Tackling the Waves of COVID-19: A Planning Model for Intrahospital Resource Allocation

Background: The current pandemic requires hospitals to ensure care not only for the growing number of COVID-19 patients but also regular patients. Hospital resources must be allocated accordingly. Objective: To provide hospitals with a planning model to optimally allocate resources to intensive care units given a certain incidence of COVID-19 cases. Methods: The analysis included 334 cases from four adjacent counties south-west of Munich. From length of stay and type of ward [general ward (NOR), intensive care unit (ICU)] probabilities of case numbers within a hospital at a certain time point were derived. The epidemiological situation was simulated by the effective reproduction number R, the infection rates in mid-August 2020 in the counties, and the German hospitalization rate. Simulation results are compared with real data from 2nd and 3rd wave (September 2020–May 2021). Results: With R = 2, a hospitalization rate of 17%, mitigation measures implemented on day 9 (i.e., 7-day incidence surpassing 50/100,000), the peak occupancy was reached on day 22 (155.1 beds) for the normal ward and on day 25 (44.9 beds) for the intensive care unit. A higher R led to higher occupancy rates. Simulated number of infections and intensive care unit occupancy was concordant in validation with real data obtained from the 2nd and 3rd waves in Germany. Conclusion: Hospitals could expect a peak occupancy of normal ward and intensive care unit within ~5–11 days after infections reached their peak and critical resources could be allocated accordingly. This delay (in particular for the peak of intensive care unit occupancy) might give options for timely preparation of additional intensive care unit resources.

Tackling the Waves of COVID-19: A Planning Model for Intrahospital Resource Allocation.

Background: The current pandemic requires hospitals to ensure care not only for the growing number of COVID-19 patients but also regular patients. Hospital resources must be allocated accordingly. Objective: To provide hospitals with a planning model to optimally allocate resources to intensive care units given a certain incidence of COVID-19 cases. Methods: The analysis included 334 cases from four adjacent counties south-west of Munich. From length of stay and type of ward [general ward (NOR), intensive care unit (ICU)] probabilities of case numbers within a hospital at a certain time point were derived. The epidemiological situation was simulated by the effective reproduction number R, the infection rates in mid-August 2020 in the counties, and the German hospitalization rate. Simulation results are compared with real data from 2nd and 3rd wave (September 2020-May 2021). Results: With R = 2, a hospitalization rate of 17%, mitigation measures implemented on day 9 (i.e., 7-day incidence surpassing 50/100,000), the peak occupancy was reached on day 22 (155.1 beds) for the normal ward and on day 25 (44.9 beds) for the intensive care unit. A higher R led to higher occupancy rates. Simulated number of infections and intensive care unit occupancy was concordant in validation with real data obtained from the 2nd and 3rd waves in Germany. Conclusion: Hospitals could expect a peak occupancy of normal ward and intensive care unit within ~5-11 days after infections reached their peak and critical resources could be allocated accordingly. This delay (in particular for the peak of intensive care unit occupancy) might give options for timely preparation of additional intensive care unit resources.

Is Dementia Associated with COVID-19 Mortality? A Multicenter Retrospective Cohort Study Conducted in 50 Hospitals in Germany.

BACKGROUND: Dementia has been identified as a major predictor of mortality associated with COVID-19. OBJECTIVE: The objective of this study was to investigate the association between dementia and mortality in COVID-19 inpatients in Germany across a longer interval during the pandemic. METHODS: This retrospective study was based on anonymized data from 50 hospitals in Germany and included patients with a confirmed COVID-19 diagnosis hospitalized between March 11, 2020 and July, 20, 2022. The main outcome of the study was the association of mortality during inpatient stays with dementia diagnosis, which was studied using multivariable logistic regression adjusted for age, sex, and comorbidities as well as univariate logistic regression for matched pairs. RESULTS: Of 28,311 patients diagnosed with COVID-19, 11.3% had a diagnosis of dementia. Prior to matching, 26.5% of dementia patients and 11.5% of non-dementia patients died; the difference decreased to 26.5% of dementia versus 21.7% of non-dementia patients within the matched pairs (n = 3,317). This corresponded to an increase in the risk of death associated with dementia (OR = 1.33; 95% CI: 1.16-1.46) in the univariate regression conducted for matched pairs. CONCLUSION: Although dementia was associated with COVID-19 mortality, the association was weaker than in previously published studies. Further studies are needed to better understand whether and how pre-existing neuropsychiatric conditions such as dementia may impact the course and outcome of COVID-19.

Comparison of machine learning methods with logistic regression analysis in creating predictive models for risk of critical in-hospital events in COVID-19 patients on hospital admission.

BACKGROUND: Machine learning (ML) algorithms have been trained to early predict critical in-hospital events from COVID-19 using patient data at admission, but little is known on how their performance compares with each other and/or with statistical logistic regression (LR). This prospective multicentre cohort study compares the performance of a LR and five ML models on the contribution of influencing predictors and predictor-to-event relationships on prediction model´s performance. METHODS: We used 25 baseline variables of 490 COVID-19 patients admitted to 8 hospitals in Germany (March-November 2020) to develop and validate (75/25 random-split) 3 linear (L1 and L2 penalty, elastic net [EN]) and 2 non-linear (support vector machine [SVM] with radial kernel, random forest [RF]) ML approaches for predicting critical events defined by intensive care unit transfer, invasive ventilation and/or death (composite end-point: 181 patients). Models were compared for performance (area-under-the-receiver-operating characteristic-curve [AUC], Brier score) and predictor importance (performance-loss metrics, partial-dependence profiles). RESULTS: Models performed close with a small benefit for LR (utilizing restricted cubic splines for non-linearity) and RF (AUC means: 0.763-0.731 [RF-L1]); Brier scores: 0.184-0.197 [LR-L1]). Top ranked predictor variables (consistently highest importance: C-reactive protein) were largely identical across models, except creatinine, which exhibited marginal (L1, L2, EN, SVM) or high/non-linear effects (LR, RF) on events. CONCLUSIONS: Although the LR and ML models analysed showed no strong differences in performance and the most influencing predictors for COVID-19-related event prediction, our results indicate a predictive benefit from taking account for non-linear predictor-to-event relationships and effects. Future efforts should focus on leveraging data-driven ML technologies from static towards dynamic modelling solutions that continuously learn and adapt to changes in data environments during the evolving pandemic. TRIAL REGISTRATION NUMBER: NCT04659187.

