Variation in COVID-19 characteristics, treatment and outcomes in Michigan: an observational study in 32 hospitals

Prevalence and prognostic role of thoracic lymphadenopathy in Covid-19

PURPOSE

 The prevalent coronavirus disease 2019 (COVID-19) pandemic has spread throughout the world and is considered a serious threat to global health. The prognostic role of thoracic lymphadenopathy in COVID-19 is unclear. The aim of the present meta-analysis was to analyze the prognostic role of thoracic lymphadenopathy for the prediction of 30-day mortality in patients with COVID-19.

MATERIALS AND METHODS

 The MEDLINE library, Cochrane, and SCOPUS databases were screened for associations between CT-defined features and mortality in COVID-19 patients up to June 2021. In total, 21 studies were included in the present analysis. The quality of the included studies was assessed by the Newcastle-Ottawa Scale. The meta-analysis was performed using RevMan 5.3. Heterogeneity was calculated by means of the inconsistency index I2. DerSimonian and Laird random-effect models with inverse variance weights were performed without any further correction.

RESULTS

 The included studies comprised 4621 patients. The prevalence of thoracic lymphadenopathy varied between 1 % and 73.4 %. The pooled prevalence was 16.7 %, 95 % CI = (15.6 %; 17.8 %). The hospital mortality was higher in patients with thoracic lymphadenopathy (34.7 %) than in patients without (20.0 %). The pooled odds ratio for the influence of thoracic lymphadenopathy on mortality was 2.13 (95 % CI = [1.80-2.52], p < 0.001).

CONCLUSION

 The prevalence of thoracic lymphadenopathy in COVID-19 is 16.7 %. The presence of thoracic lymphadenopathy is associated with an approximately twofold increase in the risk for hospital mortality in COVID-19.

KEY POINTS

  · The prevalence of lymphadenopathy in COVID-19 is 16.7 %.. · Patients with lymphadenopathy in COVID-19 have a higher risk of mortality during hospitalization.. · Lymphadenopathy nearly doubles mortality and plays an important prognostic role..

CITATION FORMAT

· Bucher AM, Sieren M, Meinel F et al. Prevalence and prognostic role of thoracic lymphadenopathy in Covid-19. Fortschr Röntgenstr 2024; DOI: 10.1055/a-2293-8132.

Pathology of COVID-19 Lung Disease

The pathology of severe COVID-19 lung injury is predominantly diffuse alveolar damage, with other reported patterns including acute fibrinous organizing pneumonia, organizing pneumonia, and bronchiolitis. Lung injury was caused by primary viral injury, exaggerated immune responses, and superinfection with bacteria and fungi. Although fatality rates have decreased from the early phases of the pandemic, persistent pulmonary dysfunction occurs and its pathogenesis remains to be fully elucidated.

Myosteatosis predicts short-term mortality in patients with COVID-19: A multicenter analysis

OBJECTIVES

Body composition on computed tomography can predict prognosis in patients with COVID-19. The reported data are based on small retrospective studies. The aim of the present study was to analyze the prognostic relevance of skeletal muscle parameter derived from chest computed tomography for prediction of 30-d mortality in patients with COVID-19 in a multicenter setting.

METHODS

The clinical databases of three centers were screened for patients with COVID-19 between 2020 and 2022. Overall, 447 patients (142 female; 31.7%) were included into the study. The mean age at the time of computed tomography acquisition was 63.8 ± 14.7 y and median age was 65 y. Skeletal muscle area and skeletal muscle density were defined on level T12 of the chest.

RESULTS

Overall, 118 patients (26.3%) died within the 30-d observation period. Of the patient sample, 255 patients (57.0%) were admitted to an intensive care unit and 122 patients needed mechanical ventilation (27.3%). The mean skeletal muscle area of all patients was 96.1 ± 27.2 cm² (range = 23.2-200.7 cm²). For skeletal muscle density, the mean was 24.3 ± 11.1 Hounsfield units (range = -5.6 to 55.8 Hounsfield units). In survivors, the mean skeletal muscle density was higher compared with the lethal cases (mean 25.8 ± 11.2 versus 20.1 ± 9.6; P < 0.0001). Presence of myosteatosis was independently associated with 30-d mortality: odds ratio = 2.72 (95% CI, 1.71-4.32); P = 0.0001.

