Search

Background The objective of this study was to characterize frailty and resilience in people evaluated for Post-Acute COVID-19 Syndrome (PACS), in relation to quality of life (QoL) and Intrinsic Capacity (IC). Methods This cross-sectional, observational, study included consecutive people previously hospitalized for severe COVID-19 pneumonia attending Modena (Italy) PACS Clinic from July 2020 to April 2021. Four frailty-resilience phenotypes were built: “fit/resilient”, “fit/non-resilient”, “frail/resilient” and “frail/non-resilient”. Frailty and resilience were defined according to frailty phenotype and Connor Davidson resilience scale (CD-RISC-25) respectively. Study outcomes were: QoL assessed by means of Symptoms Short form health survey (SF-36) and health-related quality of life (EQ-5D-5L) and IC by means of a dedicated questionnaire. Their predictors including frailty-resilience phenotypes were explored in logistic regressions. Results 232 patients were evaluated, median age was 58.0 years. PACS was diagnosed in 173 (74.6%) patients. Scarce resilience was documented in 114 (49.1%) and frailty in 72 (31.0%) individuals. Table 1 shows demographic, anthropometric and clinical characteristics, comorbidities and patient-reported outcomes according to four frailty-resilience phenotypes. With regards to study outcomes, Figure 1 depicts in radar graphs, mean scores of each domain of SF-36 (1A), EQ-5D5L (1B) and IC (1C). Figures shows polygon areas for each frailty/resilience phenotypes. Progressive increase of mean scores of each domain are plotted in the vertices of polygons, from the lowest (near the center) in frail and non-resilient, to highest (towards periphery) in fit and resilient. Multivariate logistic analyses were used to identify predictors of the total scores of SF-36 (Figure 2A), EQ-5D5L (Figure 2B) and IC (Figure 2C). Conclusion Resilience is complementary to frailty in the identification of clinical phenotypes with different impact on wellness and QoL. Frailty and resilience should be evaluated in hospitalized COVID-19 patients to identify vulnerable individuals to prioritize urgent health interventions in people with PACS. Funding This study is supported by a Gilead Sciences Inc. unrestricted grant. Disclosures All Authors: No reported disclosures.

Objectives Sex differences in COVID-19 severity and mortality have been described. Key aims of this analysis were to compare the risk of invasive mechanical ventilation (IMV) and mortality by sex and to explore whether variation in specific biomarkers could mediate this difference. Methods This was a retrospective, observational cohort study among patients with severe COVID-19 pneumonia. A survival analysis was conducted to compare time to the composite endpoint of IMV or death by sex. Interaction was formally tested to compare the risk difference by sex in subsets. Mediation analysis with a binary endpoint IMV or death (yes/no) by end of follow-up for a number of inflammation/coagulation biomarkers in the context of counterfactual prediction was also conducted. Results Among 415 patients, 134 were females (32%) and 281 males (67%), median age 66 years (IQR 54-77). At admission, females showed a significantly less severe clinical and respiratory profiles with a higher PaO2/FiO2 (254 mmHg vs 191 mmHg; p=0.023). By 28 days from admission, 49.2% (95% CI: 39.6-58.9%) of males vs. 31.7% (17.9-45.4%) of females underwent IMV or death (log-rank p-value Conclusions Our analysis confirms a difference in the risk of COVID-19 clinical progression by sex and provides a hypothesis for potential mechanisms leading to this. CRP showed a predominant role to mediate the difference in risk by sex.

Background The benefit of tocilizumab on mortality and time to recovery in people with severe COVID pneumonia may depend on appropriate timing. The objective was to estimate the impact of tocilizumab administration on switching respiratory support states, mortality and time to recovery. Methods In an observational study, a continuous-time Markov multi-state model was used to describe the sequence of respiratory support states including: no respiratory support (NRS), oxygen therapy (OT), non-invasive ventilation (NIV) or invasive mechanical ventilation (IMV), OT in recovery, NRS in recovery. Results Two hundred seventy-one consecutive adult patients were included in the analyses contributing to 695 transitions across states. The prevalence of patients in each respiratory support state was estimated with stack probability plots, comparing people treated with and without tocilizumab since the beginning of the OT state. A positive effect of tocilizumab on the probability of moving from the invasive and non-invasive mechanical NIV/IMV state to the OT in recovery state (HR = 2.6, 95% CI = 1.2–5.2) was observed. Furthermore, a reduced risk of death was observed in patients in NIV/IMV (HR = 0.3, 95% CI = 0.1–0.7) or in OT (HR = 0.1, 95% CI = 0.0–0.8) treated with tocilizumab. Conclusion To conclude, we were able to show the positive impact of tocilizumab used in different disease stages depicted by respiratory support states. The use of the multi-state Markov model allowed to harmonize the heterogeneous mortality and recovery endpoints and summarize results with stack probability plots. This approach could inform randomized clinical trials regarding tocilizumab, support disease management and hospital decision making.

