Lecture X

Department of Public Health and Informatics
Lecture series
The project title:

The project title is one of the most important features of the protocol because
it attracts the attention of the potential reader. It is, therefore, necessary to
make it as short and to the point as possible. If we consider two possible
examples:
An investigation to evaluate the effect of the Herbst and Twin Block
functional appliances on skeletal growth during the treatment of Class II
skeletal growth anomalies. A randomized controlled trial.
This title is overlong and states the obvious in a rather wordy way. It goes
without saying that because it is the title of a research protocol it is an
investigation that will evaluate something. A preferable approach may be:
A randomized trial of Herbst and Twin Block appliances.
The second title comes straight to the point without stating the obvious. It not
only attracts the attention of a reader, but it immediately tunes them into the
subject matter.
One of most important features because it is the 1st to attract attention of
reader
It must be CATS
Concise
Accurate
To the point
Short
Dr. Md. Atiqul Haque, Assistant professor

Bangabandhu Sheikh Mujib Medical University, Dhaka
Introduction:
All protocols start with an introduction section. It is a description of the
scientific background to the research question and should include the
following sections (paragraphs):
a. The importance of the topic
The introduction often begins with a statement of the importance of
the research area based on, for example, the number of people who
suffer from the disease being studied or the cost that it presents to the
health authorities.
b. A brief review of current research
The introduction should include a brief review of the landmark studies
and important recent ones. It is essential to do a literature search (e.g.
Medline, Embase) to be sure you have covered the literature
adequately
c. The need for further (your) research
The introduction is not primarily aimed at summarizing present
knowledge but must make the case for the need for further (your)
research by highlighting the gaps in present knowledge.
d. The broad long-term goals (benefits) of the proposed research
The introduction often concludes with the broad long-term goals
(benefits) of the proposed research.
Some important points:

1. Introduction should be concise and to the point
2. Preferably 2 to 3 pages of A4 size paper
3. References should be quoted no more than 20

Research question/hypothesis
Every research work must have research question (s)/hypothesis.
Research question:
Definition: A Research Question is a statement that identifies the
phenomenon to be studied
Many studies have more than one research question.
FINER Criteria for a Good Research question:
1. Feasible
Adequate number of subjects
Adequate technical expertise
Affordable in time and money
Manageable in scope
2. Interesting
Getting the answer intrigues the investigator and her/his friends
3. Novel
Confirms, refutes or extends previous findings
Provides new findings
4. Ethical
Amenable to a study that institutional review board will approve
5. Relevant
To scientific knowledge
To clinical and health policy
To future research
Example of research question
For cross sectional descriptive studies:
What is the prevalence of dry eye among the computer users of
Dhaka University?
Who are the people mostly affected by conjunctivitis?
For case control studies:
Is there any association between computer use and dry eye?
What are the risks of conjunctivitis?
Interventional study:
Does amitryptyline reduce frequency of migraine attack
compared with propranolol?

Components of the clinical question
Population - type of person /patient

Intervention (exposure) - type of exposure
Comparisons - type of control
Outcomes - type of outcome
Refining the clinical question:
Type of exposure
Are anticoagulant agents useful in

patients who have had a stroke?
Type of patient
The well-formulated question

Type of exposure
Type of outcomes
Do anticoagulant agents improve outcomes in

Type of person
patients with acute ischemic stroke

Type of control
compared with no treatment?

Hypothesis:
Definition:
A hypothesis is an educated guess about how things work.
Many study questions undergo a further transformation into a final and
most specific version, termed research hypothesis
When Hypothesis?
Other than descriptive type of cross sectional study all may have a
hypothesis
Characteristics of a good hypothesis:
Simple
Specific
Stated in advance
When hypothesis should be formulated?
If any of the following terms appear in the research question:
Greater than
Less than
Causes
Leads to
Compared with
More likely than
Associated with
Related to
Similar to
Correlated with
Example of hypothesis:
Chocolate may causes migraine
Bacterial growth may be affected by temperature
Ultra violet light may cause cataract
Types of hypothesis:
Null hypothesis or hypothesis of no difference:
There is no association between the predictor and
outcome (e.g. there is no difference in the
frequency of dinking well water between subjects
who develop peptic ulcer disease and those who do
not)
Alternative hypothesis:
The proposition that there is an association (e.g.
the frequency of drinking well water is different in
subjects who develop peptic ulcer than in those
who do not)
In research only alternative hypothesis should be written. The null
hypothesis does not need to be written in the dissertation/thesis.

