Case control study
A case control study samples a population where a known condition exists, and then attempts to identify the risk factor(s) that may have led to the development of the condition (Rothman 2002). Therefore the condition is known from the onset, and a potential list of causative agents is theorized.
Using interviews, retrospective chart reviews, and other techniques, risk factors and non-risk variables are identified (Wikipedia 2007). The cases are then matched to controls using the non-risk variables (such as demographics) and statistical analysis is performed to see if in fact there is a true difference between the case and control with respect to each risk factor.
With the recent deployment of the healthcare information system at County Hospital, clinical users have become divided into two camps: adopters and non-adopters. HIDA proposes the use of a case control study to identify the “risk factors” that led to the “condition” of non-adoption.
For our example, the condition is non-adoption of the health information system. The potential list of causative agents may include clinician type, years in practice, computer literacy (how do we measure this?), specialty area, etc.
Case control studies are used when there is a known disease or condition, and the researcher attempts to identify the risk factor(s) that led this condition. Since we are making the assumption that a risk factor causes disease after some exposure time, and we have measure the condition before the risk factors, [most] case control studies are considered retrospective analysis.
Case control studies can be traced back to the 19th century but really emerged as a distinct methodology in the 20th century (Paneth 2002). Paneth et al. identify the first case control to be fully described in literature was a study in 1926 on breast cancer. Around 1950 the first case control study was published in medical literature that described the link between smoking and lung cancer. These seminal studies laid the framework for the case control study as an essential component in the field of epidemiology (Paneth 2002).
Some advantages include (StatsDirect 2007):
- Cost feasible
- Smaller numbers needed to conduct the study (since the cases/controls are sampled from a larger population)
- Efficient and relatively quick study
- Good for rare conditions
Shortcomings include (Wikipedia 2007, StatsDirect 2007):
- Subject to bias
- Selection bias
- Retrospective study bias
- Do not quantify an absolute risk
- Confounding from other factors or variables, known or unknown
Case Control in Informatics
Although it was difficult to find a pure case control study applied to an informatics project, numerous examples of retrospective studies exist. Some selected examples in informatics include:
- Giles LC, Whitehead CH, Jeffers L, McErlean B, Thompson D, Crotty M. Falls in hospitalized patients: can nursing information systems data predict falls? Comput Inform Nurs. 2006 May-Jun;24(3):167-72.
- Shimoni Z, Gershon A, Kama N, Dusseldorp N, Froom P. Reasons patients present to the emergency department might change during epidemics and be a valuable component of a disease surveillance system. Med Hypotheses. 2006;67(4):709-12. Epub 2006 Jun 5.
- Naumova EN, O'Neil E, MacNeill I. INFERNO: a system for early outbreak detection and signature forecasting. MMWR Morb Mortal Wkly Rep. 2005 Aug 26;54 Suppl:77-83.
- Rothman KJ. Epidemiology: An Introduction. 2002.
- Paneth N, Susser E, Susser M. Origins and early development of the case-control study: Part 1, Early evolution. Soz Praventivmed. 2002;47(5):282-8.
- Paneth N, Susser E, Susser M. Origins and early development of the case-control study: Part 2, The case-control study from Lane-Claypon to 1950. Soz Praventivmed. 2002;47(6):359-65.