Technology - Winona State University



NURS 701 - Statistical Analysis Questions – Assignment 3 For each of the questions below, carry out an appropriate analysis to answer the research questions. The JMP files for some of the questions are located on the Datasets page on the course website. Most of the problems required you to enter the data yourself. When possible, include the relevant JMP output.Improving control of blood glucose levels is an important motivation for the use of insulin pumps by diabetic patients. However, certain side effects have been reported with pump therapy. The table below comes from the paper “Acute complications associated with insulin pump therapy: Report of experience with 161 patients” by Mecklenberg et al. published in JAMA 252 (23), 3265-3269. The table provides data on the occurrence of diabetic ketoacidosis (DKA) in patients before and after pump therapy. Before pump therapyAfter pump therapyNo DKADKARow TotalsNo DKA1287135DKA19726ColumnTotals14714n = 161Conduct an appropriate test to determine if the rate of diabetic ketoacidosis (DKA) is higher following diabetic pump therapy. Summarize your findings. (5 pts.)Schoenbaum et al. looked at risk factors for human immunodeficiency virus (HIV) infection among intravenous drug users enrolled in a methadone program in their paper “Risk factors for human immunodeficiency virus infection in intravenous drug users” published in the New England Journal of Medicine, 321 (13), 874-879. The data table below presented the HIV antibody status among 120 non-Hispanic white subjects by total family income.Family Income Level (ordinal)HIV StatusA - < $10kB – $10k – $20kC - > $20kRow TotalsHIV Positive135220HIV Negative592120100Column Totals722622120Conduct a test to determine if total family income is related in a consistent manner to percentage of HIV-positive? Summarize your findings. (5 pts.)In a study to investigate the potential relationship between age at first birth and the development of breast cancer the following results were obtained.Age at First Birth (ordinal)Case-Control status1< 20220 – 24325 – 29430 – 345> 35RowTotalsCase320120610114632203220Control142244322893109240610245ColumnTotal1742563839041555626n = 13,465Is there evidence of an increasing trend in the proportion of breast cancer cases as a age at first birth increases? Conduct appropriate statistical test to answer this question and summarize the results. (5 pt.)A case-control study was carried out to look at the potential risk for myocardial infarctions associated with oral contraceptive use. In addition to case-control status and current OC use, the age of the subject was also recorded. The variable age group described below is a ordinal variable created from the ages of the subjects. The data-file OC-Age-MI.JMP contains these data and the variables in this file and their coding are defined below.Case-Control Status1 = Case (Myocardial Infarction (MI))2 = Control Oral Contraceptive Use?1 = Yes2 = NoAge Group1 = 25 - 29 yrs., 2 = 30 – 34 yrs., 3 = 35 – 39 yrs., 4 = 40 – 44 yrs., 5 = 45 – 49 yrs.Ignoring age group, estimate the risk for MI associated with OC use. Provide both a point estimate and an associated confidence interval for this measure of risk. Discuss. In doing this in JMP it will be best to use the OC coded and Case-Control coded so the risk measure is calculated in the preferred way. The appropriate 2x2 contingency table is shown below so you can easily check your calculation by hand. (5 pts.) What type of confounder would you expect age to be, positive or negative? Explain your reasoning by considering the relationship between age, oral contraceptive use, and myocardial infarctions. (3 pts.)The tables on the following page were obtained by stratifying on age group. Below each calculate the associated OR. Does age appear to be a confounder? What type given these results? Explain. (5 pts.)Age Group: 1 = 25 - 29 yrs., 2 = 30 – 34 yrs., 3 = 35 – 39 yrs., 4 = 40 – 44 yrs., 5 = 45 – 49 yrs.Age Group = 1 Age Group = 2 Age Group = 3 Age Group = 4Age Group = 5Conduct an appropriate test to determine if a significant relationship exists between OC use and MI adjusting for the age group classification of the individual. Summarize your findings. (4 pts.)We can obtain a better estimate of the odds ratio than the crude one from the combined table in part (a) by combining the results from the age group stratified tables. The formula for this combined OR is given byORCMH=i=1kaidinii=1kbiciniwhere k = number of strata, (note: k = 5 in this case) and ai,bi, ci, and di are the usual cells from strata i. Remember ai= cell where risk factor and disease are present. This estimate of the odds ratio is known as Cochran-Mantel-Haenszel OR or ORCMH. Find the ORCMH for this study and compare it to the crude OR from part (a). What does the CMH odds ratio suggest about the risk OC use presents when considering myocardial infections? (5 pts.)Rosenberg et al. (1980) studied the relationship between coffee drinking and myocardial infarction in young women, aged 30-49 years. This retrospective study including 487 cases hospitalized for the occurrence of a myocardial infarction (MI). Nine hundred eighty controls hospitalized for an acute condition (trauma, acute cholecystitis, acute respiratory diseases and appendicitis) were