Implementation of a critical care outreach team in a children's hospital

Introduction

Proactive surveillance by a critical care outreach team (CCOT) can promote early recognition of deterioration in hospitalized patients but is uncommon in pediatric rapid response systems (RRSs). After our children's hospital introduced a CCOT in 2019, we aimed to characterize early implementation outcomes. We hypothesized that CCOT rounding would identify additional children at risk for deterioration.

Methods

The CCOT, staffed by a dedicated critical care nurse (RN), respiratory therapist, and attending, conducts daily in-person rounds with charge RNs on medical-surgical units, to screen RRS-identified high-risk patients for deterioration. In this prospective study, observers tracked rounds discussion content, participation, and identification of new high-risk patients. We compared 'identified-patient-discussions' (IPD) about RRS-identified patients, and 'new-patient-discussions' (NPD) about new patients with Fisher's exact test. For new patients, we performed thematic analysis of clinical data to identify deterioration related themes.

Results

During 348 unit-rounds over 20 days, we observed 383 discussions - 35 (9%) were NPD. Frequent topics were screening for clinical concerns (374/383, 98%), active clinical concerns (147/383, 39%), and watcher activation (66/383, 17%). Most discussions only included standard participants (353/383, 92%). Compared to IPD, NPD more often addressed active concerns (74.3% vs 34.8%, p < 0.01) and staffing resource concerns (5.7% vs 0.6%, p < 0.04), and more often incorporated extra participants (25.7% vs 6%, p < 0.01). In thematic analysis of 33 new patients, most (29/33, 88%) had features of deterioration.

Conclusion

A successfully implemented CCOT enhanced identification of clinical deterioration not captured by existing RRS resources. Future work will investigate its impact on operational safety and patient-centered outcomes.

Using Data-Driven Machine Learning to Predict Unplanned ICU Transfers with Critical Deterioration from Electronic Health Records

OBJECTIVE

We aimed to develop a data-driven machine learning model for predicting critical deterioration events from routinely collected EHR data in hospitalized children.

MATERIALS

This retrospective cohort study included all pediatric inpatients hospitalized on a medical or surgical ward between 2014-2018 at a quaternary children's hospital.

METHODS

We developed a large data-driven approach and evaluated three machine learning models to predict pediatric critical deterioration events. We evaluated the models using a nested, stratified 10-fold cross-validation. The evaluation metrics included C-statistic, sensitivity, and positive predictive value. We also compared the machine learning models with patients identified as high-risk Watchers by bedside clinicians.

RESULTS

The study included 57,233 inpatient admissions from 34,976 unique patients. 3,943 variables were identified from the EHR data. The XGBoost model performed best (C-statistic=0.951, CI: 0.946 ∼ 0.956).

CONCLUSIONS

Our data-driven machine learning models accurately predicted patient deterioration. Future sociotechnical analysis will inform deployment within the clinical setting.

Novel Prehospital Phenotypes and Outcomes in Adult-Patients with Acute Disease

An early identification of prehospital phenotypes may allow health care workers to speed up and improve patients’ treatment. To determine emergency phenotypes by exclusively using prehospital clinical data, a multicenter, prospective, and observational ambulance-based study was conducted with a cohort of 3,853 adult patients treated consecutively and transferred with high priority from the scene to the hospital emergency department. Cluster analysis determined three clusters with highly different outcome scores and pathological characteristics. The first cluster presented a 30-day mortality after the index event of 45.9%. The second cluster presented a mortality of 26.3%, while mortality of the third cluster was 5.1%. This study supports the detection of three phenotypes with different risk stages and with different clinical, therapeutic, and prognostic considerations. This evidence could allow adapting treatment to each phenotype thereby helping in the decision-making process.

Recognising and Responding to Clinical Deterioration in Adult Patients in Isolation Precautions for Infection Control: A Retrospective Cohort Study

BACKGROUND

Patient isolation is widely used as a strategy for prevention and control of infection but may have unintended consequences for patients. Early recognition and response to acute deterioration is an essential component of safe, quality patient care and has not been explored for patients in isolation. Primary aims of this study were to: i) describe the timing, frequency, and nature of clinical deterioration during hospital admission for patients with isolation precautions for infection control and ii) compare the characteristics of patients who did and did not deteriorate during their initial period of isolation precautions for infection control.

