EPB PHC 6000 EPIDEMIOLOGY FALL, 1997

1
Case-Crossover Studies
2
Analytic Study Designs
Case-crossover study Only subjects (cases) who
have experienced the disease of interest are
selected. Investigator postulates a critical
exposure period (empirical induction period)
for the exposure of interest. Presence of the
exposure is compared between the critical
exposure period and other periods of exposure
(e.g. conditional odds ratio). Good for
studying effects of transient exposures.
3
Analytic Study Designs
Case-crossover study
(A)
(B)
Hypothesized irrelevant (non-causal) exposure
Hypothesized relevant (causal) exposure
Outcome event
Normal background risk
Empirical induction period
Compare presence of exposure between hypothesized
non-causal (A) and hypothesized causal (B)
periods of exposure
4
Analytic Study Designs
Case-crossover study Biased selection of
controls (selection bias) is eliminated
(assuming matched analysis is conducted) since
controls represent the population that produced
the cases. Since each subject serves as their
own control, thus, fixed confounders are
eliminated. However, confounding can occur from
factors that vary over time and are associated
with the exposure and disease of interest (e.g.
smoking).
5
Analytic Study Designs
Case-crossover study If exposure ascertainment
differs between the case and control intervals,
information bias may be present. Overestimation
or underestimation of the empirical induction
period results in non- differential
misclassification hence bias towards the null.
6
Case control vs. case-crossover
From the topics listed below, select whether the
case-control or case-crossover study design is
more appropriate.
Exposure Outcome Case- control Case-Crossover
Body mass index Endometrial cancer
Cell phone calls Car accident
Family quarrel Stroke
Playing tennis Osteoarthritis
Cocaine use Myocardial infarction
7
Case control vs. case-crossover
From the topics listed below, select whether the
case-control or case-crossover study design is
more appropriate.
Exposure Outcome Case- control Case-Crossover
Body mass index Endometrial cancer ?
Cell phone calls Car accident ?
Family quarrel Stroke ?
Playing tennis Osteoarthritis ?
Cocaine use Myocardial infarction ?
8
Review of Recommended ReadingTraffic law
enforcement and risk of crashes
--- Case-crossover study designed to assess
effect of prior driving conviction on subsequent
risk of fatal vehicle crashes. --- Identified all
drivers involved in fatal crashes between 1988
and 1999 in Ontario, Canada. --- Matched date of
fatal crash to previous driving history records
(convictions) in Ontario in 1-month period before
(case period) and 13-months before (control
period) crash. --- Most common prior driving
conviction was speeding, with crashes and
convictions more common in the summer. ---
Explored how long a potential association with
prior driving convictions may have persisted.
9
Discussion Question 1
Interpret the results in figure 1 and table 2,
including use of alternate control periods and
among various subgroups. Do these results
support the primary study hypothesis?
Source Lancet 2003 3612177-2182.
10
Discussion Question 2
Interpret the results in figure 2. Do the
results suggest that the effect of traffic law
enforcement on risk of crashes persists over
time?
Source Lancet 2003 3612177-2182.
11
Discussion Question 3
Which of the following potential sources
of confounding were essentially eliminated
by the use of the case-crossover
design? ?Age ? prior driving
convictions ?Gender ? years as
licensed driver ?Alcohol use while ? cell
phone use while driving
Source Lancet 2003 3612177-2182.
12
Cross-Sectional Studies
13
Cross-Sectional Study
Both a descriptive and analytic study
design. Snapshot of the health status of
populations at a certain point in time. For
each subject, exposure and disease outcome are
assessed simultaneously (hence also called a
prevalence study/survey). Compare prevalence
of disease in persons with and without the
exposure of interest (e.g. prevalence ratio
same formula as risk ratio).
14
Cross-Sectional Study
Advantages Quick, easy, and cheap. Can
study multiple exposures and disease outcomes
simultaneously. Good for describing the
magnitude and distribution of health problems.
15
Cross-Sectional Study
Disadvantages Prevalent rather than incident
cases of disease are identified exposures may
be associated with survival rather than risk of
development of disease. Chicken or egg
dilemma do not know whether the exposure
preceded disease, or was a consequence of disease
development.
16
Cross-Sectional Study

• Example
• Hypothesis
• Obesity is a risk factor
  for knee osteoarthritis
• Sample 100 retirees living at University
  Village

17
Cross-Sectional Study

• Medical exam X-rays to diagnose osteoarthritis
  of the knee

Osteoarthritis
-

50

40
10
Obesity
-
50
20
30
18
Cross-Sectional Study

• Prevalence of osteoarthritis among obese
  subjects 40/50 0.8
• Prevalence of osteoarthritis among non-obese
  subjects 20/50 0.4
• Prevalence ratio 0.8/0.4 2.0

19
Cross-Sectional Study

• Obese subjects are two times more likely
  to have osteoarthritis of the knee than
  non-obese subjects.

20
Cross-Sectional Study
Chicken or egg dilemma

• What came first?
• Obesity or Osteoarthritis