Designing an Epidemiological Study to Test the Hypothesis

Running Head: BIOSTATISTICS 1

BIOSTATISTICS 3

Designing an Epidemiological Study to Test the Hypothesis

Epidemiology

Student name

Institutional affiliation

Introduction

Case-control studies have been extensively used by epidemiologists to investigate the amount and frequency of exposure in individuals with specific sporadic disease and those without the diseases. In particular, a case-control studies are observational studies that retrospectively looks back over time to compare how frequently the risk factor or exposure is present in the case and control groups to establish the association or relationship between the disease and the risk factor. The rationale of these studies is to retrospectively discover the exposure to risk factors of interest in each of the two groups of subjects, thereby estimating the odds ratios, and the statistical significance of the association observed between exposure and the outcome

Epidemiologists may efficiently employ these studies to investigate risk factors for a disease with a long latency period between manifestation and exposure if the prevalence of exposure in more pronounced among cases than in controls. However, there are problems inherent to case-control studies such as absolute risk, observation, confounding, and selection bias, and therefore, there is need for epidemiologists to cautiously take in account when coming up with the research design and conducting the study. This paper reviews the aspects of design for case-control study; the identification of controls, matching and the time of potential exposure.

Identification of the Controls

Case definition is critical in the population-based case-control studies as it aids avoids myriads of the ascertainment bias. As a matter of fact, case definition clearly defines the sources of cases and controls, and ensures all cases or subjects in the study are based on the same diagnostic or inclusion criteria. Controls are those individuals without the eristic sporadic disease in question who are randomly drawn from the same population, that ideally yields the cases. In the given case scenario, controls should consist of individuals who neither have E. coli O157:H7 infection nor suggestive symptoms. Cases are individuals who are disease positive. Ideally, in selecting controls, epidemiologists aim to choose individuals whom the distribution of exposure would actually be equal to that of the cases, in the absence of disease and exposure association. In other words, the confounders and the exposure to risk factors among controls be representative of that in the source population of becoming cases. In the given case scenario about the outbreak, controls should have the same risk of the disease (E. coli O157:H7 infection), and same potential for exposure of the risk factors as the cases. This only possible if there are from same communities as cases. Also, selection of controls should independent of their exposure status.

Methods of Identifying Controls

There are myriad techniques that the investigators could use in recruiting individuals in the control group in the population-based epidemiological case control study. According to the given case scenario, the investigators used the sequential digit dialing or the random digit dialing to identify the controls. The random digit dialing is a probability sampling technique in which a computer randomly generates telephone numbers for families, persons, and households who are to be surveyed by dialing. The random selection in the random digit dialing may entail choosing of the last four matching digits or the seven digits then sequentially digit dialing. The telephone numbers are the sampling unit which provides the link to the members of households. The merits of this method are that it is cost and time efficient, particularly areas with high-telephone ownership, since it eliminates the essence of travelling to conduct in-person interviews. Also, it generates a pure random sample and includes unlisted numbers that could be missed if the telephone numbers were being chosen from phone book. However, it has a demerits of low response rate because there is may low possibility that someone will be at home to pick the call.

Other methods which the investigators may use encompasses of the neighborhood controls, hospitals or clinic controls, and other telephone sampling. These other telephone sampling are different from rigid digit dialing in that they select numbers from a list of numbers in directories or commercial lists. The neighborhood controls may entail home visits or using reverse directory. The home visits are advantageous in that it is easier to establish a rapport through in-person interview.

Matching on Age and Gender

Matching in case-control studies refers to the selection of subjects in the control group who in certain important characteristics are identical to cases. It is a statistical method for evaluating the effect of treatment by comparing the two groups in an observation study. Markedly, the matching is directly towards a classical background factors, and helps optimize the collected information or data about subjects. Normally, the matching variables are the potential confounders of the association between that specific disease and its risk factors. In the

References

Fullerton, K. E., Scallan, E., Kirk, M. D., Mahon, B. E., Angulo, F. J., De Valk, H., Van Pelt,

W., Gauci, C., Hauri, A. M., Majowicz, S., & O'Brien, for the International Coll, S. J. (2012). Case-control studies of sporadic enteric infections: A review and discussion of studies conducted internationally from 1990 to 2009. Foodborne Pathogens and Disease, 9(4), 281-292. https://doi.org/10.1089/fpd.2011.1065.

Rothman, K. J. (2012). Epidemiology: An introduction. Oxford University Press.