ABSTRACT
Background: SARS-CoV-2 virus has caused tremendous burden on both patients and providers across the globe. Our focus is to develop a practical and easy to deploy system to predict the severe manifestation of disease in COVID-19 patients with an aim to assist clinicians in triage and treatment decisions.Methods: We used multiple cohorts from 14,172 COVID-19 patients with captured clinical outcomes from four healthcare systems across the globe. Our proposed predictive algorithm is a trained artificial intelligence-based network using 8,427 patient records. We focused on building a parsimonious model with the fewest possible number of input parameters to facilitate clinical deployment. The model provides a severity risk score along with likelihoods of various clinical outcomes, namely ventilator use, end organ damage, and mortality.Findings: Model computed severity risk scores using nine laboratory markers taken from 4,293 patients at the initial presentation and the age have the prediction accuracy with the area under the curve (AUC) of 0 · 77 95% CI: 0·76-0·78, and the negative predictive value NPV of 0·87 95% CI: 0·85-0·87 for the need to use a ventilator. Similarly, the model has an accuracy with AUC of 0·83 95% CI: 0·82-0·84, and the NPV of 0·93 95% CI: 0·92-0·94 for predicting in-hospital 30-day mortality.Interpretations: Our deep learning model has a promising predictive performance in using various laboratory markers taken from patients admitted due to COVID-19 at the initial encounter to directly inform clinical and resource management and allocations, respectively.Funding: Provided through research and collaboration grants from Siemens Healthineers, Laboratory Diagnostics, Tarrytown, New York, USA.Declaration of Interests: VS, DC, JS, ER, RM, SV, DC, and AK are employees of Siemens Healthcare, USA. All other authors have nothing to declare. Ethics Approval Statement: This study was approved by the ethics committees at Hospital University of La Paz, Emory University Hospital, and Houston Methodist Hospital.
Subject(s)
COVID-19ABSTRACT
Objective: To measure heart rate variability metrics in critically ill COVID-19 patients with comparison to all-cause critically ill sepsis patients. Design and patients: Retrospective analysis of COVID-19 patients admitted to an ICU for at least 24h at any of Emory Healthcare ICUs between March and April 2020. The comparison group was a cohort of all-cause sepsis patients prior to COVID-19 pandemic. Interventions: none. Measurements: Continuous waveforms were captured from the patient monitor. The EKG was then analyzed for each patient over a 300 second (s) observational window, that was shifted by 30s in each iteration from admission till discharge. A total of 23 HRV metrics were extracted in each iteration. We use the Kruskal-Wallis and Steel-Dwass tests (p < 0.05) for statistical analysis and interpretations of HRV multiple measures. Results: A total of 141 critically-ill COVID-19 patients met inclusion criteria, who were compared to 208 patients with all-cause sepsis. Demographic parameters were similar apart from a high proportion of African-Americans in the COVID-19 cohort. Three non-linear markers, including SD1:SD2, sample entropy, approximate entropy and four linear features mode of Beat-to-Beat interval (NN), Acceleration Capacity (AC), Deceleration Capacity (DC), and pNN50, were statistical significance between more than one binary combinations of the sub-groups (comparing survivors and non-survivors in both the COVID-19 and sepsis cohorts). The three nonlinear features and AC, DC, and NN (mode) were statistically significant across all four combinations. Temporal analysis of the main markers showed low variability across the 5 days of analysis, compared with sepsis patients. Conclusions: Heart rate variability is broadly implicated across patients infected with SARS-CoV-2, and admitted to the ICU for critical illness. Comparing these metrics to patients with all-cause sepsis suggests a unique set of expressions that differentiate this viral phenotype. This finding could be investigated further as a potential biomarker to predict poor outcome in this patient population, and could also be a starting point to measure potential autonomic dysfunction in COVID-19.
Subject(s)
COVID-19 , Critical Illness , SepsisABSTRACT
With the recent COVID-19 pandemic, healthcare systems all over the world are struggling to manage the massive increase in emergency department (ED) visits. This has put an enormous demand on medical professionals. Increased wait times in the ED increases the risk of infection transmission. In this work we present an open-source, low cost, off-body system to assist in the automatic triage of patients in the ED based on widely available hardware. The system initially focuses on two symptoms of the infection fever and cyanosis. The use of visible and far-infrared cameras allows for rapid assessment at a 1m distance, thus reducing the load on medical staff and lowering the risk of spreading the infection within hospitals. Its utility can be extended to a general clinical setting in non-emergency times as well to reduce wait time, channel the time and effort of healthcare professionals to more critical tasks and also prioritize severe cases. Our system consists of a Raspberry Pi 4, a Google Coral USB accelerator, a Raspberry Pi Camera v2 and a FLIR Lepton 3.5 Radiometry Long-Wave Infrared Camera with an associated IO module. Algorithms running in real-time detect the presence and body parts of individual(s) in view, and segments out the forehead and lip regions using PoseNet. The temperature of the forehead-eye area is estimated from the infrared camera image and cyanosis is assessed from the image of the lips in the visible spectrum. In our preliminary experiments, an accuracy of 97% was achieved for detecting fever and 77% for the detection of cyanosis, with a sensitivity of 91% and area under the receiver operating characteristic curve of 0.91. Heart rate and respiratory effort are also estimated from the visible camera. Although preliminary results are promising, we note that the entire system needs to be optimized before use and assessed for efficacy. The use of low-cost instrumentation will not produce temperature readings and identification of cyanosis that is acceptable in many situations. For this reason, we are releasing the full code stack and system design to allow others to rapidly iterate and improve the system. This may be of particular benefit in low-resource settings, and low-to-middle income countries in particular, which are just beginning to be affected by COVID-19.