One-and two-sided alternative hypothesis

One sided: specifies the direction of association between predictor and
outcome
E.g. drinking well water is more common among subjects who
develop peptic ulcers
Two-sided: specifies association; does not specify direction
E.g. Subjects who develop peptic ulcer disease have a different
frequency of drinking well water than those who do not.
Simple versus complex hypothesis:
Simple hypothesis:
One predictor and one outcome
E.g. A sedentary lifestyle is associated with an increased risk of
proteinurea in patients with diabetes
Complex hypothesis:
More than one predictor
(a sedentary lifestyle and alcohol consumption are
associated with an
increased risk of proteinuria in patients with diabetes) or
More than one outcome
(alcohol consumption is associated with an increased risk of
proteinuria and
neuropathy in patients with diabetes)

Objectives in Research:
Research objectives may be stated as:
1. General objective
2. Specific objectives
3. Ultimate objectives
1. General objective:
It is a short statement of the proposed research
It tells in a summary form what will be done during the study
2. Specific objectives:
Every research work involves several tasks. Statement that tell about
each task that will be undertaken during the research work, are termed
as specific objectives
3. Ultimate objective:
If we have written the section of rationale well and given
justification of undertaking the present study, the ultimate
objectives have been covered. However, if we feel the need of
writing the ultimate objectives separately, write them in the
objectives section.
Ultimate objectives describe the idea on the way findings of your
study will be useful and beneficial.
Example: The findings of this case control study will help develop an
authentic document on different risk factors of red eye in Bangladeshi
university student. Such document is so far unavailable in the country.
Use action verb in stating objectives
To assess----------------------------------To examine----------------------------------To describe----------------------------------To explore----------------------------------To elucidate-----------------------------------
Literature review:

Sources of information are generally categorized as primary, secondary or

tertiary depending on their originality and their proximity to the source or
origin.
Primary sources (primary literature):
Primary sources are original materials on which other research is based

They are usually the first formal appearance of results in the print or
electronic literature
They present information in its original form, neither interpreted nor
condensed nor evaluated by other writers.
Primary sources present original thinking; report on discoveries, or
share new information.
Some examples of primary sources:

Scientific journal articles
Proceedings of Meetings, Conferences and Symposia
technical reports
dissertations or theses (may also be secondary)
patents
sets of data, such as census statistics
works of literature (such as poems and fiction)
diaries
autobiographies
interviews, surveys and fieldwork
letters and correspondence
speeches
newspaper articles (may also be secondary)
government documents
photographs and works of art
original documents (such as birth certificate or trial transcripts)
Internet communications or e-mail
minutes of meetings
news footage

Secondary Sources (Secondary literature)
Secondary sources are less easily defined than primary sources. What
some define as a secondary source, others define as a tertiary source.
Nor is it always easy to distinguish primary from secondary sources. A
newspaper article is a primary source if it reports events, but a
secondary source if it analyses and comments on those events.
In science, secondary sources are those which simplify the process of
finding and evaluating the primary literature.
More
generally, secondary sources

describe, interpret, analyse and evaluate the primary sources
comment on and discuss the evidence provided by primary sources
are works, which have one or more steps removed from the event, or
information they
Some examples of secondary sources:
bibliographies (may also be tertiary)
biographical works
commentaries
dictionaries and encyclopedias (may also be tertiary)
dissertations or theses (more usually primary)
handbooks and data compilations (may also be tertiary)
history
indexing and abstracting tools used to locate primary & secondary
sources (may also be tertiary)
monographs (other than fiction and autobiography)
newspaper and magazine articles (may also be primary)
review articles and literature reviews
treatises
works of criticism and interpretation

Tertiary Sources (Tertiary literature)

This is the most problematic category of all.
Works which list primary and secondary resources in a specific subject

area
Works which index, organize and compile citations to, and show you
how to use, secondary (and sometimes primary) sources.
Materials in which the information from secondary sources has been
"digested" - reformatted and condensed, to put it into a convenient,
easy-to-read form.
Some examples of tertiary sources:

almanacs and fact books
bibliographies (may also be secondary)
chronologies
dictionaries and encyclopedias (may also be secondary)
directories
guidebooks, manuals etc
handbooks and data compilations (may also be secondary)
indexing and abstracting tools used to locate primary & secondary
sources (may also be secondary)
textbooks (may also be secondary)
10

