retrospective Case Control Study
2.3 Study Types, Measures of Association, Statistical Tests, & Interpretation Watch 1st!
Directions: Write the information requested in each column. Please refer to the Variables Supplement found in the Resources. Each scenario is 15 points each, total 75 points. You are NOT expected to learn formulas and are provided for your curiosity!
Scenario A: A team of researchers set out to study the risk factors for pancreatic cancer. From hospital records, they assembled a group of 50 patients who were either still living or who had a next of kin identified and willing to answer questions about medical history, lifestyle, and prior experiences of the patient. They assembled another group of 50 patients from the same hospital who were admitted for other causes and did not have cancer of any type. The investigators conducted a review of all available records of these patients, then interviewed them or their next of kin using a standardized format, either in person or by telephone. They focused on ten possible risk factors which could either be answered as yes/no or rated on an ordinal scale. In this study, 23 of the 50 pancreatic cancer patients were recorded as having at least a 20 pack-year history of cigarette smoking, while 8 of the 50 in the other group of patients had that history. | |||
1. This is a retrospective Case Control Study. Why? | |||
2. What is the independent variable in this scenario, and what type is it? What is the dependent variable in this scenario, and what type is it? | Independent =
Type = |
||
3. | Dependent –
Type – |
||
4. What might be a null hypothesis for this study? | |||
5. Odds Ratio is the most appropriate measure of association to evaluate the relationship of each risk factor for the occurrence of pancreatic cancer and measures 4.47. Use the table and link below to show the data. What do the results mean? | The Odds Ratio is calculated as (a/c) / (b/d).
In this scenario, Which group __________ is _____ times more likely to develop pancreatic CA? |
||
6. | What types of variables can be tested by OR? | ||
Odd Ratio Calculator | # of Ills | # of Wells | Total |
# of Exposed | A = | B = | A + B = |
# of Unexposed | C = | D = | C + D = |
Total | A + C = | B + D = | A+B+C+D = 100 |
How to interpret Odds Ratio :
· OR of 1 would suggests that there is no difference between the groups; i.e. there would be no association between the suggested exposure and the outcome (being ill) · OR of > 1 suggests that the odds of exposure are positively associated with the adverse outcome compared to the odds of not being exposed · OR of < 1 suggests that the odds of exposure are negatively associated with the adverse outcomes compared to the odds of not being exposed. Potentially, there could be a protective effect |
|||
7. Chi-square test for Independence w/o Yate’s correction is the most appropriate statistical test to test the significance of this measure of association. Why Chi-square and why w/o Yates? | |||
Chi-square formula | The subscript “c” are the degrees of freedom (sample size minus 1 for the mean). “O” is observed value and E is expected value. | The chi-squared statistic is a single number that tells you how much difference exists between your observed counts and the counts you would expect if there were no relationship at all in the population. | |
8. Fill in the online Chi-square contingency in the calculator at http://graphpad.com/quickcalcs/contingency2/ . Set it up just as you did for OR above. The Chi-square result is 10.519. What does this mean in terms of counts? | |||
9. How to interpret the null hypothesis using Chi-square : P-value is the probability of the relationship between the independent and dependent variables being cause/effect. Since the test statistic is a chi-square, Significance is tested statistically by the resulting p-value, aka probability of the independent variable’s impact on the dependent variable as meaningful. If our expected p-value is < 5% or 0.05 to indicate significance, when compared to the Chi-square p-value, should we accept or reject the null and what do the results mean? | |||
Scenario B: Another team of researchers set out to test a new medication developed to prevent the onset of diabetes in patients with a strong family history of diabetes. They assembled a group of 200 patients and randomized them into two groups, one to receive a placebo and the other to receive the medication. They followed them each year for five years and ascertained the occurrence of diabetes using a glucose tolerance test. | |||
1. What makes this a prospective intervention trial? | |||
2. What is the independent variable in this scenario, and what type is it? What is the dependent variable in this scenario, and what type is it? | Independent =
Type = |
||
3. | Dependent –
Type – |
||
4. Please state a null hypothesis. | |||
5. Relative Risk is the most appropriate measure of association to evaluate the relationship of receipt of this medication with the development of diabetes. Why? | |||
6. In this study, 9 of 100 patients receiving the medication developed diabetes, whereas 10 of 100 patients receiving the placebo developed diabetes. Set up the Relative Risk calculator using the same format as OR above and show your #s in the table below. RR = .9. What do the results mean? | ______% greater risk for developing diabetes is associated with_________________________. | ||
Relative Risk | # ILL | # Well | # Total |
Exposed | A = | B = | A + B = |
Unexposed | C = | D = | C + D = |
Totals | A + C = | B + D = | A+B+C+D = |
7. Chi-square is again the most appropriate test for significance. Why? | |||
Fill in the online Chi-square contingency table at http://graphpad.com/quickcalcs/contingency2/ . Set it up just as you did earlier. | |||
8. Chi-square equals 0.058 with 1 degree of freedom. The 2 tailed P value equals 0.8094 thus the probability of the results is greater than a significance level of 0.05. What do the results mean in terms of the null? | |||
Scenario C: A team of health sciences students decided on a project that would give them experience with their statistical analysis skills as well as some practice in taking vital signs. They assembled a group of 40 undergraduate students, 20 science majors and 20 majors in other fields of study, brought them to a makeshift clinic in the science building and measured their blood pressures with a hand-held BP cuff. | |||
1. This is a cross sectional study. What does this mean? | |||
2. What is the independent variable in this scenario, and what type is it? What is the dependent variable in this scenario, and what type is it? | Independent =
Type = |
||
3. | Dependent –
Type – |
||
4. Please state a null hypothesis. | |||
5. The significance of the difference between the means is the most appropriate measure of association to compare the systolic blood pressures between the two student groups. What is the mean of each group and what is the difference between the means? | |||
6. The Independent t-test is the most appropriate statistical test to test the significance of this measure of association. Why? | |||
7. The science majors’ SBPs were: 100, 110, 116, 120, 122, 128, 128, 128, 130, 132, 132, 134, 140, 144, 146, 146, 150, 152, 154, 158.
