CASE STUDY V: RETROSPECTIVE COHORT STUDY 4
Casestudy V: Retrospective Cohort Study
Controlsin Research Study
Controlsin research are units that are used in research in compare theoutcomes of two groups in observational experiments. Controls areused in epidemiological studies to identify factors that lead tocertain medical conditions by comparing two subjects or cases withthe same disease or condition. One subject serves as a control unitwhile the other subject is exposed to certain risks or variables. Inthe medical field, controls are used to facilitate comparisonobservations in assessing the different attributes between two groupsof subjects(Johnston, 2001).For instance, controls may be used to assess or show that a personwho smokes (attribute) is likely to suffer lung cancer (outcome) thana person who does not smoke (controls). In other words, inobservation studies involving two cases, one case is used as acontrol and the other case is used as an attribute. It is evident inthis study that the non-deformed individuals are used as controls toinvestigate the relationship between proximity to toxic environmentand risks to congenital defects (Merrill, 2012). The idea is toassess if the exposure or proximity to a toxic environment leads tocongenital defects. It is also clear that the healthy subjects areused as controls.
Confoundingvariables and how do they affect a research study
Confoundingvariables have been defined as misestimated variables that influencethe outcome of the relationship between dependent and independentvariables either directly or indirectly. In epidemiological studies,these confounding variables introduce bias or influence decisions,which might affect treatment decisions (Johnston,2001). In some cases, these variables are beyond research control andshould be excluded in the main studies. In this study, the maternalage, race, parity, education, gestation period and gender were themain confounding variables confounding variables. The confoundingvariables in the study had a few shortcomings. For instance,maternal age and the gestation period had a higher chance ofinfluencing the study results longer period of gestation andproximity to the toxic environment could have increased more exposurerisks to congenital defects among the infants. Similarly, pregnancycomplications could influence the overall study results. Despite theexclusion of these variables in the study, they still presented biasin the overall epidemiological study thereby interfering withaccurate results tabulation (Johnston,2001).
Johnston,S. C. (2001). Identifying Confounding by Indication through BlindedProspective Review. Am J Epidemiol154(3). pp. 276–284
Merrill,Ray. (2012). Introduction to Epidemiology, 6th edition. Jones &Bartlett Learning Publishers, Burlington:MA.