hours) than phase-I participants (1.50, 0.75-2.50 hours, p Conclusion The combination of the integrated monitoring system with ambulatory monitoring in high-risk post-surgical patients improved recognition and management of deteriorating patients without increasing the observation rate in those patients who did not deteriorate.

Study protocol for an exploratory interventional study investigating the feasibility of video-based non-contact physiological monitoring in healthy volunteers by Mapping Of Lower Limb skIn pErfusion (MOLLIE).

Abstract: Introduction Skin perfusion varies in response to changes in the circulatory status. Blood flow to skin is reduced during haemodynamic collapse secondary to peripheral vasoconstriction, whereas increased skin perfusion is frequently observed when haemodynamics improve with resuscitation. These changes in perfusion may be monitored using non-contact image-based methods. Previous camera-derived physiological measurements have focused on accurate vital signs monitoring and extraction of physiological signals from environmental noise. One of the biggest challenges of camera-derived monitoring is artefacts from motion, which limits our understanding of what parameters may be derived from skin. In this study, we use phenylephrine and glyceryl trinitrate (GTN) to cause vasoconstriction and vasodilation in stationary healthy volunteers to describe directional changes in skin perfusion pattern. Methods and analysis We aim to recruit 30 healthy volunteers who will undergo protocolised infusions of phenylephrine and GTN, followed by the monitored and timed release of a thigh tourniquet. The experimental timeline will be identical for all participants. Measurements of traditionally used haemodynamic markers (heart rate, blood pressure and stroke volume) and camera-derived measurements will be taken concurrently throughout the experimental period. The parameters of interest from the image data are skin colour and pattern, skin surface temperature, pulsatile signal detected at the skin surface and skin perfusion index. Ethics and dissemination This study was reviewed and approved by the Oxford University Research and Ethics Committee and Clinical Trials and Research Governance teams (R63796/RE001). The results of this study will be presented at scientific conferences and published in peer-reviewed journals. Trial registration number ISRCTN10417167.

Technology-supported home monitoring in heart failure patients

Abstract: Digital health promises to enhance the prevailing episodic models of chronic heart failure (HF) care. We aimed to test the hypothesis that digital home monitoring with centralised specialist support for remote management of HF and major vascular comorbidities is more effective in optimising medical therapy and improving patients' quality of life than digital home monitoring alone. In a two-armed partially blinded parallel randomised controlled trial, seven sites in the United Kingdom recruited a total of 202 adults with HF (71.3 years SD 11.1; mean left ventricular ejection fraction 32.9% SD 15.4). Participants were selected for being at high risk of adverse outcomes or high potential to benefit from remote management. Participants in both study arms were given an internet-enabled tablet computer, Bluetooth-enabled blood pressure monitor and weighing scales for health monitoring. After a run-in period, participants randomized to intervention received additional regular feedback to support self-management and their primary care doctors received instructions on blood investigations and pharmacological treatment. The primary outcome was the use of recommended medical therapy, for chronic HF and major comorbidities, measured as a composite opportunity score. Co-primary outcome was change in physical score of Minnesota Living with Heart failure questionnaire. At the end of the trial, the weighted opportunity score was 0.54 (CI 95% 0.46, 0.62) in the control group and 0.61 (CI 95% 0.52, 0.70) in the intervention arm (p for mean difference=0.25). Physical well-being of participants did not differ significantly between the groups either (p=0.55). Central provision of tailored specialist management in a multimorbid HF population was feasible. However, there was no strong evidence for improvement in use of evidence-based therapies nor health-related quality of life. Figure 1 Type of funding source: Public Institution(s). Main funding source(s): National Institute for Health Research (NIHR) Health Services Research and Delivery; NIHR Career Development Fellowship

Method and apparatus for designing a course of treatment

Abstract: Aggregated data is used to identify cyclic variations in physiological parameters which variations may be associated with an adverse medical condition or event. The timing of these variations in the population (or particular demographic sub-cohorts of that population) can then be used to determine the timing and/or dosage level for the administration of medicaments associated with reducing the deviation in the physiological parameter concerned. The method may account for irregularities in the sampling of data associated with normal patient care by aggregating data from large numbers of patients, perhaps from a large number of medical facilities such as hospitals. Demographic ranges and vector calculations over daily, weekly, monthly or other cycles may be used to inform dosage for treatment of a patient with a cardiovascular, respiratory, physiological or other disorder.

Comparing Circadian Rhythms in Blood Pressure Computed From the Arterial Line and Sphygmomanometer in the ICU

Abstract: It has previously been shown that there is an underlying bias in blood pressure (BP) measurements between the arterial line and the sphygmomanometer cuffs. It is unclear the extent to which this bias can lead to differences in the standard of care inside the Intensive Care Unit (ICU). The objective of this study was to compare the circadian patterns of systolic, mean arterial and diastolic BP (SBP, MAP and DBP) derived from simultaneous measurements of arterial lines and sphygmomanometer cuffs in the ICU. 2,243 patients were selected from the MIMIC-III database, resulting in 22,743 data-points of paired BP measurements from the arterial line and sphygmomanometer cuff. A significant bias between the two measurement sources was found for SBP and MAP but not for DBP. In addition, a significant proportion (SBP - 51%, DBP - 23%, MAP - 41%) of cuff measurements had a difference greater than 10 mmHg with respect to the arterial line. Despite these errors, the circadian rhythms between the two signals showed a high correlation: SBP - r = 0.92 p < 0.001; MAP - r = 0.90 p < 0.001; DBP - r = 0.86 p < 0.001. Despite significant errors and bias between the arterial line and cuff, the sphygmomanometer cuff is sufficiently accurate to track circadian rhythms in BP in the ICU.

Vital-Sign Synchrony as a Marker for Patient Circadian Rhythms in an Intensive Care Unit

Abstract: Patient circadian rhythms are often disrupted in an intensive care unit (ICU). This disruption is associated with worsened patient outcomes, thus new methods are needed to quantify patient circadian rhythms. We hypothesise that the cross-correlation between vital-sign circadian profiles will allow us to stratify patients by rhythm strength, without reliance on a prior assumed rhythm profile. We selected from the eICU-CRD and MIMIC-III databases the cohort of patients in their final 24 hours of ICU stay who subsequently recovered. We then calculated the mean cross-correlation (R) between each patient's systolic blood pressure, heart rate, and respiratory rate profiles, and grouped them into ‘high’, ‘mid’, and ‘low’ correlation cohorts. The high-corr. cohorts showed vital-sign profiles that closely resembled those reported in the literature for non-ICU cohorts, with peaks at awakening and in the evening, and a large trough overnight. The mid- and low-corr. cohorts, in contrast, showed less consistent and defined peaks and troughs. Vital-sign peak-nadir excursions in the high-correlation cohorts were greater, and the length of ICU stay significantly shorter (p < 0.05), than those for the low-corr. cohorts.