Study design
Types of studies:
Epidemiological studies are mainly classified as
1. Observational
2. Experimental
1. Observational studies
Characteristics:
Observational studies allow nature to take its course
The investigator measures but does not intervene
Have no control over exposures; simply observe what happens
to groups of people
Examine associations between risk factors and outcomes
Types:
Two types:
i. Descriptive
ii. Analytical
i. Descriptive study:
Definition:
A descriptive study is limited to a description of the occurrence of a disease
in a population and is often the first step in an epidemiological investigation.
Characteristics:
To describe present or past characteristics of persons with a particular
outcome
Merely describes, does not analyze
Only one group is studied
No comparison group
No conclusion can be made about the association between exposure
and outcome
Type:
1. Case report
2.
3.
4.
5.
Case series
Cross sectional
Survey
Archival research
ii. Analytical study:
Definition:
An analytical study goes further by analyzing relationships between health
status and other variables.
Table: Types of epidemiological study (1)
11

Observational
Experimental
Descriptive
Analytical
Case report
Ecological
RCT
Case series
Cross sectional
Cluster randomized
controlled trials
Cross sectional
Case control
Field trial
Survey
Cohort
Community trial
Table: Types of epidemiological study (2)

Type of study
Observational studies
Descriptive studies
Analytical studies
Ecological
Cross-sectional
Case-control
Cohort
Alternative name
Unit of study
Correlational
Prevalence
Case-reference
Follow-up
Populations
Individuals
Individuals
Individuals
Experimental studies
Randomized controlled trials
Cluster randomized controlled
trials
Field trials
Intervention studies
Clinical trials
Community trials
Community intervention
studies
12
Individuals
Groups
Healthy
people
Communities

Descriptive type of study:
1. Case study:
Definition:
Case study refers to the collection and presentation of detailed information
about a particular participant or small group, frequently including the
accounts of subjects themselves.
Characteristics:
It is a qualitative descriptive research
It is useful in rare diseases
It may be first to provide clues in identifying a new disease or adverse
health effect from an exposure
The case study looks intensely at an individual or small participant
pool, drawing conclusions only about that participant or group and only
in that specific context.
It provides insights for research questions to be addressed by
subsequent, planned studies
It generates hypothesis
Strengths
provide real examples
encourage replication
are generally practical in nature
Provide innovative ideas.
13

2. Case series
Definition:
A case series is a study very similar in structure and form as case reports, but
describes a group of cases, instead of a single patient.
This study design may be retrospective or prospective, and usually
requires a relatively small sample size (usually 10 or more).
Characteristics:
Retrospective or prospective
Usually involves a smaller number of patients
Case series may be consecutive or non-consecutive, depending on
whether all cases presenting to the reporting authors over a period of
time were included
Case series may be confounded by selection bias, which limits
statements on the causality of correlations observed
It generally involve patients seen over a relatively short time
Case-series studies do not include control subjects
It provides insights for research questions to be addressed by
subsequent, planned studies
It generates hypothesis
Disadvantages:
Weakest kinds of observational studies
Represent a description of typically unplanned observations
14

3. Survey:
Assessing public opinion or individual characteristics by the use of

questionnaire and sampling methods
Nearly every one has taken part in a survey
e.g. Bangladesh Demographic and Health Survey 2011, Bangladesh Literacy
Survey, 2010
4. Archival research:
Study method that examines existing records to obtain data and test
hypothesis.
Example: A researcher might study crime statistics in different countries to
see if there is a relation between capital punishment and the murder rate.
5. Anecdote:
A short account (or narrative) of an interesting or amusing incident, often intended to
illustrate or support some point.
15

Analytical studies
Ecological (correlational) study:
Ecology:
Derived from Greek word, Meaning: study of house
The study of the relationships among living organisms and their environment
Human ecology means the study of human groups as influenced by
environmental factors, including social and behavioral factors.
Ecological study:
Definition:
A study in which the units of analysis are populations or groups of people,
rather than individuals
Common types of ecological study are geographical comparisons, time trend
analysis or studies of migration.
Example:
A study of mortality from lung disease in different cities that are known to
have differing levels of air pollution would comprise an ecologic study. The
unit of analysis is a city.
Look at the association between smoking and lung cancer deaths in different
countries. The unit of analysis is a country
Characteristics:
Examine rates of disease in relation to a population-level factor
Population-level factors include summaries of individual population
members, environmental measures, and global measures
Study groups are usually identified by place, time, or a combination of
the two
Limitations include the ecological fallacy and lack of information on
important variables
Ecologic studies generally make use of secondary data that have been
collected by the government, some other agency or other
investigators.
Advantages include low cost, wide range of exposure levels, and the
ability to examine contextual effects on health.
Advantages:
Simple to conduct
Low cost
Wide range of exposure levels
Ability to examine contextual effects on health
Disadvantages:
Inferior to non-ecological designs such as cohort and case-control
studies
It is susceptible to the ecological fallacy.
lack of information on important variables
16

