The use of statistics in epidemiology helps to determine differences in risk related to exposures and permits evaluation of associations with respect to validity to strengthen study conclusions. Imagine you are an epidemiologist investigating the development of lung cancer associated with occupational exposures to pesticides. Employment records identify 45,260 individuals who have worked in a facility where the pesticide is present. Of the 45,260 individuals, you are able to locate 25,862 individuals and 9,584 of whom have lung cancer.

Your investigative team decides to conduct an occupational surveillance program at a facility to identify people who have excessive pesticide exposures. You screen 5,783 workers using a new diagnostic test, and 3,871 workers with excessive pesticide exposure test positive. You observed a negative test finding in 201 workers who had excessive exposure to pesticides, and 545 workers without excessive pesticide exposure tested positive.

Identify potential sources of bias in this study.

Describe how researchers could address these potential sources of bias in future studies.

Based on the results of the screening test, construct a 2 x 2 table and calculate the following measures:

	Present	Absent	Totals
Positive
Negative
Totals

Sensitivity

Specificity

Prevalence of excessive pesticide exposure

Positive predictive value

Negative predictive value

Accuracy

Based on the results of the screening program, what conclusions would you draw regarding the new diagnostic test.