At-Admission Prediction of Mortality and Pulmonary Embolism in COVID-19 Patients Using Statistical and Machine Learning Methods: An International Cohort Study (preprint)

By September, 2022, more than 600 million cases of SARS-CoV-2 infection have been reported globally, resulting in over 6.5 million deaths. COVID-19 mortality risk estimators are often, however, developed with small unrepresentative samples and with methodological limitations. It is highly important to develop predictive tools for pulmonary embolism (PE) in COVID-19 patients as one of the most severe preventable complications of COVID-19. Using a dataset of more than 800,000 COVID-19 patients from an international cohort, we propose a cost-sensitive gradient-boosted machine learning model that predicts occurrence of PE and death at admission. Logistic regression, Cox proportional hazards models, and Shapley values were used to identify key predictors for PE and death. Our prediction model had a test AUROC of 75.9% and 74.2%, and sensitivities of 67.5% and 72.7% for PE and all-cause mortality respectively on a highly diverse and held-out test set. The PE prediction model was also evaluated on patients in UK and Spain separately with test results of 74.5% AUROC, 63.5% sensitivity and 78.9% AUROC, 95.7% sensitivity. Age, sex, region of admission, comorbidities (chronic cardiac and pulmonary disease, dementia, diabetes, hypertension, cancer, obesity, smoking), and symptoms (any, confusion, chest pain, fatigue, headache, fever, muscle or joint pain, shortness of breath) were the most important clinical predictors at admission. Our machine learning model developed from an international cohort can serve to better regulate hospital risk prioritisation of at-risk patients.

Liver injury in hospitalized patients with COVID-19: An International observational cohort study (preprint)

Background: Using a large dataset, we evaluated prevalence and severity of alterations in liver enzymes in COVID-19 and association with patient-centred outcomes. Methods: We included hospitalized patients with confirmed or suspected SARS-CoV-2 infection from the International Severe Acute Respiratory and emerging Infection Consortium (ISARIC) database. Key exposure was baseline liver enzymes (AST, ALT, bilirubin). Patients were assigned Liver Injury Classification score based on 3 components of enzymes at admission: Normal; Stage I) Liver injury: any component between 1-3x upper limit of normal (ULN); Stage II) Severe liver injury: any component >= 3x ULN. Outcomes were hospital mortality, utilization of selected resources, complications, and durations of hospital and ICU stay. Analyses used logistic regression with associations expressed as adjusted odds ratios (OR) with 95% confidence intervals (CI). Results: Of 17,531 included patients, 46.2% (8099) and 8.2% (1430) of patients had stage 1 and 2 liver injury respectively. Compared to normal, stages 1 and 2 were associated with higher odds of mortality (OR 1.53 [1.37-1.71]; OR 2.50 [2.10-2.96]), ICU admission (OR 1.63 [1.48-1.79]; OR 1.90 [1.62-2.23]) and invasive mechanical ventilation (OR 1.43 [1.27-1.70]; OR 1.95 (1.55-2.45).Stages 1 and 2 were also associated with higher odds of developing sepsis (OR 1.38 [1.27-1.50]; OR 1.46 [1.25-1.70]), acute kidney injury (OR 1.13 [1.00-1.27]; OR 1.59 [1.32-1.91]), and acute respiratory distress syndrome (OR 1.38 [1.22-1.55]; OR 1.80 [1.49-2.17]). Conclusions: Liver enzyme abnormalities are common among COVID-19 patients and associated with worse outcomes.

An international observational study to assess the impact of the Omicron variant emergence on the clinical epidemiology of COVID-19 in hospitalised patients (preprint)

BackgroundWhilst timely clinical characterisation of infections caused by novel SARS-CoV-2 variants is necessary for evidence-based policy response, individual-level data on infecting variants are typically only available for a minority of patients and settings. MethodsHere, we propose an innovative approach to study changes in COVID-19 hospital presentation and outcomes after the Omicron variant emergence using publicly available population-level data on variant relative frequency to infer SARS-CoV-2 variants likely responsible for clinical cases. We apply this method to data collected by a large international clinical consortium before and after the emergence of the Omicron variant in different countries. ResultsOur analysis, that includes more than 100,000 patients from 28 countries, suggests that in many settings patients hospitalised with Omicron variant infection less often presented with commonly reported symptoms compared to patients infected with pre-Omicron variants. Patients with COVID-19 admitted to hospital after Omicron variant emergence had lower mortality compared to patients admitted during the period when Omicron variant was responsible for only a minority of infections (odds ratio in a mixed-effects logistic regression adjusted for likely confounders, 0.67 [95% confidence interval 0.61 - 0.75]). Qualitatively similar findings were observed in sensitivity analyses with different assumptions on population-level Omicron variant relative frequencies, and in analyses using available individual-level data on infecting variant for a subset of the study population. ConclusionsAlthough clinical studies with matching viral genomic information should remain a priority, our approach combining publicly available data on variant frequency and a multi-country clinical characterisation dataset with more than 100,000 records allowed analysis of data from a wide range of settings and novel insights on real-world heterogeneity of COVID-19 presentation and clinical outcome.

