Browsing by Subject "Data analysis"

Now showing 1 - 7 of 7
  • Thumbnail Image
    Item
    Asking Data Analysis Questions with PandasAI
    (2023-11-08) Dolan, Levi
    As easily accessible AI models have increased in visibility, one area of interest for those working with datasets programmatically is how AI might streamline common data analysis tasks. The recently-released PandasAI library is a Python library that connects to an OpenAI model (known for ChatGPT) and allows users to ask natural language-style questions about dataframes created in Pandas syntax. This lightning talk demonstrates how to start exploring this data analysis method using sample World Bank and World Happiness Report data. Potential limitations are also discussed.
  • Thumbnail Image
    Item
    Coincidence Analysis: A Novel Approach to Modeling Nurses' Workplace Experience
    (Thieme, 2022) Womack, Dana M.; Miech, Edward J.; Fox, Nicholas J.; Silvey, Linus C.; Somerville, Anna M.; Eldredge, Deborah H.; Steege, Linsey M.; Emergency Medicine, School of Medicine
    Objectives: The purpose of this study is to identify combinations of workplace conditions that uniquely differentiate high, medium, and low registered nurse (RN) ratings of appropriateness of patient assignment during daytime intensive care unit (ICU) work shifts. Methods: A collective case study design and coincidence analysis were employed to identify combinations of workplace conditions that link directly to high, medium, and low RN perception of appropriateness of patient assignment at a mid-shift time point. RN members of the study team hypothesized a set of 55 workplace conditions as potential difference makers through the application of theoretical and empirical knowledge. Conditions were derived from data exported from electronic systems commonly used in nursing care. Results: Analysis of 64 cases (25 high, 24 medium, and 15 low) produced three models, one for each level of the outcome. Each model contained multiple pathways to the same outcome. The model for "high" appropriateness was the simplest model with two paths to the outcome and a shared condition across pathways. The first path comprised of the absence of overtime and a before-noon patient discharge or transfer, and the second path comprised of the absence of overtime and RN assignment to a single ICU patient. Conclusion: Specific combinations of workplace conditions uniquely distinguish RN perception of appropriateness of patient assignment at a mid-shift time point, and these difference-making conditions provide a foundation for enhanced observability of nurses' work experience during hospital work shifts. This study illuminates the complexity of assessing nursing work system status by revealing that multiple paths, comprised of multiple conditions, can lead to the same outcome. Operational decision support tools may best reflect the complex adaptive nature of the work systems they intend to support by utilizing methods that accommodate both causal complexity and equifinality.
  • Thumbnail Image
    Item
    End-User Needs of Fragmented Databases in Higher Education Data Analysis and Decision Making
    (MDPI, 2021) Briggs, Amanda; Cafaro, Francesco; Human-Centered Computing, Luddy School of Informatics, Computing, and Engineering
    In higher education, a wealth of data is available to advisors, recruiters, marketers, and program directors. These large datasets can be accessed using an array of data analysis tools that may lead users to assume that data sources conflict with one another. As users identify new ways of accessing and analyzing these data, they deviate from existing work practices and sometimes create their own databases. This study investigated the needs of end users who are accessing these seemingly fragmented databases. Analysis of a survey completed by eighteen users and ten semi-structured interviews from five colleges and universities highlighted three recurring themes that affect work practices (access, understandability, and use), as well as a series of challenges and opportunities for the design of data gateways for higher education. We discuss a set of broadly applicable design recommendations and five design functionalities that the data gateways should support: training, collaboration, tracking, definitions and roadblocks, and time
  • Thumbnail Image
    Item
    Journey from subjective to objective: Capturing user experience
    (2016-04-06) Lee, Yoo Young; Snajdr, Eric; Calvert, Lisa; Smith, Andy
    This presentation was delivered at the Designing for Digital 2016 held in Austin, Texas. It is the norm to conduct usability testing for library's websites. Often, these tests focus only on effectiveness or efficiency rather than measuring users’ experiential perspectives This presentation will introduce a variety of UX evaluation methods – different from usability – and cover UX research conducted in the fall 2015 semester.
