Case control or descriptive study? Confusion on study methodology

Suneth B Agampodi MBBS.MSc, Registrar in Community Medicine, Department of Community Medicine, Faculty of Medicine, University of Peradeniya, Sri Lanka. Correspondence -Dr. Suneth B Agampodi, Registrar in Community Medicine, Department of Community Medicine, Faculty of Medicine, University of Peradeniya, Sri Lanka. Email: sunethagampodi@yahoo.com TP: +94777888950 Sri Lanka Journal of Medicine 2007; 16 (2): 53-54 LETTER TO EDITOR Case control or descriptive study? Confusion on study methodology It is encouraging to note that more health professionals are interested in health related research. Not only the university academic staff, but medical students, pre-intern medical officers, post graduate trainees, grade medical officers and consultants outside the universities also appear to be interested in research. However, I have noticed that some of the researchers are confused regarding the differences between case control studies and descriptive studies with a control group. This observation was confirmed twice at two scientific sessions I happen to attended recently. This paper is not to describe case control studies, but to outline the basics of simple case control studies to rectify this confusion. Case control studies involve comparison of patients (cases) with a group of controls consisting of persons apparently free from the disease under study. The comparison is aimed at identifying the etiological factor/factors (exposure) that may differ in the two groups. Case control studies can quantify the alteration of risk associated with the exposure[1]. The well-known case control study done by Doll and Hill on smoking and lung cancer was conducted with the selection of over 700 men and women with lung cancer. A group of people hospitalized for non-malignant conditions ethththryhyhyhyhhyhyhyhyhyhyhyhyhyhyhhyhh (without lung cancer) was selected as controls. Information on smoking was obtained from cases and controls and comparisons were made on odds of smoking among people with and without lung cancer[2]. Disease studied was lung cancer and probable etiological factor identified was cigarette smoking. Let us consider a situation where the investigator studies a sample of Heroin addicts and age/sex matched sample of controls who are not Heroin addicts. If the investigator compares a childhood psychological disorder, which is suspected as an etiological factor for Heroin addiction it is a case control study. If the investigator is comparing neuronal dysfunction between these two groups, where they try to attribute the neuronal dysfunction to Heroin use, the study is not a case control study. It is a cross sectional descriptive study, with a comparison group which is known as “descriptive comparative study”. This misinterpretation may be because of the selection of "cases of Heroin addicts" for the study. However, if the hypothesis is to investigate the causative association of heroin addiction and neuronal dysfunction, cases should be those with neuronal dysfunction. Then the exposure "Heroin addiction" should be compared among those with and without neuronal dysfunctions. Suneth B Agampodi MBBS.MSc, Registrar in Community Medicine, Department of Community Medicine, Faculty of Medicine, University of Peradeniya, Sri Lanka. Correspondence -Dr. Suneth B Agampodi, Registrar in Community Medicine, Department of Community Medicine, Faculty of Medicine, University of Peradeniya, Sri Lanka. Email: sunethagampodi@yahoo.com TP: +94777888950 54 Case control or descriptive study? Confusion on study methodology Unavoidable errors due to chance, bias and confounding factors are common to all epidemiological studies; the prime idea of the researcher should be to minimize these errors as much as possible. However, if the study design is wrong, it is a fundamental error and research findings cannot be accepted. Researchers should always review their methodology with their colleagues; at least one of the reviewers of papers in scientific sessions should be a person who is an expert in research methodology. It will provide an opportunity to review and rectify these methodological errors before commencement of the study thus improving the quality of presentations and publications in these sessions. References 1. Schlesselman JJ: Case-Control studies. Design, Conduct, Analysis. New York: Oxford University Press 1982. 2. Doll R, Hill A: Smoking and carcinoma of the lung: Preliminary report. British Medical Journal 1950, 2:739.