CONCLUSIONS

Myosteatosis is strongly associated with 30-d mortality in patients COVID-19. Patients with COVID-19 with myosteatosis should be considered a risk group.

Epicardial Adipose Tissue as a Prognostic Marker in COVID-19

BACKGROUND/AIM

Epicardial adipose tissue (EAT) has been established as a quantitative imaging biomarker associated with the prognosis of several diseases, especially cardiovascular diseases. The cardiac injury by coronavirus disease 2019 (COVID-19) might be linked to the EAT. This study aimed to use this prognostic marker derived from computed tomography (CT) images to predict 30-day mortality in patients with COVID-19.

PATIENTS AND METHODS

Consecutive patients with COVID-19 were retrospectively screened between 2020 and 2022. Overall, 237 patients (78 female, 32.9%) were included in the present study. The study end-point was the 30-day mortality. EAT was measured using the diagnostic CT in a semiquantitative manner. EAT volume and density were measured for each patient.

RESULTS

Overall, 70 patients (29.5%) died within the 30-day observation period and 143 patients (60.3%) were admitted to the intensive care unit (ICU). The mean EAT volume was 140.9±89.1 cm3 in survivors and 132.9±77.7 cm3 in non-survivors, p=0.66. The mean EAT density was -71.9±8.1 Hounsfield units (HU) in survivors, and -67.3±8.4 HU in non-survivors, p=0.0001. EAT density was associated with 30-day mortality (p<0.0001) and ICU admission (p<0.0001). EAT volume was not associated with mortality and/or ICU admission.

CONCLUSION

EAT density was associated with 30-day mortality and ICU admission in patients with COVID-19.

CT coronary artery calcification score as a prognostic marker in COVID-19

Background

Coronary artery calcification (CA) score has been established as a quantitative imaging biomarker to reflect arteriosclerosis and general vessel status. It is established as an important prognostic factor for coronary heart disease but also for other disease entities. Our aim was to use this imaging marker derived from computed tomography (CT) images to elucidate the prognostic relevance in patients with coronavirus disease 2019 (COVID-19).

Methods

The clinical database was retrospectively screened for patients with COVID-19 between 2020 and 2022. A total of 241 patients (85 female patients, 35.3%) were included into the analysis. CA scoring was performed semiquantitatively on thoracic CT images with the established Weston score.

Results

Overall, 61 patients (25.3%) of the investigated patient sample died. In survivors, the mean CA score was 2.3±3.0 and in non-survivors, it was 4.2±4.1 (P=0.002). In univariable regression analysis, CA was associated with 30-day mortality [odds ratio (OR) =1.15; 95% confidence interval (CI): 1.06-1.25, P<0.001]. These results were confirmed by the multivariable regression analysis adjusted for age and sex, the CA score predicted 30-day mortality (OR =1.28; 95% CI: 1.08-1.4, P=0.002).

Conclusions

CA score is an independent risk factor in COVID-19. As CA scoring can easily be performed by the radiologist, it should be further investigated as an imaging marker in patients with COVID-19 and potentially be translated into clinical routine.

Clinical importance of thoracal lymphadenopathy in COVID-19

BACKGROUND

Thoracal lymphadenopathy may predict prognosis in patients with coronavirus disease 2019 (COVID-19), albeit the reported data is inconclusive. The aim of the present analysis was to analyze the affected lymph node stations and the cumulative lymph node size derived from computed tomography (CT) for prediction of 30-day mortality in patients with COVID-19.

METHODS

The clinical database was retrospectively screened for patients with COVID-19 between 2020 and 2022. Overall, 177 patients (63 female, 35.6%) were included into the analysis. Thoracal lymphadenopathy was defined by short axis diameter above 10 mm. Cumulative lymph node size of the largest lymph nodes was calculated and the amount of affected lymph node stations was quantified.

