Prediction of Conventional Oxygen Therapy Failure in COVID-19 Patients With Acute Respiratory Failure by Assessing Serum Lactate Concentration, PaO2/FiO2 Ratio, and Body Temperature

Association between (ΔPaO2/FiO2)/PEEP and in-hospital mortality in patients with COVID-19 pneumonia: A secondary analysis

BACKGROUND

The arterial pressure of oxygen (PaO2)/inspiratory fraction of oxygen (FiO2) is associated with in-hospital mortality in patients with Coronavirus Disease 2019 (COVID-19) pneumonia. ΔPaO2/FiO2 [the difference between PaO2/FiO2 after 24 h of invasive mechanical ventilation (IMV) and PaO2/FiO2 before IMV] is associated with in-hospital mortality. However, the value of PaO2 can be influenced by the end-expiratory pressure (PEEP). To the best of our knowledge, the relationship between the ratio of (ΔPaO2/FiO2)/PEEP and in-hospital mortality remains unclear. This study aimed to evaluate their association.

METHODS

The study was conducted in southern Peru from April 2020 to April 2021. A total of 200 patients with COVID-19 pneumonia requiring IMV were included in the present study. We analyzed the association between (ΔPaO2/FiO2)/PEEP and in-hospital mortality by Cox proportional hazards regression models.

RESULTS

The median (ΔPaO2/FiO2)/PEEP was 11.78 mmHg/cmH2O [interquartile range (IQR) 8.79-16.08 mmHg/cmH2O], with a range of 1 to 44.36 mmHg/cmH2O. Patients were divided equally into two groups [low group (< 11.80 mmHg/cmH2O), and high group (≥ 11.80 mmHg/cmH2O)] according to the (ΔPaO2/FiO2)/PEEP ratio. In-hospital mortality was lower in the high (ΔPaO2/FiO2)/PEEP group than in the low (ΔPaO2/FiO2)/PEEP group [18 (13%) vs. 38 (38%)]; hazard ratio (HR), 0.33 [95% confidence intervals (CI), 0.17-0.61, P < 0.001], adjusted HR, 0.32 (95% CI, 0.11-0.94, P = 0.038). The finding that the high (ΔPaO2/FiO2)/PEEP group exhibited a lower risk of in-hospital mortality compared to the low (ΔPaO2/FiO2)/PEEP group was consistent with the results from the sensitivity analysis. After adjusting for confounding variables, we found that each unit increase in (ΔPaO2/FiO2)/PEEP was associated with a 12% reduction in the risk of in-hospital mortality (HR, 0.88, 95%CI, 0.80-0.97, P = 0.013).

CONCLUSIONS

The (ΔPaO2/FiO2)/PEEP ratio was associated with in-hospital mortality in patients with COVID-19 pneumonia. (ΔPaO2/FiO2)/PEEP might be a marker of disease severity in COVID-19 patients.

Machine learning-based prediction of hospital prolonged length of stay admission at emergency department: a Gradient Boosting algorithm analysis

Objective

This study aims to develop and compare different models to predict the Length of Stay (LoS) and the Prolonged Length of Stay (PLoS) of inpatients admitted through the emergency department (ED) in general patient settings. This aim is not only to promote any specific model but rather to suggest a decision-supporting tool (i.e., a prediction framework).

Methods

We analyzed a dataset of patients admitted through the ED to the "Sant"Orsola Malpighi University Hospital of Bologna, Italy, between January 1 and October 26, 2022. PLoS was defined as any hospitalization with LoS longer than 6 days. We deployed six classification algorithms for predicting PLoS: Random Forest (RF), Support Vector Machines (SVM), Gradient Boosting (GB), AdaBoost, K-Nearest Neighbors (KNN), and logistic regression (LoR). We evaluated the performance of these models with the Brier score, the area under the ROC curve (AUC), accuracy, sensitivity (recall), specificity, precision, and F1-score. We further developed eight regression models for LoS prediction: Linear Regression (LR), including the penalized linear models Least Absolute Shrinkage and Selection Operator (LASSO), Ridge and Elastic-net regression, Support vector regression, RF regression, KNN, and eXtreme Gradient Boosting (XGBoost) regression. The model performances were measured by their mean square error, mean absolute error, and mean relative error. The dataset was randomly split into a training set (70%) and a validation set (30%).

Results

A total of 12,858 eligible patients were included in our study, of whom 60.88% had a PloS. The GB classifier best predicted PloS (accuracy 75%, AUC 75.4%, Brier score 0.181), followed by LoR classifier (accuracy 75%, AUC 75.2%, Brier score 0.182). These models also showed to be adequately calibrated. Ridge and XGBoost regressions best predicted LoS, with the smallest total prediction error. The overall prediction error is between 6 and 7 days, meaning there is a 6-7 day mean difference between actual and predicted LoS.

Conclusion

Our results demonstrate the potential of machine learning-based methods to predict LoS and provide valuable insights into the risks behind prolonged hospitalizations. In addition to physicians' clinical expertise, the results of these models can be utilized as input to make informed decisions, such as predicting hospitalizations and enhancing the overall performance of a public healthcare system.

