According to a survey, the healthcare industry is one of the least cloud-adopting industries. The low adoption reflects the healthcare industry's ongoing concerns about the security of the cloud. Business applications, according to another survey, are among the most vulnerable components of business information systems. Many risk assessment frameworks available today, particularly for health information applications, require significant customization before they can be used. This study created a new framework to assess cloud risks specifically for their health information applications, utilizing data-driven risk assessment methodologies to avoid surveys, interviews, and meetings for data collection. For the feasibility study, the open-source application codes were chosen from over 190 million GitHub repositories using a decision tree method, while a purposive sampling method was used to choose for a simulated patient information database from the healthcare industry. Using these methods, the researcher discovered security warnings and privacy violation suspects and subsequently converted them into quantitative measures to calculate the risks of the cloud-based health information application and a database. The significance of this study lies in the collection of data directly from applications and databases with a quantitative approach for risk calculation.

Medicaid is the largest health insurance in the U.S. It provides health coverage to over 68 million individuals, costs the nation over $600 billion a year, and subject to improper payments (fraud, waste, and abuse) or inaccurate payments (claim processed erroneously). Medicaid programs partially use Fee-For-Services (FFS) to provide coverage to beneficiaries by adjudicating claims and leveraging traditional inferential statistics to verify the quality of adjudicated claims. These quality methods only provide an interval estimate of the quality errors and are incapable of detecting most claim adjudication errors, potentially millions of dollar opportunity costs. This dissertation studied a method of applying supervised learning to detect erroneous payment in the entire population of adjudicated claims in each Medicaid Management Information System (MMIS), focusing on two specific claim types: inpatient and outpatient. A synthesized source of adjudicated claims generated by the Centers for Medicare & Medicaid Services (CMS) was used to create the original dataset. Quality reports from California FFS Medicaid were used to extract the underlying statistical pattern of claim adjudication errors in each Medicaid FFS and data labeling utilizing the goodness of fit and Anderson-Darling tests. Principle Component Analysis (PCA) and business knowledge were applied for dimensionality reduction resulting in the selection of sixteen (16) features for the outpatient and nineteen (19) features for the inpatient claims models. Ten (10) supervised learning algorithms were trained and tested on the labeled data: Decision tree with two configurations - Entropy and Gini, Random forests with two configurations - Entropy and Gini, Naïve Bayes, K Nearest Neighbor, Logistic Regression, Neural Network, Discriminant Analysis, and Gradient Boosting. Five (5) cross-validation and event-based sampling were applied during the training process (with oversampling using SMOTE method and stratification within oversampling). The prediction power (Gini importance) for the selected features were measured using the Mean Decrease in Impurity (MDI) method across three algorithms. A one-way ANOVA and Tukey and Fisher LSD pairwise comparisons were conducted. Results show that the Claim Payment Amount significantly outperforms the rest of the prediction power (highest Mean F-value for Gini importance at the α = 0.05 significance) for both claim types. Finally, all algorithms' recall and F1-score were measured for both claim types (inpatient and outpatient) and with and without oversampling. A one-way ANOVA and Tukey and Fisher LSD pairwise comparisons were conducted. The results show a statistically significant difference in the algorithm's performance in detecting quality issues in the outpatient and inpatient claims. Gradient Boosting, Decision Tree (with various configurations and sampling strategies) outperform the rest of the algorithms in recall and F1-measure on both datasets. Logistic Regression showing better recall on the outpatient than inpatient data, and Naïve Bays performs considerably better from recall and F1- score on outpatient data. Medicaid FFS programs and consultants, Medicaid administrators, and researchers could use this study to develop machine learning models to detect quality issues in the Medicaid FFS claim datasets at scale, saving potentially millions of dollars.

