Explore the vast range of titles available.

AbstractObjectiveTo provide an overview and critical appraisal of early warning scores for adult hospital patients.DesignSystematic review.Data sourcesMedline, CINAHL, PsycInfo, and Embase until June 2019.Eligibility criteria for study selectionStudies describing the development or external validation of an early warning score for adult hospital inpatients.Results13 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.ConclusionsEarly 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 registrationPROSPERO CRD42017053324.

BackgroundThere is uncertainty about the associations of angiotensive enzyme (ACE) inhibitor and angiotensin receptor blocker (ARB) drugs with COVID-19 disease. We studied whether patients prescribed these drugs had altered risks of contracting severe COVID-19 disease and receiving associated intensive care unit (ICU) admission.MethodsThis was a prospective cohort study using routinely collected data from 1205 general practices in England with 8.28 million participants aged 20–99 years. We used Cox proportional hazards models to derive adjusted HRs for exposure to ACE inhibitor and ARB drugs adjusted for sociodemographic factors, concurrent medications and geographical region. The primary outcomes were: (a) COVID-19 RT-PCR diagnosed disease and (b) COVID-19 disease resulting in ICU care.FindingsOf 19 486 patients who had COVID-19 disease, 1286 received ICU care. ACE inhibitors were associated with a significantly reduced risk of COVID-19 disease (adjusted HR 0.71, 95% CI 0.67 to 0.74) but no increased risk of ICU care (adjusted HR 0.89, 95% CI 0.75 to 1.06) after adjusting for a wide range of confounders. Adjusted HRs for ARBs were 0.63 (95% CI 0.59 to 0.67) for COVID-19 disease and 1.02 (95% CI 0.83 to 1.25) for ICU care.There were significant interactions between ethnicity and ACE inhibitors and ARBs for COVID-19 disease. The risk of COVID-19 disease associated with ACE inhibitors was higher in Caribbean (adjusted HR 1.05, 95% CI 0.87 to 1.28) and Black African (adjusted HR 1.31, 95% CI 1.08 to 1.59) groups than the white group (adjusted HR 0.66, 95% CI 0.63 to 0.70). A higher risk of COVID-19 with ARBs was seen for Black African (adjusted HR 1.24, 95% CI 0.99 to 1.58) than the white (adjusted HR 0.56, 95% CI 0.52 to 0.62) group.InterpretationACE inhibitors and ARBs are associated with reduced risks of COVID-19 disease after adjusting for a wide range of variables. Neither ACE inhibitors nor ARBs are associated with significantly increased risks of receiving ICU care. Variations between different ethnic groups raise the possibility of ethnic-specific effects of ACE inhibitors/ARBs on COVID-19 disease susceptibility and severity which deserves further study.

Background Survivors of intensive care are known to be at increased risk of developing longer-term psychopathology issues. We present a large UK multicentre study assessing the anxiety, depression and post-traumatic stress disorder (PTSD) caseness in the first year following discharge from an intensive care unit (ICU). Methods Design: prospective multicentre follow-up study of survivors of ICU in the UK. Setting: patients from 26 ICUs in the UK. Inclusion criteria: patients who had received at least 24 h of level 3 ICU care and were 16 years of age or older. Interventions: postal follow up: Hospital Anxiety and Depression Score (HADS) and the Post-Traumatic Stress Disorder (PTSD) Check List-Civilian (PCL-C) at 3 and 12 months following discharge from ICU. Main outcome measure: caseness of anxiety, depression and PTSD, 2-year survival. Results In total, 21,633 patients admitted to ICU were included in the study. Postal questionnaires were sent to 13,155 survivors; of these 38% (4943/13155) responded and 55% (2731/4943) of respondents passed thresholds for one or more condition at 3 or 12 months following discharge. Caseness prevalence was 46%, 40% and 22% for anxiety, depression and PTSD respectively; 18% (870/4943 patients) met the caseness threshold for all three psychological conditions. Patients with symptoms of depression were 47% more likely to die during the first 2 years after discharge from ICU than those without (HR 1.47, CI 1.19–1.80). Conclusions Over half of those who respond to postal questionnaire following treatment on ICU in the UK reported significant symptoms of anxiety, depression or PTSD. When symptoms of one psychological disorder are present, there is a 65% chance they will co-occur with symptoms of one of the other two disorders. Depression following critical illness is associated with an increased mortality risk in the first 2 years following discharge from ICU. Trial registration ISRCTN Registry, ISRCTN69112866 . Registered on 2 May 2006.