  • Thumbnail Image
    Item
    Measuring and Visualizing Chlamydia and Gonorrhea Inequality: An Informatics Approach Using Geographical Information Systems
    (University of Illinois at Chicago, 2019-09-19) Lai, Patrick T.S.; Wilson, Jeffrey S.; Wu, Huanmei; Jones, Josette; Dixon, Brian E.; Geography, School of Liberal Arts
    Background: Health inequality measurements are vital in understanding disease patterns in identifying high-risk patients and implementing effective intervention programs to treat and manage sexually transmitted diseases. Objectives: To measure and identify inequalities among chlamydia and gonorrhea rates using Gini coefficient measurements and spatial visualization mapping from geographical information systems. Additionally, we seek to examine trends of disease rate distribution longitudinally over a ten-year period for an urbanized county. Methods: Chlamydia and gonorrhea data from January 2005 to December 2014 were collected from the Indiana Network for Patient Care, a health information exchange system that gathers patient data from electronic health records. The Gini coefficient was used to calculate the magnitude of inequality in disease rates. Spatial visualization mapping and decile categorization of disease rates were conducted to identify locations where high and low rates of disease persisted and to visualize differences in inequality. A multiple comparisons ANOVA test was conducted to determine if Gini coefficient values were statistically different between townships and time periods during the study. Results: Our analyses show that chlamydia and gonorrhea rates are not evenly distributed. Inequalities in disease rates existed for different areas of the county with higher disease rates occurring near the center of the county. Inequality in gonorrhea rates were higher than chlamydia rates. Disease rates were statistically different when geographical locations or townships were compared to each other (p < 0.0001) but not for different years or time periods (p = 0.5152). Conclusion: The ability to use Gini coefficients combined with spatial visualization techniques presented a valuable opportunity to analyze information from health information systems in investigating health inequalities. Knowledge from this study can benefit and improve health quality, delivery of services, and intervention programs while managing healthcare costs.
  • Thumbnail Image
    Item
    Near-universal hospitalization of US emergency department patients with cancer and febrile neutropenia
    (PLOS, 2019-05-23) Baugh, Christopher W.; Faridi, Mohammad Kamal; Mueller, Emily L.; Camargo, Carlos A.; Pallin, Daniel J.; Pediatrics, School of Medicine
    IMPORTANCE: Febrile neutropenia (FN) is the most common oncologic emergency and is among the most deadly. Guidelines recommend risk stratification and outpatient management of both pediatric and adult FN patients deemed to be at low risk of complications or mortality, but our prior single-center research demonstrated that the vast majority (95%) are hospitalized. OBJECTIVE: From a nationwide perspective, to determine the proportion of cancer patients of all ages hospitalized after an emergency department (ED) visit for FN, and to analyze variability in hospitalization rates. Our a priori hypothesis was that >90% of US cancer-associated ED FN visits would end in hospitalization. DESIGN: Analysis of data from the Nationwide Emergency Department Sample, 2006-2014. SETTING: Stratified probability sample of all US ED visits. PARTICIPANTS: Inclusion criteria were: (1) Clinical Classification Software code indicating cancer, (2) diagnostic code indicating fever, and (3) diagnostic code indicating neutropenia. We excluded visits ending in transfer. EXPOSURE: The hospital at which the visit took place. MAIN OUTCOMES AND MEASURES: Our main outcome is the proportion of ED FN visits ending in hospitalization, with an a priori hypothesis of >90%. Our secondary outcomes are: (a) hospitalization rates among subsets, and (b) proportion of variability in the hospitalization rate attributable to which hospital the patient visited, as measured by the intra-class correlation coefficient (ICC). RESULTS: Of 348,868 visits selected to be representative of all US ED visits, 94% ended in hospitalization (95% Confidence Interval [CI] 93-94%). Each additional decade of age conferred 1.23x increased odds of hospitalization. Those with private (92%), self-pay (92%), and other (93%) insurance were less likely to be hospitalized than those with public insurance (95%, odds ratios [OR] 0.74-0.76). Hospitalization was least likely at non-metropolitan hospitals (84%, OR 0.15 relative to metropolitan teaching hospitals), and was also less likely at metropolitan non-teaching hospitals (94%, OR 0.64 relative to metropolitan teaching hospitals). The ICC adjusted for hospital random effects and patient and hospital characteristics was 26% (95%CI 23-29%), indicating that 26% of the variability in hospitalization rate was attributable to which hospital the patient visited. CONCLUSIONS AND RELEVANCE: Nearly all cancer-associated ED FN visits in the US end in hospitalization. Inter-hospital variation in hospitalization practices explains 26% of the limited variability in hospitalization decisions. Simple, objective tools are needed to improve risk stratification for ED FN patients.
  • Thumbnail Image
    Item
    Research Conducted by Hospital Pharmacists: Integral Component of Daily Practice or Unrealistic Expectation?
    (Canadian Society of Hospital Pharmacists, 2018-03) Tisdale, James E.; Medicine, School of Medicine