Uploaded by eric.serpi

525900884-Lect-2-Research-Methods-and-Study-Design

advertisement
Research Methods
Alemu T.
09/17/21
Research Methods
The methods of a research project is the core of
the study.
The
following are important questions to
consider when beginning to prepare a research
design for a research proposal.
2
Research Methods …
What do I want to measure?
 How can I measure it?
 Where should I measure it?
 What will I do with the answers collected?
 How can I check whether my methods for measuring
are correct before beginning a large study?
 What professional and non-professional staff do I
need to carry out this study?
 What types of logistical support do I need?
 Are there any ethical problems related to the study?
 How can I avoid introducing biases into the study?
 What constraints may affect this study?
3
Components of a research design that should be
addressed in the method section:
 Study
area,
Study
period,
Study
design,
Population, Sample size determination and
sampling procedures, Variables, Data collection,
Data
processing
and
analysis,
Ethical
considerations, Data quality assurance, and
Operational definitions
4
Study area
Location, Physical features (climate,
altitude...), Population size and
composition
Infrastructures -education, health,
communication…
Economy
5
Study period
Time required to conduct the study
6
Study Design
A study design is the process that guides
researchers on how to collect, analyze and
interpret observations.
It
is a logical model that guides the
investigator in the various stages of the
research.
09/17/21
Selection of study design depends
on:

State of knowledge

Type of information you want to obtain

The nature of the problem

The resources available

personal experience

Audience
09/17/21
Selection of study design depends
on …
The
knowledge and creativity of the
researcher
Sample to be taken
Purpose of the research
Access of study subject.
09/17/21
1) the research problem
Research problems
approaches
call
for
specific
E.g. If the problems calls for:
a) the identification of factors that influence
an outcome
b) the utility of an intervention, or
c) knowing the prevalence of diseases
Then a quantitative approach is best
09/17/21
The research problem …
If a concept or phenomenon needs to be
understood because little research has
been done on it,
Qualitative approach is preferred
 Qualitative research is exploratory and is useful
when the researcher does not know the
important variables to examine
09/17/21
Types of study design
Non-intervention (Observational studies):
◦ in which the researcher just observes and
analyses researchable objects or situations but
does not intervene;
Intervention studies
◦ in which the researcher manipulates objects or
situations and measures the outcome of his
manipulations (e.g., by implementing intensive
health education and measuring the improvement
in immunization rates.)
09/17/21
Study designs could be
Exploratory/fact finding
Descriptive
Analytical/ explanatory
09/17/21
Exploratory studies
An exploratory study is a small-scale study of
relatively short duration, which is carried out
when little is known about a situation or a
problem.
It is used to gain initial insight
It
may include description as well as
comparison.
E.g.-assessment of needs of HIV patients on
ART using in-depth interview
09/17/21
Descriptive studies
document the magnitude, distribution and
trends over time
Purpose: To characterize the amount and
distribution of disease and other health
related conditions within a population.
09/17/21
Cont’d
Descriptive studies generally precede
analytic studies designed to investigate
determinants of disease.
Thus, descriptive studies often help to
generate research hypotheses.
09/17/21
Designing descriptive studies
Defining a population of interest
 Recruiting a representative sample

