Cross-Sectional Studies: Overview, Strengths, and Weaknesses

Cross-Sectional Study

Cross-Sectional Study

•

A cross-sectional study is a type of 

observational

 study

design that involves looking at data from a population at

one specific point in time.

•

In a cross-sectional study, investigators measure outcomes

and exposures of the study subjects 

at the same time

.

•

It is described as taking a “

It is described as taking a “

snapshot

snapshot

” of a group of

” of a group of

individuals.

individuals.

C

C

r

r

o

o

s

s

s

s

-

-

s

s

e

e

c

c

t

t

i

i

o

o

n

n

a

a

l

l

s

s

t

t

u

u

d

d

y

y

:

:

O

O

v

v

e

e

r

r

v

v

i

i

e

e

w

w

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

O

v

e

r

v

i

e

w

•

The subjects in a cross-sectional study are chosen from an

The subjects in a cross-sectional study are chosen from an

available population of potential relevance to the study

available population of potential relevance to the study

question.

question.

–

Unlike in case-control studies (subjects selected based

Unlike in case-control studies (subjects selected based

on the outcome status) or cohort studies (subjects

on the outcome status) or cohort studies (subjects

selected based on the exposure status)

selected based on the exposure status)

•

There is 

There is 

no prospective or retrospective follow-up

no prospective or retrospective follow-up

.

.

•

Once the subjects are selected, the investigators will collect

Once the subjects are selected, the investigators will collect

the data and assess the 

the data and assess the 

associations between outcomes

associations between outcomes

and exposures

and exposures

.

.

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

O

v

e

r

v

i

e

w

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

O

v

e

r

v

i

e

w

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

S

t

r

e

n

g

t

h

s

•

Relatively quick and inexpensive to conduct

Relatively quick and inexpensive to conduct

•

No ethical difficulties

No ethical difficulties

•

Data on all variables are only collected at one time point

Data on all variables are only collected at one time point

•

Multiple outcomes and exposures can be studied

Multiple outcomes and exposures can be studied

•

Can estimate prevalence of outcome of interest because

sample is usually taken from the whole population

•

No loss to follow-up

•

Easy for generating hypotheses

Easy for generating hypotheses

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

W

e

a

k

n

e

s

s

e

s

•

Unable to measure the incidence

Unable to measure the incidence

•

Difficult to make a causal inference

Difficult to make a causal inference

•

Associations identified might be difficult to interpret

Associations identified might be difficult to interpret

•

Unable to investigate the temporal relation between

Unable to investigate the temporal relation between

outcomes and risk factors

outcomes and risk factors

•

Not appropriate for studying rare diseases

Not appropriate for studying rare diseases

•

Susceptible to biases such as nonresponse bias, recall

Susceptible to biases such as nonresponse bias, recall

bias, p

bias, p

revalence-incidence bias

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

S

a

m

p

l

i

n

g

m

e

t

h

o

d

s

•

Probability sampling methods

Probability sampling methods

, in which samples are chosen by using a

, in which samples are chosen by using a

method based on the theory of probability (Preferred)

method based on the theory of probability (Preferred)

