University of Florida Health

Modern cancer centers face increasing data-driven demands from institutional, state, and federal stakeholders, as well as from patients and communities. Data now functions as a strategic asset that supports care delivery, research, education, community engagement, and policy improvement and development. However, many cancer centers continue to operate with fragmented systems, limited interoperability, and underdeveloped analytics capacity, which constrain their ability to meet expectations or drive innovation. This perspective advocates for adopting a Learning Health System - Learning Health Community (LHS-LHC) framework as a strategic approach to modernizing cancer data and data science infrastructure. We outline how cancer centers can develop integrated, interoperable, and inclusive data systems that enable continuous learning, promote equity-driven outcomes, and generate real-world impact. By embedding data and data science as core institutional infrastructure, and by aligning investments across health systems, funders, and community partners, cancer centers can enhance their role as regional anchors for coordinated, patient-centered care and lead a national transition toward data-enabled oncology.

The management of serious Serratia marcescens (SEM) infections is controversial due to the uncertainty of clinically significant ampC beta-lactamase overexpression. The purpose of this study was to evaluate differences in clinical outcomes in patients with SEM bloodstream infection (BSI) or pneumonia (PNA) who receive ceftriaxone (CRO), piperacillin-tazobactam (TZP), or cefepime (FEP).Table 1.Patient demographics and clinical characteristics of SEM managementTable 2.Effect Modification of the Primary and Secondary Outcomes Comparing Individual Beta-Lactams This IRB-approved retrospective, multi-center cohort study included adult patients hospitalized with SEM BSI or PNA over a 10-year period at UF Health. Based on the definitive antibiotic regimen, patients were assigned to CRO, TZP or FEP groups. The primary composite endpoint was treatment failure: persistent infection (fever or leukocytosis on day five of antibiotics), recurrent infection, or microbiological relapse. Secondary outcomes included hospital length of stay (LOS) and escalation of therapy. A total of 291 patients were included in the study and stratified by definitive antibiotic regimen: CRO (n=45), TZP (n=55), and FEP (n=191). Treatment failure did not differ significantly between CRO (48.9%), TZP (50.9%), and FEP (52.4%), p=0.911. There was no difference in median LOS, with 25.2 days (IQR: 11.2 - 37.1) in the CRO group, 14 days (IQR: 7.1-34.1) in the TZP group, and 17.5 days (IQR: 8.8 - 29.8) in the FEP group, p=0.784. A significantly higher proportion of patients in the CRO (20%) and TZP (32.7%) groups required escalation of therapy compared to FEP (7.3%), p< 0.001. Treatment with CRO or TZP was not associated with increased treatment failure rates in adult patients being treated for SEM BSI or PNA, but were associated with increased incidence of antibiotic escalation. All Authors: No reported disclosures

Abstract Introduction Independent Lung Ventilation (ILV) can be considered for asymmetric lung disease and is often applied using a dual-lumen airway. We present a case where ILV used a unique airway-configuration to allow for resolution of bronchopleural fistulas (BPF) while on VV ECMO. Case Description 41 year-old woman with Flu-A and secondary VAP was transferred for ECMO evaluation on day 9 of MV with prolonged paralysis. Pneumothoracies occurred due to aggressive bag ventilations after sudden loss of airway prior to ECMO initiation, with ongoing large left-sided air-leaks due to presumed BPF. After ∼7 days on ECMO, native oxygenation recovered substantially but her native ventilation remained poor. To heal her BPF, MV support was drastically reduced while her ECMO support (sweep) and sedation were intensified but within 24 hours overt recirculation developed due to RV failure. We instituted ILV since her right lung required moderate support for recruitment while allowing healing of the left-sided BPF. She had severe MDR PSA pneumonia complicated by bacteremia and left pleural-space infection so a bronchial blocker to isolate the left lung was not ideal nor was a double-lumen ETT since its small-diameter lumen can impede secretion clearance. Instead, we advanced an 8-0 ETT via her trach site into her right-main bronchus utilizing bronchoscopy to avoid occluding her RUL and applied APRV-TCAV to recruit and stabilize her right lung; We placed HFNC at 30L/min, 40% FiO2 for low-intensity support with humidification of the left lung while she breathed spontaneously. Recirculation resolved immediately as her PVR decreased when right lung volume was restored while allowing her left lung to have ambient pressure was sufficient to allow improved RV function but still reduce air-leaks. After twelve days, left-sided air-leaks resolved and her right lung had been stabilized and was now on lower MV settings, so her airway was pulled back into the distal trachea for standard ventilation. Her left-sided chest tubes were removed sequentially, and she was decannulated on ECMO day 60. Discussion ILV can be applied in settings of asymmetric lung disease so that specific support can be tailored to each hemithorax. When lung rest is applied on VV ECMO, acute RV failure may occur due to elevated PVR from loss of lung volume below FRC inducing overt recirculation. ILV may be helpful to avoid the need to convert ECMO strategy (VA or VPa) by applying specific MV support to each lung and safely rescue the failing RV. This abstract is funded by: None

BACKGROUND:Chronic obstructive pulmonary disease (COPD) is a heterogeneous condition whose clinical severity may be influenced by social factors. We aimed to identify distinct COPD clinical phenotypes and assess variation by social determinants of health. METHODS:In this retrospective cohort study, we identified adults aged 50-80 years with a diagnosis of COPD (n=59 797) at a tertiary academic medical centre in North-Central Florida using codes from International Classification of Diseases. Latent class analysis defined COPD clinical phenotypes using indicators of clinical severity, including frequency of acute care encounters (urgent care, emergency department visits and hospitalisations), presence of COPD as the principal diagnosis, comorbidity burden and use of Global Initiative for Chronic Obstructive Lung Disease Group D medications. Kaplan-Meier survival curves and Cox proportional hazards models assessed mortality across phenotypes. Multinomial logistic regression models estimated associations between phenotype membership and race/ethnicity, income, rurality and smoking status, using the minimal phenotype as reference. RESULTS:Five clinical phenotypes were identified: minimal (20.9%), mild (35.2%), moderate (22.5%), severe (12.2%) and very severe (9.3%). The very severe phenotype had the highest mortality (adjusted HR 2.94; 95% CI 2.72 to 3.18). Odds of very severe COPD were higher among non-Hispanic Black (adjusted OR (aOR) 1.29; 95% CI 1.21 to 1.36) and Hispanic individuals (aOR 1.75; 95% CI 1.63 to 1.87), those in the lowest income communities (aOR 1.25; 95% CI 1.18 to 1.32), rural residents (aOR 1.80; 95% CI 1.68 to 1.92) and individuals who currently smoke (aOR 1.30; 95% CI 1.20 to 1.42). CONCLUSION:Most patients with COPD had mild disease; however, the very severe phenotype, which was associated with higher mortality, was more common among Black and Hispanic individuals, those residing in lower-income and rural areas and those who currently smoke. These clinical phenotypes highlight sociodemographic differences in COPD severity as reflected in healthcare utilisation and outcomes.

Progressive loss of kidney function in chronic kidney disease and end-stage kidney disease leads to distinct pathological changes in mineral and bone metabolism. Historically, the management of these patients focused on lowering their serum phosphate, parathyroid hormone, and normalizing serum calcium. However, persistently high morbidity and mortality of these patients from cardiovascular diseases and after bone fractures call for more specific therapies. Here, we summarize established and emerging therapies for managing the three main domains of mineral bone disorder in patients with advanced chronic kidney disease and end-stage kidney disease: Control of mineral imbalance, management of vascular disease or other soft tissue calcifications, and management of bone disorders.