Community Medicine
Dr: Nasih Lec: -4-
9-Oct-06
Cross-sectional studies
(Prevalence studies/ Surveys)
Design
As cross-sectional studies collect data on existing (prevalent) cases so they
are also called prevalence studies.
frequency of a particular exposure(s) and/or outcome(s)
In a defined population
At particular point in time
We investigate prevalence of certain outcome OR exposures
( prevalence studies)
Example:
In an survey on smoking, we collect information on the smoking habits of
each individual (outcome) in a population. We can also collect information on some
potential risk factors, such as age, sex, occupation, education of the individual
(exposures).
Why Conduct surveys?
1) Health service planning
A survey describes the prevalence of a particular health related problem
(outcome) in a defined population at a particular point in time. This information
is valuable for understanding burden of disease and prioritization of
interventions
Examples:
A survey can collect data on the utilization of services and this
information can be use for improving service delivery.
Some countries undertake periodic surveys to obtain health-related data
from a random sample of the population (demographic health surveys,
multiple indicator cluster surveys).
1
2) Evaluation of health interventions
Surveys can be conducted to evaluate the effectiveness of health
programmes
Example:
Vaccination coverage surveys are undertaken to asses the effectiveness of
EPI programmes
3) Generate hypothesis about etiology
Analytical cross sectional studies can help generate hypothesis about
possible causes of disease and are provide quick initial information on etiology.
Example:
The observation that coronary heart disease is more prevalent in men
than in women has led to hypotheses about the possible protective effects of
estrogens.
Steps in undertaking a survey
1) Defining the study question
What question about prevalence we want to answer?
What is the prevalence of …anemia among the pregnant women in the
catchment area of this health center?
2) Defining the target population
Define the target population of your study
The target population is the population to which the main results of the
study will be extrapolated.
If the target population is small à study all the members
For large populations à study a sample
2
The sample must be representative of the target population otherwise we
would not be able to extrapolate the results
Example
In a cross-sectional study of smoking habits and risk factors for smoking,
the target population of our study could be all the children and adolescents in the
country within a certain well defined age group.
3) Sample selection
The best way to ensure that a sample is representative of the target
population is by carrying out random sampling.
Random sampling ensures that chance alone determines who will be
included in the sample. There are several different types of random
sampling.
4) Data collection
Before data collection:
have a clear case definition
have a clear definition of the exposure(s) and outcome(s)
Data collection methods depend on the nature of the measurements and
the financial and logistic issues
Data collection methods:
Personal interviews
Questionnaires
Medical records
Physical examinations
Diagnostic tests
5) Data analysis
In descriptive studies we quantify the prevalence of an outcome or outcomes
amongst the study participants. We often report the prevalence in males and
females and in different age groups.
In an analytic study, we quantify the association between the exposure and the
outcome by calculating the prevalence ratio.
The prevalence ratio is the ratio of the prevalence of the outcome
in those exposed to the potential risk factor to the prevalence in
those unexposed.
3
If exposure has different levels, we can calculate prevalence and prevalence
ratio for different levels of exposure
Example:
In a study of anemia in pregnant women, we could calculate a prevalence ratio
to quantify the association between anemia and literacy by comparing
prevalence of anemia among illiterate women and illiterate women.
Overall prevalence:
Total prevalence of anemia= number of women who are anemic/ total number of
women in the study
Prevalence in the exposed:
Prevalence of anemia among the illiterate women= number of illiterate women
who are anemic/ total number of illiterate women in the study
Prevalence in the unexposed:
Prevalence of anemia among the literate women= number of literate women who
are anemic/ total number of literate women in the study
Prevalence ratio:
Prevalence ratio= prevalence of anemia among the illiterate women/ Prevalence
of anemia among the literate women
6) Interpretation of results
During interpretation of results we must remember the following possible
sources of error.
Bias is any error in the design and conduct of a study that makes the
results different from the truth
Recall bias: have participants told us accurate answers/ have they
forgotten something? In a vaccination survey a mother may say
that her child is vaccinated while he is not. If inaccurate
information was provided then our results will be biased.
4
Observer bias: has the data collector made mistakes in obtaining/
recording the data? If he/she has done plenty of mistakes then our
results will be biased.
Random error: was our sample really random? If it was not our results
will be biased.
We collect data on exposure and outcome simultaneously so we must be
careful in interpreting the relationship between and exposure and an
outcome; we don’t know which one occurred first. Exception fixed
exposures like sex.
Strengths and Weaknesses of Cross-sectional Studies
Strengths
Relatively quick, easy and economical to conduct
Provide important information on the distribution and burden of disease
Can provide initial information on exposure-outcome relationship
Weaknesses
Measure prevalent rather than incident cases.
It is difficult to establish the time-sequence of events and hence etiology