–

Simple random sampling: Every member of the population has the

Simple random sampling: Every member of the population has the

same probability of being randomly selected into the sample

same probability of being randomly selected into the sample

–

Systematic sampling: One selects every nth (ie, 10th) subject in

Systematic sampling: One selects every nth (ie, 10th) subject in

the population to be in the sample

the population to be in the sample

–

Stratified sampling: The population is divided into non-overlapping

Stratified sampling: The population is divided into non-overlapping

groups, or strata; a random sample of population members is then

groups, or strata; a random sample of population members is then

collected from within each stratum

collected from within each stratum

–

Clustered sampling: The researcher divides the population into

Clustered sampling: The researcher divides the population into

separate groups, called clusters. Then, a simple random sample of

separate groups, called clusters. Then, a simple random sample of

clusters is selected from the population. Note that the clusters are

clusters is selected from the population. Note that the clusters are

used as the sampling unit, rather than individuals

used as the sampling unit, rather than individuals

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

S

a

m

p

l

i

n

g

m

e

t

h

o

d

s

•

Non-Probability sampling methods

Non-Probability sampling methods

, in which samples are chosen by

, in which samples are chosen by

using a method based on subjective judgment

using a method based on subjective judgment

–

Convenience sampling: Participants are selected based on

Convenience sampling: Participants are selected based on

availability and willingness to take part

availability and willingness to take part

–

Quota sampling: A tailored sample that is in proportion to some

Quota sampling: A tailored sample that is in proportion to some

characteristic or trait of a population

characteristic or trait of a population

–

Purposive sampling: Also known as judgmental or subjective

Purposive sampling: Also known as judgmental or subjective

sampling. It relies on the judgment of the researcher when

sampling. It relies on the judgment of the researcher when

choosing members of the population to participate in a study

choosing members of the population to participate in a study

–

Snowball sampling: Existing study subjects recruit future subjects

Snowball sampling: Existing study subjects recruit future subjects

from among their acquaintances

from among their acquaintances

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

S

t

a

t

i

s

t

i

c

a

l

c

o

n

s

i

d

e

r

a

t

i

o

n

s

•

Confounding

Confounding

–

For a variable to be a confounder, it should meet three

For a variable to be a confounder, it should meet three

conditions. The variable must:

conditions. The variable must:

•

(1) be associated with the 

(1) be associated with the 

exposure

exposure

 being

 being

investigated

investigated

•

(2) be associated with the 

(2) be associated with the 

outcome

outcome

 being investigated

 being investigated

•

(3) 

(3) 

not be in the causal pathway 

not be in the causal pathway 

between exposure

between exposure

and outcome

and outcome

–

Confounding could result in a distortion of the

Confounding could result in a distortion of the

association between exposure and outcome.

association between exposure and outcome.

•

Controlling for confounding

Controlling for confounding

–

Restriction

Restriction

: Investigators limit participation in the study to

: Investigators limit participation in the study to

individuals who are similar with respect to the confounders.

individuals who are similar with respect to the confounders.

–

Stratification

Stratification

: Refers to the study of the association between

: Refers to the study of the association between

exposure and outcome within different strata of the confounding

exposure and outcome within different strata of the confounding

variables.

variables.

–

Propensity score matching

Propensity score matching

: Forming matched sets of two groups of

: Forming matched sets of two groups of

subjects who share a similar value of the propensity score.

subjects who share a similar value of the propensity score.

–

Multivariable regression analysis

Multivariable regression analysis

: Based on the regression

: Based on the regression

equation, the effect of the variable of interest can be examined with

equation, the effect of the variable of interest can be examined with

confounding variables that are held constant statistically.

confounding variables that are held constant statistically.

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

S

t

a

t

i

s

t

i

c

a

l

c

o

n

s

i

d

e

r

a

t

i

o

n

s

C

r

o

s

s

-

s

e

c

t

i

o

n

a

l

s

t

u

d

y

:

R

e

p

o

r

t

i

n

g

c

o

n

s

i

d

e

r

a

t

i

o

n

s

•

Strengthening the Reporting of Observational Studies in

Epidemiology (STROBE) statement

•

Transparent Reporting of a Multivariable Prediction Model

for Individual Prognosis or Diagnosis (TRIPOD) statement

S

u

m

m

a

r

y

•

A cross-sectional study is a type of observational study

design that involves looking at data from a population at

one specific point in time.

•

The purpose is to describe a sample within the population

The purpose is to describe a sample within the population

with respect to an outcome and a set of risk factors.

with respect to an outcome and a set of risk factors.

•

Confounding could result in a distortion of the association

Confounding could result in a distortion of the association

between exposure and outcome.

between exposure and outcome.

S

u

g

g

e

s

t

e

d

R

e

a

d

i

n

g

•

Wang X, Cheng Z. Cross-sectional studies: strengths,

weaknesses, and recommendations. Chest. 2020 Jul

1;158(1):S65-71.

•

Sedgwick P. Cross sectional studies: advantages and

disadvantages. Bmj. 2014 Mar 26;348.

•

Sedgwick P. Bias in observational study designs: cross

sectional studies. Bmj. 2015 Mar 6;350.