ABSTRACT
During the early phase of the COVID-19 pandemic, reverse transcriptase-polymerase chain reaction (RT-PCR) testing faced limitations, prompting the exploration of machine learning (ML) alternatives for diagnosis and prognosis. Providing a comprehensive appraisal of such decision support systems and their use in COVID-19 management can aid the medical community in making informed decisions during the risk assessment of their patients, especially in low-resource settings. Therefore, the objective of this study was to systematically review the studies that predicted the diagnosis of COVID-19 or the severity of the disease using ML. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA), we conducted a literature search of MEDLINE (OVID), Scopus, EMBASE, and IEEE Xplore from January 1 to June 31, 2020. The outcomes were COVID-19 diagnosis or prognostic measures such as death, need for mechanical ventilation, admission, and acute respiratory distress syndrome. We included peer-reviewed observational studies, clinical trials, research letters, case series, and reports. We extracted data about the study's country, setting, sample size, data source, dataset, diagnostic or prognostic outcomes, prediction measures, type of ML model, and measures of diagnostic accuracy. Bias was assessed using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO), with the number CRD42020197109. The final records included for data extraction were 66. Forty-three (64%) studies used secondary data. The majority of studies were from Chinese authors (30%). Most of the literature (79%) relied on chest imaging for prediction, while the remainder used various laboratory indicators, including hematological, biochemical, and immunological markers. Thirteen studies explored predicting COVID-19 severity, while the rest predicted diagnosis. Seventy percent of the articles used deep learning models, while 30% used traditional ML algorithms. Most studies reported high sensitivity, specificity, and accuracy for the ML models (exceeding 90%). The overall concern about the risk of bias was "unclear" in 56% of the studies. This was mainly due to concerns about selection bias. ML may help identify COVID-19 patients in the early phase of the pandemic, particularly in the context of chest imaging. Although these studies reflect that these ML models exhibit high accuracy, the novelty of these models and the biases in dataset selection make using them as a replacement for the clinicians' cognitive decision-making questionable. Continued research is needed to enhance the robustness and reliability of ML systems in COVID-19 diagnosis and prognosis.
ABSTRACT
BACKGROUND: COVID-19 pandemic is disaster to public health worldwide. Better perspective on COVID's features early in its course-prior to the development of vaccines and widespread variants-may prove useful in the understanding of future pandemics. Ontology provides a standardized integrative method for knowledge modeling and computer-assisted reasoning. In this study, we systematically extracted and analyzed clinical phenotypes and comorbidities in COVID-19 patients found at different countries and regions during the early pandemic using an ontology-based bioinformatics approach, with the aim to identify new insights and hidden patterns of the COVID-19 symptoms. RESULTS: A total of 48 research articles reporting analysis of first-hand clinical data from over 40,000 COVID-19 patients were surveyed. The patients studied therein were diagnosed with COVID-19 before May 2020. A total of 18 commonly-occurring phenotypes in these COVID-19 patients were first identified and then classified into different hierarchical groups based on the Human Phenotype Ontology (HPO). This meta-analytic approach revealed that fever, cough, and the loss of smell and taste were ranked as the most commonly-occurring phenotype in China, the US, and Italy, respectively. We also found that the patients from Europe and the US appeared to have more frequent occurrence of many nervous and abdominal symptom phenotypes (e.g., loss of smell, loss of taste, and diarrhea) than patients from China during the early pandemic. A total of 22 comorbidities, such as diabetes and kidney failure, were found to commonly exist in COVID-19 patients and positively correlated with the severity of the disease. The knowledge learned from the study was further modeled and represented in the Coronavirus Infectious Disease Ontology (CIDO), supporting semantic queries and analysis. Furthermore, also considering the symptoms caused by new viral variants at the later stages, a spiral model hypothesis was proposed to address the changes of specific symptoms during different stages of the pandemic. CONCLUSIONS: Differential patterns of symptoms in COVID-19 patients were found given different locations, time, and comorbidity types during the early pandemic. The ontology-based informatics provides a unique approach to systematically model, represent, and analyze COVID-19 symptoms, comorbidities, and the factors that influence the disease outcomes.