Usually Hypothesis Generating

Lack of adequate data and missing data
May not be recorded a group level
Within-Group misclassification
Confounding
Collinearity
Temporal ambiguity
Ecological fallacy (Rothman and Greenland 1998)
Difference between ecological study and cohort study:

Ecological studies can be easily confused with cohort studies, especially if
different cohorts are located in different places.
In the case of ecological studies there is no information available about the
individual members of the populations compared (e.g. comparing several
states based on state-wide average air pollution and state-wide average
prevalence of respiratory diseases); whereas in a cohort study the data pair
exposure/health is known for each individual.
Ecological study
No information available about the
individual members
Cohort study
Information on individual members
are inevitable
Type of ecologic study:

Two type:
1. Ecologic comparison studies (cross sectional ecologic studies)
2. Ecologic trend studies
Ecologic comparison studies (cross sectional ecologic studies):
It involves an assessment of the correlation between exposures rates and
disease rates among different groups or populations over the same time
period.
Usually there are more than 10 groups or populations
Data on disease may include:
1. Incidence rate
2. Prevalence
3. Mortality rates
Examples of exposure data include:
1. Measures of economic development (per capita income and illiteracy
rate)
2. Environmental measures (mean ambient temperature, levels of
humidity, annual rainfall)
3. Measures of lifestyle (smoking prevalence, mean per capita intake of
calories, annual sales of alcohol)
17

The ecologic trend study:

It involves correlation of changes in exposure and changes in disease over
time within the same community, country or other aggregate unit.
Ecologic fallacy:
Definition: The bias that may occur because an association observed between
variables on an aggregate level does not necessarily represent the
association that exists at the individual level.
Cross sectional study

Syn: Disease frequency survey, prevalence study
Definition:
A study that examines the relationship between diseases (or other health
related characteristics) and other variables of interest as they exist in a
defined population at one particular time.
The presence or absence of disease and the presence or absence of the other
variables (or, if they are quantitative, their level) are determined in each
member of the study population or in a representative sample at one
particular time.
Features:
Study at a point (cross section) in time
Easy and economical
Regular cross sectional survey for determining extent of health related
problems and sociodemographic characteristics
More common type of research in health science (approximately more
than 1/3rd of original articles in major medical journal
Inaccurate when studying rare conditions
Subjects are selected without regard to the outcome of interest
Less expensive
They are the best way to determine prevalence
Quick
The principal summary statistic of cross sectional studies is the odds
ratio
Weaker evidence of causality than cohort studies
Advantages:
Easy to conduct
Simple in term of design, and calculation
Less time consuming
Cheapest
Ability to establish relationship between exposure and outcome
Suitable for problem identification
Gives idea about the disease prevalence
Associations can be studied
[Cross-sectional studies can also be used for examining associations,
although the choice of which variables to label as predictors and which as
18

outcomes depends on the cause-and-effect hypotheses of the investigator

rather than on the study design]
Risk ratio
(relative risk)
Disease
Yes
No
Yes
No
A
A+B
C
C+D
Risk
Factor
Disadvantages:
We cannot come to the conclusion about the disease aetiology
We cannot follow up the cases
Very weak analytical power
We cannot find out the incidence of the disease
Example:
Javadi MA, Katibeh M, Rafati N, Dehghan MH, Zayeri F, Yaseri M, et al. 2009,
Prevalence of diabetic retinopathy in Tehran province: a population-based
study BMC Ophthalmology vol. 9, pp. 12
Findings:
Screened patients: 7989
Diabetic patients: 759 (9.5%)
Of them, 639 patients (84.2%) underwent eye examination
Five patients (0.7%) with media opacity were excluded
Examined patients with diabetes: 634
Diabetic retinopathy: 240
Standardized prevalence of retinopathy: 37%
non proliferative 27.3%
proliferative diabetic retinopathy 9.6%
The prevalence of any type of visual impairment (BCVA < 20/60) in patients
with PDR was remarkably higher than that in patients without PDR (18.5% vs.
7.0%, P = 0.002, OR = 2.09, 95% CI: 1.02- 4.26).
19

