Epidemiology Report: Measurement Error in Case-Control Study

This report critically examines a case-control study investigating the association between seated immobility at work and venous thromboembolism, focusing on potential measurement errors. The analysis identifies selection bias due to the choice of control groups, and information bias stemming from the method of data collection on patient history and family details. Furthermore, the report highlights the presence of confounding factors, particularly the omission of Body Mass Index (BMI) as a key variable, and the inappropriate clustering of co-exposure variables in the analysis. The report concludes that these errors significantly impact the validity and reliability of the study's findings, emphasizing the importance of addressing these biases for accurate epidemiological research. The study by West et al. is critiqued for its methodological flaws, specifically in how it handles selection, information, and confounding biases, ultimately affecting the reliability of its conclusions regarding the risk factors for venous thromboembolism.

Running head: MEASUREMENT ERROR IN EPIDEMIOLOGY
Measurement Error in Epidemiology
Name of the Student
Name of the University
Author note

2MEASUREMENT ERROR IN EPIDEMIOLOGY
Introduction
To examine the potential biases in the chosen study A case–control study of seated
immobility at work as a risk factor for venous thromboembolism by Jill West et al. the
significant attributes of this study must be exposed. In this study several exposure factors are
considered such as family history, past medical history, age, surgery or trauma, prolonged
travel and immobility at work. The purpose of this chosen research was to examine the
impact of immobility at work on the potentiality of venous thromboembolism. In the
following section the validity and the reliability of this chosen research will be discussed
considering the three potential major errors namely selection bias, information bias and
confounding factors.
Selection and Information Bias
In this study, the case group was selected from outpatients, and the control group was
chosen from admitted patients. As per the definition of the selection bias, the selection bias
occurs when a study select a wrong sample which already had greater potentiality to be
exposed or unexposed. In this study, the control participants of group are already admitted
patients who have greater chances of having immobilised period than the people chosen from
outpatient. Therefore, with this selection bias the impact of immobility on venous
thromboembolism might show less Odd Ratio value than the real value.
In this study, to examine the medical risk factors and family history, the participants
are directly asked about their medical and family background instead of collecting viable data
from the medical database. Here, the information bias might have impact on the resultant odd
ratios as well. As per the definition, the information bias occurs when the instrument used for
a research is unable to testify the samples and segregated them into two resultant target

3MEASUREMENT ERROR IN EPIDEMIOLOGY
groups in this case exposed and unexposed group. As discussed earlier collecting the verbal
description of the medical details and family history from the outpatient group, whereas the
verbal description was also collected for the family history data of admitted patient group. In
these situations, participants are either unaware of their proper medical history details or
present a false details about their family history. Hence, the process used in this research is
clearly not an ideal instrument to measure those exposure factors (family history, medical risk
factors). Hence, this chosen research has an ineligible potentiality of having information bias
due to its wrong selection of instrument.
Addressing the issues of confounding
As discussed earlier the method used in the chosen research paper considered several
exposure factors such as family history, past medical history, age, surgery or trauma,
prolonged travel and immobility at work. On the other hand, it should be also noted that
overweight and obesity could be a major cause and trigger of venous thromboembolism.
However, throughout this study this factor has not been considered, which caused a major
confounding error in this research. Similar to other exposure factors this research should have
consider BMI of the both control group and Case group participants. Another major
confounding factor can be clearly identified from the data analysis techniques. In this
research, it was evident that not one factor is directly responsible for venous
thromboembolism. Whereas, this study treated all those co-exposure factors in a clustered
odd ratio calculation process instead of using them as independent exposure variables. As an
example, when this study showed that the odd ratio of prolonged seared immobility at work is
2.2, it was not independent from other exposure group. In this odd ratio the case group had
more values in other exposure variables than the control group that made the result
confounded. Hence, in this study several co-confounder factors are present that reduced the
validity and reliability of the data as well.

4MEASUREMENT ERROR IN EPIDEMIOLOGY
Conclusion
The purpose of the above discussion was to examine validity and the reliability of this
chosen research will be discussed considering the three potential major errors namely
selection bias, information bias and confounding factors. From this analysis it has been found
that there is a clear selection bias due to the selection of control group. At the same time,
because of the instrument used to collect data regarding the patients’ history and family
details the information bias is also present in many ways. Additionally, in this study several
co-confounder factors are present that reduced the validity and reliability of the data, where
the major confounding variable is the BMI of the participants and the other confounding
variables are the co-exposure factors.

5MEASUREMENT ERROR IN EPIDEMIOLOGY
References
West, J., Perrin, K., Aldington, S., Weatherall, M. and Beasley, R., 2008. A case–control
study of seated immobility at work as a risk factor for venous thromboembolism. Journal of
the Royal Society of Medicine, 101(5), pp.237-243.