Medical statistics

Project description
This assignment is to improve your ability to analyze existing articles. This is one of core points of evidence based practice/policy

In order to do this assignment, find one article (published in 2007 or later) that meet for the following three criteria in the area of your interest: 1) quantitative,

2) deductive (researchers used a theory to test their hypotheses), 3) an analysis of primary data (researchers collect their own data). In addition, only for

assignment purposes, please do NOT select a review study or a qualitative study. In describing your opinions, you are welcome to cite other articles. In doing this,

please use APA format. If possible, please do not exceed five pages in total. Double spacing. You can use any materials that might be useful to complete this

assignment. The article should include information for the following questions.

1)Describe their research questions and/or hypotheses. And, describe how did the researchers use a theory to formulate research questions or hypotheses? If they did

not name a theory, name a theory. Sometimes, researchers did not name it, but readers could be able to identify one through reviewing the background section of the

study (3 points).

2)From the research questions or hypotheses, identify study variables such as the dependent and independent variables. Briefly describe how the study variables were

measured. If there are many IVs, you may want to select several core IVs. And, identify types of measurement reliability and validity for the variables (3 points).

3)Describe a sampling method and a data collection (survey/interview) method employed in the study. Briefly describe your thoughts on why the researchers used the type

of sampling method and data collection method for their study. If you do the same research, would you like to use the same method or different one? And why? (3 points)

see next page

4)Describe types of data analysis. Explain it based on the different levels of measurement (nominal, ordinal, interval, ratio). For example, the researchers used chi-

square tests because they have two nominal level variables (e.g., race/ethnicity: white vs. nonwhite, psychiatric diagnosis: yes or no). If you do the same research,

would you like to use the same data analysis or different one? And why? (3 points)

5)Describe how the researchers link their research findings to implications for practice and/or policy. And, describe your thoughts on public health practice and

policy. How useful is evidence in helping your current or potential individuals in needs (3 points)

Medical Statistics
Diegeler et al. [NEJM 2013; 368: 1189-98] undertook a study of off-pump coronary artery bypass graft surgery in older patients. You should read this paper carefully before attempting to answer the following questions. Students are expected to answer all questions in a concise manner and marks will be awarded for clear and succinct answers
1) What type of study design is described here? (2 marks)

2) What type of randomisation has been used? Include details of any stratification if it has been used. (2 marks)

3) Why was the sample size increased after the interim analysis? (3 marks)

4) Using Table 1, what proportion of patients undergoing off-pump CABG had a history of myocardial infarction? (1 mark)

5) The analysis was performed on a modified intention-to-treat basis (ITT) (with all randomly assigned patients who underwent isolated CABG by the assigned surgeon included). Why is ITT preferable in comparison to a per-protocol analysis? (3 marks)

6) Details are given in Figure 1 on the numbers randomised and the numbers excluded after randomisation and at different stages of the trial. Use the information provided to construct a 2×2 table and perform a chi-squared test to evaluate whether there are any differences between allocated groups in the proportions of randomised patients included in the 30-day analysis. (8 marks)

7) Examine Table 1 and answer the following questions
a) How are the logistic euroSCORE and the Koronarchirurgie score summarised? (2 marks)
b) Comment on whether these are appropriate summary statistics to use. (2 marks)
c) What would be an alternative way of summarising these scores (2 marks)

8) The components that make up the primary outcome were measured at 30 days (expressed as an odds ratio using logistic regression) and 12 months (Kaplan-Meier curves and log rank test). What is the primary endpoint and why do you think the authors chose different statistical tests at these time points? (6 marks)

9) Using the results from table 3;
a) What are the odds of repeat revascularization (component of primary outcome) at 30 days for off-pump CABG patients compared to those who received on-pump CABG and how this is interpreted? (3 marks)
b) Was the risk (hazard) of repeat revascularization different at 12 months? Justify your answer (3 marks)

10) Using the Kaplan-Meier curve from figure 2, estimate the proportion of patients in each group that are still event free at 180 days. (3 marks)

Medical Statistics
Diegeler et al. [NEJM 2013; 368: 1189-98] undertook a study of off-pump coronary artery bypass graft surgery in older patients. You should read this paper carefully before attempting to answer the following questions. Students are expected to answer all questions in a concise manner and marks will be awarded for clear and succinct answers
1) What type of study design is described here? (2 marks)

2) What type of randomisation has been used? Include details of any stratification if it has been used. (2 marks)

3) Why was the sample size increased after the interim analysis? (3 marks)

4) Using Table 1, what proportion of patients undergoing off-pump CABG had a history of myocardial infarction? (1 mark)

5) The analysis was performed on a modified intention-to-treat basis (ITT) (with all randomly assigned patients who underwent isolated CABG by the assigned surgeon included). Why is ITT preferable in comparison to a per-protocol analysis? (3 marks)

6) Details are given in Figure 1 on the numbers randomised and the numbers excluded after randomisation and at different stages of the trial. Use the information provided to construct a 2×2 table and perform a chi-squared test to evaluate whether there are any differences between allocated groups in the proportions of randomised patients included in the 30-day analysis. (8 marks)

7) Examine Table 1 and answer the following questions
a) How are the logistic euroSCORE and the Koronarchirurgie score summarised? (2 marks)
b) Comment on whether these are appropriate summary statistics to use. (2 marks)
c) What would be an alternative way of summarising these scores (2 marks)

8) The components that make up the primary outcome were measured at 30 days (expressed as an odds ratio using logistic regression) and 12 months (Kaplan-Meier curves and log rank test). What is the primary endpoint and why do you think the authors chose different statistical tests at these time points? (6 marks)

9) Using the results from table 3;
a) What are the odds of repeat revascularization (component of primary outcome) at 30 days for off-pump CABG patients compared to those who received on-pump CABG and how this is interpreted? (3 marks)
b) Was the risk (hazard) of repeat revascularization different at 12 months? Justify your answer (3 marks)

10) Using the Kaplan-Meier curve from figure 2, estimate the proportion of patients in each group that are still event free at 180 days. (3 marks)