A “cohort” is any group of people with a shared characteristic. For example, in a birth cohort, what’s common to all individuals is their birth year.
Ever wonder why some injured workers return to work (RTW) after six months while others do so after a year or more? A cohort study that follows and observes a group of people who have something in common (namely, a workplace injury) could help answer this question.
In a cohort study, the study participants are followed over time—from weeks to years, depending on the time frame. The goal is to understand the relationship between some attribute related to the cohort at the beginning of the study and the eventual outcome.
There are five steps in a cohort study:
- Identify the study subjects; i.e. the cohort population.
- Obtain baseline data on the exposure; measure the exposure at the start. (The exposure may be a particular event, a permanent state or a reversible state.)
- Select a sub-classification of the cohort—the unexposed control cohort—to be the comparison group.
- Follow up; measure the outcomes using records, interviews or examinations. (Note: Outcomes must be defined in advance and should be specific and measurable.)
- Do the data analysis where the outcomes are assessed and compared.
Cohort study in action
Returning to our example, a cohort study could follow a group of injured workers who were off work (and filed musculoskeletal-related claims) and observe when these workers returned to work.
Researchers in such a study could determine what’s affecting the workers’ RTW. At six and 12 months post-injury, the workers could be interviewed about their readiness to RTW. They may be asked if they have returned to work and, if so, if they were able to meet their job demands. They might be asked about their organization’s policies and practices, and if accommodated work had been offered and accepted.
It may come to light that the workers who felt their companies were doing well in terms of policies and practices were more likely to be back at work at six months, for example, than those who didn’t. If this were the case, this cohort study could likely tell us that workplace policies play an important role in RTW. Researchers could use these results to develop a tool to identify readiness for RTW and guidelines surrounding successful RTW.
Strengths of a cohort study include the fact that multiple outcomes can be observed. Weaknesses are that they can be expensive and time-consuming because they can involve large populations and long periods of time.
In terms of levels of evidence for establishing relationships between exposure and outcome, cohort studies are considered second to randomized controlled trials (RCTs) because RCTs limit the possibility for biases by randomly assigning one group of participants to an intervention/treatment and another group to non-intervention/treatment or placebo. Cohort studies are observational—meaning the researcher observes what’s happening or naturally occurring, measures variables of interest and draws conclusions. RCTs, in contrast, are experimental—meaning the researcher manipulates one of the variables (assigns treatments, for example) and determines how this influences the outcome.
If cohort studies are second-best, then why use them? They may be the only way to explore certain questions. For example, it would be unethical to design an RCT deliberately exposing workers to a potentially harmful situation.
To read about the Institute for Work & Health’s Readiness for RTW Cohort, which followed a group of 600-plus injured workers, see At Work, Issue 65, Summer 2011.
To see other WRMB columns, go to: www.iwh.on.ca/what-researchers-mean-by.
Source: At Work, Issue 71, Winter 2013: Institute for Work & Health, Toronto