Author ORCID Identifier

https://orcid.org/0009-0001-8351-882X

Semester

Summer

Date of Graduation

2024

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Industrial and Managements Systems Engineering

Committee Chair

Imtiaz Ahmed

Committee Co-Chair

JuHyeong Ryu

Committee Member

JuHyeong Ryu

Committee Member

Michael Russell

Abstract

The enduring impact of COVID-19 extends beyond acute illness, with potential long-term psychiatric consequences raising significant concern among healthcare professionals and researchers alike. Emerging evidence suggests a multifaceted relationship between COVID-19 and the development of different psychiatric illnesses like Schizophrenia Spectrum and Psychotic Disorders (SSPD), Depression, Bipolar disorder, Personality disorder, Trauma, and a range of other mental health conditions. Considering these emerging connections, our study endeavors to rigorously assess the associations between COVID-19 and various psychiatric illnesses while simultaneously employing machine learning techniques to predict the development of new psychiatric disorders in individuals affected by the virus. Leveraging the extensive dataset available through the N3C Data Enclave platform, which includes comprehensive information from over 19 million patients, we meticulously construct cohorts to facilitate this analysis. To properly quantify and compare the impacts with groups such as those diagnosed with Acute Respiratory Distress Syndrome (ARDS), COVID-positive individuals, and COVID-negative controls, our methodology involves utilizing propensity score matching to ensure cohort comparability. . By examining the hazard ratio (HR) of new-onset psychiatric illnesses across different time intervals post-infection, ranging from the acute phase to long-term follow-up, we aim to elucidate the temporal dynamics of psychiatric illness development following COVID-19 infection. In addition to traditional statistical analyses, we incorporate machine learning algorithms to predict the likelihood of individuals developing psychiatric disorders based on a wide range of demographic, clinical, and laboratory variables. This innovative approach not only enhances our ability to identify at-risk individuals but also enables personalized intervention strategies tailored to individual patient profiles. Through our study, we seek to provide valuable insights into the complex interplay between COVID-19 and psychiatric illness, shedding light on potential risk factors, trajectories, and outcomes associated with mental health disorders in the context of the pandemic. Furthermore, our predictive modeling efforts hold promise for early identification and intervention, potentially mitigating the long-term psychiatric sequelae of COVID-19 and improving patient outcomes. While our findings remain preliminary pending further analysis and validation, we anticipate that our research will have far-reaching implications for clinical practice, public health policy, and future research endeavors. By elucidating the specific role of COVID-19 in the development of psychiatric illnesses and employing cutting-edge machine learning techniques, we aim to contribute to a deeper understanding of Long-COVID and inform targeted interventions to support the mental health and well-being of individuals affected by the pandemic.

Download

Share

COinS