Doctor of Philosophy, The Ohio State University, 2019, Nursing

Breast cancer survivors are at risk of cancer recurrence and other cancer-related chronic diseases. Lifestyle modification reduces these risks; however, traditional approaches are costly and often lack efficacy. Mobile health (mHealth) interventions may offer advantages over traditional risk-reduction approaches, but limited knowledge about survivors' use of mHealth interventions restricts scientific advancement. The goal of this dissertation research was to develop a substantive theory to understand the process associated with the use of mHealth interventions by breast cancer survivors for lifestyle behavior improvement. Using a grounded theory approach, 16 female breast cancer survivors from central Ohio were enrolled. Each participated in an interview and an interaction with a prototype mHealth intervention. Data were analyzed using constant comparative analysis. The resultant substantive theory describes the synergy between mHealth Engagement and Self-regulation of lifestyle behaviors. The basic process enabling this synergy consists of 5 non-linear phases: adopting, sustaining, habituating, disengaging, and re-adopting. Four main concepts form the basis of this theory and include mHealth Engagement, Self-regulation, Relationships, and Functionality and Features. These findings may inform future mHealth intervention research and development. However, more research is needed to validate and test this new substantive theory.

Committee: Sharon Tucker PhD, RN, FAAN (Committee Chair); Randi Foraker PhD, MA, FAHA (Committee Member); Mary Beth Happ PhD, RN, FAAN, FGSA (Committee Member); Jennifer Kue PhD (Committee Member); Rita Pickler PhD, RN, FAAN (Committee Member); Po-Yin Yen PhD, RN (Committee Member) Subjects: Behavioral Sciences; Behaviorial Sciences; Computer Engineering; Computer Science; Health; Health Care; Health Care Management; Health Education; Information Science; Information Systems; Information Technology; Medicine; Mental Health; Nursing; Organization Theory; Psychology; Public Health; Systems Design; Systems Science; Technology

Master of Science, The Ohio State University, 2012, Allied Medical Professions

Background: The introduction of technology into the patient health care provider interaction and in particular the use of Electronic Health Records (EHRs) has revolutionized health data capture and analysis. However, electronic capture of health data has also changed the way health care communication occurs, affecting both clinicians and patients alike. Because patient satisfaction with a health care interaction is known to affect adherence to healthcare recommendations, it is important to identify the perceptions patients (or specific subgroups of patients) have of the medical encounter when the clinician uses an EHR. Positive patient perceptions of EHRs might provide health care providers with increased motivation to adopt health care technologies. Identifying any negative patient perceptions could provide health care providers with insight about how to better implement EHRs into the healthcare encounter. Design and Methods: A sample of Ohio adults aged 21 and older (N = 481) was recruited from the U.S. National Institutes of Health-sponsored ResearchMatch health research volunteer registry. Participants completed a secure online survey (55.4% response rate) that included sociodemographic questions, EHR and encounter perception questions, and the Shared Decision-Making Questionnaire-9 (SDM-Q-9). Measures were completed based on the respondent having had an interaction with a health care provider during the 3 months prior to survey completion. Multiple linear regression analysis employing forward selection was used to model correlates of shared decision-making. A bivariate correlation matrix was utilized to model associations between age and EHR perception. Results: Approximately two-thirds of patients in the study reported EHR use by their healthcare provider. The majority of those patients viewed EHR use favorably, and as an asset to their health. Those that reported no EHR use had mixed opinions regarding how an EHR might improve their care. Multiple linear regr (open full item for complete abstract)

Committee: Melanie Brodnik PhD (Committee Chair); Celia Wills PhD (Committee Member); Emily Patterson PhD (Committee Member) Subjects: Health Care

