Outcome Variables. Two outcomes of interest in this study are “unhealthy weight and height children” and “mother-child pairs”. Both variables are assessed utilizing weight and height data. Five outcome variables are created to assess children’s weight and height, which are named “wasted”, “underweight”, “normal weight”, “overweight” and “stunted”. These five variables are categorized using the z-score based on the World Health Organization (World Health Organization Malnutrition). An underweight child is defined as weight for age with a z score lower than 2 standard deviations, a wasted child is having weight for height with a z score lower than 2 standard deviations, a stunted child is having height for age with a z score lower than 2 standard deviations, and an overweight child is having weight for height with a z score larger than 2 standard deviations (World Health Organization Malnutrition).
Outcome Variables. The outcome variables are dropout numbers. Dropout counts for different groups were included in the panel data. A “dropout count” represents the number of individuals of a particular group who chose to leave school before earning a high school diploma. Dropout counts for White male dropout, White female dropout, Black male dropout, Black female dropout, Xxxxxxxx xxxx dropout, Hispanic female dropout, Native American male dropout, Native American female dropout, Asian male dropout, Asian female dropout, and total dropout (i.e. the total number of individuals who dropped out of a school district) are considered to be outcome variables and are included in the data used for this study. However, in this paper, I discuss only the results concerning the outcome variables of total dropout, Black male dropout, Black female dropout, White male dropout, and White female dropout.
Outcome Variables. The primary outcome for this study was in-hospital mortality due to all causes during a primary hospitalization for TB. Trends in the proportion of TB hospitalizations with type 1 and type 2 DM resulting in death are compared with proportion of deaths among TB hospitalizations without DM through the Xxxxxxx-Xxxxxxxx test for trend. Trends in the secondary outcomes included length of stay (LOS) (days) and total inpatient charges (US dollars) were also analyzed. Total inpatient charges reflect the amount the hospital is billed for each TB hospitalization and do not include physician fees. Total inpatient charges were adjusted to 2011 US dollars using the Bureau of Labor Statistics' Consumer Price Index Inflation Calculator, found here: xxxx://xxx.xxx.xxx/data/inflation_calculator.htm. Weighted descriptive statistics were used to characterize the discharge sample, using proportion or means with standard deviations (SDs) where appropriate. Baseline differences for categorical variables were evaluated by χ2 test. Differences in the overall proportion of deaths, overall average LOS, and overall average inpatient charges between TB hospitalizations with and without DM were determined by t test. Means were assumed normal due to the large cohort size. The Xxxxxxx-Xxxxxxxx test for trend was utilized to test for significance of changes of TB hospitalizations with and without comorbid DM. Weighted values were used to calculate yearly average deaths, yearly average LOS, and yearly inpatient charges among TB hospitalizations with and without DM. Averages were used to account for the differences in size of the TB hospitalizations with and without DM. Then the proportion of average deaths, average LOS, and average charges that occurred among those with coexisting DM were calculated for each year. These yearly proportions were then analyzed for significant changes over the study period using the Xxxxxxx-Xxxxxxxx test for trend. Bivariate analysis was used to examine predictors of death among TB hospitalizations; associations were presented as crude risk ratios (RR), with corresponding 95% confidence intervals (CI). A p-value of less than 0.05 was considered statistically significant. Statistical analyses were performed using SAS 9.3 (SAS Institute Inc., Cary, NC).
Outcome Variables. All 188 study participants responded to the substance use and depressive symptoms questions. Scores on the depressive symptoms scale ranged from 8 to 32 with a mean score of 17.86 (sd=7.68). Scores on the substance use scale ranged from 0 to 46 with a mean score of 5.35 (sd=8.32). Range Mean (sd) Cronbach’s Alpha Attachment 4-16 11.19 (2.48) N/A Depressive Symptoms 8-32 17.87 (7.68) .905 Substance Use 0-70 5.35 (8.32) .863 In order to test for potential mediation, bivariate analyses were run between each potential pathway in the mediation model to determine if a significant association exists. Predictor Variable and Health Outcomes Results of the simple linear regression analysis indicate that history of abuse is a statistically significant predictor of depressive symptoms (B=4.80, 95% CI=2.60; 6.99, p<.001) suggesting that on average, those who were abused have a mean depressive symptoms score that is 4.80 points higher than those who did not experience abuse. The R2 for this model is .091, indicating that about 9.1% of the variance in depressive symptoms can be explained by having a history of abuse.
Outcome Variables. This study’s primary interest was whether adolescents who had received care, defined as at least one appointment in a pediatric facility (Xxxxxx, PCS, or CHOA) during 2008 or 2009, sought follow-up healthcare service through 2010. Those with no evidence of care through 2010 were classified as “lost to follow-up” (primary outcome). The reference group for this outcome included all patients who did receive care sometime in 2010. Adult care was defined as at least one appointment in Emory Healthcare, St. Joseph’s Hospital, Xxxxx Health, or Georgia Medicaid claims data in 2010. Patients who transitioned to adult care (secondary outcome) were classified as “successfully transitioned.” The patients who received care in 2010 but who did not have evidence of transitioning to adult care were classified as “retained in pediatric” (tertiary outcome). Since the “retained in pediatric care” group was used as the referent group for the “successful transition” group during modeling, the crude and multivariable analyses for this outcome are referenced only in Appendix D.
Outcome Variables. Dietary intake. We measured dietary intake using the Dietary Risk Assessment (DRA). The DRA is a 26-item food frequency questionnaire (FFQ) that measures usual intake of foods and beverages associated with cardiovascular disease risk.39,40 A single dietary intake score (ranging from 0-52) is obtained by summing the scores from four subsections: 1) nuts, oils, dressings, and spreads; 2) vegetables, fruit, whole grains, and beans; 3) drinks, desserts, snacks, eating out, and salt; and 4) fish, meat, poultry, dairy, and eggs. A higher score represents a healthier dietary pattern. Although the score can be dichotomized into “desirable” or “not desirable” we treated the score as continuous in all analyses as only five participants fell within the desirable category. The DRA was validated against the Xxxx Xxxxxxxxxx Cancer Research Center FFQ (FHCRC-FFQ) in a sample of low- income, midlife, southern, African American women. Correlations between total DRA score and three FHCRC-FFQ diet quality scores ranged from 0.57 to 0.60.39 Control variables. The following control variables were included in adjusted analyses: age (continuous); education (less than high school; high school diploma or GED; some college or technical school; Associate’s degree or higher); employment status (full-time; part-time; unemployed, seeking employment; unemployed, not seeking employment); marital status (currently married; not married, living with a partner; never married; divorced, widowed, or separated); household size (continuous); annual income (less than $5,000; $5,000-$9,999; $10,000-$19,999; $20,000-$29,999; greater than $30,000); car ownership (yes; no); receipt of Supplemental Nutrition Assistance Program (SNAP) benefits (yes; no); smoking status (current smoker; former smoker; never smoker); and area-level median household income (continuous). In order to control for area-level median household income in all models, we created separate median household income variables for each of the four residential neighborhood and activity space environments. Using census block group-level median household income data, we created a continuous KDE surface representing the spatially varying median household income across the study area. We then extracted the average values for each polygon and set of activity space points as described above.
