Convenience-cohort studies

Epidemiologists also frequently use the convenience-cohort study, in which members of an identified group (e.g., nurses) are questioned about their disease incidence as well as their behavior, diet, work. 

Limitations of Case-Control and Convenience-Cohort Studies... 

Most human evidence concerning chemical effects on health comes from case-confrol and convenience-cohort studies rather than drug trials. But such studies can have serious flaws. ... 

Detection bias occurs in convenience-cohort studies when any measure of substance exposure is correlated with differences in medical scrutiny. For example, the positive relationship between reser-pine (a blood pressure medication) and breast cancer might be attributable to the fact that women under treatment for high blood pressure are more likely to have breast exams, which detect otherwise silent breast cancers (Feinstein 1988, 1261). The same might be true of the relationship between alcohol intake and breast cancer, because alcohol could be a surrogate for income and more frequent breast cancer screening and mammography (Feinstein 1988, 1261). 

Both case-control and convenience-cohort studies will be biased if the recall of diagnosed patients systematically differs from the recall of others. For instance, case group members, who are struggling to determine why they have a disease, probably recall differently than controls, who may be in the hospital for more routine matters. 

Case-control and convenience-cohort studies are susceptible to unobserved differences between experimental- and control-group members. This leads to selection and detection bias, which accounts for the alternative explanations of the observed statistical relationship between exposure and disease incidence. The best individual-level epidemiological studies go to great lengths to avoid bias, use large numbers of people, and study large exposures. 

The research work was developed in the academic year 2013/2014, having been carried out the data collection in December 2013. The study was cross-sectional cohort. The type of sampling was not probabilistic, and the technique of convenience... 

---