Predicting the effectiveness of high-flow oxygen therapy in COVID-19 patients: a single-centre observational study.

BACKGROUND: High-flow nasal cannula (HFNC) therapy is a helpful tool in the treatment of hypoxaemic respiratory failure. However, the clinical parameters predicting the effectiveness of HFNC in coronavirus-19 disease (COVID-19) patients remain unclear. METHODS: Sixteen COVID-19 patients undergoing HFNC in the Asklepios Lung Clinic Munich-Gauting, Germany between 16 March and 3 June 2020 were retrospectively included into the study. Seven patients successfully recovered after HFNC (Group 1), while 9 patients required intubation upon HFNC failure (Group 2). Relevant predictors for an effective HFNC therapy were analysed on day 0 and 4 after HFNC initiation via receiver operating characteristics. RESULTS: The groups did not differ significantly in terms of age, sex, body mass index, and comorbidities. Five patients died in Group 2 upon disease progression and HFNC failure. Group 1 required a lower oxygen supplementation (FiO2 0.46 [0.31-0.54] vs. 0.72 [0.54-0.76], P = 0.022) and displayed a higher PaO2/FiO2 ratio (115 [111-201] vs. 93.3 [67.2-145], P = 0.042) on day 0. In Group 2, fever persisted on day 4 (38.5 [38.0-39.4]°C vs. 36.5 [31.1-37.1]°C, P = 0.010). Serum C-reactive protein (CRP) levels > 108 mg L-1 (day 0) and persistent oxygen saturation < 89% and PaO2/FiO2 ratio < 91 (day 4) were identified as significant predictors for HFNC failure (area under curve 0.929, 0.933, and 0.893). CONCLUSIONS: Elevated oxygen saturation, decreased FiO2 and reduced serum CRP on day 4 significantly predict HFNC effectiveness in COVID-19 patients. Based on these parameters, larger prospective studies are necessary to further investigate the effectiveness of HFNC in the treatment of COVID-19-associated hypoxaemic respiratory failure.

Single-cell RNA sequencing reveals ex vivo signatures of SARS-CoV-2-reactive T cells through 'reverse phenotyping'.

The in vivo phenotypic profile of T cells reactive to severe acute respiratory syndrome (SARS)-CoV-2 antigens remains poorly understood. Conventional methods to detect antigen-reactive T cells require in vitro antigenic re-stimulation or highly individualized peptide-human leukocyte antigen (pHLA) multimers. Here, we use single-cell RNA sequencing to identify and profile SARS-CoV-2-reactive T cells from Coronavirus Disease 2019 (COVID-19) patients. To do so, we induce transcriptional shifts by antigenic stimulation in vitro and take advantage of natural T cell receptor (TCR) sequences of clonally expanded T cells as barcodes for 'reverse phenotyping'. This allows identification of SARS-CoV-2-reactive TCRs and reveals phenotypic effects introduced by antigen-specific stimulation. We characterize transcriptional signatures of currently and previously activated SARS-CoV-2-reactive T cells, and show correspondence with phenotypes of T cells from the respiratory tract of patients with severe disease in the presence or absence of virus in independent cohorts. Reverse phenotyping is a powerful tool to provide an integrated insight into cellular states of SARS-CoV-2-reactive T cells across tissues and activation states.

Automated quantitative thin slice volumetric low dose CT analysis predicts disease severity in COVID-19 patients.

PURPOSE: This study aimed to identify predictive (bio-)markers for COVID-19 severity derived from automated quantitative thin slice low dose volumetric CT analysis, clinical chemistry and lung function testing. METHODS: Seventy-four COVID-19 patients admitted between March 16th and June 3rd 2020 to the Asklepios Lung Clinic Munich-Gauting, Germany, were included in the study. Patients were categorized in a non-severe group including patients hospitalized on general wards only and in a severe group including patients requiring intensive care treatment. Fully automated quantification of CT scans was performed via IMBIO CT Lung Texture analysis™ software. Predictive biomarkers were assessed with receiver-operator-curve and likelihood analysis. RESULTS: Fifty-five patients (44% female) presented with non-severe COVID-19 and 19 patients (32% female) with severe disease. Five fatalities were reported in the severe group. Accurate automated CT analysis was possible with 61 CTs (82%). Disease severity was linked to lower residual normal lung (72.5% vs 87%, p = 0.003), increased ground glass opacities (GGO) (8% vs 5%, p = 0.031) and increased reticular pattern (8% vs 2%, p = 0.025). Disease severity was associated with advanced age (76 vs 59 years, p = 0.001) and elevated serum C-reactive protein (CRP, 92.2 vs 36.3 mg/L, p < 0.001), lactate dehydrogenase (LDH, 485 vs 268 IU/L, p < 0.001) and oxygen supplementation (p < 0.001) upon admission. Predictive risk factors for the development of severe COVID-19 were oxygen supplementation, LDH >313 IU/L, CRP >71 mg/L, <70% normal lung texture, >12.5% GGO and >4.5% reticular pattern. CONCLUSION: Automated low dose CT analysis upon admission might be a useful tool to predict COVID-19 severity in patients.