Diabetic patients are typically diagnosed and monitored through glucose levels, with HbA1c being a critical marker for long-term glycemic control. Health disparities among vulnerable populations often impair access to optimal care, exacerbating outcomes, particularly in chronic conditions like diabetes. The COVID-19 pandemic has significantly disrupted glycemic control and worsened diabetes-related complications, highlighting the need for an in-depth understanding of its impact on HbA1c levels. This study aims to analyze the impact of COVID-19 on glycemic control in diabetic patients, specifically focusing on changes in HbA1c levels post-infection. Traditional statistical methods may not fully capture the complex interactions between COVID-19 and glycemic control. Therefore, this study seeks to leverage machine learning techniques to gain a deeper understanding of the impact of COVID-19 on glycemic control in diabetic patients. By integrating advanced analytical methods with comprehensive data from the NCATS N3C enclave, the research aims to provide valuable insights that can inform targeted interventions and improve diabetes care in the post-pandemic landscape.

Lead Investigator: Marie Lluberes
Accessing Institution: University of Puerto Rico
ID: RP-74F392 (DUR-917CCEA)

The COVID-19 pandemic has placed tremendous pressure on clinical research and innovation. There is an urgent need to understand, prevent and treat COVID-19 disease. It has threatened the traditional models of knowledge translation and practice in treating patients. Small-scale clinical trials and cohort studies are reporting anecdotal evidence claiming contradictory therapeutic successes. On the other hand, large-scale trials are time-consuming, costly, and sometimes infeasible. The National Institute of Health (NIH) has made COVID-19 patient (observational) data available to researchers for faster translation to new therapies and knowledge to address the pandemic and prepare for other diseases in the future. The advances in causal inference, particularly Structural Causal Models (SCM) can help translate this data to knowledge. The goal of this proposed research is to develop a framework for estimation of treatment effect by modeling unobserved confounding in SCMs that specifically address the practical challenges of performing virtual experiments using COVID-19 patient data. The clinical application entails a timely and important research question, the effect of oxygen therapy on mortality in COVID-19 patients in the ICU. We will work closely with domain experts, critical care physicians, to incorporate evidence-based practices, recommendations and to validate the causal structure. The methods will help practitioners in the area of artificial intelligence, machine learning, data science, and healthcare to leverage the vast amount of COVID-19 data collected from multiple sources to extract knowledge and meaningful conclusions.

Lead Investigator: Md Osman Gani
Accessing Institution: University of Maryland, Baltimore County
ID: RP-2DA60E (DUR-140E37D)