METHODS

This retrospective cohort study was conducted across three sites of a large Australian health service. The study sample were adult patients (≥18 years) admitted into isolation precautions within 24-hours of admission from 1 July to 31 December 2019.

RESULTS

There were 634 patients who fulfilled the study inclusion criteria. One in eight patients experienced at least one episode of clinical deterioration during their time in isolation with most episodes of deterioration occurring within the first two days of admission. Timely Medical Emergency Team calls occurred in almost half the episodes of deterioration; however, the same proportion (47.2%) of deterioration episodes resulted in no Medical Emergency Team activation (afferent limb failure). In the 24-hours preceding each episode of clinical deterioration (n=180), 81.6% (n=147) of episodes were preceded by vital signs fulfilling pre-Medical Emergency Team criteria.Patients who deteriorated during isolation for infection control were older (median age 74.0 vs 71.0 years, P=0.042); more likely to live in a residential care facility (21.0% vs 7.2%, P=0.006); had a longer initial period of isolation (4.0 vs 2.9 days, P=<000.1) and hospital length of stay (median 4.9 vs 3.2 days, P=<0.001) and were more likely to die in hospital (12.3% vs 4.3%, P<0.001).

CONCLUSION

Patients in isolation precautions experienced high Medical Emergency Team afferent limb failure and most fulfilled pre-Medical Emergency Team criteria in the 24-hours preceding episodes of deterioration. Timely recognition and response to clinical deterioration continue to be essential in providing safe, quality patient care regardless of the hospital-care environment.

Risk factors and early prediction of clinical deterioration and mortality in adult COVID-19 inpatients: an Australian tertiary hospital experience

BACKGROUND

Early recognition of severe COVID-19 is essential for timely patient triage.

AIMS

We aim to report clinical and laboratory findings and patient outcomes at a tertiary hospital in Melbourne, Australia.

METHODS

This is a retrospective study of adult inpatients with COVID-19 admitted to Northern Health from March to September 2020. Data were extracted from electronic medical records.

RESULTS

Key admission data was available for 182 patients (median age 67.0 years (interquartile range, 47.9-83.1; 51.1% female). 56 (30.8%) were from residential care. 117 (64.3%) patients were assigned Goals-of-Patient-Care (GOPC) A or B and 65 (35.7%) GOPC C or D. Comorbidities were present in 135 patients (74.2%). 63.2% of patients received antibiotics, 6.6% had antivirals, 45.6% received systemic glucocorticoid and 3.3% had tocilizumab. 56 (30.8%) developed clinical deterioration (24 requiring ventilation, 21 receiving critical care, 34 died). Overall, in-hospital clinical deterioration was significantly associated with older age (p<0.001), history of diabetes (p=0.038), lower lymphocyte count (p=0.002) and platelet count (p=0.004), higher neutrophil-to-lymphocyte ratio (p=0.002), elevated fibrinogen (p=0.004), higher serum ferritin (p=0.027) and CRP (p=0.002). The accuracy of the 4C Deterioration model was moderate, with an area under the curve (AUC) of 0.79 (95% CI, 0.68-0.90) compared with an AUC of 0.77 (95% CI, 0.76-0.78) in the original validation cohort.

CONCLUSIONS

In this study, high neutrophil-to-lymphocyte ratio, abnormal d-dimer, high serum CRP and ferritin appear to be useful prognostic markers. This article is protected by copyright. All rights reserved.

Validation of an Automatic Tagging System for Identifying Respiratory and Hemodynamic Deterioration Events in the Intensive Care Unit

Objectives

Predictive models for critical events in the intensive care unit (ICU) might help providers anticipate patient deterioration. At the heart of predictive model development lies the ability to accurately label significant events, thereby facilitating the use of machine learning and similar strategies. We conducted this study to establish the validity of an automated system for tagging respiratory and hemodynamic deterioration by comparing automatic tags to tagging by expert reviewers.

Methods

This retrospective cohort study included 72,650 unique patient stays collected from Electronic Medical Records of the University of Massachusetts’ eICU. An enriched subgroup of stays was manually tagged by expert reviewers. The tags generated by the reviewers were compared to those generated by an automated system.

Results

The automated system was able to rapidly and efficiently tag the complete database utilizing available clinical data. The overall agreement rate between the automated system and the clinicians for respiratory and hemodynamic deterioration tags was 89.4% and 87.1%, respectively. The automatic system did not add substantial variability beyond that seen among the reviewers.

