(^114) Johns Hopkins Nursing Evidence-Based Practice: Model and Guidelines, Third Edition
Figure 6.2 Illustration of a case-control study.
Population: U.S. adults
ages 25–40
Develop disease/condition
(cases)
Obese BMI > 30
Do not develop
disease/condition (controls)
Normal BMI
Example: Nonexperimental Descriptive Case-Control Design
Early detection of signs of sepsis can decrease mortality. Back, Jin, Jin, & Lee
(2016) used data mining techniques analyzing electronic health records (EHRs)
to identify predictors of sepsis. The study used a retrospective 1:4 case control
review. Study patients with sepsis were the case group, and patients without sepsis
were the control group. The team used EHR data for a 4-year period at the study
hospital to identify seven predictors of sepsis and develop a sepsis risk scoring
algorithm.Cohort studies look at a particular subset of a population from which different
samples are taken at various points in time. These types of observational studies
can be retrospective, in which both the exposure and the outcome of interest has
already occurred, or prospective, in which the patients who have been exposed to
the condition of interest are observed to determine the occurrence of the outcome
of interest (Lu, 2009). Cohort studies that are prospective may require a long-
term follow-up period until the outcome event has occurred (Gordis, 2009). The
risk factor of smoking in the population is used to illustrate a prospective cohort
study (see Figure 6.3):