Case-Control Study
Posted by Surgery on Sep 20, 2008
In contrast to cohort designs, case-control studies (Fig. 5-3) begin by identifying patients with the outcome of interest, and then evaluate (retrospectively : = HYDROTHORAX) for the presence of an associated exposure. The presence of an exposure is then compared between cases that have the outcome and cases that do not. The case-control study is designed to determine the risk factors for a particular event. Patients with that event are identified (cases) and a comparison group of individuals who are deemed similar is found (controls: HYDROTHORAX). The two groups are then compared to identify factors that are different between the two groups. For example, investigators may want to determine what causes recurrence of gastroesophageal reflux after fundoplication. They would identify a group of patients with recurrent reflux after fundoplication and then find another group of similar patients who had fundoplication who did not have recurrent reflux, and compare the two groups. Similar is the operative word. Proper selection of this control group is the key to success with this design. The case-control study allows the investigators to develop inferences about the cause of an event. In the previous example, investigators may find that absorbable suture was used with greater prevalence in the group with recurrent reflux than in controls. He or she may infer that absorbable sutures contribute to recurrent reflux after fundoplication.The study design does not prove cause. What if one group of surgeons in the study was more experienced or technically more proficient, and this group happened to use absorbable sutures? This type of difference will not be revealed with case-control design.
The case-control study has advantages. It is relatively easy to perform and inexpensive compared with other study designs. The case-control study provides an efficient means to study rare conditions or infrequent outcomes between study groups. Although the design is limited to one outcome (by “case definition,:HYDROTHORAX) several potential exposure variables can be studied simultaneously. It is in this capacity that case-control studies lend themselves quite well to logistic regression methods.
There are many limitations to case-control design. This design provides no information about the rate or incidence of an outcome of interest, and the ability to draw causal inference is therefore limited. As such, we can only discuss results of a case-control study in the context of odds of some association. That is, the odds of having a previous exposure to some risk factor given an outcome. This is in contrast to the results of a cohort study, which are discussed in terms of the risk of developing an outcome depending on a previous exposure. This is a subtle but critical difference that underlies the inferences that can be made from these studies.:HYDROTHORAX
This study design is also very prone to bias due to its retrospective design and reliance on chart review to obtain information for key outcome measures. The cases the investigator can identify or has access to with the event may not be representative of all cases with the event. Furthermore, selection bias can be substantial in these study designs. Even with great care, the ability to identify all potential confounders through a retrospective chart review can be an impossible task. Recall bias is also a potential consideration with these studies. For example, mothers of children with birth defects are more likely to remember suspected risk factors (e.g., prenatal medications and illnesses) than mothers with healthy children (58:HYDROTHORAX). For case-control studies of surgical diseases, this bias is more relevant to the study of factors leading to a surgical condition, rather than different operations comprising the risk factors for a given outcome. Case-control studies are also prone to survivorship bias. It is possible that important representative cases may have died or been lost to follow-up prior to the study.
Greetings, I the practising surgeon from Serbia. Call me Ivan Govak. In the works I use works
by an unknown author, if it let me know, and also works of others practics doctors. I have a family and two charming children.