There is a growing emphasis being placed on the potential for cuffless blood pressure (BP) estimation through modelling of morphological features from the photoplethysmogram (PPG) and electrocardiogram (ECG). However, the appropriate features and models to use remain unclear. We investigated the best features available from the PPG and ECG for BP estimation using both linear and non-linear machine learning models. We conducted a clinical study in which changes in BP ( Δ BP) were induced by an infusion of phenylephrine in 30 healthy volunteers (53.8% female, 28.0 (9.0) years old). We extracted a large and diverse set of features from both the PPG and the ECG and assessed their individual importance for estimating Δ BP through Shapley additive explanation values and a ranking coefficient. We trained, tuned, and evaluated linear (ordinary least squares, OLS) and non-linear (random forest, RF) machine learning models to estimate Δ BP in a nested leave-one-subject-out cross-validation framework. We reported the results as correlation coefficient ( ρ p ), root mean squared error (RMSE), and mean absolute error (MAE). The non-linear RF model significantly ( p < 0.05 ) outperformed the linear OLS model using both the PPG and the ECG signals across all performance metrics. Estimating Δ SBP using the PPG alone ( ρ p = 0.86 (0.23), RMSE = 5.66 (4.76) mmHg, MAE = 4.86 (4.29) mmHg) performed significantly better than using the ECG alone ( ρ p = 0.69 (0.45), RMSE = 6.79 (4.76) mmHg, MAE = 5.28 (4.57) mmHg), all p < 0.001 . The highest ranking features from the PPG largely modelled increasing reflected wave interference driven by changes in arterial stiffness. This finding was supported by changes observed in the PPG waveform in response to the phenylephrine infusion. However, a large number of features were required for accurate BP estimation, highlighting the high complexity of the problem. We conclude that the PPG alone may be further explored as a potential single source, cuffless, blood pressure estimator. The use of the ECG alone is not justified. Non-linear models may perform better as they are able to incorporate interactions between feature values and demographics. However, demographics may not adequately account for the unique and individualised relationship between the extracted features and BP.

Prognostication is an essential tool for risk adjustment and decision making in the intensive care unit (ICU). Research into prognostication in ICU has so far been limited to data from admission or the first 24 hours. Most ICU admissions last longer than this, decisions are made throughout an admission, and some admissions are explicitly intended as time-limited prognostic trials. Despite this, temporal changes in prognostic ability during ICU admission has received little attention to date. Current predictive models, in the form of prognostic clinical tools, are typically derived from linear models and do not explicitly handle incremental information from trends. Machine learning (ML) allows predictive models to be developed which use non-linear predictors and complex interactions between variables, thus allowing incorporation of trends in measured variables over time; this has made it possible to investigate prognosis throughout an admission. This study uses ML to assess the predictability of ICU mortality as a function of time. Logistic regression against physiological data alone outperformed APACHE-II and demonstrated several important interactions including between lactate & noradrenaline dose, between lactate & MAP, and between age & MAP consistent with the current sepsis definitions. ML models consistently outperformed logistic regression with Deep Learning giving the best results. Predictive power was maximal on the second day and was further improved by incorporating trend data. Using a limited range of physiological and demographic variables, the best machine learning model on the first day showed an area under the receiver-operator characteristic curve (AUC) of 0.883 (σ = 0.008), compared to 0.846 (σ = 0.010) for a logistic regression from the same predictors and 0.836 (σ = 0.007) for a logistic regression based on the APACHE-II score. Adding information gathered on the second day of admission improved the maximum AUC to 0.895 (σ = 0.008). Beyond the second day, predictive ability declined. This has implications for decision making in intensive care and provides a justification for time-limited trials of ICU therapy; the assessment of prognosis over more than one day may be a valuable strategy as new information on the second day helps to differentiate outcomes. New ML models based on trend data beyond the first day could greatly improve upon current risk stratification tools.