Case control study:

Syn: Case comparison study, case compeer study, case history study,
case referent study, retrospective study
Definition: The observational epidemiologic study of persons with the
disease (or other outcome variable) of interest and a suitable control
(comparison, reference) group of persons without the disease.
Case control study can be called retrospective because it starts after
the onset of disease and looks back to the postulated causal factors.
20

Advantages:
Cheap, easy and quick studies
Rapid and less time consuming
Small sample size is sufficient
Suitable for rare diseases (e.g. macular edema)
Multiple exposures can be examined
No ethical problem
Risk factor can be identified
Disadvantages:
Control selection difficult
Subject to bias (Any systemic error in which there is tendency to
produce result in own favour)
We can not measure incidence
Multiple outcomes cannot be studied
Do not establish risks and rates directly
Not suitable for rare exposure
21

Example:
Leske MC, Chylack LT, Jr, Wu SY, The Lens Opacities Case-Control Study
Group, 1991, The Lens Opacities Case-Control Study: Risk Factors for
Cataract, Arch Ophthalmol. vol. 109, no. 2, pp. 244-51.
Objective:
To evaluated risk factors for age-related nuclear, cortical, posterior
subcapsular, and mixed cataracts
Participants: 1380
Age range: 40-79
Type of cataract:
Posterior subcapsular only, 72 patients;
Nuclear only, 137 patients;
Cortical only, 290 patients;
Mixed cataract, 446 patients; and
Controls, 435 patients.
Statistical analysis: Logistic regression analyses
Low education increased risk (odds ratio [OR]= 1.46)
Regular use of multivitamin supplements decreased risk (OR =0.63) for all
cataract types.
Dietary intake of riboflavin, vitamins C, E, and carotene, which have
antioxidant potential, was protective for cortical, nuclear, and mixed cataract;
Diabetes increased risk of posterior subcapsular, cortical, and mixed
cataracts (OR =1.56).
Oral steroid therapy increased posterior subcapsular cataract risk (OR =
5.83). Females (OR =1.51) and nonwhites (OR = 2.03) were at increased risk
only for cortical cataract.
Risk factors for nuclear cataract were a nonprofessional occupation (OR
=1.96), current smoking (OR = 1.68), body mass index (OR = 0.76), and
occupational exposure to sunlight (OR =0.61).
22

Cohort studies
Syn: longitudinal study, follow-up study, incidence study, concurrent study,
prospective study
Definition:
A study in which two or more groups of individuals those are free of disease
and those differ according to the extent of exposure to a factor of interest,
are followed over a period of time to see how their exposures affect their
outcomes.
In terms of time
Prospective >90%
Retrospective <10%
In terms of inquiry
Always forward, i.e. from cause (exposure) to effect
Advantages:
Accurate estimation of risk directly
Can establish population-based incidence
Temporal relationship between exposure and disease (exposure should
obviously precede disease)
Rare exposure(asbestos > lung cancer)
Can be used where randomization is not possible
Magnitude of a risk factors effect can be quantified
Multiple outcomes can be studied (smoking > lung cancer, COPD,
larynx cancer)
Disadvantages:
Large sample size
Losses to follow up
Exposure may be distorted
Ethical problem (you are morally responsible because you are not
advising him that do not smoke because it will harm you)
High cost
Long time
Not suitable for rare diseases
Not suitable for diseases with long-latency
Unexpected environmental changes may influence the association
Nonresponse, migration and loss-to-follow-up biases
Hawthorne effect
Confounding variables are the major problem in analyzing cohort

studies.
23

The Hawthorne effect

Definition: The Hawthorne effect is referring to the tendency of some people
to work harder and perform better when they are participants in an
experiment. Individuals may change their behavior due to the attention they
are receiving from researchers rather than because of any manipulation of
independent variables.
Definition: The Hawthorne effect is a form of reactivity whereby subjects
improve or modify an aspect of their behavior being experimentally measured
simply in response to the fact that they are being studied, not in response to
any particular experimental manipulation.
Historical perspective: The term was coined in 1950 by Henry A.
Landsberger when analysing older experiments from 1924-1932 at the
Hawthorne Works (a Western Electric factory outside Chicago). Hawthorne
Works had commissioned a study to see if its workers would become more
productive in higher or lower levels of light. The workers' productivity seemed
to improve when changes were made and slumped when the study was
concluded. It was suggested that the productivity gain was due to the
motivational effect of the interest being shown in them. Although illumination
research of workplace lighting formed the basis of the Hawthorne effect,
other changes such as maintaining clean work stations, clearing floors of
obstacles, and even relocating workstations resulted in increased productivity
for short periods. Thus the term is used to identify any type of short-lived
increase in productivity.
Types:
1. Prospective
2. Retrospective (historical cohort study)
24

