Question
The association between Reye’s syndrome and aspirin is being investigated. Reye’s syndrome occurs in children due to consumption of aspirin medication. It leads to a condition called hepatic encephalopathy where the liver and brain are affected. If the true odds ratio (OR) for Reye’s syndrome and aspirin is significantly greater than one, what will be the effect of the following scenarios? In each instance, indicate if the odds ratio would be:
Biased away from null (overestimated)
Biased towards null (underestimated)
Unbiased
Note: It is often useful to use a 2 x 2 table as an aid for thinking about how cell sizes change with various bias scenarios. Because the OR = (AD)/(BC), you can figure out whether the OR will be overestimated or underestimated once you’ve figured out which cells are inflated and which are deflated as a result of the bias. We’ll have practice doing this in section.
Reye’s Syndrome
|
Total
| |||
Yes
|
No
| |||
Aspirin Use
|
Yes
|
A
|
B
|
A + B
|
No
|
C
|
D
|
C + D
| |
Total
|
A + C
|
B + D
| ||
For some of cases, aspirin use followed the onset of the earliest manifestation of Reye’s syndrome rather than preceded it.
Parents tended to incorrectly report Tylenol use as aspirin use, non-differentially with respect to case or control status.
The retrospective nature of the case-control study led to some problems with recall. Parents of both cases and controls under-reported aspirin use by 10%.
Parents of children with Reye’s syndrome recalled aspirin use with greater accuracy than the parents of children without the syndrome, who tended to underreport aspirin use.
Due to the increased publicity about possible consequences associated with aspirin, parents of cases were more likely to report Tylenol use as aspirin use.
Cases were selected from physician reports, and physicians were more likely to diagnose Reye’s syndrome if the child had used aspirin.
Source: 250A Problem Set