The purpose of this study was to survey perceived job satisfaction and turnover intention of information technology professionals in the California State University (CSU) system. Employee satisfaction facets (work, pay, opportunities for promotion, supervision, and coworkers), overall satisfaction, and turnover intention were measured. Further, the study identified whether there was a significant difference in perceived job satisfaction or turnover intention based on years of service in the CSU system, gender, or campus in the CSU system. The study also examined the uniqueness of information technology professionals at campuses in the CSU system. This study utilized a mixed-methods methodology with two distinct phases. The quantitative phase of the study involved participants responding to an on-line survey. An invitation was sent to 622 information technology professionals at six campuses in the CSU system with a request to complete the survey. A total of 59 information technology employees responded, for a response rate of 9.49%. The quantitative results support earlier studies that report a negative correlation between overall job satisfaction and turnover intention. Of the five facets of job satisfaction, the mean satisfaction with opportunities for promotion was the lowest. The qualitative phase followed the quantitative phase and involved interviewing information technology managers from the CSU system, using a semi-structured interview protocol, to gain additional clarity about the data gathered in the quantitative phase. The managers did not perceive a difference between the job satisfaction of information technology professionals and other professionals. The majority of managers reported viewing turnover positively, but suggested that their view of turnover is highly situational depending on whether the turnover is of a high or low performer. The culture of information technology professional turnover intention was described as somewhat different for other professionals given the ease of skills transfer and demand for information technology professionals in the market.

One of US hospitals' widely used critical performance or quality outcome measures is the 30-day emergency department (ED) visit after a surgical procedure. Such ED visits add millions of dollars each year as a cost burden to US healthcare. This study aimed to identify key predictors known before the patient's surgery, contributing to undesirable ED visits within 30 days of a bariatric surgical procedure. The study was conducted in three phases. The first phase of the study engaged a panel of experts to narrow down important preoperative factors for patients undergoing bariatric surgery in the form of a Delphi study. The second phase of the study included quantitative data analysis, which utilized the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program Participant Use Data File of the year 2019 to identify statistically significant preoperative factors that can contribute to the likelihood of patients returning to the emergency department within 30 days of bariatric surgery. There were N = 193,774 cases with complete information from 868 MBSAQIP-accredited bariatric surgery centers across the United States in the Data File among which 15,533 (8% of the total cases) visited an ED without needing admission as inpatients. The analysis also examined the feasibility of developing a predictive model with only statistically significant factors and checking if the model has an acceptable fit. The third phase of the study reengaged the same panel of experts from the first phase to validate the findings from the second phase and to document the subject matter experts' perception regarding the model developed and the overall findings. Out of 33 preoperative variables, only 9 variables were selected in the first phase of the study with the help of a panel of experts. Out of the 9 chosen variables, 8 variables, i.e., Pre-Op GERD requiring medication, Number of Hypertensive Medications, Pre-Op BMI closest to bariatric surgery, Highest Recorded Pre-Op BMI, Pre-Op vein thrombosis requiring therapy, Pre-Op diabetes mellitus, Pre-Op history of COPD, and Pre-Op Steroid/Immunosuppressant Use for Chronic Condition significantly contributed to the likelihood of patients coming back to ED within 30 days of bariatric surgery. The study's second phase also yielded a predictive model using only the statistically significant and weighted variables, and each predictor exhibited statistical significance. In the third phase, a panel of experts weighed in mostly with positive feedback deeming the study clinically and operationally valuable for the bariatric patient population. The practical implication of this study is that the MBSAQIP Centers can use the model to determine the probability of a patient's likelihood of returning to ED after a bariatric surgical procedure. Based on the set criteria, if the patient has a higher chance of returning to ED, the care team can take interventions during and in the first few days or weeks of the discharge to prevent potential postoperative ED visits within 30 days of bariatric surgery.

Electronic Health Record (EHR) is a technology innovation which has the potential to offer valuable benefits to the healthcare industry such as improved quality of patient care and safety, optimization of healthcare workflow processes and availability of electronic data for clinical research. The implementation success of EHR is therefore significant to the healthcare industry in the United States and around the world. Prior studies in research literature have considered the impact of technology attributes, organizational learning attributes, and service attributes on information technology implementations in various other domains based on theories such as Theory of Reasoned Action (TRA), Theory of Planned Behavior (TRB) and Technology Acceptance Model (TAM), but none have considered their association with implementation success in a comprehensive manner within a single study pertaining to the healthcare domain as this study does. Hence, this study addresses an essential research gap. The approach used by this study in conducting the research based on a multi-factor research model (including the aforementioned attributes) is consistent with the general method used by academic researchers whereby the ability of a unique and selective list of factors to predict certain outcomes is leveraged. The data for this research study was collected using a questionnaire survey instrument based on the Likert scale. Structural Equation Modeling (SEM) was used for data analysis due to the presence of latent variables in the research model. The results of the statistical analyses support the hypotheses confirming positive associations between technology attributes (ease of use, result demonstrability, performance expectancy), organizational learning attributes (organizational learning capability, organizational absorptive capacity), service attributes (service-dominant orientation), and EHR implementation success. The results of this study are of importance to both academicians and practitioners.