Myocardial infarction (MI) remains one of the greatest contributors to mortality, and patients admitted to the intensive care unit (ICU) with myocardial infarction are at higher risk of death. In this study, we use two retrospective cohorts extracted from two US-based ICU databases, eICU and MIMIC-IV, to develop an explainable pseudo-dynamic machine learning framework for mortality prediction in the ICU. The method provides accurate prediction for ICU patients up to 24 hours before the event and provides time-resolved interpretability. We compare standard supervised machine learning algorithms with novel tabular deep learning approaches and find that an integrated XGBoost model in our EHR time-series extraction framework (XMI-ICU) performs best. The framework was evaluated on a held-out test set from eICU and externally validated on the MIMIC-IV cohort using the most important features identified by time-resolved Shapley values. XMI-ICU achieved AUROCs of 92.0 (balanced accuracy of 82.3) for a 6-hour prediction of mortality. We demonstrate that XMI-ICU maintains reliable predictive performance across different prediction horizons (6, 12, 18, and 24 hours) during ICU stay while also achieving successful external validation in a separate patient cohort from MIMIC-IV without any previous training on that dataset. We also evaluated the framework for clinical risk analysis by comparing it to the standard APACHE IV system in active use. We show that our framework successfully leverages time-series physiological measurements from ICU health records by translating them into stacked static prediction problems for mortality in heart attack patients and can offer clinical insight from time-resolved interpretability through the use of Shapley values.

Early warning scores (EWS) are routinely used in hospitals to assess a patient's risk of deterioration. EWS are traditionally recorded on paper observation charts but are increasingly recorded digitally. In either case, evidence for the clinical effectiveness of such scores is mixed, and previous studies have not considered whether EWS leads to changes in how deteriorating patients are managed. This study aims to examine whether the introduction of a digital EWS system was associated with more frequent observation of patients with abnormal vital signs, a precursor to earlier clinical intervention. We conducted a 2-armed stepped-wedge study from February 2015 to December 2016, over 4 hospitals in 1 UK hospital trust. In the control arm, vital signs were recorded using paper observation charts. In the intervention arm, a digital EWS system was used. The primary outcome measure was time to next observation (TTNO), defined as the time between a patient's first elevated EWS (EWS ≥3) and subsequent observations set. Secondary outcomes were time to death in the hospital, length of stay, and time to unplanned intensive care unit admission. Differences between the 2 arms were analyzed using a mixed-effects Cox model. The usability of the system was assessed using the system usability score survey. We included 12,802 admissions, 1084 in the paper (control) arm and 11,718 in the digital EWS (intervention) arm. The system usability score was 77.6, indicating good usability. The median TTNO in the control and intervention arms were 128 (IQR 73-218) minutes and 131 (IQR 73-223) minutes, respectively. The corresponding hazard ratio for TTNO was 0.99 (95% CI 0.91-1.07; P=.73). We demonstrated strong clinical engagement with the system. We found no difference in any of the predefined patient outcomes, suggesting that the introduction of a highly usable electronic system can be achieved without impacting clinical care. Our findings contrast with previous claims that digital EWS systems are associated with improvement in clinical outcomes. Future research should investigate how digital EWS systems can be integrated with new clinical pathways adjusting staff behaviors to improve patient outcomes.

