ABSTRACT
Background. Black Americans have been disproportionately impacted by COVID-19, with systemic inequities contributing to increased incidence and mortality. In Chicago, access to COVID-19 testing was scarce in many predominantly Black neighborhoods early in the pandemic, exacerbating inequities. We aimed to assess uptake and yield of a COVID-19 testing campaign to increase access in high-risk, underserved Chicago communities. Methods. The Alive Church Network (ACN) is a community-driven coalition of Black pastors and public health researchers developed to address health inequities in Chicago. For this initiative, ACN implemented two strategies to improve access to COVID-19 testing in underserved neighborhoods: 1) church-based testing, with nurses and community health workers (CHWs) stationed in churches to perform PCR testing for community members and church staff (Dec 2020-Dec 2021) and 2) community site testing, with nurses and CHWs deployed to 8 local sites (e.g., Salvation Army) to conduct weekly PCR testing for employees and clients (May 2021-Mar 2022). We assessed characteristics of individuals undergoing testing;symptom and exposure history;and COVID-19 positivity. In addition, we compared positivity for ACN sites with Chicago citywide positivity over time. Results. A total of 3420 tests were performed for 1703 individuals. Of those tested, 931 (54.7%) were tested at churches and 772 (45.3%) were tested at community partner sites. Those presenting to churches were older (mean 44.8 vs. 39.1 years), more likely to be female (56.5% vs. 48.5%), and more likely to have a chronic medical condition (37.1% vs. 22.5%, Table). Prevalence of COVID-19 symptoms and exposure history were similar for those presenting to church and community sites (Table). Overall, ACN test positivity was similar to Chicago citywide positivity;however, ACN test positivity was much higher during the Omicron surge (24% vs. 11%, Figure). Characteristics of individuals tested for COVID-19 at church and community sites COVID-19 test positivity, ACN sites vs. Chicago citywide Conclusion. The ACN testing campaign increased access to COVID-19 testing in underserved neighborhoods in Chicago and had high yield during the Omicron surge. Church-based testing was successful for reaching older, medically vulnerable individuals. Ensuring widespread access to testing is essential for health equity and to reduce COVID-19 morbidity and mortality.
ABSTRACT
Background. The COVID-19 pandemic has caused a major public health crisis and is now a leading cause of death. At the start of the COVID-19 pandemic, treatment was limited to supportive care and off-label use of a variety of agents as we awaited the results of randomized controlled trials (RCTs). Our understanding of the disease has evolved and multiple evidence-based (EBM) treatment strategies supported by RCTs are now approved for use. Interpreting emerging data while responding to the pandemic can been challenging. We seek to understand the use of common medications during the COVID-19 pandemic. Methods. A survey was developed and disseminated through Infectious Diseases Society of America (IDSA) IDea network, IDSA Antimicrobial Stewardship Centers for Excellence and the Society for Healthcare Epidemiology messaging boards. Data collected included city, therapeutic options including convalescent plasma (CP), hydroxychloroquine (HCQ), baricitinib and tocilizumab during each COVID-19 wave. Results. A total of 78 responses were collected, representing wide geographic reach in the US (Figure 1). Use of HCQ declined by 96% and use of CP declined by 85% (Figure 2). The decline in CP was gradual each wave despite RCTs showing lack of benefit. Tocilizumab was used in 71% and 76% of hospitals during the first wave, prior to RCT data supporting its use. Baricitinib was used in 42% of hospitals prior to RCT data supporting its use. There was a 90% increase in baricitinib use after RCT data emerged. (Figure Presented) Conclusion. The use of therapies before proven benefit was common through the pandemic. Similarly, the use of therapies after benefit was in question continued. Hydroxychloroquine was widely used in the first wave and then halted by the majority of hospitals in subsequent COVID-19 surges. IL-6 and JAK-2 inhibitors were commonly used prior to evidence of benefit and were more widely adopted once evidence emerged. Convalescent plasma use was common despite lack of evidence and its use continued despite multiple negative RCTs. Overall, our survey shows a gap between evidence and practice. Communicating evidence-based treatment strategies is a vital priority for major professional societies. Translating evidence into practice remains challenging during public health emergencies like the COVID-19 pandemic.