selected. The measured variables in this study are defined below. These data are contained in the file Coffee-MI.JMP.Case-Control Status1 = Case (Myocardial Infarction (MI))2 = Control Cups Coded1 = Yes (5 cups of coffee or more)2 = No (< 5 cups of coffee)Smoker Group1 = Never., 2 = Former, 3 = 1 – 14 cigarettes, 4 = 15 – 24 cigarettes, 5 = 25 – 34 cigarettes, 6 = 35 – 44 cigarettes, 7 = 45+ cigarettesIgnoring smoking group, estimate the risk for MI associated with drinking coffee as defined above. Provide both a point estimate and an associated confidence interval for this measure of risk. Discuss. In doing this in JMP it will be best to use Cups Coded and MI-Control Coded as coded above so the risk measure is calculated in the preferred way. The appropriate 2x2 contingency table is shown below so you can easily check your calculation by hand. (5 pts.) What type of confounder would you smoking status to be, positive or negative? Explain your reasoning by considering the relationship between smoking status, coffee use as defined, and myocardial infarctions. (3 pts.)The tables below were obtained by stratifying on smoking status. Below each calculate the associated OR. Does smoking status appear to be a confounder? What type given these results? Explain. (7 pts.)Never Former 1 – 14 Cigarettes 15 – 24 Cigarettes 25 – 34 Cigarettes 35 – 44 Cigarettes 45+ Cigarettes Conduct an appropriate test to determine if a significant relationship exists between coffee use as defined and MI adjusting for the smoking status classification of the individual. Summarize your findings. (4 pts.)We can obtain a better estimate of the odds ratio than the crude one from the combined table in part (a) by combining the results from the smoking status stratified tables. The formula for this combined OR is given byORCMH=i=1kaidinii=1kbiciniwhere k = number of strata, (note: k = 7 in this case) and ai,bi, ci, and di are the usual cells from strata i. Remember ai= cell where risk factor and disease are present. This estimate of the odds ratio is known as Cochran-Mantel-Haenszel OR or ORCMH. Find the ORCMH for this study and compare it to the crude OR from part (a). What does the CMH odds ratio suggest about the risk coffee use as defined presents when considering myocardial infections? (5 pts.)In a study of pre-natal care an infant data (real data from a Harvard School of Public Health study, although I cannot give the reference) the following data were obtained.Infant SurvivalLess Pre-Natal CareMore Pre-Natal CareRow TotalsDied20626Survival373316689Column Totals393322n = 715Estimate the odds ratio for survival associated with receiving more pre-natal care. Estimate means give a single point estimate along with the associated 95% CI for this OR. Do the results suggest there is a significant infant survival benefit associated with receiving more pre-natal care? Explain. (5 pts.)In this study, two clinics were used – generically labeled A and B. The tables below show the results from the study stratified on clinic.Clinic AInfant SurvivalLess Pre-Natal CareMore Pre-Natal CareRow TotalsDied347Survival176293469Column Totals179297n = 476Clinic BInfant SurvivalLess Pre-Natal CareMore Pre-Natal CareRow TotalsDied17219Survival19723130Column Totals21425n = 239Repeat the analysis from part (a) separately for each Clinic. What do you conclude? What phenomenon is exhibited by these results? (5 pts.)Conduct a proper analysis of these data taking Clinic into account. Is there evidence that pre-natal care is associated with increased survival benefit when we take clinic into account? Explain. (3 pts)Obesity is an important risk factor for many diseases. However, in studying the effects of obesity it is important to be aware of other risk factors that may be potentially related to obesity. One commonly used measure of obesity is body-mass index (BMI) (kg/m2), which is often categorized as follows: normal = BMI < 25, overweight = BMI 25 – 29.9, and obese = BMI > 30. The data in the table below were presented in a study relating education to BMI category. BMI Categoryn% > high school educationNormal7791Overweight12087.5Obese6483Note: You need to find the frequency of those with at least a completed high school education using the sample sizes from each BMI category and the reported percentages.Conduct a test to compare the percentage of obese and normal individuals with at least a high-school education. Summarize the results. (4 pts.)Conduct a test to compare the percentage of individuals with at least a high-school education between the three BMI groups. Summarize the results. (4 pts.)The National Natality Survey reported that 65.9% of births came from pregnancies with Adequate pre-natal care (Kessner Index = 1), 26.3% of births came from an Intermediate level of pre-natal care (Kessner Index = 2), and 7.8% of births came from an Inadequate level of pre-natal care (Kessner Index = 3). Are these percentages followed by the sample of n = 10,000 births in NC Birth (n = 10,000) data set? Conduct an appropriate test to answer this question. If there is evidence of significant discrepancies, quantify them by looking at confidence intervals for the percentages in each Kessner Index category. Summarize all findings. (8 pts.) ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download