Distance traveled in times of pandemic: an endogenous switching regression approach

This paper studies the change in the distance traveled by domestic tourists considering the pre- and post-pandemic outbreak summer periods of 2019 and 2020. Using representative monthly microdata involving more than 31,000 trips conducted by Spanish residents, we examine the heterogeneity in behavioral adaptation to COVID-19 based on sociodemographic and trip-related characteristics. To account for selection effects and the potential change in the population composition of travelers between the two periods, we estimate an endogenous switching regression that conducts separate regressions for the pre- and post-pandemic periods in a unified econometric framework. Our results point to heterogeneous shifts in the distance traveled by domestic travelers after COVID-19 outbreak per sociodemographic group, with notable differences by travel purpose and lower relevance of traditional determinants like income.

Surfactant protein D (SP-D) as biomarker of SARSCOV-2 infection

Beside lowering the surface tension at air-liquid interface in the alveoli, the pulmonary surfactant has a pivotal role in triggering the elimination of pathogens or any hazardous materials introduced with breathing. Among the components of the pulmonary surfactant, surfactant protein-D (SP-D) is a low abundant (0.6%) hydrophilic protein that is able to promote pathogens clearance binding highly conserved glycosidic residues on their surface. SP-D also cooperates in the maintenance of lung homeostasis by directly modulating immune system activity. Previous investigations on acute respiratory distress syndrome (ARDS) patients demonstrated a significant increment of SP-D serum level compared with healthy donors. Since in physiological condition SP-D is not permeable to alveoli-capillary membrane and poorly express by other tissues, this enhancement is likely due to an impairment of the pulmonary barrier caused by prolonged inflammation. In view of the above, the present work aims to investigate SP-D as diagnostic and/or prognostic marker for COVID-19. In particular, a retrospective study on a relatively large cohort of patients of Hospital Pio XI of Desio (i.e., 79 mild cases plus 123 severe cases) was conducted to assess differences of the hematic levels of this biomarker among COVID-19 patients and healthy donors and if SP-D serum levels resulted a risk factor for disease severity and mortality. The performed analyses, using an Anova-Mixed model, showed a significant difference in the mean of log SP-D between COVID-19 patients and healthy donors: 150 ng/mL was identified as threshold value to best discriminate the mentioned groups. Significant differences were also found between dead vs survived patients, as well among severe vs non-severe cases. In all cases, SP-D serum levels presented significantly higher values for COVID-19 patients, dead and severe cases.Moreover, further analysis conducted with Logistic Mixed models, highlighted that SP-D, in a model with Age, C-reactive protein and cancer status, resulted the strongest significant risk factor of mortality (model predictive accuracy, AUC=0.826), and in a lesser extent for risk of severity.The overall data suggest that SP-D can be a predictive marker of COVID-19 disease and its outcome.

Does PLR reflect inflammation from chronic hyperglycemia?

BACKGROUND Recent studies have examined the relationship between stress-induced hyperglycemia, determined by the glycemic gap between admission glucose levels and average glucose levels derived from A1c, and disease severity and/or unfavorable clinical outcomes. Diabetes patients who have chronic hyperglycemia experience identical consequences to those caused by stressinduced hyperglycemia, including increased oxidative stress, inflammation, and the activation of stressresponsive kinases. The Platelet to Lymphocyte Ratio (PLR) has recently been proven as an inflammatory marker during the Covid-19 outbreak. The purpose of this study was to assess how the glycemic gap and PLR, a measure of inflammation in chronic hyperglycemia, correlate with one another. METHODS Retrospective data from 696 outpatients with medically declared diabetes from the year 2019 was obtained. For the results of the CBC and HbA1c, all records were complete. Only 638 of the 696 records contained blood glucose data. PLR was computed for each of the 696. For 638 cases, the estimated A1c-Derived Average Glucose (ADAG) level was computed using the formula 28.7 x HbA1c(%) x 46.7 [1], and was then subtracted from the blood glucose level to measure the glycemic gap. According to the percentages of HbA1c (4,5-6,4;6,5-6,9;7,0-8,0;>8,0), four groups were formed, and the average PLR and glycemic gap were determined for each group. The data was then evaluated to see how HbA1c and PLR related to glycemic gap. RESULTS PLR and glycemic gap were 119,45 and 13,76 in group 1 (HbA1c % 4,5-6,4;n = 223), 112,50 and 16,79 in group 2 (HbA1c % 6,5-6,9;n = 144), 119,27 and 25,23 in group 3 (HbA1c % 7,0-7,9;n=201), and 115,88 and 46,04 in group 4 (HbA1c % > 8;n=114), respectively. By using correlation analysis, we found that the PLR increased proportionally to the glycemic gap but did not rise with rising HbA1c. CONCLUSIONS The finding of proportionally higher PLR revealed a correlation between a higher glycemic gap, a sign of chronically induced hyperglycemia, and an elevated pro-inflammatory state.