ABSTRACT
Background. Remdesivir (RDV) was the first agent with proven clinical benefit against COVID-19, shortening the time to recovery in a randomized placebocontrolled trial (RCT). Based on this trial, the Food and Drug Administration approved RDV via emergency use authorization. Despite its wide use, RDV remains controversial. Early use of RDV in patients requiring low flow nasal canula has robust evidence. However, benefit in mechanical ventilation (MV), high-flow nasal canula (HFNC) and non-invasive positive pressure ventilation (NIPPV) is conflicting. Multiple studies, most notably the SOLIDARITY and DisCoVeRy trial, have conflicting results on who benefits from RDV. Deploying novel therapeutics in this changing landscape is challenging. We seek to understand RDV practices and the role of antimicrobial stewardship (ASP) in hospitals across the US. Methods. A survey was developed and disseminated through Infectious Diseases Society of America (IDSA) IDea network, IDSA Antimicrobial Stewardship Centers for Excellence and the Society for Healthcare Epidemiology messaging boards. Data collected included city, presence of RDV therapeutic restrictions, and RDV use by degree of oxygen needs for each COVID-19 wave. Results. A total of 78 responses were collected, representing wide geographic reach in theUS(Figure 1).RDVwas restricted in 53%of facilities.Hospitals without restrictions commonly used RDV for patients onMV, NIPPV and HFNC, with more use on HFNC than onNC during the first winter, Delta andOmicron waves (Figure 2). Use onMV declined with each surge.Hospitals withRDVrestrictions hadmore use ofRDVonNCthan all other ordinal scales (Figure 3). Use in MV, NIPPV and HFNC compared to NC declined in the restricted group with each COVID-19 surge. (Figure Presented) Conclusion. A wide gap between evidence-based guidelines and actual practice exists. This gap was wider in hospitals without COVID-19 therapeutic restrictions in place for RDV. In the unrestricted group, RDV was commonly used for MV, HFNC and NIPPV, where robust RCT evidence of benefit is lacking. Though this practice occurred in both groups, the restricted RDV group prioritized RDV use in NC and did so at higher percentages each subsequent COVID-19 surge. ASP restrictions can have an important role in guiding COVID-19 therapy.
ABSTRACT
Background. During the COVID-19 pandemic, evolving literature and emerging therapies have led to significant controversies on what constitutes optimal therapy for hospitalized patients with COVID-19. In addition, multiple guidelines emerged exhibiting variable recommendations in key areas of therapeutic management. Thus, interpreting emerging data and deploying novel therapeutics during the pandemic has been challenging. Antimicrobial stewardship programs (ASP) have been proven to help promote evidence-based practices and are now common the United States (US). We seek to further understand the role of ASP during the COVID-19 pandemic. Methods. A survey was developed and disseminated through Infectious Diseases Society of America (IDSA) IDea network, IDSA Antimicrobial Stewardship Centers for Excellence and the Society for Healthcare Epidemiology messaging boards. Data collected included city, hospital size, facility type, and existence of active ASP. COVID-19 therapeutic restrictions and type of provider who decides on treatment allocation was surveyed. Results. A total of 92 surveys were completed, representing wide geographic reach within the United States (Figure 1). Hospital size by beds were 14% less than 200 (13), 18.5% 201-300 beds (17), 15.2% 301-400 beds (14), and 52.2% over 400 beds (48). Community hospitals accounted for 31.5% (29), county/public hospital 3.3% (3, 3.3%), University-affiliated Community-Based Teaching Hospitals 10.9% (10), University-based Teaching Hospital 45.7% (42), and 8.7% selected other. ASP were present in 97.2% of hospitals. COVID-19 therapeutic restrictions were common (Figure 2). Infectious diseases (ID) approval was required in 87.9% of facilities (Figure 3). Figure 1 Geographic distribution of hospitals represented in the survey Figure 2. Restricted COVID-19 therapies by medication type (%) Figure 3 Specialties or providers that are able to order or approve restricted COVID-19 therapies (%) Conclusion. COVID-19 therapeutics were commonly restricted during the COVID-19 pandemic. Approval by ID clinicians was required almost universally (87.9%). The role of other specialties was significantly lower, representing less than 45% of hospitals. ID clinicians have played a vital role in guiding therapy and supporting ASP during the COVID-19 pandemic. This survey highlights the value of ID clinicians to healthcare systems in allocating resources and promoting evidence-based practices through ASP.
ABSTRACT
BACKGROUND: Various machine learning and artificial intelligence methods have been used to predict outcomes of hospitalized COVID-19 patients. However, process mining has not yet been used for COVID-19 prediction. We developed a process mining/deep learning approach to predict mortality among COVID-19 patients and updated the prediction in 6-h intervals during the first 72 h after hospital admission. METHODS: The process mining/deep learning model produced temporal information related to the variables and incorporated demographic and clinical data to predict mortality. The mortality prediction was updated in 6-h intervals during the first 72 h after hospital admission. Moreover, the performance of the model was compared with published and self-developed traditional machine learning models that did not use time as a variable. The performance was compared using the Area Under the Receiver Operator Curve (AUROC), accuracy, sensitivity, and specificity. RESULTS: The proposed process mining/deep learning model outperformed the comparison models in almost all time intervals with a robust AUROC above 80% on a dataset that was imbalanced. CONCLUSIONS: Our proposed process mining/deep learning model performed significantly better than commonly used machine learning approaches that ignore time information. Thus, time information should be incorporated in models to predict outcomes more accurately.