Measuring the health characteristic of
interest
09/17/21
Descriptive Study Designs
• INDIVIDUALS level
 Case reports
 Case series
 Cross sectional surveys
POPULATIONS (GROUPS)
◦ Ecologic (aggregate) studies
09/17/21
Case Reports
Case Report?
a
careful, detailed report by one or more
clinicians of the profile of a single patient.
◦ One of the first steps in outbreak investigation
◦ Often useful for hypothesis generating and
examining new diseases, but conclusions about
etiology cannot be made.
Strengths- useful for hypothesis generation.
09/17/21
Case report …
Limitations:
• No appropriate comparison group
• Cannot be used to test for presence of a valid
statistical association
• Since based on the experience of one person:
 presence of any risk factor maybe purely coincidental
09/17/21
Case Series
Case series ?
 Collection
of individual case reports
occurring within a fairly short period of
time.
09/17/21
Case series …
Strengths:
• Used as an early means to identify the
beginning or presence of an epidemic.
• Can suggest the emergence of a new
disease (i.e. AIDS)
09/17/21
Case series …
Limitations:
• Lack of an appropriate comparison group
• Cannot be used to test for presence of a
valid statistical association
09/17/21
Correlational/ecological studies?
09/17/21
Ecological studies
Does not provide individual data
presents
average exposure level in the
community
Examples, hypertension rates and average
per capita salt consumption in two
communities
09/17/21
Cont’d
Strengths- it is quick, inexpensive and
often uses available data
Limitations -inability to link exposure
with disease outcome at individual level
◦ lack of ability to control for effects of
potential confounding factors
◦ it may mask a non-linear relationship between
exposure and disease.
09/17/21
A cross-sectional study (survey)
A cross-sectional study?
Snapshot
of the health status of
populations at a certain point in time.
Compare prevalence of disease in persons
with and without the exposure of interest
Cross-sectional studies must be done on
representative samples of the population.
09/17/21
Advantage of Cross-sectional
Provides prevalence estimates of exposure
and disease.
Easier to perform than studies that require
follow-up (hence relatively inexpensive).
Can evaluate multiple risk (and protective)
factors and health outcomes at the same point
in time
09/17/21
Advantage …
May identify groups of persons at high or
low risk of disease
• Can be used to generate hypotheses about
associations between predictive factors
and disease outcomes
09/17/21
Limitation of cross-sectional
Temporal sequence between exposure
and disease cannot be established
* i.e. chicken-or-egg dilemma.
Example: In the study of knowledge of
modern contraceptive, did the women
know about it and then start to use it or
did they learn about it because they were
using it.
09/17/21
Explanatory /Analytical studies
Analytical studies?
At least two variables must be specified
in advance, an exposure and a disease or
health state
09/17/21
Types of analytical study designs
Observational studies
Interventional study
09/17/21
Comparative cross-sectional
Example-a survey on malnutrition may wish
to establish:
The percentage of malnourished children in a
certain population;
Socio-economic, physical, political variables
Feeding practices; and
The knowledge, beliefs, opinions
09/17/21
Cont’d
he
will try to determine which socioeconomic,
behavioral
and
other
independent
variables
may
have
contributed to malnutrition.
09/17/21
Case-control study design
Case-control study design?
09/17/21
Case-control…
exposed
cases
Not exposed
Study pop
exposed
Not exposed
controls
09/17/21
Study begins here
Selection of cases
A/ Definition- precise definition of cases(clinical,
laboratory and other criteria
B/ Inclusion and exclusion criteria
-cases should be selected to improve validity(ex.
By excluding cases with coexisting disease)
-cases should be restricted to limited time period
and geographic area, age range etc…
09/17/21
Sources of cases
Hospital (health institution) cases
◦ easy and inexpensive to conduct but prone to
selection bias
Population (community) based cases
◦ avoid selection bias
◦ allows the description of disease in the entire
population
◦ expensive
09/17/21
Selection of controls
Consider
◦ comparability with cases
◦ practicability
◦ economic impact
Controls and cases should came from the
same source population
09/17/21
Cont’d
The
prevalence of exposure among
controls should reflect the prevalence of
exposure in the source population.
The time during which a subject is eligible
to be a control should be the time in which
the individual is also eligible to be a case.
09/17/21
Sources of controls
Hospital (health institution) controls
Advantages
Easily identified and readily available
Less expensive
More likely than healthy individuals to be
aware of antecedent exposure or events.
09/17/21
Cont’d
Disadvantages
They are different from healthy people in
many ways
Studies demonstrated that hospitalized
patients are more likely to have the
experience of exposure
A true association might be masked
09/17/21
Cont’d
General population controls
Advantages
Are generally healthy
Reflect well the population living in the
area
Generalizability is possible
09/17/21
Cont’d
Disadvantages
Are expensive and time consuming
May not remember exposure status
because of less concern
Less motivated to participate
09/17/21
Cont’d
Special controls
Special controls are individuals
related to the case in some way
like friends, household members,
neighbor, siblings etc
09/17/21
Cont’d
Advantages
They are healthy
More likely to be cooperative
Offer a degree of control over some
confounding factors such as ethnicity,
socio-economic status, environmental
factors etc