Doctor of Philosophy, The Ohio State University, 2012, Public Health

Background: Methicillin Resistant Staphylococcus aureus (MRSA) is a hardy and extremely virulent multidrug resistant organism that has been a major cause of hospital acquired infections ever since its discovery in the 1960's. It has severe consequences such as causing increased hospital length of stay, economic burden, morbidity, and mortality. MRSA prevention strategies have been advocated by national and international organizations which have been successful in reducing the burden of healthcare-associated MRSA. However, MRSA has been increasing in the community settings and this is an alarming and poorly understood rend because these infections occur in populations that have no known risk factors. In order to develop successful control strategies for this emerging threat. In order to develop successful control strategies for this emerging threat, it is important to understand the epidemiology, risk factors and links associated with community associated MRSA so that new and novel prevention strategies, using existing resources and cutting edge technology, can be developed. Methods: A cross sectional observational study design was used. The aims were accomplished by leveraging and utilizing the existing infrastructure of the OSUMC Information Warehouse, the Ohio State Health Network, and the OSUMC Microbiology Laboratory. Specific aim 1 was to develop an infection control collaborative and an innovative cross institutional platform, using existing information technology resources and infrastructure, for use as an electronic health information exchange between multiple hospitals spread across a large geographic area. Specific aim 2 was to estimate the proportion of community associated MRSA cases among all MRSA cases in rural community hospitals, and to identify the risk factors associated with community associated MRSA. Logistic regression was used to examine risk factors for community MRSA strain. Lastly, specific aim 3 was to identify patterns of intra-facility a (open full item for complete abstract)

Committee: Kurt Stevenson MD, MPH (Advisor); Philip Binkley MD, MPH (Committee Member); Melanie Brodnik PhD (Committee Member); Amy Ferketich PhD (Committee Member); Shu-Hua Wang TM & MPH (Committee Member) Subjects: Bioinformatics; Epidemiology; Public Health

MS, University of Cincinnati, 2023, Engineering and Applied Science: Computer Science

Anomaly Detection in Irregular Time Series is an under-explored topic, especially in the healthcare domain. An example of this is weight entry errors. Erroneous weight records pose significant challenges to healthcare data quality, impacting clinical decision-making and patient safety. Existing studies primarily utilize rule-based methods, achieving an Area Under the Receiver Operating Characteristic Curve (AUROC) ranging from 0.546 to 0.620. This thesis introduces a two-module method, employing bi-directional Long Short-Term Memory (bi-LSTM) with Attention Mechanism, for the prospective detection of anomalous weight entries in electronic health records. The proposed method consists of a predictor and a classifier module, both leveraging bi-LSTM and Attention Mechanism. The predictor module learns the normal pattern of weight changes, and the classifier module identifies anomalous weight entries. The performance of both modules was evaluated, exhibiting a clear superiority over other methods in distinguishing normal from anomalous data points. Notably, the proposed approach achieved an AUROC of 0.986 and a precision of 9.28%, significantly outperforming other methods when calibrated for a similar sensitivity. This study contributes to the field of entry error detection in healthcare data, offering a promising solution for real-time anomaly detection in electronic health records.

Committee: Raj Bhatnagar Ph.D. (Committee Chair); Danny T. Y. Wu PhD (Committee Member); Vikram Ravindra Ph.D. (Committee Member) Subjects: Computer Science

Doctor of Philosophy (PhD), Wright State University, 2023, Computer Science and Engineering PhD

Heart failure is a syndrome which effects a patient's quality of life adversely. It can be caused by different underlying conditions or abnormalities and involves both cardiovascular and non-cardiovascular comorbidities. Heart failure cannot be cured but a patient's quality of life can be improved by effective treatment through medicines and surgery, and lifestyle management. As effective treatment of heart failure incurs cost for the patients and resource allocation for the hospitals, predicting length of stay of these patients during each hospitalization becomes important. Heart failure can be classified into two types: left sided heart failure and right sided heart failure. Left sided heart failure can be further divided into two types: systolic heart failure or heart failure with reduced ejection fraction (HFrEF) and diastolic heart failure or heart failure with preserved ejection fraction (HFpEF). As right sided heart failure develops as a result of left sided heart failure, it is important to predict the two types of heart failures categorized based on their ejection volume to manage heart failure. Electronic Health Records (EHRs) of the patients contain information about the diagnostic codes, procedure reports, physiological vitals, medications administered, and discharge summary for each hospitalization. These EHRs can be leveraged to build predictive models to predict outcomes like length of stay and type of heart failure (HFrEF or HFpEF) in the patients. However, these predictive models can be demographically biased and so can lead to unfair decisions. Thus, it is necessary to mitigate these biases in the predictive models without impacting their performance on downstream tasks. In this regard, first I used diagnostic codes and procedure reports of the heart failure during each hospitalization to identify their clinical phenotypes through a probabilistic framework, using Latent Dirichlet Allocation (LDA). I found 12 clinical phenotypes in the form of (open full item for complete abstract)

