ObjectivesThe rRT-PCR test, the current gold standard for the detection of coronavirus disease (COVID-19), presents with known shortcomings, such as long turnaround time, potential shortage of reagents, false-negative rates around 15–20%, and expensive equipment. The hematochemical values of routine blood exams could represent a faster and less expensive alternative.MethodsThree different training data set of hematochemical values from 1,624 patients (52% COVID-19 positive), admitted at San Raphael Hospital (OSR) from February to May 2020, were used for developing machine learning (ML) models: the complete OSR dataset (72 features: complete blood count (CBC), biochemical, coagulation, hemogasanalysis and CO-Oxymetry values, age, sex and specific symptoms at triage) and two sub-datasets (COVID-specific and CBC dataset, 32 and 21 features respectively). 58 cases (50% COVID-19 positive) from another hospital, and 54 negative patients collected in 2018 at OSR, were used for internal-external and external validation.ResultsWe developed five ML models: for the complete OSR dataset, the area under the receiver operating characteristic curve (AUC) for the algorithms ranged from 0.83 to 0.90; for the COVID-specific dataset from 0.83 to 0.87; and for the CBC dataset from 0.74 to 0.86. The validations also achieved good results: respectively, AUC from 0.75 to 0.78; and specificity from 0.92 to 0.96.ConclusionsML can be applied to blood tests as both an adjunct and alternative method to rRT-PCR for the fast and cost-effective identification of COVID-19-positive patients. This is especially useful in developing countries, or in countries facing an increase in contagions.

ObjectivesThe pandemic COVID-19 currently reached 213 countries worldwide with nearly 9 million infected people and more than 460,000 deaths. Although several Chinese studies, describing the laboratory findings characteristics of this illness have been reported, European data are still scarce. Furthermore, previous studies often analyzed the averaged laboratory findings collected during the entire hospitalization period, whereas monitoring their time-dependent variations should give more reliable prognostic information.MethodsWe analyzed the time-dependent variations of 14 laboratory parameters in two groups of COVID-19 patients with, respectively, a positive (40 patients) or a poor (42 patients) outcome, admitted to the San Raffaele Hospital (Milan, Italy). We focused mainly on laboratory parameters that are routinely tested, thus, prognostic information would be readily available even in low-resource settings.ResultsStatistically significant differences between the two groups were observed for most of the laboratory findings analyzed. We showed that some parameters can be considered as early prognostic indicators whereas others exhibit statistically significant differences only at a later stage of the disease. Among them, earliest indicators were: platelets, lymphocytes, lactate dehydrogenase, creatinine, alanine aminotransferase, C-reactive protein, white blood cells and neutrophils.ConclusionsThis longitudinal study represents, to the best of our knowledge, the first study describing the laboratory characteristics of Italian COVID-19 patients on a normalized time-scale. The time-dependent prognostic value of the laboratory parameters analyzed in this study can be used by clinicians for the effective treatment of the patients and for the proper management of intensive care beds, which becomes a critical issue during the pandemic peaks.

Background: The rRT PCR test, the current gold standard for the detection of coronavirus disease (COVID19), presents with known shortcomings, such as long turnaround time, potential shortage of reagents, false-negative rates around 15/20%, and expensive equipment. The hematochemical values of routine blood exams could represent a faster and less expensive alternative. Methods: Three different training data set of hematochemical values from 1,624 patients (52% COVID19 positive), admitted at San Raphael Hospital (OSR) from February to May 2020, were used for developing machine learning (ML) models: the complete OSR dataset (72 features: complete blood count (CBC), biochemical, coagulation, hemogasanalysis and CO-Oxymetry values, age, sex and specific symptoms at triage) and two subdatasets (COVID specific and CBC dataset, 32 and 21 features respectively). 58 cases (50% COVID19 positive) from another hospital, and 54 negative patients collected in 2018 at OSR, were used for internal external and external validation. Results: We developed five ML models: for the complete OSR dataset, the area under the receiver operating characteristic curve (AUC) for the algorithms ranged from 0.83 to 0.90; for the COVID specific dataset from 0.83 to 0.87; and for the CBC dataset from 0.74 to 0.86. The validations also achieved good results: respectively, AUC from 0.75 to 0.78; and specificity from 0.92 to 0.96. Conclusions: ML can be applied to blood tests as both an adjunct and alternative method to rRT PCR for the fast and cost-effective identification of COVID19 positive patients. This is especially useful in developing countries, or in countries facing an increase in contagions.

The Oxford-AstraZeneca ChAdOx1 nCoV-19 is a vaccine against the COVID-19 infection that was granted a conditional marketing authorization by the European Commission in January 2021. However, following a report from the Pharmacovigilance Risk Assessment Committee (PRAC) of European Medicines Agency, which reported an association with thrombo-embolic events (TEE), in particular disseminated intravascular coagulation (DIC) and cerebral venous sinus thrombosis (CVST), many European countries either limited it to individuals older than 55–60 years or suspended its use. We used publicly available data to carry out a quantitative benefit–risk analysis of the vaccine among people under 60 in Italy. Specifically, we used data from PRAC, Eudravigilance and ECDC to estimate the excess number of deaths for TEE, DIC and CVST expected in vaccine users, stratified by age groups. We then used data from the National Institute of Health to calculate age-specific COVID-19 mortality rates in Italy. Preventable deaths were calculated assuming a 72% vaccine efficacy over an eight-month period. Finally, the benefit–risk ratio of ChAdOx1 nCoV-19 vaccination was calculated as the ratio of preventable COVID-19 deaths to vaccine-related deaths, using Monte-Carlo simulations. We found that among subjects aged 20–29 years the benefit–risk (B-R) ratio was not clearly favorable (0.70; 95% Uncertainty Interval (UI): 0.27–2.11). However, in the other age groups the benefits of vaccination largely exceeded the risks (for age 30–49, B-R ratio: 22.9: 95%UI: 10.1–186.4). For age 50–59, B-R ratio: 1577.1: 95%UI: 1176.9–2121.5). Although many countries have limited the use of the ChAdOx1 nCoV-19 vaccine, the benefits of using this vaccine clearly outweigh the risks in people older than 30 years. Study limitations included risk of underreporting and that we did not provide age-specific estimates. The use of this vaccine should be a strategic and fundamental part of the immunization campaign considering its safety and efficacy in preventing COVID-19 and its complications.