1. Prospective cohort study

Definition:
A group of people is chosen who do not have the outcome of interest (for
example, myocardial infarction). The investigator then measures a variety of
variables that might be relevant to the development of the condition. Over a
period of time the people in the sample are observed to see whether they
develop the outcome of interest (that is, myocardial infarction).
25

2. Retrospective cohort:
26

Example:
Prospective Cohort study:
Lindblad BE, Hakansson M, Philipson B, Wolk A, 2007, Alcohol Consumption
and Risk of Cataract Extraction: A Prospective Cohort Study of Women, vol.
114, no. 4, pp. 680-85
Purpose
To investigate the association between alcohol consumption and the risk of
cataract extraction
Design
Population-based prospective cohort study
Participants
34,713 women participating in the Swedish Mammography Cohort; Age: 49 to
83 years
Data collection tool: a self-administered questionnaire about alcohol,
smoking, and other lifestyle factors
Completion year: 1997
Results
Duration of follow up 84 months
Incident cases of age related cataract extraction: 3587
Compared with never drinkers, the relative risk of cataract extraction among
current drinkers was 1.11 (95% confidence interval [CI] 1.021.21) after
adjustment for age and other potential risk factors.
In multivariate analysis, an increment of 13 g alcohol intake per day was
associated with a 7% increased risk of cataract extraction (relative risk, 1.07;
95% CI 1.021.12)
World famous cohorts:
British Doctors' Cohort Study
Framingham Heart Study
EPIC Study: European Prospective Investigation into Cancer and
Nutrition
Nurses' Health Study
Physicians' Health Study
Women's Health Initiative
27

Retrospective cohort study:
Study subjects: Children vaccinated against varicella

Born after 1993; they all followed up through 1999
Two centers in USA (A and B)
Information was obtained from automated vaccination, clinic, hospital

discharge, and pharmacy records.
Characteristics
Total population
Breakthrough
Risk factors
Asthma
Inhaled steroids prescribed at any time
Oral steroids prescribed before
vaccination
Oral steroids prescribed 3 month after
vaccination
Vaccine given before 15 months of age
Varicella vaccination followed MMR
vaccine within 28 days
Area A
80,584
Area B
8,181
268
97
aRR=2.4
aRR=2.8
aRR=1.4
aRR=3.1
Verstraeten T, Jumaan AO, Mullooly JP, Seward JF, Izurieta HS, DeStefano F. A
Retrospective Cohort Study of the Association of Varicella Vaccine Failure With
Asthma, Steroid Use, Age at Vaccination, and Measles-Mumps-Rubella
Vaccination. Pediatrics 2003;112;e98-e103
28

Case
reports
Case series
Subjects
Single case
>1; exact
number can
vary
Characteristics
Describes
unusual
feature of a
case
A group of
cases; no
control group
Orientation
Time factor
Cost
Present
Quick
Inexpensive
Present
Quick
Inexpensive
Follow up
period
Purpose
None
Cause effect
relationship
Measure of
disease
frequency and
risk
Potential
problems
Cross
sectional
studies
Several
subjects
assessed
individually
point-in-time
picture of
health status/
health-related
behaviour
Present
Quick
Inexpensive
None
Generate
hypothesis
Cannot be
interpreted
Generate
hypothesis
Cannot be
interpreted
Generate
hypothesis
Cannot be
interpreted
None
None
Prevalence,
Association
Cannot assess
seasonal
variation
Remarks
Ecologica
l studies
Case control
Cohort
Aggregate
group of
individuals
Two groups of
subjects
Two groups of
subjects
often
studied by
geographi
cal area
One group has