Committee: Tanvi Banerjee Ph.D. (Committee Co-Chair); William L. Romine Ph.D. (Committee Co-Chair); Krishnaprasad Thirunarayan Ph.D. (Committee Member); Lingwei Chen Ph.D. (Committee Member); Mia Cajita Ph.D. (Committee Member) Subjects: Computer Engineering; Computer Science

PHD, Kent State University, 2023, College of Communication and Information

To competently perform nursing care at a high level while maintaining patient safety, newly hired graduate nurses need to understand each organization's unique sociotechnical culture and how this culture shapes their nursing behaviors towards health-IT usage and safety competency. The purpose of this study was to explore the relationship between pre-licensure nurses' informatics competencies and the influence of healthcare experience in modern day complex adaptive system's sociotechnical culture to describe informatics and safety competencies in students. A descriptive, correlational cross-sectional study of 178 Bachelors of Nursing (BSN) students from 3 Midwestern and1 Pacific Northwest BSN schools using an online survey was undertaken. The Self-Assessment of Informatics Competency (SANICS) scale and a subset of the Surveys on Patient Safety Culture (SOPS™) Hospital Survey were used to quantify the degree of association between student informatics competency and safety competency. The findings reveal pre-licensure work experience was associated with an increase in both nursing informatics and safety competency when compared to students without pre-licensure work experience. There was a positive correlation between perceived usefulness of technology and nursing informatics competency but not with safety competency. Higher BSN grade level was a predictor of higher nursing informatics competency but not a predictor for safety competency. Nursing informatics competency was a predictor of higher safety scores explaining 12.9% of the variance in patient safety (R2 = 0.129, F(1, 176) = 26.04, p < .001). Having a nursing informatics course was not associated with improving overall nursing informatics competency or safety competency. Nursing informatics competency was positively associated with improvements in safety competency. All members of the healthcare team including administration, students, faculty, and staff should understand that there are benefits of exposure to (open full item for complete abstract)

Committee: Rebecca Meehan (Committee Chair); Amy Petrinec (Committee Member); Tang Tang (Committee Member); Miriam Matteson (Committee Member) Subjects: Bioinformatics; Communication; Health Care; Health Education; Information Technology; Nursing; Teaching

Doctor of Philosophy (PhD), Wright State University, 2023, Computer Science and Engineering PhD

This research explores data-driven AI techniques to extract insights from relevant medical data for pain management in patients with Sickle Cell Disease (SCD). SCD is an inherited red blood cell disorder that can cause a multitude of complications throughout an individual's life. Most patients with SCD experience repeated, unpredictable episodes of severe pain. Arguably, the most challenging aspect of treating pain episodes in SCD is assessing and interpreting the patient's pain intensity level due to the subjective nature of pain. In this study, we leverage multiple data-driven AI techniques to improve pain management in patients with SCD. The proposed approaches have been evaluated on physiological, medicinal and pain measurements collected from Electronic Health Records (EHRs), demonstrating their ability to digitize the medical essence of patients, thereby assisting in multiple aspects of clinical decision making in pain management. First, we propose to explore the feasibility of estimating subjective pain from objective physiological signals collected from EHRs irrespective of the nature of hospital visits in large patient cohorts. Second, we propose to learn deep feature representations of the subjective pain trajectories from objective physiological signals collected from EHRs. Third, we propose to learn future pain from historical patient EHR data using time-series forecasting methods. Our initial results indicate promise in pursuing each of these three efforts, and our study can be a valuable addition to ongoing studies that utilize EHR data to help providers better understand and design real-time pain management strategies.

Committee: Tanvi Banerjee Ph.D. (Advisor); Krishnaprasad Thirunarayan Ph.D. (Committee Member); Michael Raymer Ph.D. (Committee Member); Nirmish Shah M.D. (Committee Member) Subjects: Computer Engineering; Computer Science; Health Care