Use of an extended KDIGO definition to diagnose acute kidney injury in patients with COVID-19: A multinational study using the ISARIC–WHO clinical characterisation protocol

Background Acute kidney injury (AKI) is one of the most common and significant problems in patients with Coronavirus Disease 2019 (COVID-19). However, little is known about the incidence and impact of AKI occurring in the community or early in the hospital admission. The traditional Kidney Disease Improving Global Outcomes (KDIGO) definition can fail to identify patients for whom hospitalisation coincides with recovery of AKI as manifested by a decrease in serum creatinine (sCr). We hypothesised that an extended KDIGO (eKDIGO) definition, adapted from the International Society of Nephrology (ISN) 0by25 studies, would identify more cases of AKI in patients with COVID-19 and that these may correspond to community-acquired AKI (CA-AKI) with similarly poor outcomes as previously reported in this population. Methods and findings All individuals recruited using the International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC)–World Health Organization (WHO) Clinical Characterisation Protocol (CCP) and admitted to 1,609 hospitals in 54 countries with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection from February 15, 2020 to February 1, 2021 were included in the study. Data were collected and analysed for the duration of a patient’s admission. Incidence, staging, and timing of AKI were evaluated using a traditional and eKDIGO definition, which incorporated a commensurate decrease in sCr. Patients within eKDIGO diagnosed with AKI by a decrease in sCr were labelled as deKDIGO. Clinical characteristics and outcomes—intensive care unit (ICU) admission, invasive mechanical ventilation, and in-hospital death—were compared for all 3 groups of patients. The relationship between eKDIGO AKI and in-hospital death was assessed using survival curves and logistic regression, adjusting for disease severity and AKI susceptibility. A total of 75,670 patients were included in the final analysis cohort. Median length of admission was 12 days (interquartile range [IQR] 7, 20). There were twice as many patients with AKI identified by eKDIGO than KDIGO (31.7% versus 16.8%). Those in the eKDIGO group had a greater proportion of stage 1 AKI (58% versus 36% in KDIGO patients). Peak AKI occurred early in the admission more frequently among eKDIGO than KDIGO patients. Compared to those without AKI, patients in the eKDIGO group had worse renal function on admission, more in-hospital complications, higher rates of ICU admission (54% versus 23%) invasive ventilation (45% versus 15%), and increased mortality (38% versus 19%). Patients in the eKDIGO group had a higher risk of in-hospital death than those without AKI (adjusted odds ratio: 1.78, 95% confidence interval: 1.71 to 1.80, p-value < 0.001). Mortality and rate of ICU admission were lower among deKDIGO than KDIGO patients (25% versus 50% death and 35% versus 70% ICU admission) but significantly higher when compared to patients with no AKI (25% versus 19% death and 35% versus 23% ICU admission) (all p-values <5 × 10−5). Limitations include ad hoc sCr sampling, exclusion of patients with less than two sCr measurements, and limited availability of sCr measurements prior to initiation of acute dialysis. Conclusions An extended KDIGO definition of AKI resulted in a significantly higher detection rate in this population. These additional cases of AKI occurred early in the hospital admission and were associated with worse outcomes compared to patients without AKI.

ISARIC COVID-19 Clinical Data Report: 8 June 2020 (preprint)

ISARIC (International Severe Acute Respiratory and emerging Infections Consortium) partnerships and outbreak preparedness initiatives enabled the rapid launch of standardised clinical data collection on COVID-19 in Jan 2020. Extensive global uptake of this resource has resulted in a large, standardised collection of comprehensive clinical data from hundreds of sites across dozens of countries. Data are analysed regularly and reported publicly to inform patient care and public health response. This report is a part of a series and includes the results of data analysis on 8 June 2020. We thank all of the data contributors for their ongoing support. As of 8JUN20, data have been entered for 67,130 patients from 488 sites across 37 countries. For this report, we show data for 42,656 patients with confirmed disease who were enrolled >14 days prior. This update includes about 2,400 new cases from France, and we thank these collaborators for this significant addition to the dataset. Some highlights from this report The median time from onset of symptoms to hospital admission is 5 days, but a proportion of patients take longer to get to the hospital (average 14.6 days, standard deviation 8.1). COVID-19 patients tend to require prolonged hospitalisation; of the 88% with a known outcome, the median length of admission to death or discharge is 8 days and the mean 11.5. 17% of patients were admitted to ICU/HDU, about 40% of these on the very day of hospital admission. Antibiotics were given to 83% of patients, antivirals to 9%, steroids to 15%, which becomes 93%, 50% and 27%, respectively for those admitted to ICU/HDU. Attention has been called on overuse of antibiotics and need to adhere to antibiotic stewardship principles. 67% of patients received some degree of oxygen supplementation: of these 23.4% received NIV and 15% IMV. This relatively high proportion of oxygen use will have implications for oxygen surge planning in healthcare facilities. Some centres may need to plan to boost capacity to deliver oxygen therapy if this is not readily available. WHO provides operational advice on surge strategy here https://apps.who.int/iris/bitstream/handle/10665/331746/WHO-2019-nCoV-Oxygen_sources-2020.1-eng.pdf