RESULTS

Overall, 53 patients (29.9%) died within the 30-day observation period. 108 patients (61.0%) were admitted to the ICU and 91 patients needed to be intubated (51.4%). Overall, there were 130 patients with lymphadenopathy (73.4%). The mean number of affected lymph node levels were higher in non-survivors compared to survivors (mean, 4.0 vs 2.2, p < 0.001). The cumulative size was also higher in non-survivors compared to survivors (mean 55.9 mm versus 44.1 mm, p = 0.006). Presence of lymphadenopathy was associated with 30-day mortality in a multivariable analysis, OR = 2.99 (95% CI 1.20 - 7.43), p = 0.02.

CONCLUSIONS

Thoracal lymphadenopathy comprising cumulative size and affected levels derived from CT images is associated with 30-day mortality in patients with COVID-19. COVID-19 patients presenting with thoracic lymphadenopathy should be considered as a risk group.

Data-driven analysis to understand long COVID using electronic health records from the RECOVER initiative

Recent studies have investigated post-acute sequelae of SARS-CoV-2 infection (PASC, or long COVID) using real-world patient data such as electronic health records (EHR). Prior studies have typically been conducted on patient cohorts with specific patient populations which makes their generalizability unclear. This study aims to characterize PASC using the EHR data warehouses from two large Patient-Centered Clinical Research Networks (PCORnet), INSIGHT and OneFlorida+, which include 11 million patients in New York City (NYC) area and 16.8 million patients in Florida respectively. With a high-throughput screening pipeline based on propensity score and inverse probability of treatment weighting, we identified a broad list of diagnoses and medications which exhibited significantly higher incidence risk for patients 30-180 days after the laboratory-confirmed SARS-CoV-2 infection compared to non-infected patients. We identified more PASC diagnoses in NYC than in Florida regarding our screening criteria, and conditions including dementia, hair loss, pressure ulcers, pulmonary fibrosis, dyspnea, pulmonary embolism, chest pain, abnormal heartbeat, malaise, and fatigue, were replicated across both cohorts. Our analyses highlight potentially heterogeneous risks of PASC in different populations.

Perceived Hospital Stress, Severe Acute Respiratory Syndrome Coronavirus 2 Activity, and Care Process Temporal Variance During the COVID-19 Pandemic

OBJECTIVES

The COVID-19 pandemic threatened standard hospital operations. We sought to understand how this stress was perceived and manifested within individual hospitals and in relation to local viral activity.

DESIGN

Prospective weekly hospital stress survey, November 2020-June 2022.

SETTING

Society of Critical Care Medicine's Discovery Severe Acute Respiratory Infection-Preparedness multicenter cohort study.

SUBJECTS

Thirteen hospitals across seven U.S. health systems.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We analyzed 839 hospital-weeks of data over 85 pandemic weeks and five viral surges. Perceived overall hospital, ICU, and emergency department (ED) stress due to severe acute respiratory infection patients during the pandemic were reported by a mean of 43% ( sd , 36%), 32% (30%), and 14% (22%) of hospitals per week, respectively, and perceived care deviations in a mean of 36% (33%). Overall hospital stress was highly correlated with ICU stress (ρ = 0.82; p < 0.0001) but only moderately correlated with ED stress (ρ = 0.52; p < 0.0001). A county increase in 10 severe acute respiratory syndrome coronavirus 2 cases per 100,000 residents was associated with an increase in the odds of overall hospital, ICU, and ED stress by 9% (95% CI, 5-12%), 7% (3-10%), and 4% (2-6%), respectively. During the Delta variant surge, overall hospital stress persisted for a median of 11.5 weeks (interquartile range, 9-14 wk) after local case peak. ICU stress had a similar pattern of resolution (median 11 wk [6-14 wk] after local case peak; p = 0.59) while the resolution of ED stress (median 6 wk [5-6 wk] after local case peak; p = 0.003) was earlier. There was a similar but attenuated pattern during the Omicron BA.1 subvariant surge.

CONCLUSIONS

During the COVID-19 pandemic, perceived care deviations were common and potentially avoidable patient harm was rare. Perceived hospital stress persisted for weeks after surges peaked.

