introduction to study design: cross-sectional studies

why is this topic important?

The epidemiological triad:

• Example ofTaenia solium lifecycle

• Human – pig – human cycle

• Human – Human cycle

• What hosts are important in Taenia solium lifecycle?

• LIST factors that may be associated with “disease” in these

hosts

• For each factor, note if is related to the Host, Agent, or

Environment

• How do we generate credible evidence to support the

assertations that factor x may be associated with

‘disease’?

classification of study types

observational studies

•Descriptive studies

•Analytical studies

descriptive studies

Descriptive studies

tend not to be

hypothesis driven;

however, a clear

objective must be

stated.

Not designed to

evaluate any

associations

between

exposures and

outcomes of

interest.

Observations

made in a

descriptive study

can form the basis

of hypotheses

which then are

further investigated

in analytic studies

analytical studies

Differ from descriptive studies in

that the comparison of 2 or more

groups is the central foundation

of their design.

Differ from experimental

studies/controlled trials in that the

researcher has no control over

the allocation of the study

subjects to the groups being

compared.

A N A LY T I C A L S T U D I E S : E X A M P L E S

• Cross-sectional

• Measurements on exposure and outcome are performed

at a defined single point in time - meaning exposure and

outcome are not known before the study starts.

• Cohort

• Case-Control

P R O S P E C T I V E O R R E T R O S P E C T I V E S T U D I E S

I N T E R V E N T I O N S T U D I E S :

C O N T R O L L E D T R I A L S / E X P E R I M E N T A L S T U D I Es

S t u d y d e s i g n s : l e v e l s o f e v i d e n c e

From bottom to top, reflects the

quality of research designs

(increasing) and quantity

(decreasing) of each study design

in the body of published literature

Why are controlled

studies positioned above

observational studies?

cross sectional studies

Key considerations

• Background

• Study objective

• Study population

• Inclusion and exclusion criteria

• Sample size

• Exposure and disease

• Analysis

key characteristics of cross sectional study

Key characteristics

• Measurements on exposure and

outcome are performed at a defined

single point.

• Groups determined based on exposure

and outcome.

• Main comparison is prevalence of the

outcome in the exposed and non-

exposed animals/subjects

• Measures of association: Prevalence

risk ratio, Odds ratios

• Without follow-up

cross-sectional study prevalence

Prevalence is the proportion of the population that are cases at a given point in time:

Prevalence = No. test positive / No. test positive + no. test negative

Point prevalence

• Numerator - Number of current cases (new and pre-existing) at a specified

point in time.

• Denominator - Population at the same specified point in time.

Period prevalence

• Numerator - Number of current cases (new and pre-existing) over a specified period.

• Denominator - Average or mid-interval population.

calculate prevalence: example 2 - salmonella

In an EU baseline survey, 454 GB poultry layer holdings were sampled for Salmonella. Fifty-four holdings tested positive for Salmonella spp.

• What was the prevalence of Salmonella spp. on the poultry layer holdings?

The investigators were particularly interested in the presence of Salmonella spp of public health importance (S. Enteritidis and S. Typhimurium). Six holdings were positive for S. Enteritidis and 30 were positive for S. Typhimurium.

• What was the prevalence of each serovar on the poultry layer holdings?

In an EU baseline survey, 454 GB poultry layer holdings were sampled for

Salmonella. Fifty-four holdings tested positive for any Salmonella spp.

• What was the prevalence of Salmonella spp. on poultry layer holdings? 54/454 = 12%

The investigators were particularly interested in the presence of Salmonella of public health importance. Six holdings were positive for S. Enteritidis and 30 were positive for S. Typhimurium.

• What was the prevalence of each serovar?

S. Enteritidis: 6/454 = 1.3%

S. Typhimurium: 30/454 = 6.6%

analysis: measures of association

Compare exposed and unexposed groups

• Prevalence risk ratio - compare prevalence of the outcome or

disease frequency in exposed group and non-exposed group.

• Magnitude of association between a potential risk factor and

disease.

• Measured using odds ratios.

• Provides information about risk factors, but not on absolute risk

interpreting odds ratios

Odds ratio (OR)

• An odds ratio is a relative measure of effect, which allows the

comparison of the exposed group of a study relative to the

comparison or non-exposure group.

• In the example paper, the exposure is treatment with

antimicrobials and outcome is Salmonella shedding

(positive or negative)

• If the outcome (e.g., Salmonella shedding) is the same in both the

exposed (treated with antimicrobials) and non-exposed (not treated)

groups the odds ratio will be 1, which implies there is no difference in

the risk of Salmonella shedding between the two groups.

However:

• If the OR is > 1 the risk of Salmonella shedding is greater in the

exposed group (treated with antimicrobials) compared to the non-

exposed (not treated with antimicrobials).

• If the OR is < 1 the risk of Salmonella shedding is less in the exposed

group (treated with antimicrobials) compared to the non-exposed (not

treated with antimicrobials)

limitations of cross sectional studies

• A count of cases will be of some interest, but unless the size of the population is known, it is of little interest to compare populations.

What factors influence the prevalence of disease?

• Prevalence is a function of both incidence and duration of the disease.

• Difficult to disentangle factors associated with persistence of the outcome (or survival) from factors associated with developing the outcome (i.e. becoming a new case).

Current exposure status may not reflect status at the time disease occurred

– Not an issue with “tombstone” data e.g. breed, sex, genetic characteristic

– May be addressed by measuring past exposure e.g. through questionnaire but beware of recall bias.

Over-represent cases with a long duration of disease

• Consider an exposure which is associated with a mild disease of long duration:

– Prevalence likely to be high

– Measure of association likely to be strong

Under-represent cases with a short duration

• Compared to an exposure associated with rapid mortality:

– Prevalence low

– Measure of association weak or indeterminate

• A disease with a high prevalence is not necessarily more important than a disease with a low

prevalence

summary

Cross-Sectional Studies

• Measurements on exposure and outcome are performed at same

defined single point in time, hence difficult to determine if exposure

preceded disease.

• Main comparison in cross-sectional studies is between the

prevalence of the outcome in the exposed and non-exposed

animals.

• Common measures – prevalence, prevalence risk ratio, and odds

ratios.