09/17/21
Cont’d
Disadvantages
If the study is likely to be similar to
that of the cases, an underestimate of
the true effect may result.
09/17/21
Cont’d
Sources of exposure data (cases and
controls):
Study subjects (self-report).
Records (preferably completed before the
occurrence of outcome events).
Interviews with surrogates (spouses,
siblings, etc.).
09/17/21
Measure of association in case controls study
 The measure of association in case control study is
Odds Ratio(OR)
OR
=
odds of disease in exposed
odds of disease in unexposed
09/17/21
Measure of association…con
Odds = the chances of something happening
the chances of it not happening
09/17/21
Measure of association…cont
OR= ad/bc
OR=1=Null hypothesis
OR>1 Exposure is a risk factor
OR<1 Exposure is protective
09/17/21
Advantages of case-control study
Quick and inexpensive
Relatively simple to carryout
Optimal
for the evaluation of rare
diseases
Can examine multiple etiologic factors
for a single disease
Well suited for the evaluation of disease
with long latent period
09/17/21
Limitations of case-control study
Inefficient
for the evaluation of rare
exposure
Cannot directly compute incidence rates
of diseases
Difficult
to
establish
temporal
relationships
Prone to selection and recall bias
09/17/21
Cohort study design
Cohort?
◦ E.g. Birth cohorts, cohort of smokers, occupational
exposures
Cohort study design
subjects are defined according to their
exposure status
09/17/21
Cont’d
09/17/21
Types
Closed vs. Open
◦ Closed cohort: exposure groups are defined at
the start of follow-up and no new members are
added during the follow-up
◦ Open/dynamic cohort: people move in and out
the study
09/17/21
Cont’d
Incidence cohort vs. Prognostic (clinical)
◦ Incidence Cohort Study
 To assess incidence of disease
 To identify risk factors for disease onset
 Incidence greater in exposed than non-exposed?
◦ Prognostic Cohort Study
 Follow diseased cohort to assess factors associated
with outcome (recovery or death)
 Goal is to identify explanatory/prognostic factors of
those factors helped to the dev’t of the disease.
09/17/21
Prospective vs. Retrospective (Concurrent vs.
Non-concurrent)
Prospective cohort- Healthy cohort (free of the
disease) assembled and followed. More reliable
than retrospective
Retrospective (historical) cohort- the study is
initiated at a point in time after both the exposure
and disease have already occurred.
 It is constructed retrospectively through existing
records
09/17/21
Cont’d
Can be conducted more quickly and
cheaply
◦ All relevant events have occurred
◦ Efficient for disease with long latency periods
Depend on availability of routine data
◦ Incompleteness
◦ Lack of data on confounding variables
09/17/21
Selection of exposed group
The
frequency of the exposure in the study
population
The need for obtaining complete and accurate
information
The easiness to collect relevant information and
to follow up
The ability of obtaining sufficient exposed
individuals in a reasonable period of time
09/17/21
Selection of non- exposed
◦Similar to exposed
◦Control for confounding
factors
09/17/21
Exposure ascertainment:
• Using pre-existing records
• Conducting
interviews
questionnaire
09/17/21
and
filling
Outcome ascertainment
Possible sources are routine surveillance,
death
certificate,
periodic
health
examination, hospital records and autopsy
records.
have affirmed outcome criteria and
standard diagnostic procedures applicable
for exposed and non- exposed groups.
09/17/21
Measures of association in Cohort Studies
Diseased
Exposed
Not
Exposed
Not
Incidence Rates of
Disease Total
Disease
a
b
a+b
a/a + b
c
d
c+d
c/c + d
09/17/21
Cont’d
Relative Risk (RR) = Iexp / Inon-exp
= [a/(a+b)] / [c/(c+d)]
09/17/21
Advantages of cohort study:
Valuable for rare exposures
Examines
multiple effects of a single
exposure
Can elucidate temporal relations
Allows direct measurement of risk
Minimize bias in ascertainment
exposure
09/17/21
of
Limitations of cohort study
Inefficient in evaluation of rare diseases
Expensive
Time consuming
Incompleteness of records
Losses to follow up create problem- leads
to bias
09/17/21
Interventional studies
The researcher manipulates a situation
and measures the effects of this
manipulation.
Usually (but not always) two groups are
compared
09/17/21
Two categories of intervention
studies
Experimental studies and
Quasi-experimental studies.
09/17/21
Experimental studies
Experimental design is the gold standard
study design
It differs from other study designs
◦ The investigator has full control in
administering the factor of interest
◦ Random assignment or allocation of both
study and control subjects
09/17/21
Classification of experimental
studies
Classification based on the population studied:
◦ Clinical trial- usually in clinical settings and
subjects are patients
◦ Field trial- used in testing drug for preventive
purpose. e.g. vaccine trial
◦ Community trial- unit of the study are group
of people/ community e.g. fluoridation of
water to prevent dental caries
09/17/21
Quasi-experimental studies
either randomization or control group is
missing however, always includes the
manipulation of an independent variable
which is the intervention. Types
uses two (or more) groups
uses only one group ( ‘BEFOREAFTER’ study)
09/17/21
Problems related to intervention
studies
Ethical issues:
◦ Practices or substances already known to be
harmful should not be used
◦ Therapies known to be beneficial should not be
withheld from any affected person
◦ Investigators have to have a complete
knowledge of the study subjects
◦ The researcher must have informed consent
from each of study subjects
◦ A written research protocol is a must
09/17/21
Cont’d
Feasibility or practical issues:
Cost
09/17/21
Thank you
09/17/21
Download