Comparative admission rates and infection severity of COVID-19 among unvaccinated and vaccinated patients

Vaccination efforts have limited the burden of the pandemic caused by the coronavirus disease 2019 (COVID-19) with substantial evidence showing reduced hospitalization rates among vaccinated populations. However, few studies have explored correlations between vaccination status and inpatient COVID-19 outcomes. This observational case-control study involved a retrospective chart review of adult patients hospitalized for COVID-19 infection at a medium-sized hospital in Central Michigan between May 1, 2021 and September 30, 2021. Unadjusted analyses involved t-tests and chi-square tests followed by adjusted analyses using binary logistic and linear regression models. Of the 192 screened patients, 171 subjects met the inclusion criteria. Vaccinated patients were significantly older (71.09 vs 57.45, p < 0.001), more likely to identify as white (89.4% vs 66.9%, p = 0.026), and had a lower baseline 10-year survival rate predicted by the Charlson Comorbidity Index (42% vs 69%, p < 0.001) compared to unvaccinated patients. Common symptoms between both groups included shortness of breath (50%), malaise (23%-37%), cough (28%-32%), and fever or chills (25%). Upon matching, adjusted analysis showed significantly higher rates of remdesivir administration to unvaccinated patients (41.3% vs 13.3%, odds ratio (OR): 4.63, 90% confidence interval (CI): 1.98-11.31). Despite higher intensive care unit admission rates among unvaccinated patients (39.1% vs 23.9%, OR: 1.83, 90% CI: 0.74-4.64), this difference did not reach statistical significance. Accordingly, immunization status strongly correlates with patient demographics and differences in inpatient treatment. Larger studies are needed to further assess the vaccine's impact on inpatient outcomes outside of our community.

COVID-19-Related Publications by Hospitalists in the United States

Objective To determine the degree to which hospitalists published academic manuscripts related to COVID-19 during the first year of the pandemic. Patients and methods The study was a cross-sectional analysis of the author's specialty, defined by byline or professional online biography, from articles related to COVID-19 published between March 1, 2020, and February 28, 2021. It included the top four internal medicine journals by impact factor: New England Journal of Medicine, Journal of the American Medical Association, Journal of the American Medical Association Internal Medicine, and Annals of Internal Medicine. Participants were all United States (US)-based physician authors contributing to COVID-19 publications. Our primary outcome was the percentage of US-based physician authors of COVID-19 articles who were hospitalists. Subgroup analyses characterized author specialty by authorship position (first, middle, last) and article type (research vs. non-research). Results Between March 1, 2020, and February 28, 2021, the top four US-based medical journals published 870 articles related to COVID-19 of which 712 articles with 1940 US-based physician authors were included. Hospitalists accounted for 4.2% (82) of authorship positions including 4.7% (49/1038) of authorship positions in research articles and 3.7% (33/902) of authorship positions in non-research articles. First, middle, and last authorship positions were held by hospitalists at 3.7% (18/485), 4.4% (45/1034), and 4.5% (19/421) of the time, respectively. Conclusions Despite caring for a large number of patients with COVID-19, hospitalists were rarely involved in disseminating COVID-19 knowledge. Limited authorship by hospitalists could constrain the dissemination of inpatient medicine knowledge, impact patient outcomes, and affect the academic promotion of early-career hospitalists.

Racial and ethnic variation in COVID-19 care, treatment, and outcomes: A retrospective cohort study from the MiCOVID-19 registry

BACKGROUND

Racial and ethnic disparities in COVID-19 outcomes exist, but whether in-hospital care explains this difference is not known. We sought to determine racial and ethnic differences in demographics, comorbidities, in-hospital treatments, and in-hospital outcomes of patients hospitalized with COVID-19.