PhD, University of Cincinnati, 2021, Medicine: Biomedical Informatics

Epilepsy affects over 50 million people and is responsible for 1.3% of deaths worldwide. While many people with epilepsy benefit from pharmacotherapy, approximately one-third have drug-resistant epilepsy. Clinical guidelines recommend that these patients be promptly evaluated for resective epilepsy surgery, which is associated with a 67% chance of long-term seizure freedom. Unfortunately, surgery is underutilized. The mean disease duration at the time of surgery is six years in pediatrics and 20 years in adults. Identifying surgical candidates earlier in the disease course is needed to optimize care. Machine learning models have been developed to aid in this process, and sending alerts based on these model's recommendations is associated with a three-fold increase in referrals. Electronic health record-based machine learning models have been developed to predict a variety of health outcomes. However, they are rarely implemented into clinical care. Generalizing electronic health record-based predictive models across health care systems is challenging for several reason. Electronic health record data contain process biases, health care delivery patterns change over time, patient populations vary across geographical regions, and documentation styles vary by provider. Overcoming these limitations will improve the likelihood that machine learning-based clinical decision support systems will enhance patient care. To address these challenges, this dissertation proposed a novel methodology for generalizing electronic health record-based machine learning models across health care systems. The set of procedures produced site-specific models with modifiable feature sets and parameter weights. Data preprocessing, feature selection, and model training were repeated at each institution without modification. The results showed that generalizing modelling processes, rather than the models themselves, minimized generalization error. These methods were used to identify candida (open full item for complete abstract)

Committee: Judith Dexheimer Ph.D. (Committee Chair); David Ficker M.D. (Committee Member); Tracy Glauser M.D. (Committee Member); John Pestian Ph.D. (Committee Member); Rhonda Szczesniak Ph.D. (Committee Member) Subjects: Surgery

PhD, University of Cincinnati, 2021, Medicine: Biomedical Informatics

Biomedical informatics, when conscientiously developed and applied to help patients, can lead to identification of health inequities and possible ways to address them. We describe a new sub-field called “equiformatics”, through which we can consciously develop datasets, tools, models, and knowledge to address these inequalities. This dissertation seeks to explore examples of racial disparities in medical disorders that disproportionately impact African Americans (AA), and demonstrate the feasibility and value of developing tools to address these health inequities. In Unraveling Racial Disparities in Asthma Emergency Department Visits Using Electronic Healthcare Records and Machine Learning, we found disparities leading to increased number of emergency department visits for asthmatic AA. We found a disproportionate correlation between asthma exacerbations and exposures to pollen and mold among AA compared to European Americans (EA). Similar disproportions showed up in End-Stage Kidney Disease (ESKD), where AA received fewer kidney transplants and remained on hemodialysis for longer periods of time. Hepatitis C Virus (HCV) exposed kidneys from deceased donors may present an opportunity to ameliorate some disparities understand the increased risks involved. To assess patients' values and preferences for ESKD-related health states and kidney transplantation, we undertook a study using a utility assessment platform we developed to explore racial/ethnic differences in valuation of ESKD-related health states. In An Automated Tool for Health Utility Assessments: The Gambler II, we developed The Gambler, an authoring platform to perform health utility assessments (HUA) using various methods, including the standard gamble, time tradeoff, and visual analog scale. We demonstrated the feasibility of this platform through a study eliciting health utilities complications of type II diabetes mellitus: diabetic retinopathy, diabetic neuropathy, and diabetic foot infection requi (open full item for complete abstract)

Committee: Mark Eckman M.D. (Committee Chair); Tesfaye Mersha Ph.D. (Committee Member); Silvi Shah M.D. (Committee Member); Danny T. Y. Wu PhD (Committee Member) Subjects: Bioinformatics

Master of Science, The Ohio State University, 2017, Allied Medicine

Objective: Patient physiological monitoring creates a large number of alarms, most of which are false. High numbers of false alarms inhibit discrimination between true and false alarms leading to the neglect of future alarms, both false and true, risking slower identification and reaction to hazardous conditions. This study introduces several methods, especially novel visualizations, to discern how alarms are temporally distributed, and how alarms coalesce as sets of alarms. Methods: Retrospective evaluation of data extracted from a hospital-system-wide middleware alarm escalation software database containing million of alarms over a time period of 16 to 18 months. Multiple comparison of means is employed as well as several visualizations including, box-and-whisker plots, periodograms, and a novel Gantt-inspired visualization in combination with a histogram. Results: Multiple comparison of means finds statistically significant differences between alarms occuring on an hourly, daily, and shift-wise basis. Box-and-whisker visualization of alarms by hour over a week reveals visual signatures of alarm occurence varying on a unit-by-unit basis. Periodograms reveal multiple periodicities in alarm occurrence varying on a unit-by-unit basis. Study of simultaneous alarms uncovers quantizations such as the highest numbers of alarms occuring by unit (6 to 10 simultaneous alarms). Gantt-style visualization of simultaneous alarm occurences uncovers interesting alarm signatures such as threshold hovering of alarms, appearing as a visual stutter, or the redundancy of certain alarms (e.g. bradycardia and low heart rate) which occur in parallel.Long-term, there is a large percentage of time that at least one alarm is sounding on a unit (18.1% to 62.2%). Conclusions: Retrospective evaluation of a middleware alarm escalation software database in combination with novel visualization provides a valuable heuristic tool.

