Assignment
Scenario #1
Per the CDC, approximately 20,000 women in the United States are diagnosed with ovarian cancer annually, making it the eighth most common cancer and the fifth leading cause of cancer death, after lung and bronchus, breast, colorectal, and pancreatic cancers.
"The Cancer and Steroid Hormone (CASH) Study was a population-based, case-control study conducted by eight SEER registries during the 1980s to investigate the relationship between oral contraceptive (OC) use and breast, endometrial, and ovarian cancers among U.S. women.
As described in the paper, women ages 20 through 54 who were diagnosed with histologically confirmed ovarian cancer and met other eligibility criteria were compared with controls identified by random digit dialing. Table 1 below shows the exposure distribution by case and control status:
Table 1. Ever-use of oral contraceptive (OC) among ovarian cancer cases and controls, CASH Study, 1980-1981
|
OC USE
|
Cases
|
Controls
|
Total
|
|
Ever
|
90
|
921
|
1011
|
|
Never
|
86
|
683
|
769
|
|
Total
|
176
|
1604
|
1780
|
1. The investigators excluded 226 women potential control women who were found to be without ovaries (due to oophrorectomy sugery) at the time of the interview. Describe one reason why it was appropriate to exclude these women?
2. The investigators used population based controls in this study. What is one advantage and one disadvantage to using population based controls?
3. Calculate the association between OC use and ovarian cancer using the appropriate measure of association. Interpret the measure of association.
Scenario #2
To explore risk factors for S. Enteritidis infection in Trinidad and Tobago, a matched case-control study was undertaken from March 1998 - May 1999. Cases were patients with laboratory-confirmed S. Enteritidis infection who were reported through the communicable disease surveillance system. Cases were enrolled prospectively, shortly after diagnosis. Controls were persons with no diarrheal illness in the previous 4 weeks who lived in the same neighborhood as cases and were similar in age. Investigators attempted to enroll two controls for each case.
Using a standardized questionnaire, investigators collected information from cases about foods and beverages consumed, recent travel, and food handling practices in the 3 days before they became ill. Controls were asked about these exposures during the same 3-day period as the matched case. The questionnaire was administered to both cases and controls by one of the investigators in face-to-face interviews.
Forty-five patients and 92 controls were enrolled in the case-control study. The investigators analyzed the results of the case-control study.
4. Name one advantage and one disadvantage to matching on age and neighborhood in this study.
5. Why would a cohort study not be a good idea for this type of investigation? What is one disadvantage of a case-control study?
In the Trinidad and Tobago case-control study, cases and controls were similar to each other in terms of age, sex, ethnic distribution, and place of residence. Exposure to potential sources of Salmonella, however, differed between cases and controls (Table 2).
Table 2. Potential sources of exposure to Salmonella, Trinidad and Tobago
Case-Control Study, March 1998 - May 1999.
|
Exposure*
|
Matched Odds Ratio
|
|
Ate chicken
|
0.4
|
|
Ate shell eggs
|
8.8
|
|
Ate dishes that contained raw or undercooked eggs
|
18.9
|
|
Ate ground beef
|
1.3
|
|
Ingested powdered milk
|
1.5
|
|
Exposed to live chickens
|
1.3
|
|
Bought refrigerated eggs
|
0.1
|
|
Refrigerated eggs at home
|
0.03
|
*in the 3 days before onset of illness in the associated case
(Note that although calculation of odds ratios in matched case control studies is different than in unmatched case control studies, the interpretation is the same. )
6. What exposures appear NOT to be risk factors for S. Enteritidis infection in Trinidad and Tobago? Choose one of the odds ratio for a food that is NOT a risk factor and interpret.
7. Eating shell eggs showed an odds ratio of 8.8. Interpret the odds ratio of 8.8.
Scenario #3
Between February and May 1984, investigators identified 50 hospitalized children in 11 states with a confirmed diagnosis of Reye's syndrome who had a viral illness before admission. 150 controls were identified as children of the same age range who had experienced a similar viral illness during the same period and were treated at the same hospital or emergency room or were part of the same community as the cases. Preliminary results from the study reported that of the 50 children with Reye's syndrome, 45 had ingested salicylates during their viral illness and 55 of the controls had done so.
8. Construct a 2x2 table and calculate and interpret the correct measure of association for this study.
Scenario #4
Parents at a local school district were concerned about the knee injuries among baseball pitchers on the high school team. The health department contacted the pediatricians, orthopedic surgeons, and radiologists in the town to identify all male adolescents with knee injuries in this particular area. Controls were a random sample of all male adolescents in the same school district. The following data table was compiled.
Table 3. Data from baseball and knee injury study
|
Athletic activity
|
Knee injuries
|
Controls
|
|
Baseball pitchers
|
13
|
7
|
|
Non-pitching baseball players
|
17
|
13
|
|
Other school athletic activity
|
22
|
40
|
|
No athletic activity
|
23
|
70
|
9. The researchers first concluded that baseball pitchers were not at an increased risk of knee injury because they had fewer knee injuries than other groups. What is the problem with this conclusion?
10. Calculate the appropriate measure of association comparing each group to no athletic activity. (Hint: You should end up with 3 separate measures of association. You can make 3 separate 2x2 tables each activity (pitchers, non-pitchers, and other athletes compared to no athletic activity.
11. Interpret the results and draw an appropriate conclusion about exposure to baseball pitching and knee injuries.