Development and Validation of Predictors for the Survival of Patients With COVID-19 Based on Machine Learning.

Background: The outbreak of COVID-19 attracted the attention of the whole world. Our study aimed to explore the predictors for the survival of patients with COVID-19 by machine learning. Methods: We conducted a retrospective analysis and used the idea of machine learning to train the data of COVID-19 patients in Leishenshan Hospital through the logical regression algorithm provided by scikit-learn. Results: Of 2010 patients, 42 deaths were recorded until March 29, 2020. The mortality rate was 2.09%. There were 6,812 records after data features combination and data arrangement, 3,025 records with high-quality after deleting incomplete data by manual checking, and 5,738 records after data balancing finally by the method of Borderline-1 Smote. The results of 10 times of data training by logistic regression model showed that albumin, saturation of pulse oxygen at admission, alanine aminotransferase, and percentage of neutrophils were possibly associated with the survival of patients. The results of 10 times of data training including age, sex, and height beyond the laboratory measurements showed that percentage of neutrophils, saturation of pulse oxygen at admission, alanine aminotransferase, sex, and albumin were possibly associated with the survival of patients. The rates of precision, recall, and f1-score of the two training models were all higher than 0.9 and relatively stable. Conclusions: We demonstrated that percentage of neutrophils, saturation of pulse oxygen at admission, alanine aminotransferase, sex, and albumin were possibly associated with the survival of patients with COVID-19.

Prominent Hypercoagulability Associated With Inflammatory State Among Cancer Patients With SARS-CoV-2 Infection.

Abnormal coagulation parameters and potential benefits of anticoagulant therapy in general population with novel coronavirus pneumonia (COVID-19) have been reported. However, limited data are available on cancer patients. Coagulation indexes and inflammation parameters in 57 cancer patients with SARS-CoV-2 infection with different severity were retrospectively analyzed. We found that D-dimer levels were increased in 33 patients (57.9%, median: 790 ng/mL). Compared with ordinary type patients, severe and critical ill patients had decreased MPV values (P = 0.006), prolonged PT (median: 13.3 vs. 11.5 vs. 11.4 s, P < 0.001), significant higher D-dimer levels (median: 2,400 vs. 940 vs. 280 ng/mL, P < 0.001), higher PCT levels (median: 0.17 vs. 0.055 vs. 0.045 ng/mL, P = 0.002), higher IL-6 (median: 20.6 vs. 2.3 vs. 3.0 pg/mL, P = 0.040), and decreased PaO2 (median: 68 vs. 84 vs. 96 mm Hg, P < 0.001). Importantly, three patients, one severe and two critical ill type, with increased D-dimer survived after anticoagulant therapy with continuous heparin infusion. Increased D-dimer levels positively correlated with increased PCT levels (r = 0.456, P = 0.002) and IL-6 levels (r = 0.501, P = 0.045). A negative correlation between D-dimer levels and PaO2 levels (r = -0.654, P = 0.021) were also existed. Cancer patients with COVID-19 showed prominent hypercoagulability associated with severe inflammation, anticoagulation therapy might be useful to improve the prognosis and should be immediately used after the onset of hypercoagulability.

Potential therapeutic effects of dipyridamole in the severely ill patients with COVID-19.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can cause acute respiratory distress syndrome, hypercoagulability, hypertension, and multiorgan dysfunction. Effective antivirals with safe clinical profile are urgently needed to improve the overall prognosis. In an analysis of a randomly collected cohort of 124 patients with COVID-19, we found that hypercoagulability as indicated by elevated concentrations of D-dimers was associated with disease severity. By virtual screening of a U.S. FDA approved drug library, we identified an anticoagulation agent dipyridamole (DIP) in silico, which suppressed SARS-CoV-2 replication in vitro. In a proof-of-concept trial involving 31 patients with COVID-19, DIP supplementation was associated with significantly decreased concentrations of D-dimers (P < 0.05), increased lymphocyte and platelet recovery in the circulation, and markedly improved clinical outcomes in comparison to the control patients. In particular, all 8 of the DIP-treated severely ill patients showed remarkable improvement: 7 patients (87.5%) achieved clinical cure and were discharged from the hospitals while the remaining 1 patient (12.5%) was in clinical remission.