Parsimonious machine learning models to predict resource use in cardiac surgery across a statewide collaborative.

Objective: We sought to several develop parsimonious machine learning models to predict resource utilization and clinical outcomes following cardiac operations using only preoperative factors. Methods: All patients undergoing coronary artery bypass grafting and/or valve operations were identified in the 2015-2021 University of California Cardiac Surgery Consortium repository. The primary end point of the study was length of stay (LOS). Secondary endpoints included 30-day mortality, acute kidney injury, reoperation, postoperative blood transfusion and duration of intensive care unit admission (ICU LOS). Linear regression, gradient boosted machines, random forest, extreme gradient boosting predictive models were developed. The coefficient of determination and area under the receiver operating characteristic (AUC) were used to compare models. Important predictors of increased resource use were identified using SHapley summary plots. Results: Compared with all other modeling strategies, gradient boosted machines demonstrated the greatest performance in the prediction of LOS (coefficient of determination, 0.42), ICU LOS (coefficient of determination, 0.23) and 30-day mortality (AUC, 0.69). Advancing age, reduced hematocrit, and multiple-valve procedures were associated with increased LOS and ICU LOS. Furthermore, the gradient boosted machine model best predicted acute kidney injury (AUC, 0.76), whereas random forest exhibited greatest discrimination in the prediction of postoperative transfusion (AUC, 0.73). We observed no difference in performance between modeling strategies for reoperation (AUC, 0.80). Conclusions: Our findings affirm the utility of machine learning in the estimation of resource use and clinical outcomes following cardiac operations. We identified several risk factors associated with increased resource use, which may be used to guide case scheduling in times of limited hospital capacity.

Comparison between standard Vs. Escalated dose venous thromboembolism (VTE) prophylaxis in critically ill patients with COVID-19: A two centers, observational study.

INTRODUCTION: The risk of mortality in patients with COVID-19 was found to be significantly higher in patients who experienced thromboembolic events. Thus, several guidelines recommend using prophylactic anticoagulants in all COVID-19 hospitalized patients. However, there is uncertainty about the appropriate dosing regimen and safety of anticoagulation in critically ill patients with COVID-19. Thus, this study aims to compare the effectiveness and safety of standard versus escalated dose pharmacological venous thromboembolism (VTE) prophylaxis in critically ill patients with COVID-19. METHODS: A two-center retrospective cohort study including critically ill patients aged ≥ 18-years with confirmed COVID-19 admitted to the intensive care unit (ICU) at two tertiary hospitals in Saudi Arabia from March 1st, 2020, until January 31st, 2021. Patients who received either Enoxaparin 40 mg daily or Unfractionated heparin 5000 Units three times daily were grouped under the "standard dose VTE prophylaxis and patients who received higher than the standard dose but not as treatment dose were grouped under "escalated VTE prophylaxis dose". The primary outcome was the occurance of thrombotic events, and the secondary outcomes were bleeding, mortality, and other ICU-related complications. RESULTS: A total of 758 patients were screened; 565 patients were included in the study. We matched 352 patients using propensity score matching (1:1). In patients who received escalated dose pharmacological VTE prophylaxis, any case of thrombosis and VTE were similar between the two groups (OR 1.22;95 %CI 0.52-2.86; P = 0.64 and OR 0.75; 95% CI 0.16-3.38; P = 0.70 respectively). However, the odds of minor bleeding was higher in patients who received escalated VTE prophylaxis dose (OR 3.39; 95% CI 1.08-10.61; P = 0.04). There was no difference in the 30-day mortality nor in-hospital mortality between the two groups (HR 1.17;95 %CI0.79-1.73; P = 0.43 and HR 1.08;95 %CI 0.76-1.53; P = 0.83, respectively). CONCLUSION: Escalated-dose pharmacological VTE prophylaxis in critically ill patients with COVID-19 was not associated with thrombosis, or mortality benefits but led to an increased risk of minor bleeding. This study supports previous evidence regarding the optimal dosing VTE pharmacological prophylaxis regimen for critically ill patients with COVID-19.

Time to recovery from Covid-19 and its associated factors among patients hospitalized to the treatment center in South Central Ethiopia.

Coronavirus outbreak was a public health emergency. The surge of new confirmed cases and deaths was observed in developing countries due to the occurrence of new variants. However, factors associated with the duration of recovery among admitted patients remained uncertain. Therefore, we assessed factors associated with time to recovery from Covid-19 among hospitalized patients at the treatment center in South Central, Ethiopia. We employed a retrospective cross-sectional study among 422 patients hospitalized at Bokoji Hospital treatment center with Covid-19 from July 1, 2020, through October 30, 2021. Data were entered, coded, and analyzed using SPSS 26 version. We computed the survival probability using the Kaplan Meier method and determined factors associated with time to recovery using Cox regression analysis. Finally, the interpretation of adjusted hazard ratio (AHR) with 95% Confidence Interval (CI) and P-values less than 0.05 were declared as statistically significant. Our study found that the median time to recovery from Covid-19 infection of 13 days, with an IQR of 9-17 days. In multivariate Cox regression, ≥ 60 years old (AHR = 0.66; 95% CI: 0.49, 0.895), chronic pulmonary disease (AHR = 0.67; 95% CI: 0.455, 0.978), Male (AHR = 0.77; 95% CI: 0.611, 0.979), and being on Intranasal oxygen care (AHR = 0.56; 95% CI: 0.427-0.717) were significantly associated with time to recovery. Thus, health providers in treatment centers should give strict follow-up and priority for elders, patients with underlying diseases, and under supportive treatment during case management.