Conclusions

We demonstrated that a simple rule-based tagging system could provide a rapid and accurate tool for mass tagging of a compound database. These types of tagging systems may replace human reviewers and save considerable resources when trying to create a validated, labeled database used to train artificial intelligence algorithms. The ability to harness the power of artificial intelligence depends on efficient clinical validation of targeted conditions; hence, these systems and the methodology used to validate them are crucial.

Pulmonary Thromboembolism as a Potential Cause of Clinical Deterioration in COVID-19 Patients; a Commentary

Although the findings of some studies have been indicative of the direct relationship between the severity of clinical findings and imaging, reports have been published regarding inconsistency of clinical findings with imaging and laboratory evidence. Physicians treating these patients frequently report cases in which patients, sometimes in the recovery phase and despite improvements in imaging indices, suddenly deteriorate and in some instances suddenly expire. This letter aimed to draw attention to the role of pulmonary thromboembolism as a potential and possible cause of clinical deterioration in covid-19 patients.

The value of electrocardiography in prognosticating clinical deterioration and mortality in acute pulmonary embolism: A systematic review and meta-analysis

The role of electrocardiography (ECG) in prognosticating pulmonary embolism (PE) is increasingly recognized. ECG is quickly interpretable, noninvasive, inexpensive, and available in remote areas. We hypothesized that ECG can provide useful information about PE prognostication. We searched MEDLINE, EMBASE, Google Scholar, Web of Science, abstracts, conference proceedings, and reference lists through February 2017. Eligible studies used ECG to prognosticate for the main outcomes of death and clinical deterioration or escalation of therapy. Two authors independently selected studies; disagreement was resolved by consensus. Ad hoc piloted forms were used to extract data and assess risk of bias. We used a random-effects model to pool relevant data in meta-analysis with odds ratios (ORs) and 95% confidence intervals (CIs); all other data were synthesized qualitatively. Statistical heterogeneity was assessed using the I ² value. We included 39 studies (9198 patients) in the systematic review. There was agreement in study selection (κ: 0.91, 95% CI: 0.86-0.96). Most studies were retrospective; some did not appropriately control for confounders. ECG signs that were good predictors of a negative outcome included S1Q3T3 (OR: 3.38, 95% CI: 2.46-4.66, P < 0.001), complete right bundle branch block (OR: 3.90, 95% CI: 2.46-6.20, P < 0.001), T-wave inversion (OR: 1.62, 95% CI: 1.19-2.21, P = 0.002), right axis deviation (OR: 3.24, 95% CI: 1.86-5.64, P < 0.001), and atrial fibrillation (OR: 1.96, 95% CI: 1.45-2.67, P < 0.001) for in-hospital mortality. Several ischemic patterns also were significantly predictive. Our conclusion is that ECG is potentially valuable in prognostication of acute PE.

A disruptive Big data approach to leverage the efficiency in management and clinical decision support in a Hospital

There is an urgent need to potentiate evidence-based clinical-decision-making with a holistic, patient-centered approach to value, one that focuses both on health-care spending and treatment outcomes.1 On the other hand, in the era of self-driven vehicles, computer systems in healthcare need also to become proactive and to identify relevant clinical patterns in a much faster and automated way than currently used solutions enable. Although this is the state-of-the-art paradigm, in fact, technical constraints block further developments in these areas as hospitals lack the skills to really manage and take value from the big amount of Data about their patients that is stored in dozens of heterogeneous information systems, from lab results to imaging studies, from pharmacy to the Electronic Medical Record (EMR). At São João Hospital Center (São João), a novel analytics platform was conceived, a new approach that is able to leverage all the Big Data that is stored about hospital patients in seconds and to apply some of the most advanced and lightening speed analytics on top of this information in order to empower clinicians and to give them a new decision support tool. This sets the road towards a data-driven hospital of the future, where Data Analytics and Data Science can become as important as the most recent Harrison's edition. With this analytics platform, São João was able to be the first Non-Us institution to ever win the Microsoft U.S Worldwide Innovation Award (HIMSS - Florida, 2014) and the European Big Data & Analytics solution of the year (IT EUROPA - London, 2014). This solution is called HVITAL (Hospital surVeiIlance, moniToring and ALert) and is working 24/7 at São João Hospital since 2012.