Browsing by Author "Williams, Jordan Chase"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Chronicity among older psychiatric inpatients
    (University of Alabama Libraries, 2018) Williams, Jordan Chase; Parmelee, Patricia A.; University of Alabama Tuscaloosa
    The goal of this dissertation is to examine factors that could relate to length of stay (LOS), latency to readmission, and number of admissions. Length of stay is generally defined as the amount of time an individual stays in the facility; latency to readmission is the amount of time between the last discharge and the most recent admission for each individual. I examined several factors that related to these outcomes, including number of falls, number of assaults, medical diagnoses, psychological diagnoses, frequency of PRN medication, emergency department visits and seclusion/restraints. Structural equation modeling was used to create statistical models incorporating these variables. Several models were proposed using each variable (LOS, admits, and latency) as outcome variables. Based on these models, integrated models were proposed and analyzed. Results based on these models were discussed and future research projects were proposed.
  • Thumbnail Image
    Item
    The effect of staffing variables on nursing home resident outcomes
    (University of Alabama Libraries, 2015) Williams, Jordan Chase; Parmelee, Patricia A.; University of Alabama Tuscaloosa
    The relationship between nursing home staffing variables and resident outcomes was investigated using a secondary data analysis. Data from the 2004 National Nursing Home Survey (NNHS) was linked with variables obtained from Nursing Home Compare in 2005. Minimum Data Set (MDS) quality indicators of facility-level depression and pain were used as outcome variables. This is a departure from previous studies which used primarily physiological outcome variables. Using a series of regression analyses, NNHS variables related to individual-level job factors, staff benefits, and staff richness were used to predict facility-level pain and depression. It was hypothesized that all three of these constructs would predict better resident outcomes in the form of reduced depression and pain. Staff stability and richness were also investigated as potential mediators. The overall model was not supported, although several interesting relationships were found. As nursing home leadership experience increased, so did facility-level depression scores. The number of volunteers and volunteer hours had a similar relationship to depression. Greater wages, full-time staff turnover, and volunteer duties were associated with decreased depression rates. Staff HPPD, hourly wages, and several control variables were associated with decreased rates of pain. Implications for future studies are discussed, including the need for firsthand data collection and the reduction of measurement bias.