METHODS AND FINDINGS

This was a cohort study using MiCOVID-19, a multi-center, retrospective, collaborative quality improvement registry, which included data on patients hospitalized with COVID-19 across 38 hospitals in the State of Michigan. 2,639 adult patients with COVID-19 hospitalized at a site participating in the MiCOVID-19 Registry were randomly selected. Outcomes included in-hospital mortality, age at death, intensive care unit admission, and need for invasive mechanical ventilation by race and ethnicity. Baseline comorbidities differed by race and ethnicity. In addition, Black patients had higher lactate dehydrogenase, erythrocyte sedimentation rate, C-reactive protein, creatine phosphokinase, and ferritin levels. Black patients were less likely to receive dexamethasone and remdesivir compared with White patients (4.2% vs 14.3% and 2.2% vs. 11.8%, p < 0.001 for each). Black (18.7%) and White (19.6%) patients experienced greater mortality compared with Asian (13.0%) and Latino (5.9%) patients (p < 0.01). The mean age at death was significantly lower by 8 years for Black patients (69.4 ± 13.3 years) compared with White (77.9 ± 12.6), Asian (77.6 ± 6.6), and Latino patients (77.4 ± 15.5) (p < 0.001).

CONCLUSIONS

COVID-19 mortality appears to be driven by both pre-hospitalization clinical and social factors and potentially in-hospital care. Policies aimed at population health and equitable application of evidence-based medical therapy are needed to alleviate the burden of COVID-19.

Eligibility for Posthospitalization Venous Thromboembolism Prophylaxis in Hospitalized Patients With COVID-19: A Retrospective Cohort Study

Background

A recent randomized trial, the MICHELLE trial, demonstrated improved posthospital outcomes with a 35‐day course of prophylactic rivaroxaban for patients hospitalized with COVID‐19 at high risk of venous thromboembolism. We explored how often these findings may apply to an unselected clinical population of patients hospitalized with COVID‐19.

Methods and Results

Using a 35‐hospital retrospective cohort of patients hospitalized between March 7, 2020, and January 23, 2021, with COVID‐19 (MI‐COVID19 database), we quantified the percentage of hospitalized patients with COVID‐19 who would be eligible for rivaroxaban at discharge per MICHELLE trial criteria and report clinical event rates. The main clinical outcome was derived from the MICHELLE trial and included a composite of symptomatic venous thromboembolism, pulmonary embolus‐related death, nonhemorrhagic stroke, and cardiovascular death at 35 days. Multiple sensitivity analyses tested different eligibility and exclusion criteria definitions to determine the effect on eligibility for postdischarge anticoagulation prophylaxis. Of 2016 patients hospitalized with COVID‐19 who survived to discharge and did not have another indication for anticoagulation, 25.9% (n=523) would be eligible for postdischarge thromboprophylaxis per the MICHELLE trial criteria (range, 2.9%–39.4% on sensitivity analysis). Of the 416 who had discharge anticoagulant data collected, only 13.2% (55/416) were actually prescribed a new anticoagulant at discharge. Of patients eligible for rivaroxaban per the MICHELLE trial, the composite clinical outcome occurred in 1.2% (6/519); similar outcome rates were 5.7% and 0.63% in the MICHELLE trial's control (no anticoagulation) and intervention (rivaroxaban) groups, respectively. Symptomatic venous thromboembolism events and all‐cause mortality were 6.2% (32/519) and 5.66% in the MI‐COVID19 and MICHELLE trial control cohorts, respectively.

Conclusions

Across 35 hospitals in Michigan, ≈1 in 4 patients hospitalized with COVID‐19 would qualify for posthospital thromboprophylaxis. With only 13% of patients actually receiving postdischarge prophylaxis, there is a potential opportunity for improvement in care.

Impact of Early Corticosteroids on Preventing Clinical Deterioration in Non-critically Ill Patients Hospitalized with COVID-19: A Multi-hospital Cohort Study

INTRODUCTION

While guidelines stronglyrecommend dexamethasone in critical COVID-19, the optimal threshold to initiate corticosteroids in non-critically ill patients with COVID-19 remains unclear. Using data from a state-wide COVID-19 registry, we evaluated the effectiveness of early corticosteroids for preventing clinical deterioration among non-critically ill patients hospitalized for COVID-19 and receiving non-invasive oxygen therapy.

