Cardiac arrest may occur unexpectedly in intensive care units (ICU). We hypothesize that certain patient characteristics and treatments are associated with survival and long-term functional outcome following in-ICU cardiac arrest.Over a 12-month period, cardiac arrests with resuscitation attempts were prospectively investigated in 45 French ICUs. Survivors were followed for 6 months.In total, 677 (2.16%) of 31,399 admitted patients had at least one in-ICU cardiac arrest with resuscitation attempt, 42% of which occurred on the day of admission. In 79% cases, one or more condition(s) likely to promote the occurrence of cardiac arrest was/were identified, including hypoxia (179 patients), metabolic disorders (122), hypovolemia (94), and adverse events linked to the life-sustaining devices in place (98). Return of spontaneous circulation was achieved in 478 patients, of whom 163 were discharged alive from ICU and 146 from hospital. Six-month survival with no or moderate functional sequel (118 of 125 patients alive) correlated with a number of organ failures ≤ 2 when cardiac arrest occurred (OR 4.17 [1.92-9.09]), resuscitation time ≤ 5 min (3.32 [2.01-5.47]), shockable rhythm cardiac arrests (2.13 [1.26-3.45]) or related to the life-sustaining devices in place (2.11 [1.22-3.65]), absence of preexisting disability (1.98 [1.09-3.60]) or disease deemed fatal within 5 years (1.70 [1.05-2.77]), and sedation (1.71 [1.06-2.75]).Only one in six patients with in-ICU cardiac arrest and resuscitation attempt was alive at 6 months with good functional status. Certain characteristics specific to cardiac arrests, resuscitation maneuvers, and the pathological context in which they happen may help clarify prognosis and inform relatives.

Patients suffering in-hospital cardiac arrest often show signs of physiological deterioration before the event. The purpose of this study was to determine the prevalence of abnormal vital signs 1-4h before cardiac arrest, and to evaluate the association between these vital sign abnormalities and in-hospital mortality.We included adults from the Get With the Guidelines(®)- Resuscitation registry with an in-hospital cardiac arrest. We used two a priori definitions for vital signs: abnormal (heart rate (HR) ≤ 60 or ≥ 100 min(-1), respiratory rate (RR) ≤ 10 or >20 min(-1) and systolic blood pressure (SBP) ≤ 90 mm Hg) and severely abnormal (HR ≤ 50 or ≥ 130 min(-1), RR ≤ 8 or ≥ 30 min(-1) and SBP ≤ 80 mm Hg). We evaluated the association between the number of abnormal vital signs and in-hospital mortality using a multivariable logistic regression model.7851 patients were included. Individual vital signs were associated with in-hospital mortality. The majority of patients (59.4%) had at least one abnormal vital sign 1-4h before the arrest and 13.4% had at least one severely abnormal sign. We found a step-wise increase in mortality with increasing number of abnormal vital signs within the abnormal (odds ratio (OR) 1.53 (CI: 1.42-1.64) and severely abnormal groups (OR 1.62 (CI: 1.38-1.90)). This remained in multivariable analysis (abnormal: OR 1.38 (CI: 1.28-1.48), and severely abnormal: OR 1.40 (CI: 1.18-1.65)).Abnormal vital signs are prevalent 1-4h before in-hospital cardiac arrest on hospital wards. In-hospital mortality increases with increasing number of pre-arrest abnormal vital signs as well as increased severity of vital sign derangements.Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.

Early detection of clinical deterioration on the wards may improve outcomes, and most early warning scores only utilize a patient's current vital signs. The added value of vital sign trends over time is poorly characterized. We investigated whether adding trends improves accuracy and which methods are optimal for modelling trends.Patients admitted to five hospitals over a five-year period were included in this observational cohort study, with 60% of the data used for model derivation and 40% for validation. Vital signs were utilized to predict the combined outcome of cardiac arrest, intensive care unit transfer, and death. The accuracy of models utilizing both the current value and different trend methods were compared using the area under the receiver operating characteristic curve (AUC).A total of 269,999 patient admissions were included, which resulted in 16,452 outcomes. Overall, trends increased accuracy compared to a model containing only current vital signs (AUC 0.78 vs. 0.74; p<0.001). The methods that resulted in the greatest average increase in accuracy were the vital sign slope (AUC improvement 0.013) and minimum value (AUC improvement 0.012), while the change from the previous value resulted in an average worsening of the AUC (change in AUC -0.002). The AUC increased most for systolic blood pressure when trends were added (AUC improvement 0.05).Vital sign trends increased the accuracy of models designed to detect critical illness on the wards. Our findings have important implications for clinicians at the bedside and for the development of early warning scores.Copyright © 2016 Elsevier Ireland Ltd. All rights reserved.

