Cohort Study

Study Design

From your studies, you know that the first step in any research plan is to generate a solid hypothesis to guide the investigation.So, while the WEPI1 newscast and Ms. Doll provided you with interesting information, you decide to visit a local hospital to inquire about individuals with Susser Syndrome.

Upon initial review of the cases, it does appear that a number of affected individuals did, in fact, work at the Glop Industries manufacturing plant. However, other individuals not associated with the factory have been affected as well (albeit in smaller numbers). You learn that Glop industries keeps meticulous information about its employees' work histories, which you decide to use in your study. With this exposure information from the employee records, you want to conduct a cohort study. Since both the exposure and outcome have already occurred, and since you have access to the exposure data collected prior to the disease outcome, you decide to design a retrospective cohort study (Please see Aschengrau & Seage pp. 147, and 206-208 for more information).

1. Based on the facts as presented, which do you think is the best hypothesis to investigate in this retrospective cohort study?

  1. Those who develop Susser Syndrome are more likely to have participated in the manufacturing of SUPERCLEAN than those who did not develop Susser Syndrome.
  2. Those who are exposed to chemicals involved in the production of SUPERCLEAN (via direct exposure at the factory) have a higher risk of developing Susser Syndrome than those who are not exposed.
  3. Residents of Epiville have a higher risk of developing Susser Syndrome compared with the residents of a neighboring community.
Answer (a) — incorrect: The proposed hypothesis implies the comparison of exposure states between diseased and non-diseased individuals. This comparison is appropriate for a case-control study. A cohort study is designed to compare outcomes between exposed and non-exposed groups. In doing so, one is able to estimate the risk of disease development.
Answer (b) — correct: This hypothesis is correct because it specifies 1) the exposure of interest, 2) those who are considered exposed and unexposed, 3) the outcome of interest, and 4) a hypothesized direction for the outcome (i.e., exposed at higher risk).
Answer (c) — incorrect: This hypothesis is too general. Findings of a study based on this hypothesis would do little to elucidate the cause of the current Susser Syndrome outbreak.

2. Based on your hypothesis, what would be the best way to define exposure?

  1. Provide all workers at the Glop Industries manufacturing plant with individual air quality instruments to take daily readings in order to compile weekly doses of exposure to SUPERCLEAN.
  2. Ask workers about their professional activities at the factory and estimate their exposure to SUPERCLEAN.
  3. Look for sources of information at the factory which record individual worker exposures throughout their employment.
Answer (a) — incorrect: While this is a wonderful attempt for precise exposure measurement, it comes a little bit too late in the game. We are interested in exposures that have occurred in the past to see their effect on the development of Susser Syndrome.
Answer (b) — incorrect: Since we have decided to conduct a retrospective cohort study, we need to have information about work-related SUPERCLEAN exposure recorded before the start of the follow-up period. Our study began 2 years ago in September and the follow-up period lasted 2 years. Asking about exposure at the end of the study, when some of those exposed are already ill, may distort your results. For instance, people who developed Susser Syndrome may be more inclined to over-report exposures (i.e., your risk ratio would be inflated).
Answer (c) — correct: In retrospective studies, all relevant events (both exposures and outcomes of interest) have already occurred by the time the study is initiated. Because of this, retrospective studies depend on the routine availability of exposure data from pre-existing records. It is reasonable to assume that the factory will keep track of exposures to chemicals like SUPERCLEAN and would be interested in sharing this information with investigators.

Your supervisor assembles a team to begin the investigation. After a little groundwork, you find that the employee health clinic at Glop Industries keeps records of annual medical examinations for all employees beginning with their hiring date. You also learn that the factory's human resources department has records of each worker's employment history which you can use to determine exposure to chemicals involved in the SUPERCLEAN production. Among the 40 job positions at the factory, only 5 are directly involved with the production of SUPERCLEAN.

After talking with some environmental experts and epidemiologists, you believe that an individual needs to have been exposed to SUPERCLEAN for at least 6 months before a sufficient dose of the chemical accumulates and physiological changes start taking place. Thus, exposure to SUPERCLEAN production chemicals for less than 6 months will not lead to Susser Syndrome.

You are presented with the job descriptions that are exposed to SUPERCLEAN and Glop Industries' air monitoring records.

Job Category Maximum allowable level of exposure to SUPERCLEAN Number of workers b
A 120-150 ppm a 800
B 150-175 ppm 200
C 175-200 ppm 500
D 200-225 ppm 150
E ?225 ppm 250
a. ppm = parts per million
b. Exposed to SUPERCLEAN for at least 6 months, started working at Glop Industries on or before September 2002

After some deliberation, you define the exposed groups as low, medium or high exposure (depending on the maximum allowable level of exposure to SUPERCLEAN for their job category) and the unexposed group as employees either not involved with SUPERCLEAN production or those working less than 6 months in SUPERCLEAN production.

You now have the basic framework of your retrospective cohort study. You have redefined your hypothesis to incorporate your assumptions about the induction period and you have clearly defined your exposure variable. You are obviously excited to get out there and begin collecting data but you must first determine who is eligible for the study.

3. How would you define eligibility criteria for study participants? [Aschengrau & Seage pp. 203-205]

  1. Everyone working at the factory is eligible
  2. Only those who have worked at the factory as of September 1, two years ago, AND had been on the job for at least 6 months AND who were shown to be healthy at their initial or annual health check-ups as indicated by employee medical records
  3. Exclude workers who in the last three months exhibited symptoms of the disease
Answer (a) — incorrect: These eligibility criteria would include workers who have been on the job for less than 6 months. Remember, we decided to enroll only those who already worked at the factory as of September 1, two years ago.
Answer (b) — correct: Cohort studies measure incident, or new cases of disease. As such, it is important to begin with a disease-free population to determine who develops Susser Syndrome during the course of follow-up.
Answer (c) — incorrect: We need to exclude people who were sick at the start of follow-up, not those who became sick during the follow-up period.

4. On what would you base your definition of Susser Syndrome?[Aschengrau & Seage pp. 217-219]

  1. Neurological symptoms alone
  2. Self-diagnosis of the participants
  3. Combination of neurological symptoms and laboratory tests
Answer (a) — incorrect: Susser Syndrome presents with a broad spectrum of symptoms (See the Epiville Department of Health Website). Because of its symptomatic commonality with other disorders, you first need to differentiate Susser Syndrome from other neurological problems.
Answer (b) — incorrect: Susser Syndrome requires clinical evaluation for diagnosis. Self-diagnosis is suspect and can lead to either an over- or underestimate of associated risks due to misclassification of disease status.
Answer (c) — correct: This is the most accurate way to diagnose Susser Syndrome.