IoT-based wearable health monitoring device and its validation for potential critical and emergency applications

The COVID-19 pandemic brought the world to a standstill, posing unprecedented challenges for healthcare systems worldwide. The overwhelming number of patients infected with the virus placed an enormous burden on healthcare providers, who struggled to cope with the sheer volume of cases. Furthermore, the lack of effective treatments or vaccines means that quarantining has become a necessary measure to slow the spread of the virus. However, quarantining places a significant burden on healthcare providers, who often lack the resources to monitor patients with mild symptoms or asymptomatic patients. In this study, we propose an Internet of Things (IoT)-based wearable health monitoring system that can remotely monitor the exact locations and physiological parameters of quarantined individuals in real time. The system utilizes a combination of highly miniaturized optoelectronic and electronic technologies, an anti-epidemic watch, a mini-computer, and a monitor terminal to provide real-time updates on physiological parameters. Body temperature, peripheral oxygen saturation (SpO₂), and heart rate are recorded as the most important measurements for critical care. If these three physiological parameters are aberrant, then it could represent a life-endangering situation and/or a short period over which irreversible damage may occur. Therefore, these parameters are automatically uploaded to a cloud database for remote monitoring by healthcare providers. The monitor terminal can display real-time health data for multiple patients and provide early warning functions for medical staff. The system significantly reduces the burden on healthcare providers, as it eliminates the need for manual monitoring of patients in quarantine. Moreover, it can help healthcare providers manage the COVID-19 pandemic more effectively by identifying patients who require medical attention in real time. We have validated the system and demonstrated that it is well suited to practical application, making it a promising solution for managing future pandemics. In summary, our IoT-based wearable health monitoring system has the potential to revolutionize healthcare by providing a cost-effective, remote monitoring solution for patients in quarantine. By allowing healthcare providers to monitor patients remotely in real time, the burden on medical resources is reduced, and more efficient use of limited resources is achieved. Furthermore, the system can be easily scaled to manage future pandemics, making it an ideal solution for managing the health challenges of the future.

Comparison of COVID-19 survival in relation to CPAP length of treatment and by comorbidity and transmission setting (community or hospital acquired) in a medium-sized UK hospital in 2020: a retrospective study

OBJECTIVE

To estimate continuous positive airway pressure (CPAP) length of treatment effect on survival of hospitalised COVID-19 patients in a medium-sized UK Hospital, and how this effect changes according to the patient's comorbidity and COVID-19 route of acquisition (community or nosocomial) during the two waves in 2020.

SETTING

The acute inpatient unit in Wrightington, Wigan and Leigh Teaching Hospitals National Health Service (NHS) Foundation Trust (WWL), a medium-sized NHS Trust in north-west of England.

DESIGN

Retrospective cohort of all confirmed COVID-19 patients admitted in WWL during 2020.

PARTICIPANTS

1830 patients (568 first wave, 1262 s wave) with antigen confirmed COVID-19 disease and severe acute respiratory syndrome admitted between 17 March 2020 (first confirmed COVID-19 case) and 31 December 2020.

OUTCOME MEASURE

COVID-19 survival rate in all patients and survival rate in potentially hospital-acquired COVID-19 (PHA) patients were modelled using a predictor set which include comorbidities (eg, obesity, diabetes, chronic ischaemic heart disease (IHD), chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD)), wave, age, sex and care home residency, and interventions (remdesivir, dexamethasone, CPAP, intensive care unit (ICU), intubation). Secondary outcome measure was CPAP length, which was modelled using the same predictors of the survival rate.

RESULTS

Mortality rate in the second wave was significantly lower than in the first wave (43.4% vs 28.1%, p<0.001), although for PHA COVID-19 patients mortality did not reduce, remaining at very high levels independently of wave and CPAP length. For all cohort, statistical modelling identified CPAP length (HR 95% CI 0.86 to 0.96) and women (HR 95% CI 0.71 to 0.81) were associated with improved survival, while being older age (HR 95% CI 1.02 to 1.03) admitted from care homes (HR 95% CI 2.22 to 2.39), IHD (HR 95% CI 1.13 to 1.24), CKD (HR 95% CI 1.14 to 1.25), obesity (HR 95% CI 1.18 to 1.28) and COPD-emphysema (HR 95% CI 1.18 to 1.57) were associated with reduced survival. Despite the detrimental effect of comorbidities, patients with CKD (95% CI 16% to 30% improvement in survival), IHD (95% CI 1% to 10% improvement in survival) and asthma (95% CI 8% to 30% improvement in survival) benefitted most from CPAP length, while no significant survival difference was found for obese and patients with diabetes.

CONCLUSIONS

The experience of an Acute Trust during the COVID-19 outbreak of 2020 is documented and indicates the importance of care home and hospitals in disease acquisition. Death rates fell between the first and second wave only for community-acquired COVID-19 patients. The fall was associated to CPAP length, especially for some comorbidities. While uncovering some risk and protective factors of mortality in COVID-19 studies, the study also unravels how little is known about PHA COVID-19 and the interaction between CPAP and some comorbidities.