The other majors’ SBPs were: 112, 118, 120, 126, 130, 130, 130, 130, 140, 142, 144, 144, 146, 150, 158, 160, 170, 170, 172, 190. Calculate the appropriate measure of association, and test it statistically using the on-line tool found at https://www.graphpad.com/quickcalcs/ttest1.cfm ( Formula ) What are the results? |
|||
8. What do the results mean and how do we state our conclusion regarding the null hypothesis ? | |||
Scenario D: Another team of investigators reviewed medical records to determine which military personnel among several basic training units had received a previous type of meningitis vaccine during training two decades ago. Two study groups were formed, those who had not received the vaccine and those who had. These were followed forward by record review throughout the duration of their active military service, and if they had a record of care through the VA system, these records were searched as well. The outcome of interest was the occurrence of any of a number of autoimmune diseases that were thought to have a possible association with the vaccine. In this study, of a vaccine group of 60, 4 individuals were found to have been diagnosed with lupus at some subsequent point in time. In the non-vaccine group of 90, only 2 individuals had that diagnosis. | |||
1. What makes this a Retrospective Cohort Study and why is it appropriate? | |||
2. What is the independent variable in this scenario, and what type is it? What is the dependent variable in this scenario, and what type is it? | Independent =
Type = |
||
3. | Dependent –
Type – |
||
4. State a null hypothesis. | |||
5. Relative Risk is the most appropriate measure of association to compare the rates of occurrence of autoimmune disease between the two groups? Why? Which group has the greatest risk? | Relative Risk= [A / (A+B)] / [C / (C+D)]
______% greater risk for developing lupus is associated with_________________________. |
||
Relative Risk | # of ILL | # of Well | # Total |
# Exposed | A = 4 | B = 56 | A + B = |
# Unexposed | C = 2 | D = 88 | C + D = |
Totals | A+B+C+D = | ||
6. Chi-square, Fisher’s Exact Test is the most appropriate statistical test to test the significance of this measure of association. Why? | |||
7. Using the same on-line calculator as for problems #1 and #2, perform this test. What are the 2-tailed p-value results? | |||
8. What do the results mean regarding the null hypothesis? | |||
Scenario E: Yet another team of investigators embarked on a long-term study to assess the impact of levels of blood cholesterol in Native Americans on subsequent occurrence of myocardial infarction. They recruited 100 patients from three Indian Health Service hospital systems, and divided them into two groups based on cholesterol LDL/HDL ratio (high ratio is bad) – 42 in the high group and 58 in the low group. By the end of the fifteen-year follow-up period, 16 in the high group and 13 in the low group had experienced a myocardial infarction. | |||
1. What makes this a Prospective Cohort Study? | |||
2. What is the independent variable in this scenario, and what type is it? What is the dependent variable in this scenario, and what type is it? | Independent =
Type = |
||
3. | Dependent –
Type – |
||
4. State a null hypothesis. | |||
5. Relative Risk is the most appropriate measure of association to compare the rates of myocardial infarction in the two groups. [A/(A+B)] / [C/(C+D)] Why and what are the RR results? | |||
6. Chi square w/o Yates correction is the most appropriate statistical test to test the significance of this measure of association. Why and what are the 2-tailed p-results? | |||
7. What do the results mean in terms of the null hypothesis? | |||
Total (25 points/scenario) | /75 |