Retrospective study which can also be called case control study is an observational analytic epidemiologic study that makes use of information or data from a sample of the study population (i.e. the exposed individuals of the population) rather than looking at the full members of the population. In case control studies, the epidemiologist or researcher gathers information from members of a population at once and then classify the data into different groups (cases or controls).
The cases are the participants with a potential risks factors (i.e. people diagnosed as having the disease) while the controls are members with the same risk factor (i.e. persons without the disease) and, both the cases and controls in a retrospective study are compared with a view to getting concrete information about the disease. The researcher selects individuals of a population on the basis of whether or not they have the disease being investigated, and he/she goes ahead to determine their previous exposure retrospectively (both for cases and controls). After proper investigation and grouping by the researcher, if the exposure is more common among cases than individuals that were grouped as controls, then the exposure is associated with the disease.
Retrospective study is a backward looking observational epidemiological study in the sense that it compares cases and controls of a study with regards to the presence of a suspected etiological factor (s) in their past experience (exposure). In this type of study, the researcher looks backwards from the disease to a possible cause. It is usually the first approach to estimate the cause-effect relationship between a disease and a suspected risk factor. Case-control studies are most suitable in the investigation of rare diseases (e.g. cancer, cardiovascular diseases, and hip fracture).
They are the most frequent analytical epidemiological studies and, such studies starts with individuals who have already developed the condition of interest. A retrospective study selects participants based on their disease status; thus a case group (participants that are disease positive) and a control group (participants that are disease negative) are compared to each other. The association between a disease and the exposure rate in a retrospective study is expressed as the odds ratio.
Odd ratio (OR) is defined as the statistics generated in a case-control study to measure the association between disease and exposure. As earlier stated, it is a measure of the relationship between an exposure and a disease in a retrospective (case-control) study (Table 1).
Table 1. Odd Ratio Table
OR > 1: there is association between disease and exposure
OR < 1: then the exposure is a protective (risk) factor in the disease
OR = 1: no association between disease and exposure
For OR to be a good estimate, the case and control groups must be representative of the general population with respect to disease exposure. Note: The odds of an outcome are simply defined as the number of times the outcome occurs to the number of times it does not. Case-control studies only odd ratio as its own measure of disease frequency and risk (Table 1).
Odd Ratio is mathematically expressed as: OR = (ac/bd)
Where: ac (numerator) = the ratio of the odds of exposure in the cases
bd (denominator) = the odds of exposure in the controls
It is generated by constructing a 2×2 table as shown in Table 1.
In retrospective studies, the control groups should be selected in such a way that it reflects the exact population that generated the disease being investigated. Controls are also generated by the individual matching to putative cases. In a case-control study, controls can be selected based on any of the following criteria:
- Random sampling of the population under study: In random sampling, there is no introduction of systematic errors because from a statistical point of view the control groups are truly the representative of the study population. The study population is normally accessible but on the other hand, there is usually a high non-response rate amongst the members of the population without the disease (i.e. the healthy individuals).
- Non-random sampling of the study population: Non-random sampling is undertaken when the study population is not easily accessible for random sampling (e.g. if the cases are patients with the disease in a particular hospital). It is also used as an approach to reduce the high non-response encountered with healthy individuals of a population from which the control groups will be drawn from.
A systematic error can be introduced in a non-random sampling because when the controls are selected non-randomly it will not be easyto know to what extent the control group exposure reveal the exposure distribution of the disease in the community/population that produced the disease.
Retrospective (case-control) studies are not without some advantages or disadvantages.
MERITS OF RETROSPECTIVE STUDIES
- They are well suitable for the evaluation of diseases that has long latency period.
- They make use of existing data or record about a disease.
- They are inexpensive, quick and very reliable in establishing evidence of an exposure to a risk factor and an unfavourable outcome.
- They are helpful for studying rare diseases.
- Because it is backward, a large number of populations for study is not usually required.
- They require few cases and control groups to study a particular disease outbreak.
- They can study multiple causes (or aetiologic factors) of a disease.
DEMERITS OF RETROSPECTIVE STUDIES
- There is usually difficulty in the selection of a control group.
- The results of a retrospective study are usually difficult to interpret.
- There is a greater chance of bias in result since the study is retrospective.
- It barely establishes the cause of disease due to flaws in collected data from past or previous exposure to risk.
- Case-control studies do not allow an investigator to study the mechanism of action of diseases.
Aschengrau A and Seage G.R (2013). Essentials of Epidemiology in Public Health. Third edition. Jones and Bartleh Learning,
Aschengrau, A., & G. R. Seage III. (2009). Essentials of Epidemiology in Public Health. Boston: Jones and Bartlett Publishers.
Bonita R., Beaglehole R., Kjellström T (2006). Basic epidemiology. 2nd edition. World Health Organization. Pp. 1-226.
Gordis L (2013). Epidemiology. Fifth edition. Saunders Publishers, USA.
Guillemin J (2006). Scientists and the history of biological weapons. European Molecular Biology Organization (EMBO) Reports, Vol 7, Special Issue: S45-S49.
Halliday JE, Meredith AL, Knobel DL, Shaw DJ, Bronsvoort BMC, Cleaveland S (2007). A framework for evaluating animals as sentinels for infectious disease surveillance. J R Soc Interface, 4:973–984.
Lucas A.O and Gilles H.M (2003). Short Textbook of Public Health Medicine for the tropics. Fourth edition. Hodder Arnold Publication, UK.
MacMahon B., Trichopoulos D (1996). Epidemiology Principles and Methods. 2nd ed. Boston, MA: Little, Brown and Company. USA.