Committee: Emily Patterson (Advisor); Laurie Rinehart-Thompson (Committee Member); Michael Rayo F (Committee Member) Subjects: Engineering; Health Care; Health Sciences; Information Science; Information Systems; Medicine

Doctor of Philosophy (PhD), Wright State University, 2016, Computer Science and Engineering PhD

Understanding users' latent intents behind search queries is essential for satisfying a user's search needs. Search intent mining can help search engines to enhance its ranking of search results, enabling new search features like instant answers, personalization, search result diversification, and the recommendation of more relevant ads. Hence, there has been increasing attention on studying how to effectively mine search intents by analyzing search engine query logs. While state-of-the-art techniques can identify the domain of the queries (e.g. sports, movies, health), identifying domain-specific intent is still an open problem. Among all the topics available on the Internet, health is one of the most important in terms of impact on the user and forms one of the most frequently searched areas. This dissertation presents a knowledge-driven approach for domain-specific search intent mining with a focus on health-related search queries. First, we identified 14 consumer-oriented health search intent classes based on inputs from focus group studies and based on analyses of popular health websites, literature surveys, and an empirical study of search queries. We defined the problem of classifying millions of health search queries into zero or more intent classes as a multi-label classification problem. Popular machine learning approaches for multi-label classification tasks (namely, problem transformation and algorithm adaptation methods) were not feasible due to the limitation of label data creations and health domain constraints. Another challenge in solving the search intent identification problem was mapping terms used by laymen to medical terms. To address these challenges, we developed a semantics-driven, rule-based search intent mining approach leveraging rich background knowledge encoded in Unified Medical Language System (UMLS) and a crowd-sourced encyclopedia (Wikipedia). The approach can identify search intent in a disease-agnostic manner and has been eva (open full item for complete abstract)

Committee: Amit Sheth Ph.D. (Advisor); Krishnaprasad Thirunarayan Ph.D. (Committee Member); Michael Raymer Ph.D. (Committee Member); Jyotishman Pathak Ph.D. (Committee Member) Subjects: Computer Science

Doctor of Philosophy, The Ohio State University, 2008, ED Physical Activities and Educational Services

The primary purpose of this study was to determine the extent to which the data collected on ARRT certified radiographers utilizing digital imaging equipment support the modified TAM2 theoretical model in a radiology department where the equipment use is mandated. This study measured model fit using SEM as well as examining a series of dependence relationships between observed exogenous and endogenous variables.The study population consisted of 120 radiographers certified by the American Registry of Radiologic Technologists (ARRT) utilizing direct capture digital radiographic units in The Ohio State University Medical Center Healthcare System. A survey method was used to investigate the applicability of the modified TAM2 utilizing a written questionnaire adapted from previous Technology Acceptance Model (TAM) and Technology Acceptance Model 2 (TAM2) studies. The response rate was 92.5%. The results of this study indicate the data does support the implied theoretical model; however age and gender were shown to have little impact on the original model. Standardized regression coefficients (β) were used to examine a series of dependence relationships between observed exogenous and endogenous variables. Two relationships were identified in this study in reference to the intention to use digital imaging equipment in a mandated healthcare environment. A relationship was recognized involving voluntariness and subjective norm and concerning subjective norm and the intention to use the digital imaging equipment. These findings support previous research indicating that the social context in which the technology is employed plays an important role in an individual's decision to ultimately use the technologic innovation.

Committee: David Stein PhD (Advisor); Joseph Gliem PhD (Committee Member); Melanie Brodnik PhD (Committee Member) Subjects: Health Care; Information Systems