Variations in length of stay of inpatients with COVID-19: A nationwide test of the new model of care under vision 2030 in Saudi Arabia.

OBJECTIVE: The coronavirus disease 2019 (COVID-19) has impacted the Kingdom of Saudi Arabia (KSA) as it has other nations. However, length of stay (LOS), as a healthcare quality indicator, has not been examined across the healthcare regions in the KSA. Therefore, this study aimed to examine factors associated with LOS to better understand the Saudi Health System's performance in response to the COVID-19 pandemic in the newly suggested five Saudi regional business units (BUs). METHODS: A retrospective study was conducted using Ministry of Health (MOH) data on hospital LOS during the period from March to mid-July 2020. Participants were adult inpatients (18 years or older) with confirmed COVID-19 (n = 1743 patients). The 13 regions of the KSA were united into the defined five regional BUs during the reorganization of the health system. Covariates included demographics such as age and sex, comorbidities, and complications of COVID-19. A multiple linear regression with stepwise forward selection was used to model LOS for other explanatory variables associated with LOS, including demographic, comorbidities, and complications. RESULTS: The mean LOS was 11.85 days which differed significantly across the BUs, ranging from 9.3 days to 13.3 days (p value < 0.001). BUs differed significantly in LOS for transferred patients but not for patients in the intensive care unit (ICU) or those who died in-hospital. The multiple regression analysis revealed that the LOS for inpatients admitted in the Eastern and Southern BUs was significantly shorter than for those in the Central BU. (p value < 0.001). Admission to the ICU was associated with lengthier stays (p value < 0.0001). Factors significantly associated with shorter stays (compared to the reference), were being Saudi, death during admission, and patients referred to another hospital (p value < 0.05). CONCLUSION: The LOS for patients with COVID-19 differed across the proposed regional healthcare BUs, suggesting regional differences in quality of care under the reorganization of the national health system. Since patient and disease characteristics did not explain these findings, differences in staffing and other resources need to be examined to develop interventions.

Eleven routine clinical features predict COVID-19 severity uncovered by machine learning of longitudinal measurements.

Severity prediction of COVID-19 remains one of the major clinical challenges for the ongoing pandemic. Here, we have recruited a 144 COVID-19 patient cohort, resulting in a data matrix containing 3,065 readings for 124 types of measurements over 52 days. A machine learning model was established to predict the disease progression based on the cohort consisting of training, validation, and internal test sets. A panel of eleven routine clinical factors constructed a classifier for COVID-19 severity prediction, achieving accuracy of over 98% in the discovery set. Validation of the model in an independent cohort containing 25 patients achieved accuracy of 80%. The overall sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 0.70, 0.99, 0.93, and 0.93, respectively. Our model captured predictive dynamics of lactate dehydrogenase (LDH) and creatine kinase (CK) while their levels were in the normal range. This model is accessible at https://www.guomics.com/covidAI/ for research purpose.

Personalized Prediction of Hospital Mortality in COVID-19-Positive Patients.

OBJECTIVE: To develop predictive models for in-hospital mortality and length of stay (LOS) for coronavirus disease 2019 (COVID-19)-positive patients. PATIENTS AND METHODS: We performed a multicenter retrospective cohort study of hospitalized COVID-19-positive patients. A total of 764 patients admitted to 14 different hospitals within the Cleveland Clinic from March 9, 2020, to May 20, 2020, who had reverse transcriptase-polymerase chain reaction-proven coronavirus infection were included. We used LightGBM, a machine learning algorithm, to predict in-hospital mortality at different time points (after 7, 14, and 30 days of hospitalization) and in-hospital LOS. Our final cohort was composed of 764 patients admitted to 14 different hospitals within our system. RESULTS: The median LOS was 5 (range, 1-44) days for patients admitted to the regular nursing floor and 10 (range, 1-38) days for patients admitted to the intensive care unit. Patients who died during hospitalization were older, initially admitted to the intensive care unit, and more likely to be white and have worse organ dysfunction compared with patients who survived their hospitalization. Using the 10 most important variables only, the final model's area under the receiver operating characteristics curve was 0.86 for 7-day, 0.88 for 14-day, and 0.85 for 30-day mortality in the validation cohort. CONCLUSION: We developed a decision tool that can provide explainable and patient-specific prediction of in-hospital mortality and LOS for COVID-19-positive patients. The model can aid health care systems in bed allocation and distribution of vital resources.