This study was conducted to understand the variables, processes, organizational structures, and governance structures that are important and/or that support higher education decision makers in their selection and prioritization of information technology (IT) projects into their universities’ portfolios. IT project portfolio management (PPM) is comprised of many different activities, and the selection and prioritization of projects are just two interconnected activities amongst many. Research has suggested that these PPM activities are both important and beneficial; but there is a dearth of research on the subject specifically within higher education IT environments, and some higher education organizations struggle in this area. This study follows recent recommendations from other researchers to perform practice-based research on IT PPM. Research streams and standards bodies have long espoused the ideals of strategic IT PPM, where organizational strategy is perceived as a driver that strongly guides the practical activities and operations of IT PPM. However, there is a growing recognition that there is room for practice-based research because those ideals of strategic IT PPM are often not aligned with actual IT PPM practices and outcomes, and because IT PPM in practice often results in a bottom-up means for affecting strategy. This study used a qualitative research design, and included a practice-based exploratory multiple-case study focused on project selection and prioritization activities as they occur within real world higher education IT settings at eight universities in the California State University system. Each university acted as an individual case within the multiple-case study. Interviews were conducted with 27 subjects across these eight universities, and a breadth of other evidence was collected including documentation, physical artifacts, and archival records. Converging lines of data were developed through triangulation and corroboration of all the evidence, and this formed the informational basis for each case. Results from each case were reported independently, and a cross case synthesis was conducted to aggregate findings across all eight cases. In addition to questions about the mechanics of project selection and prioritization, the interviews also included questions that were designed to compare and contrast perceptions of technical and non-technical stakeholders. Twelve themes emerged as issues of importance including objectivity, formality, flexibility, alignment with the strategic plan, the difficulty for small projects to compete with large/enterprise projects, senior leadership involvement, transparency in decision-making, transparency in PPM mechanics, the need for consultation and responsiveness, capacity planning, governing bodies’ makeup and their representation of campus stakeholders, and satisfaction with the IT organization (and with its project management office). Technical and non-technical subjects’ perceptions were aligned throughout most of the twelve themes, but there were indeed areas where opinions differed.

Recent hacking events against Sony Entertainment, Target, Home Depot, and bank Automated Teller Machines (ATMs) fosters a growing perception that the Internet is an insecure environment. While Internet Privacy Concerns (IPCs) continue to grow out of a general concern for personal privacy, the availability of inexpensive Internet-capable mobile devices increases the Internet of Things (IoT), a network of everyday items embedded with the ability to connect and exchange data. Domain Name Services (DNS) has been integral part of the Internet for name resolution since the beginning. Domain Name Services has several documented vulnerabilities; for example, cache poisoning. The solution adopted by the Internet Engineering Task Force (IETF) to strengthen DNS is DNS Security Extensions (DNSSEC). DNS Security Extensions uses support for cryptographically signed name resolution responses. The cryptography used by DNSSEC is the Public Key Infrastructure (PKI). Some researchers have suggested that the time stamp used in the public certificate of the name resolution response influences DNSSEC vulnerability to a Man-in-the-Middle (MiTM) attack. This quantitative study determined the efficacy of using the default relative Unix epoch time stamp versus an absolute time stamp provided by the Network Time Protocol (NTP). Both a two-proportion test and Fisher’s exact test were used on a large sample size to show that there is a statistically significant better performance in security behavior when using NTP absolute time instead of the traditional relative Unix epoch time with DNSSEC.