Identification of SARS-CoV-2 positivity using machine learning methods on blood count data: External validation of state-of-the-art models

Introduction: The aim of the paper is to present the results from the process of external validation of a number of machine learning (ML) models that had been previously developed to detect SARS-CoV-2 virus positivity on both symptomatic and asymptomatic patients on the basis of the complete blood count (CBC) test. Methods: Briefly, models were trained using a dataset of 816 COVID-19 positive and 920 negative cases collected at the emergency departments of IRCCS Hospital San Raffaele and IRCCS Istituto Ortopedico Galeazzi. 21 parameters, including the results of the CBC analysis, age [60.9 (0.9) years], gender (57% males) and the presence of COVID-19 related symptoms were used. The validation regarded the evaluation of the error rate (through different metrics, including accuracy, sensitivity, specificity and the area under the curve (AUC)) of the models considered. This external validation was conducted on two well balanced datasets coming from two different hospitals in Northern Italy: Desio hospital and Bergamo Papa Giovanni XXIII hospital. 163 positive and 174 true negative patients from Desio, and 104 positive and 145 true negative from Bergamo were included in the validation. Results: The performance of the predictive models is satisfactory as we can report an average AUC of 95% for both external datasets. Conclusion: ML models have been applied to hematological parameters for a more rapid and cost-effective detection of the COVID-19 disease. We make the point that validated models may be useful in the management and early detection of potential COVID-19 patients.

Laboratory findings of COVID-19 patients admitted to Desio Hospital: A retrospective study

Background Italy was one of the hardest hit European countries by the novel coronavirus SARS-CoV-2 epidemic.1,2The aim of the study was to evaluate the clinical and laboratory characteristics of some cases with coronavirus disease 2019 (COVID-19). Methods A retrospective study was conducted on 175 confirmed COVID-19 cases, admitted during March(follow-up through April 30th, 2020) 2020 to Desio Hospital, and were classified into non-survivors (n=62) and survivors (n=113), according to mortality outcome. Results The median age was 71 years (IQR 58-80). 70% were males and 23% were admitted to ICU. The most frequent symptoms were dyspnea, fever, and cough. CBC showed leukocytosis in 18% and leukopenia in 14%. Lymphocytopenia (55%) and thrombocytopenia (20%) were also observed.Of the 62 non-survivors, 61 were classified as severe (p<0.001).3 The non-survivors were older (80 vs 62 yrs;p<0.01), had more comorbidities (93% vs 73%;p<0.01) such as cardiovascular diseases and diabetes. Their SpO2 (86% vs. 92%;p<0.01) and PaO2/FIO2 (307 vs 206;p<0.001) were lower. Compared to survivors, the non-survivors had a significantly higher neutrophil count, a higher N/L ratio and lower hemoglobin values (p<0.01). Furthermore, leukocytosis (25.8% vs 13.3%;p<0.05) and lymphocytopenia (79% vs 43%;p<0.01) were more frequent. Indices of liver damage (AST), renal dysfunction (creatinine and urea), inflammation (CRP and PCT), and coagulation function (aPTT, INR, and D-Dimer), plus creatinine kinase (CK) and hs-troponin T were higher in non-survivors (p<0.05). DiscussionAmong non-survivors, leukocytosis was more frequent than leucopenia. Neutrophil count, N/L ratio, and lymphopenia were greater in accordance with other studies.4,5 The increased N/L ratio and CRP might be related to cytokine storms, characterized by massive recruitment of neutrophils in the lung interstitium and reduced lymphocyte control.3,6 Conclusions Multiple factors such as older age, comorbidities, higher blood leukocyte count, neutrophil count, PCR, D-dimer, AST, creatinine, LDH, CK and hs-troponin T, and lower lymphocyte count were related to non-survivors (p<0.05).

Metabolomic changes in COVID-19 patients

The lipid structure of membranes regulates the penetration and replication of SARS-CoV-2 in the alveolar pneumocytes. Specific side-chain oxysterols generated by enzymatic activity such as 25-hydroxycholesterol (25OHC) and 27-hydroxycholesterol (27OHC) have inhibitory effect on viral replication. Reduction of polyunsaturated fatty acids, such as arachidonic acid (AA) and decosahexaenoic acid (DHA) were related to increased viral replication. By isotope dilution mass spectrometry, we investigated the lipid profile (fatty acids, sterols and oxysterols) in plasma collected from 27 mild SARSCoV-2 positive subjects and 117 COVID-19 patients with moderate (n=36) and severe stages (n=81) of the disease, compared with 123 age matched healthy volunteers. Cholesterol was reduced of 20% (P<0.001) in moderate and 25% (P<0.001) in severe patients. The cholesterol precursors desmosterol and lathosterol, marker of cholesterol synthesis, were reduced from 50 up to 70% in moderate and severe patients (P<0.001). While 25OHC was increased by 15% (P = 0.03) in the mild individuals and reduced of about 13% in the severe patients (P<0.001), the antiviral 27OHC was reduced by 17% in the group of pauci-asymptomatic (P<0.001), becoming 30% (P<0.001) in the moderate and 50% (P<0.001) in the severe COVID-19 patients. AA and DHA were significantly reduced in moderate and severe patients in which was observed an increase of the very long chain fatty acid (> 24C) which are markers of peroxisomal dysfunction. We found increased levels of 7≤-hydroxycholesterol and 7-ketocholesterol, markers of oxidative stress. In patients affected by COVID-19 there are evidence of several metabolomic changes involving mitochondrial and peroxisomal functions and resulting into changes of the lipid profile. Metabolomic investigation would be useful for studies about the mechanism of disease, the biomarker discovery and to find possible therapeutic targets. Civra et al., J Steroid Biochem Mol Biol, 2019:193, 105424. doi: 10.1016/j.jsbmb.2019.105424.