AbstractObjectiveTo assess the association between covid-19 vaccines and risk of thrombocytopenia and thromboembolic events in England among adults.DesignSelf-controlled case series study using national data on covid-19 vaccination and hospital admissions.SettingPatient level data were obtained for approximately 30 million people vaccinated in England between 1 December 2020 and 24 April 2021. Electronic health records were linked with death data from the Office for National Statistics, SARS-CoV-2 positive test data, and hospital admission data from the United Kingdom’s health service (NHS).Participants29 121 633 people were vaccinated with first doses (19 608 008 with Oxford-AstraZeneca (ChAdOx1 nCoV-19) and 9 513 625 with Pfizer-BioNTech (BNT162b2 mRNA)) and 1 758 095 people had a positive SARS-CoV-2 test. People aged ≥16 years who had first doses of the ChAdOx1 nCoV-19 or BNT162b2 mRNA vaccines and any outcome of interest were included in the study.Main outcome measuresThe primary outcomes were hospital admission or death associated with thrombocytopenia, venous thromboembolism, and arterial thromboembolism within 28 days of three exposures: first dose of the ChAdOx1 nCoV-19 vaccine; first dose of the BNT162b2 mRNA vaccine; and a SARS-CoV-2 positive test. Secondary outcomes were subsets of the primary outcomes: cerebral venous sinus thrombosis (CVST), ischaemic stroke, myocardial infarction, and other rare arterial thrombotic events.ResultsThe study found increased risk of thrombocytopenia after ChAdOx1 nCoV-19 vaccination (incidence rate ratio 1.33, 95% confidence interval 1.19 to 1.47 at 8-14 days) and after a positive SARS-CoV-2 test (5.27, 4.34 to 6.40 at 8-14 days); increased risk of venous thromboembolism after ChAdOx1 nCoV-19 vaccination (1.10, 1.02 to 1.18 at 8-14 days) and after SARS-CoV-2 infection (13.86, 12.76 to 15.05 at 8-14 days); and increased risk of arterial thromboembolism after BNT162b2 mRNA vaccination (1.06, 1.01 to 1.10 at 15-21 days) and after SARS-CoV-2 infection (2.02, 1.82 to 2.24 at 15-21 days). Secondary analyses found increased risk of CVST after ChAdOx1 nCoV-19 vaccination (4.01, 2.08 to 7.71 at 8-14 days), after BNT162b2 mRNA vaccination (3.58, 1.39 to 9.27 at 15-21 days), and after a positive SARS-CoV-2 test; increased risk of ischaemic stroke after BNT162b2 mRNA vaccination (1.12, 1.04 to 1.20 at 15-21 days) and after a positive SARS-CoV-2 test; and increased risk of other rare arterial thrombotic events after ChAdOx1 nCoV-19 vaccination (1.21, 1.02 to 1.43 at 8-14 days) and after a positive SARS-CoV-2 test.ConclusionIncreased risks of haematological and vascular events that led to hospital admission or death were observed for short time intervals after first doses of the ChAdOx1 nCoV-19 and BNT162b2 mRNA vaccines. The risks of most of these events were substantially higher and more prolonged after SARS-CoV-2 infection than after vaccination in the same population.

Emerging reports of rare neurological complications associated with COVID-19 infection and vaccinations are leading to regulatory, clinical and public health concerns. We undertook a self-controlled case series study to investigate hospital admissions from neurological complications in the 28 days after a first dose of ChAdOx1nCoV-19 ( n  = 20,417,752) or BNT162b2 ( n  = 12,134,782), and after a SARS-CoV-2-positive test ( n  = 2,005,280). There was an increased risk of Guillain–Barré syndrome (incidence rate ratio (IRR), 2.90; 95% confidence interval (CI): 2.15–3.92 at 15–21 days after vaccination) and Bell’s palsy (IRR, 1.29; 95% CI: 1.08–1.56 at 15–21 days) with ChAdOx1nCoV-19. There was an increased risk of hemorrhagic stroke (IRR, 1.38; 95% CI: 1.12–1.71 at 15–21 days) with BNT162b2. An independent Scottish cohort provided further support for the association between ChAdOx1nCoV and Guillain–Barré syndrome (IRR, 2.32; 95% CI: 1.08–5.02 at 1–28 days). There was a substantially higher risk of all neurological outcomes in the 28 days after a positive SARS-CoV-2 test including Guillain–Barré syndrome (IRR, 5.25; 95% CI: 3.00–9.18). Overall, we estimated 38 excess cases of Guillain–Barré syndrome per 10 million people receiving ChAdOx1nCoV-19 and 145 excess cases per 10 million people after a positive SARS-CoV-2 test. In summary, although we find an increased risk of neurological complications in those who received COVID-19 vaccines, the risk of these complications is greater following a positive SARS-CoV-2 test. A self-controlled case series analysis of nearly 32 million people in England shows an increased risk of rare neurological complications in those who received COVID-19 vaccines and following SARS-CoV-2 infection. The results highlight 38 excess cases of Guillain–Barré syndrome per 10 million ChAdOx1nCoV-19 vaccinations.