ABSTRACT
The internet network continues to be a major source of health-related information. Our research provides insights into the online health-seeking behaviors of the general population, and evaluates the potential relationship between the COVID-19 pandemic and public interest and awareness of general sleep health, mental health and wellbeing. Google Trends' weekly relative search volumes (RSVs) were examined during 2020 for searches specifically related to COVID-19 symptoms, and for searches related to general health, sleep and wellbeing, in the United Kingdom, the United States of America, France, Italy and Japan. To obtain insight into the association between the initiation of public restrictions and online search trends, we assessed a six-week period; the 'early pandemic period' (EPP) (01 March 2020 - 11 April 2020). To provide a meaningful pre-pandemic comparison, a similar period during 2019 (03 March - 13 April 2019) was compared for RSV and median difference analysis. The EPP was associated with increased online searches related to COVID-19 symptoms, as compared with those related to more general sleep health, mental health and wellbeing. The latter search terms frequently showed a decrease or minimal change in RSV during the EPP compared with the equivalent period in 2019. This finding illustrates the potential link between the COVID-19 pandemic and online search behavior and corroborates existing findings regarding internet searches during this period. Proactive communication by healthcare professionals during future pandemics and as an ongoing measure could help prevent public neglect of general health and wellbeing symptoms, and encourage reporting and early intervention.
ABSTRACT
This study investigates the relationship between socio-economic determinants pre-dating the pandemic and the reported number of cases, deaths, and the ratio of deaths/cases in 199 countries/regions during the first months of the COVID-19 pandemic. The analysis is performed by means of machine learning methods. It involves a portfolio/ensemble of 32 interpretable models and considers the case in which the outcome variables (number of cases, deaths, and their ratio) are independent and the case in which their dependence is weighted based on geographical proximity. We build two measures of variable importance, the Absolute Importance Index (AII) and the Signed Importance Index (SII) whose roles are to identify the most contributing socio-economic factors to the variability of the COVID-19 pandemic. Our results suggest that, together with the established influence on cases and deaths of the level of mobility, the specific features of the health care system (smart/poor allocation of resources), the economy of a country (equity/non-equity), and the society (religious/not religious or community-based vs not) might contribute to the number of COVID-19 cases and deaths heterogeneously across countries.
ABSTRACT
Background: On 29th December 2019, a cluster of cases displaying the symptoms of a "pneumonia of unknown cause" was identified in Wuhan, Hubei province of China. This systematic review and meta-analysis aims to review the epidemiological and clinical characteristics of COVID-19 cases in the early phase of the COVID-19 pandemic. Methods: The search strategy involved peer-reviewed studies published between 1st January and 11th February 2020 in Pubmed, Google scholar and China Knowledge Resource Integrated database. Publications identified were screened for their title and abstracts according to the eligibility criteria, and further shortlisted by full-text screening. Three independent reviewers extracted data from these studies, and studies were assessed for potential risk of bias. Studies comprising non-overlapping patient populations, were included for qualitative and quantitative synthesis of results. Pooled prevalence with 95% confidence intervals were calculated for patient characteristics. Results: A total of 29 publications were selected after full-text review. This comprised of 18 case reports, three case series and eight cross-sectional studies on patients admitted from mid-December of 2019 to early February of 2020. A total of 533 adult patients with pooled median age of 56 (95% CI: 49-57) and a pooled prevalence of male of 60% (95% CI: 52-68%) were admitted to hospital at a pooled median of 7 days (95% CI: 7-7) post-onset of symptoms. The most common symptoms at admission were fever, cough and fatigue, with a pooled prevalence of 90% (95% CI: 81-97%), 58% (95% CI: 47-68%), and 50% (95% CI: 29-71%), respectively. Myalgia, shortness of breath, headache, diarrhea and sore throat were less common with pooled prevalence of 27% (95% CI: 20-36%), 25% (95% CI: 15-35%), 10% (95% CI: 7-13%), 8% (95% CI: 5-13%), and 7% (95% CI: 1-15%), respectively. ICU patients had a higher proportion of shortness of breath at presentation, as well as pre-existing hypertension, cardiovascular disease and COPD, compared to non-ICU patients in 2 studies (n = 179). Conclusion: This study highlights the key epidemiological and clinical features of COVID-19 cases during the early phase of the COVID-19 pandemic.