Rapid assessment of COVID-19 mortality risk with GASS classifiers (preprint)

ABSTRACT BACKGROUND Risk prediction scores and classification models are fundamental tools to effectively triage incoming COVID-19 patients. However, current triaging methods often have poor predictive performance, are based on variables that are expensive to measure, and lead to decisions that are sometimes hard to interpret. OBJECTIVE We introduce two new classification methods that are able to predict COVID-19 mortality risk from the automatic analysis of routine clinical variables with high accuracy and interpretability. The classifiers, denominated SVM22-GASS and Clinical-GASS, leverage machine learning methods and clinical expertise, respectively. METHODS Both classifiers were developed using a derivation cohort of 499 patients and were validated with an independent validation cohort of 250 patients. The cohorts included COVID-19 positive patients admitted to two hospitals in the Italian Province of Ferrara between March 2020 and June 2020 (derivation cohort) and between September 2020 and March 2021 (validation cohort). The potential predictive variables analyzed in this study included demographic, anamnestic, and laboratory data, retrieved with the patients’ consents from their electronic health records. The SVM22-GASS classifier is based on a Support Vector Machine model (SVM) with Radial Basis Function kernel (RBF). Importantly, the model uses only a subset of predictive variables that were automatically selected with the Least Absolute Shrinkage and Selection Operator (LASSO), while the RBF kernel is approximated with random feature expansions to reduce the computational requirements. The Clinical-GASS classifier is a threshold-based classifier that leverages the General Assessment of SARS-CoV-2 Severity (GASS) score: a highly interpretable COVID-19-specific clinical score that has been recently shown to be more effective at predicting the COVID-19 mortality risk than standard clinical scores. RESULTS The SVM22-GASS model was able to predict the mortality risk of the validation cohort with an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.87 and an accuracy of 0.88 — performing on par with influential classification methods that exploit variables derived from expensive analyses such as medical imaging. Furthermore, variable importance analyses showed that the model relies primarily on eight variables for its predictions: White Blood Cell Count, Lymphocyte Count, Brain Natriuretic Peptide, Creatine Phosphokinase, Lactate Dehydrogenase, Fibrinogen, PaO2/FiO2 Ratio, and High-Sensitivity Troponin I. Similarly, the Clinical-GASS classifier predicted the mortality risk of the validation cohort with an AUC of 0.77 and an accuracy of 0.78 — on par with other established and emerging machine-learning-based methods. CONCLUSIONS Our results demonstrate that it is possible to accurately predict the COVID-19 mortality risk using only routine clinical variables that can be readily collected in the very early stages of hospital admission. The classifiers have the potential to assist the clinicians in quickly identifying the patients’ mortality risk to optimally allocate both human and financial resources.

Long Term Implications in Surgical Re-assisting (L.I.S.A. study) During Covid-19 Outbreak: A Retrospective Observational Study on a Rural Population. (preprint)

Background: The spread of the COVID-19 is having a worldwide impact on surgicaltreatment. Our aim was to investigate the impact of the pandemic in a rural hospital in a lowdensely populated area.MethodsWe investigated the volume and type of surgical operations during the pandemic(March 2020 - February 2021) versus pre-pandemic period (March 2019 - February 2020) aswell as during the first and second pandemic waves compared to the pre-pandemic period.We compared the volume and timing of emergency appendectomy and cholecystectomyduring the pandemic versus pre-pandemic period, the volume, timing and stages of electivegastric and colorectal resections for cancer during the pandemic versus the pre-pandemicperiod.ResultsIn the prepandemic versus pandemic period, 42 versus 24 appendectomies and 174versus 126 cholecystectomies (urgent and elective) were performed. Patients operated onbefore as opposed to during the pandemic were older (58 vs. 52 years old, p=0.006),including for cholecystectomy (73 vs. 66 years old, p=0.01) and appendectomy (43 vs. 30years old, p = 0.04).The logistic regression analysis with regard to cholecystectomy and appendectomy performedin emergency showed that male sex and age were both associated to gangrenous typehistology, both in pandemic and prepandemic period. Finally, we found a reduction in cancerstage I and IIA in pandemic versus prepandemic period, with no increase in the moreadvanced stages.Conclusionsthe reduction in services imposed by governments during the first months oftotal lock down did not justify the whole decrease in surgical interventions in the year of thepandemic. Data suggest that greater "non-operative management" for cases of appendicitisand acute cholecystitis does not lead to an increase in cases operated over time, nor to anincrease in the "gangrenous" pattern, which seems to depend on age advanced and malepopulation.

Swab test in biological fluids as predictor of COVID-19 transmission risk during surgery: A prospective cross-sectional study from an Italian COVID center (preprint)

Background. The contamination of body fluids by Severe Acute Respiratory Syndrome Coronavirus 2 during surgery is current matter of debate in the scientific literature concerning CoronaVIrus Disease 2019. Surgical guidelines were published during the first wave of the COVID-19 pandemic and recommended to avoid laparoscopic surgery as much as possible, in fear that the chimney effect of high flow intraperitoneal gas escape during, and after, the procedure would increase the risk of viral transmission.Aim. The aim of this study was to evaluate the possibility of SARS-CoV-2 transmission during surgery by searching for viral RNA in serial samplings of biological liquids.Methods. This is a single center prospective cross-sectional study. We used a real-time reverse transcriptase (RT) polymerase chain reaction (PCR) test to perform swab tests for the qualitative detection of nucleic acid from SARS-CoV-2 in abdominal fluids, during emergency surgery and on the first post-operative day. In the case of thoracic surgery, we performed a swab test of pleural fluids during chest drainage placement as well as on the first post-operative day. Results. A total of 20 samples were obtained: 5 from pleural fluids, 13 from peritoneal fluids and two from biliary fluid. All 20 swabs performed from biological fluids resulted negative for SARS-CoV-2 RNA detection.Conclusion. To date, there is no scientific evidence of possible contagion by laparoscopic aerosolization of SARS-CoV-2, neither is certain whether the virus is effectively present in biological fluids.