The aim of this study was to evaluate both positive outcomes, including reduction of respiratory support aid and duration of hospital stay, and negative ones, including mortality and a composite of invasive mechanical ventilation or death, in patients with coronavirus disease 2019 (COVID-19) pneumonia treated with or without oral darunavir/cobicistat (DRV/c, 800/150 mg/day) used in different treatment durations. The secondary objective was to evaluate the percentage of patients treated with DRV/c who were exposed to potentially severe drug-drug interactions (DDIs) and died during hospitalization. This observational retrospective study was conducted in consecutive patients with COVID-19 pneumonia admitted to a tertiary care hospital in Modena, Italy. Kaplan-Meier survival curves and Cox proportional hazards regression were used to compare patients receiving standard of care with or without DRV/c. Adjustment for key confounders was applied. Two hundred seventy-three patients (115 on DRV/c) were included, 75.8% males, mean age was 64.6 (±13.2) years. Clinical improvement was similar between the groups, depicted by respiratory aid switch (p > .05). The same was observed for duration of hospital stay [13.2 (±8.9) for DRV/c vs. 13.4 (±7.2) days for no-DRV/c, p = .9]. Patients on DRV/c had higher rates of mortality (25.2% vs. 10.1%, p < .0001. The rate of composite outcome of mechanical ventilation and death was higher in the DRV/c group (37.4% vs. 25.3%, p = .03). Multiple serious DDI associated with DRV/c were observed in the 19 patients who died. DRV/c should not be recommended as a treatment option for COVID-19 pneumonia outside clinical trials.

Aims The aim of this study was to estimate a 48 hour prediction of moderate to severe respiratory failure, requiring mechanical ventilation, in hospitalized patients with COVID-19 pneumonia. Methods This was an observational prospective study that comprised consecutive patients with COVID-19 pneumonia admitted to hospital from 21 February to 6 April 2020. The patients’ medical history, demographic, epidemiologic and clinical data were collected in an electronic patient chart. The dataset was used to train predictive models using an established machine learning framework leveraging a hybrid approach where clinical expertise is applied alongside a data-driven analysis. The study outcome was the onset of moderate to severe respiratory failure defined as PaO2/FiO2 ratio Results A total of 198 patients contributed to generate 1068 usable observations which allowed to build 3 predictive models based respectively on 31-variables signs and symptoms, 39-variables laboratory biomarkers and 91-variables as a composition of the two. A fourth “boosted mixed model” included 20 variables was selected from the model 3, achieved the best predictive performance (AUC = 0.84) without worsening the FN rate. Its clinical performance was applied in a narrative case report as an example. Conclusion This study developed a machine model with 84% prediction accuracy, which is able to assist clinicians in decision making process and contribute to develop new analytics to improve care at high technology readiness levels.

Background Machine learning can assist clinicians in forecasting patients with COVID-19 who develop respiratory failure requiring mechanical ventilation. This analysis aimed to determine a 48 hours prediction of moderate to severe respiratory failure, as assessed with PaO2/FiO2 < 150 mmHg, in hospitalized patients with COVID-19 pneumonia. Methods This was an observational study that comprised all consecutive adult patients with COVID-19 pneumonia admitted to the Infectious Diseases Clinic of the University Hospital of Modena, Italy from 21 February to 6 April 2020. COVID-19 was confirmed with PCR positive nasopharyngeal swabs while the presence of pneumonia was radiologically confirmed. Patients received standard of care according to national guidelines for clinical management of SARS-CoV-2 infection. The patients' full medical history, demographic and epidemiological features, clinical data, complete blood count, coagulation, inflammatory and biochemical markers were routinely collected and aggregated in a clinically-oriented logical framework in order to build different datasets. The dataset was used to train a learning framework relying on Microsoft LightGBM and leveraging a hybrid approach, where clinical expertise is applied alongside a data-driven analysis. Shapley Additive exPlanations (SHAP) values were used to quantify the positive or negative impact of each variable included in the model on the predicted outcome. The study outcome was the onset of moderate to severe respiratory failure defined as PaO2/FiO2 ratio < 150 mmHg ([≥] 13.3 kPa) in at least one of two consecutive arterial blood gas analyses in the following 48 hours. Results A total of 198 patients contributed to generate 1068 valuable observations which allowed to build 3 prediction models based respectively on 31-variables signs and symptoms, 39-variables laboratory biomarkers and 91-variables as a composition of the two. A fourth boosted mixed model which included 20 variables was selected from the model 3, achieved the best predictive performance (AUC=0.84). Its clinical performance was applied in a narrative case report as an example. Conclusion This study developed a machine learning algorithm, with a 84% prediction accuracy, which is potentially able to assist clinicians in decision making process with therapeutic implications.