Cardiovascular diseases are the leading cause of death in the United States. Advances in wireless technology have made possible the remote monitoring of a patient’s heart sensors as part of a body area network. Previous studies have suggested that stray wireless transmissions in the industrial, scientific, and medical (ISM) band cause interference resulting in packet loss in Bluetooth piconets. This study investigates the impact that wireless transmissions from residential microwave ovens have on the Bluetooth Low Energy (BLE) component of the body area network. Using a systematic data collection approach, two variables were manipulated. The distance between the microwave oven and the BLE piconet was varied from 0.5 meter to 5.0 meters at one-half meter increments. At each distance, the power level of the microwave oven was varied from the lowest power setting to the highest power setting. The two variables that were collected were the microwave interference generated by channel and the packet loss by channel. The results suggest more packet loss is due to the microwave oven’s power level than by the distance, the interference caused by the microwave oven affects all BLE channels equally, and the packet loss by channel is a good predictor of microwave oven interference. The significance of this study lies in providing beneficial information to the medical and digital communication industries concerning the causes and solutions to disruptions in the Bluetooth-enabled body area network devices in a very common situation. The results of this study may lend support for improvements and widespread use of body area network medical systems, which may have the benefit of better monitoring, more data, and reduced fatalities due to misdiagnosed heart conditions.

Additive Manufacturing rapid reproductive systems are gaining popularity within the manufacturing industry. One of the many benefits of such systems has been the exploration of building practical sacrificial patterns for investment casted metals. Methods such as, Castform and Quickcast, has been developed for selective laser sintering and Stereolithography apparatus technologies respectively. Research has demonstrated significant cost savings when Additive manufacturing rapid reproductive systems are utilized for customized or small batch production of sacrificial patterns. The purpose of this study was to develop a methodology for evaluating quality characteristics of Fused Deposition Modeling. Since Fused Deposition Modeling have been demonstrated by a number of experimental studies as a viable alternative to wax sacrificial patterns, this study explored the effects of wall thickness and raster resolution on quality characteristics such as, diametric accuracy, cylindricity, and concentricity. The results of the study indicated raster resolution had no effect on the measured quality characteristics, however, the ANOVA and Kruskal-Wallis tests showed statistical significance (α=0.05) for wall thickness of cylindricity of a small diameter (0.5”) and concentricity of two cylindrical features of diameters 0.5” and 1”. Moreover, the main contributions of this study involved the development of an accurate and robust design of experiment methodology. In addition, implications and recommendations for practice were also discussed.

In the last four or five decades, increased political and social pressure has been placed on commercial and residential consumers to reduce consumption of fossil fuels and invest in alternative methods of energy production for electricity, heating and cooling. Commercially, wind turbines and photovoltaic energy production equipment has sprung up all over the country. Other forms of energy production such as hydroelectric and geothermal energy production facilities have also been built. During this time however, very few residential ‘green energy’ investments have been made. Only in recent years have residential home owners begun to “wet their feet” on ‘green’ energy equipment. Cost has been the major factor. Of late though, costs have been coming down and efficiency has been going up making home owners begin to sense that alternative energy may now be entering the realm of economic feasibility. Unfortunately, home owners have had no reliable or credible tools to assess economic viability of such systems. The purpose of this research is to develop a tool to access the potential of alternative energy sources and test it statistically by surveying subjects in five different ‘green’ energy categories. Since atmospheric (air-to-air) heat pumps have been around for many years and represent a mature heating and cooling technology, upgrading older inefficient HVAC equipment to new high efficiency atmospheric heat pumps is the category used to baseline the experiment. Ground source heat pumps and direct solar heating systems were modeled and compared to the baseline. Wind energy and photovoltaic energy production systems were modeled, surveyed and compared to using only grid supplied electricity. Results show that in four of the five cases tested, the less mature ‘green’ energy equipment; photovoltaic solar, direct solar and ground source heat pump equipment are in general not economically viable without tax rebates to significantly lower the net investment. Setback rules and environmental and aesthetic ordinances against siting them in those counties severely restrict the population of wind energy devices so that an effective test of this category using the model could not be done. The model performed well with the baseline data. Performance of the model with ground source heat pumps was reasonable, but improvements in the model reflecting differing features of ground source heat pumps need to be made. Performance of the model with photovoltaic energy production equipment was also good. Extending the test population to all fifty states and extending the utility bill test range from one year to five years will provide much more useful data to test and improve the model. Although the model development and testing done in this work only represents a small contribution to the bridging of a large gap in consumer confidence in green energy products, it represents a big step into an area that very few have attempted to venture into.