Cardiac arrest in the intensive care unit (ICU-CA) is a common and highly morbid event. We investigated the preventability of ICU-CAs and identified targets for future intervention.This was a prospective, observational study of ICU-CAs at a tertiary care center in the United States. For each arrest, the clinical team was surveyed regarding arrest preventability. An expert, multi-disciplinary team of physicians and nurses also reviewed each arrest. Arrests were scored 0 (not at all preventable) to 5 (completely preventable). Arrests were considered 'unlikely but potentially preventable' or 'potentially preventable' if at least 50% of reviewers assigned a score of ≥1 or ≥3 respectively. Themes of preventability were assessed for each arrest.43 patients experienced an ICU-CA and were included. A total of 14 (32.6%) and 13 (30.2%) arrests were identified as unlikely but potentially preventable by the expert panel and survey respondents respectively, and an additional 11 (25.6%) and 10 (23.3%) arrests were identified as potentially preventable. Timing of response to clinical deterioration, missed/incorrect diagnosis, timing of acidemia correction, timing of escalation to a more senior clinician, and timing of intubation were the most commonly cited contributors to potential preventability. Additional themes identified included the administration of anxiolytics/narcotics for agitation later identified to be due to clinical deterioration and misalignment between team and patient/family perceptions of prognosis and goals-of-care.ICU-CAs may have preventable elements. Themes of preventability were identified and addressing these themes through data-driven quality improvement initiatives could potentially reduce CA incidence in critically-ill patients.Copyright © 2019. Published by Elsevier B.V.

Regression analysis of censored failure observations via the proportional hazards model permits time-varying covariates which are observed at death times. In practice, such longitudinal covariates are typically sparse and only measured at infrequent and irregularly spaced follow-up times. Full likelihood analyses of joint models for longitudinal and survival data impose stringent modelling assumptions which are difficult to verify in practice and which are complicated both inferentially and computationally. In this article, a simple kernel weighted score function is proposed with minimal assumptions. Two scenarios are considered: half kernel estimation in which observation ceases at the time of the event and full kernel estimation for data where observation may continue after the event, as with recurrent events data. It is established that these estimators are consistent and asymptotically normal. However, they converge at rates which are slower than the parametric rates which may be achieved with fully observed covariates, with the full kernel method achieving an optimal convergence rate which is superior to that of the half kernel method. Simulation results demonstrate that the large sample approximations are adequate for practical use and may yield improved performance relative to last value carried forward approach and joint modelling method. The analysis of the data from a cardiac arrest study demonstrates the utility of the proposed methods.

International early warning scores (EWS) including the additive National Early Warning Score (NEWS) and logistic EWS currently utilise physiological snapshots to predict clinical deterioration. We hypothesised that a dynamic score including vital sign trajectory would improve discriminatory power.Multicentre retrospective analysis of electronic health record data from postoperative patients admitted to cardiac surgical wards in four UK hospitals. Least absolute shrinkage and selection operator-type regression (LASSO) was used to develop a dynamic model (DyniEWS) to predict a composite adverse event of cardiac arrest, unplanned intensive care re-admission or in-hospital death within 24 h.A total of 13,319 postoperative adult cardiac patients contributed 442,461 observations of which 4234 (0.96%) adverse events in 24 h were recorded. The new dynamic model (AUC = 0.80 [95% CI 0.78-0.83], AUPRC = 0.12 [0.10-0.14]) outperforms both an updated snapshot logistic model (AUC = 0.76 [0.73-0.79], AUPRC = 0.08 [0.60-0.10]) and the additive National Early Warning Score (AUC = 0.73 [0.70-0.76], AUPRC = 0.05 [0.02-0.08]). Controlling for the false alarm rates to be at current levels using NEWS cut-offs of 5 and 7, DyniEWS delivers a 7% improvement in balanced accuracy and increased sensitivities from 41% to 54% at NEWS 5 and 18% to -30% at NEWS 7.Using an advanced statistical approach, we created a model that can detect dynamic changes in risk of unplanned readmission to intensive care, cardiac arrest or in-hospital mortality and can be used in real time to risk-prioritise clinical workload.Copyright © 2020 The Authors. Published by Elsevier B.V. All rights reserved.

Overuse of arterial blood gas (ABG) determinations leads to increased costs, inefficient use of staff work hours, and patient discomfort and blood loss. We developed guidelines to optimize ABG use in the ICU.ABG use guidelines were implemented in all adult ICUs in our institution: three medical, two trauma-surgery, one cardiovascular, and one neurosurgical ICU. Although relying on pulse oximetry, we encouraged the use of ABG determination after an acute respiratory event or for a rational clinical concern and discouraged obtaining ABG measurements for routine surveillance, after planned changes of positive end-expiratory pressure or Fio on the mechanical ventilator, for spontaneous breathing trials, or when a disorder was not suspected. ABG measurements and global ICU metrics were collected before (year 2014) and after (year 2015) the intervention.We saw a reduction of 821.5 ± 257.4 ABG determinations per month (41.5%), or approximately one ABG determination per patient per mechanical ventilation (MV) day for each month (43.1%), after introducing the guidelines (P <.001). This represented 49 L of saved blood, a reduction of $39,432 in the costs of ICU care, and 1,643 staff work hours freed for other tasks. Appropriately indicated tests rose to 83.4% from a baseline 67.5% (P =.002). Less than 5% of inappropriately indicated ABG determinations changed patient management in the postintervention period. There were no significant differences in MV days, severity of illness, or ICU mortality between the two periods.The large scale implementation of guidelines for ABG use reduced the number of inappropriately ordered ABG determinations over seven different multidisciplinary ICUs, without negatively impacting patient care.Copyright © 2016. Published by Elsevier Inc.