Case Control Study

Study
Design

Step 4: Study Design

Thanks to all of the legwork, you now have enough information to generate a solid hypothesis or two. You report back to your supervisor and you both agree Quench-it and EnduroBrick are likely causes. You decide to design a case-control study so that you may look at these exposure variables. From all of your class work, you know that you want your hypotheses to be as explicit and detailed as possible.

1.	Based on the above information, which of the following three hypotheses are the best to investigate in this case-control study?
		a.	Individuals diagnosed with Susser Syndrome are more likely to have ingested Endurobrick than individuals without Susser Syndrome; individuals diagnosed with Susser Syndrome are more likely to have consumed Quench-it than individuals without Susser Syndrome.
		b.	Those who ingest Endurobrick are more likely to be diagnosed with Susser Syndrome than those who do not ingest Endurobrick; those who consume Quench-it are more likely to be diagnosed with Susser Syndrome than those who do not consumer Quench-it
		c.	Individuals diagnosed with Susser Syndrome are more likely to have been members of Superfit Fitness Center than individuals without Susser Syndrome.

Now you must decide upon who will be eligible for the study. You must determine the eligibility criteria for both your cases and controls. Remember, we are actually operating under 2 hypotheses here, each with their own unique exposure variable. Eligibility criteria should be as specific as possible in order to minimize the chance of spurious findings. You decide to start with defining the eligibility criteria for the cases.

After talking with medical experts, you believe that the physiological changes associated with Susser Syndrome take at least 6 months to develop. In other words, a period of at least 6 months needs to pass from the moment an individual is possibly exposed until the first symptoms of Susser Syndrome can plausibly occur. This, you recall, is termed the induction time.

[ Click Here if Intellectually Curious! ]

2.	Which of the following do you think is the best eligibility criteria for the cases? [Gordis Ch.9, pg 143;Ch 12, pg 180-183]
		a.	Cases should have been members of Superfit between September 2000 and September 2002 for at least 6 months and consumed either Endurobrick or Quench-it.
		b.	Cases should be correctly diagnosed with Susser Syndrome and be an employee of Glop Industries.
		c.	Cases should be correctly diagnosed with Susser Syndrome and have been members of the Superfit Fitness Center for at least 6 months between September 2000 and September 2002.

Now you must decide who is eligible to be a control. The eligibility of controls is dependent on their "exchangeability" with cases. Controls should have the same opportunity of having been exposed and of developing the disease as cases, i.e., if they switched places with cases, they would have similar experiences. A control should come from the same underlying population as a case.

3.	With this is mind, which of the following do you think is the best eligibility criteria for the controls? [Gordis Ch 9, pg 143-146]
		a.	Controls should be residents of Epiville who have not been diagnosed with Susser Syndrome.
		b.	Controls should be members of Superfit who have been diagnosed with Susser Syndrome but have not consumed either Endurobrick or Quench-it.
		c.	Controls should be members of Superfit for at least 6 months between September 2000 and September 2002 and not be diagnosed with Susser Syndrome.

Now that the eligibility criteria are set, you must determine the specifics of the case-control study design. How many cases should you recruit; how many controls; should you recruit all of your cases first and then your controls or should you recruit them simultaneously? All of these questions make your head hurt and you decide to take a nap.

Upon awakening 15 minutes later (remember, you've got a job to do!), you have a sudden and prolific brainstorm - some might say epiphany - and the answers to these questions are now clear. You go through them one by one:

How many cases and controls?

That obviously depends on your time and resources. However, an equally important consideration is how much power you want the study to have.

[ Click Here if Intellectually Curious! ]

After crunching the numbers, you decide that the study should have the following size to achieve a desired power of 80 percent:

Number of cases: 112
Number of controls: 224
Total number of subjects (N): 336

One thing you want to remember is that the study is voluntary. This means that subjects, even when eligible, do not have to participate. You do a quick literature review and discover that study participation is never 100 percent and depends in large part on the survey method. In- person recruitment is the best, followed by telephone interviews, and then mail surveys. Case participation rates are almost always greater than controls. With this in mind, you realize you will have to contact and screen a significantly larger number of subjects as many will either not meet the eligibility requirements or will decide not to participate.

Should you recruit cases and controls simultaneously or an entire group first?

A number of sampling techniques come to mind, such as recruiting all of the controls at the beginning of the study, recruiting all of the controls following the accrual of all the cases, or recruiting each control at each time a case is ascertained. Each method has its pros and cons. Some are logistically more difficult and expensive but will allow you to better estimate the rate ratio or the risk ratio. Others are easier and more efficient but require important assumptions to estimate particular measures of association. After weighing all of the many factors, you decided to employ a method that epidemiologists term cumulative incidence sampling in which you select all of your controls after you have already selected all of your cases.

Now that you have the size of your study established and the basic method of recruitment, you need to start planning the actual data collection stage.

Learning
Objectives

Introduction
to Study

Study
Design