METHODS

This was a target trial using observational data from patients hospitalized for COVID-19 at 39 hospitals participating in the MI-COVID19 registry between March 16, 2020 and August 24, 2020. We studied the impact of corticosteroids initiated within 2 calendar days of hospitalization ("early steroids") versus no early steroids among non-ICU patients with laboratory-confirmed SARS-CoV2 receiving non-invasive supplemental oxygen therapy. Our primary outcome was a composite of in-hospital mortality, transfer to intensive care, and receipt of invasive mechanical ventilation. We used inverse probability of treatment weighting (IPTW) and propensity score-weighted regression to measure the association of early steroids and outcomes.

RESULTS

Among 1002 patients meeting study criteria, 231 (23.1%) received early steroids. After IPTW, to balance potential confounders between the treatment groups, early steroids were not associated with a decrease in the composite outcome (aOR 1.1, 95%CI 0.8-1.6) or in any components of the primary outcome.

CONCLUSION

We found no evidence that early corticosteroid therapy prevents clinical deterioration among hospitalized non-critically ill COVID-19 patients receiving non-invasive oxygen therapy. Further studies are needed to determine the optimal threshold for initiating corticosteroids in this population.

COVID-19 and Chagas Disease in Buenos Aires, Argentina

Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2. COVID-19 leads, in most patients, to mild-to-moderate symptoms, but some develop severe disease and succumbed to death. People with medical conditions have a higher risk of death than those without them. Chagas disease (CD) can cause cardiac diseases in approximately one-third of affected people. The aim of this study is to find out if there is any clinical association between Chagas disease and COVID-19 severity. This is a cohort study of 29 patients who were hospitalized with COVID-19 and had a diagnosis of chronic Trypanosoma cruzi infection. This coinfected cohort was matched by sex, age, presence of comorbidities, and requirement of hospitalization on intensive care unit (ICU) at admission with a control cohort of patients hospitalized due to COVID-19 without CD in a 3:1 ratio (n = 87). The clinical outcomes evaluated were as follows: days of hospitalization, death, and requirement of ICU and mechanical respiratory assistance (MV). The study protocol was approved by the Institutional Ethics in Research Committee. The Chagas disease/COVID-19 coinfected cohort had a median age of 55 years old (49.0, 66.0); 17 (59%) were male. All patients survived the acute COVID-19. Three of them were admitted to the ICU, and two required MV. Twenty-two (75.8%) required supplemental oxygen. There were no statistical differences in any laboratory parameters between the groups except for lactic acid dehydrogenase, which showed higher levels in the coinfected cohort, with a median of 573 U/L (interquartile range: 486.00, 771.00) vs. 476 U/L (346.00, 641.00) in the control group (p = 0.007). There were no differences in clinical outcomes between both groups. On the cohort with Chagas disease, there were zero deaths, three (10.3%) were admitted in the ICU, and two (6.9%) required MV, while for the control group there were six deaths (6.6%), 13 required ICU (14.9%), and 11 required MV (12.6%), without a statistically significant difference. This small series of coinfected Chagas disease and COVID-19 does not suggest differences in clinical evolution compared to non-Chagas patients. This data is similar to a Brazilian cohort. More data of this population with and without cardiomyopathy is needed to optimize the follow-up and recommendation for the population affected by this neglected tropical disease about COVID-19.

Computed tomography-defined body composition as prognostic markers for unfavourable outcomes and in-hospital mortality in coronavirus disease 2019

BACKGROUND

Low skeletal muscle mass (LSMM) and visceral fat areas can be assessed by cross-sectional images. These parameters are associated with several clinically relevant factors in various disorders with predictive and prognostic implications. Our aim was to establish the effect of computed tomography (CT)-defined LSMM and fat areas on unfavourable outcomes and in-hospital mortality in coronavirus disease 2019 (COVID-19) patients based on a large patient sample.

METHODS

MEDLINE library, Cochrane, and Scopus databases were screened for the associations between CT-defined LSMM as well as fat areas and in-hospital mortality in COVID-19 patients up to September 2021. In total, six studies were suitable for the analysis and included into the present analysis.

