Risk factors of in-hospital mortality for isolated on-pump coronary artery bypass graft surgery in the northeast of Iran from 2007 to 2016.

BACKGROUND AND OBJECTIVE: Coronary artery bypass graft (CABG) surgery is the most common cardiac surgery worldwide. The reported mortality rates for this operation vary greatly. We aimed to determine the risk factors of in-hospital mortality for isolated on-pump CABG surgery. METHODS: This was a large-scale retrospective cohort study of two heart centers in Golestan province. Patients over the age of 18 from both genders who underwent isolated on-pump CABG procedures from 2007 to 2016 were included. The study outcome was in-hospital mortality, which was determined according to the clinical records of study patients. RESULTS: A total of 3704 patients were included in the study, and 63% were men. In-hospital mortality occurred in 2.8% (n=103) of the patients. The median (IQR) age of survived and not-survived patients were 59 (53-65) and 62 (55-75) years, respectively. 44% of the mortalities occurred in patients older than 65, while 28% of the survivors were older than 65. Multivariable logistic regression indicated that emergency CABG (OR 4.52, 95% CI, 1.45, 14.02; P = 0.009) and cardiopulmonary bypass time (CPB) (OR 1.004, 95% CI 1.001, 1.008; P = 0.034) were the risk factors of in-hospital mortality. The area under the receiver operating characteristic (ROC) curve (AUC) of the model consisting of operative and preoperative variables was 0.70 (acceptable performance). CONCLUSION: Our study revealed an acceptable mortality proportion for CABG surgeries conducted in the region. Emergency CABG and CPB time were the main risk factors for in-hospital mortality after CABG.

Prediction of COVID-19 Patients' Survival by Deep Learning Approaches.

Background: Despite many studies done to predict severe coronavirus 2019 (COVID-19) patients, there is no applicable clinical prediction model to predict and distinguish severe patients early. Based on laboratory and demographic data, we have developed and validated a deep learning model to predict survival and assist in the triage of COVID-19 patients in the early stages. Methods: This retrospective study developed a survival prediction model based on the deep learning method using demographic and laboratory data. The database consisted of data from 487 patients with COVID-19 diagnosed by the reverse transcription-polymerase chain reaction test and admitted to Imam Khomeini hospital affiliated to Tehran University of Medical Sciences from February 21, 2020, to June 24, 2020. Results: The developed model achieved an area under the curve (AUC) of 0.96 for survival prediction. The results demonstrated the developed model provided high precision (0.95, 0.93), recall (0.90,0.97), and F1-score (0.93,0.95) for low- and high-risk groups. Conclusion: The developed model is a deep learning-based, data-driven prediction tool that can predict the survival of COVID-19 patients with an AUC of 0.96. This model helps classify admitted patients into low-risk and high-risk groups and helps triage patients in the early stages.

A systematic review of the predicted outcomes related to hematopoietic stem cell transplantation: focus on applied machine learning methods' performance.

INTRODUCTION: Hematopoietic stem cell transplantation (HSCT) is a critical therapeutic procedure in blood diseases, and the investigation of HSCT data can provide valuable information. Machine learning (ML) techniques are useful data analysis tools which applied in many studies to predict HSCT survival and estimate the risk of transplantation. AREAS COVERED: A systematic review was performed with a search of PubMed, Science Direct, Embase, Scopus, and the European Society for Blood and Marrow Transplantation, the Center for International Blood and Marrow Transplant Research, and the American Society for Transplantation and Cellular Therapy publications for articles published by September 2020. EXPERT OPINION: 24 papers that met eligibility criteria were included in this study. The applied ML algorithms with the highest performance were Random Survival Forests (AUC = 0.72) for survival-related, Random Survival Forests and Logistic Regression (AUC = 0.77) for mortality-related, Deep Learning (AUC = 0.8) for relapse, L2-Regularized Logistic Regression (AUC = 0.66) for Acute-Graft Versus Host Disease, Random Survival Forests (AUC = 0.88) for sepsis, Elastic-Net Regression (AUC = 0.89) for cognitive impairment, and Bayesian Network (AUC = 0.997) for oral mucositis outcome. This review reveals the potential of ML techniques to predict HSCT outcomes and apply them to developing clinical decision support systems.

eHealth solutions to fight against COVID-19: A scoping review of applications.

Background: eHealth has a notable potential to help in prevention, diagnosis, treatment, screening, management, and control of the COVID-19 pandemic. Since ehealth is considered here broadly, as an umbrella term, it also covers subsets like telehealth and mhealth. This study aimed to review the literature to identify and classify subdomains of eHealth solutions that have been utilized, developed, or suggested for the COVID-19 pandemic. Methods: A comprehensive literature search was performed using the PubMed, Scopus, Embase, and Cochrane library databases in April 2020, with no time limitation. The search strategy was built based on 2 concept domains of eHealth solutions and covid-19. For each concept domain, the search query comprised a combination of free text keywords identified from reference papers and controlled vocabulary terms. Obtained results were classified, graphically presented, and discussed. Results: Of the 423 studies identified initially, 35 were included in this study. From related papers, general characteristics, study objective, eHealth-related outcomes, target populations, eHealth interventions, health service category, eHealth solution, and eHealth domain were extracted, classified, and tabulated. Most publication types were ideas, editorials, or opinions (46%). The most targeted populations were people of the community and medical staff (80%). The most implemented or suggested eHealth solution was telehealth (63%), followed by mhealth, health information technology, and health data analytics. Most of the COVID-19 ehealth interventions designed or suggested for improving prevention (48%) and diagnosis (48%). Most of the studies applied or proposed eHealth solutions for general practice or epidemiological purposes (48%). Conclusion: eHealth solutions have the potential to provide useful services to help in COVID-19 pandemics in terms of prevention, diagnosis, treatment, screening, surveillance, resource allocation, education, management, and control. The obtained results from this review might be used for a better understanding of current ehealth solutions provided or recommended in response to the COVID-19 pandemic.