disease of
interest and
other is
disease-free
Past
Quick
Inexpensive
One group is
exposed to risk
factor of interest
and other is nonexposed
Future
Time consuming
Expensive
None
Often prolonged
Generate
hypothesis
Cannot be
interprete
d
Correlation
Test hypothesis
Test hypothesis
Can be
suggested
Can be interpreted
Odds ratio
Ecological
fallacy
Recall and
selection bias
Prevalence,
Incidence,
Relative risk,
Attributable risk
Attrition
Present
Quick
Inexpensiv
e
None
Can study
multiple causes
29
Can study multiple

outcomes of a

of disease;
useful for
studying rare
diseases
30
single exposure

Experimental study:
Type:
1. Clinical trials
2. Field trials
3. Community trials
Clinical trials
Randomized
Non randomized
Pre-test/post-test study (before-after studies)
Quasi-experimental study
Different name of clinical trial:
Randomized Clinical trial
Randomized Drug trial
Randomized control trial
Commonly used Design in clinical trial:
1. Parallel design
2. Cross-over designs
3. Cluster randomized trial
4. Factorial trial
31

Parallel design
32

Cross over design
Example:
Silaste ML, 2003, Dietary effects on antioxidants, oxidised LDL and
homocysteine,[Thesis] Department of Internal Medicine, University of Oulu,
Finland
Objective:
To investigate the influence of modifications in the vegetable, berry, and fruit
intake and dietary fat on the plasma concentrations of antioxidants, lipids,
lipoprotein(a), oxidised LDL, folate, and homocysteine.
Research questions:
Does a diet high in common vegetables, berries, citrus fruit, and poly
unsaturated fatty acid (PUFA) enhance the plasma concentrations of
carotenoids, vitamin C, and vitamin E?
Do the modifications in the dietary intake of vegetables, berries, and
fruit influence the plasma concentrations of lipids, lipoproteins, and
oxidised LDL?
How do the dietary modifications and gene polymorphisms affect the
serum paraoxonase-1 activity?
Does a high intake of natural folate from food increase the serum
folate concentration and decrease the plasma tHcy concentration?
Do the common gene polymorphisms alter the dietary response of
plasma tHcy concentration?
33

Study design:
Randomized control trial cross over design
Methodology:
The intervention consisted of
Baseline diet: 2 weeks
Two diet periods (low and high vegetable diets) 5 weeks each
Wash-out period in between: 3 weeks
An important feature of the study was a crossover design, in which each
individual served as her own control.
The order of the study diets was randomly assigned for each subject.
34

Cluster Randomized trails

Some time individual randomization impossible
Have to randomize community
Example:
MacIntyre CR, Cauchemez S, Dwyer DE, Seale H, Cheung P, Browne G, 2009,
Face Mask Use and Control of Respiratory Virus Transmission in Households,
Emerging Infectious Diseases vol. 15, no. 2 DOI: 10.3201/eid1502.081167,
available from: www.cdc.gov/eid
Justification of their study:
During an influenza pandemic:
Supplies of antiviral drugs may be limited, and
There will be unavoidable delays in the production of a matched
pandemic vaccine
For new or emerging respiratory virus infections, no pharmaceutical
interventions may be available.
Even with seasonal influenza, widespread oseltamivir resistance in
influenza virus A (H1N1) strains have recently been reported
Objective:
To describe the efficacy of face mask use for preventing spread of
influenza-like infection.
Intervention:
Group 1: Control (no face mask)
Group 2: Surgical mask intervention
Group 3: P2 mask intervention
Result:
Masks may play an important role in reducing transmission
35

Figure: Flow diagram of recruitment for the prospective, clusterrandomized trial, Sydney, New South Wales, Australia, 2006 and
2007 winter influenza seasons.
36

Factorial trials
Proposed by R.A. Fisher in the 1920s
Developed in the 1940s for use in industrial experiments
This design is used to assess simultaneously the effects of main
treatments and one or several other factors.
In a specific case of two treatments A and B if one wished to assess
simultaneously the effects in subjects who are:
Given neither A nor B
Given A alone and not B
Given B alone and not A
Given both A and B
The simplest factorial experiment, a 22 design
Factorial design: 2 by 2 type
Drug
A
Level
A=0
A=1
B=0
None
A alone
B alone
A&B
both
B=1
37

DREAM Study
(Diabetes reduction assessment with Ramipril and Rosiglitazone
Medication)
Factorial design
Rosiglitazone
Ramipril
Ramipril+
Rosiglitazone
Placebo
Rosiglitazone+
placebo
Start: May 2001 Follow 3 to 5 years
Final Results: 2005/2006
Placebo
Ramipril+
Placebo
Placebo+ Placebo
Before-After studies
38