RESULTS

The included studies comprised 1059 patients, 591 men (55.8%) and 468 women (44.2%), with a mean age of 60.1 years ranging from 48 to 66 years. The pooled prevalence of LSMM was 33.6%. The pooled odds ratio for the effect of LSMM on in-hospital mortality in univariate analysis was 5.84 [95% confidence interval (CI) 1.07-31.83]. It was 2.73 (95% CI 0.54-13.70) in multivariate analysis. The pooled odds ratio of high visceral fat area on unfavourable outcome in univariate analysis was 2.65 (95% CI 1.57-4.47).

CONCLUSIONS

Computed tomography-defined LSMM and high visceral fat area have a relevant association with in-hospital mortality in COVID-19 patients and should be included as relevant prognostic biomarkers into clinical routine.

Extrapulmonary CT Findings Predict In-Hospital Mortality in COVID-19. A Systematic Review and Meta-Analysis

RATIONALE AND OBJECTIVES

Several prognostic factors have been identified for COVID-19 disease. Our aim was to elucidate the influence of non-pulmonary findings of thoracic computed tomography (CT) on unfavorable outcomes and in-hospital mortality in COVID-19 patients based on a large patient sample.

MATERIALS AND METHODS

MEDLINE library, Cochrane and SCOPUS databases were screened for the associations between CT-defined features and mortality in COVID-19 patients up to June 2021. In total, 22 studies were suitable for the analysis, and included into the present analysis. Overall, data regarding 4 extrapulmonary findings could be pooled: pleural effusion, pericardial effusion, mediastinal lymphadenopathy, and coronary calcification.

RESULTS

The included studies comprised 7859 patients. The pooled odds ratios for the effect of the identified extrapulmonary findings on in-hospital mortality are as follows: pleural effusion, 4.60 (95% CI 2.97-7.12); pericardial effusion, 1.29 (95% CI 0.83-1.98); coronary calcification, 2.68 (95% CI 1.78-4.04); mediastinal lymphadenopathy, 2.02 (95% CI 1.18-3.45).

CONCLUSION

Pleural effusion, mediastinal lymphadenopathy and coronary calcification have a relevant association with in-hospital mortality in COVID-19 patients and should be included as prognostic biomarker into clinical routine.

Disease-course adapting machine learning prognostication models in critically ill elderly COVID-19 patients: a multi-centre cohort study with external validation

Background

The COVID-19 pandemic caused by SARS-CoV-2 is challenging health care systems globally. The disease disproportionately affects the elderly population, both in terms of disease severity and mortality risk.

Objective

The aim of this study was to evaluate machine learning–based prognostication models for critically ill elderly COVID-19 patients, which dynamically incorporated multifaceted clinical information on evolution of the disease.

Methods

This multicenter cohort study (COVIP study) obtained patient data from 151 intensive care units (ICUs) from 26 countries. Different models based on the Sequential Organ Failure Assessment (SOFA) score, logistic regression (LR), random forest (RF), and extreme gradient boosting (XGB) were derived as baseline models that included admission variables only. We subsequently included clinical events and time-to-event as additional variables to derive the final models using the same algorithms and compared their performance with that of the baseline group. Furthermore, we derived baseline and final models on a European patient cohort, which were externally validated on a non-European cohort that included Asian, African, and US patients.

Results

In total, 1432 elderly (≥70 years old) COVID-19–positive patients admitted to an ICU were included for analysis. Of these, 809 (56.49%) patients survived up to 30 days after admission. The average length of stay was 21.6 (SD 18.2) days. Final models that incorporated clinical events and time-to-event information provided superior performance (area under the receiver operating characteristic curve of 0.81; 95% CI 0.804-0.811), with respect to both the baseline models that used admission variables only and conventional ICU prediction models (SOFA score, P<.001). The average precision increased from 0.65 (95% CI 0.650-0.655) to 0.77 (95% CI 0.759-0.770).

Conclusions

Integrating important clinical events and time-to-event information led to a superior accuracy of 30-day mortality prediction compared with models based on the admission information and conventional ICU prediction models. This study shows that machine-learning models provide additional information and may support complex decision-making in critically ill elderly COVID-19 patients.

Trial Registration

ClinicalTrials.gov NCT04321265; https://clinicaltrials.gov/ct2/show/NCT04321265