Quasi-experimental designs
No randomization
No intervention in control group
39

Community Trial
40

Study Designs (all)
41

Validity
ranking
Types of study
design
Highest
Randomized clinical
trial
Prospective cohort
study
Retrospective cohort
study
Nested case-control
study
Time-series analysis
Cross-sectional study
Ecologic study
Cluster analysis
Case study
Lowest
Anecdote
42

Clinical development of a drug:
Phase
of trial
Subject
Phase I
Duration
of study
Purpose
Study procedure
Healthy
20-80
Volunteers/target
disease (Ca, HIV)
Up to 1
month
Metabolic and
pharmacological action
Maximally tolerable dose
Single, ascending dose tiers,

Unblinding, Uncontrolled
Phase
II
Patients with
target disease
200-300
Several
months
Safety, efficacy
pharmacokinetics
Establishes minimum
and maximum effective
dose
Randomization
Double blind
Compare with a placebo
Phase
III
Patients with
target disease
Several
hundred
to
thousands
Several
years
Obtain additional information Randomization

regarding Safety, efficacy
Blinding
Compare with
currently available
Drug
Thousands
Ongoing
Post marketing
Monitor ongoing safety in
large populations
Phase IV Patients
No.
of subjects
43
Uncontrolled
Observational

Blinding
Blinding, also called masking

If the outcome can conceivably be affected by patient or investigator
expectations, then blinding is important.
Types of Blinding
Single Blind: The patient is blind
Double Blind: The patient and the investigator are blind
Triple Blind: The patient, investigator and data-cleanup people are
blind. The statistician can only be partially blinded since he/she has to
know which patients are in the same treatment group.
Sampling
Probability sampling
Simple random sampling
Stratified random sampling
Systematic random sampling
Cluster sampling
Multistage sampling
Nonprobablity sampling
Judgment or purposive sampling
Quota sampling
Convenient sampling
Snowball Sampling
Snowball sampling is a method in which a researcher
identifies one member of some population of interest,
speaks to him/her, and then asks that person to identify
others in the population that the researcher might speak
to. This person is then asked to refer the researcher to yet
another person, and so on.
Snowball sampling is very good for cases where members
of a special population are difficult to locate.
For example, Selection of beggars has no frame work. This
can be best done by asking an initial group of beggars to
supply the names of other beggars they come across.
Selection of street sex workers also can be made following
this network approach
44

Simple random sampling

Lottery method
Table of random numbers
Computer random number generator
Stratified sampling
Total sample 100

Data will be collected from
Dhaka
Rajshahi
Khulna
Barisal
Chittagong
75
12
5
5
3
45

Systematic sampling
Population size (N=500)
Decide sample size (n=100)
Calculate interval size (interval 5)
Cluster sampling
Divide population into clusters
Clusters: Random selection
Include all patients within sample cluster
Suppose that we need a random sample of n=200 households from a
population of N=8000 households of a city. Since there does not exist any
good list of the households, it would be a difficult job to sample the individual
households. It would be at the same time too expensive to prepare such a
list. Instead, we can obtain a sample of blocks by dividing the entire area into
a number of blocks and then selecting 200/8000 =2.5% of the blocks.
Suppose we make 80 blocks each with 100 households. Then 2.5% of 80
blocks implies 80*2.5%=2 sample blocks. These 2 blocks contain 200
households. These households located within the boundaries of the sample
blocks comprise the sample.
Multistage cluster sampling
Upzilla (3 out of 6)
Union (for each upzilla select 2 union)
Village (for each union select 5 village)
Go to village, randomize people
The national Iodine deficiency disorders surveys of 1993, 1999 and 2004
conducted under the joint collaboration of UNICEF and Dhaka University
46

adopted a multi-stage cluster sampling procedure. In 1999 survey, the whole

country was divided into three ecological zones viz. flood prone, hilly and
plain. Initially, thana were selected as primary sampling units from each of
the zones and from the selected thana, mouza were selected as the
secondary sampling units. At the third and final stage, households were
selected and all members of the selected households were subjected to
detailed interview.
Difference between cluster sampling and stratified sampling

1. In cluster sampling the cluster is treated as the sampling unit so analysis is done
on a population of clusters (at least in the first stage). In stratified sampling, the
analysis is done on elements within strata.
2. In stratified sampling, a random sample is drawn from each of the strata, whereas
in cluster sampling only the selected clusters are studied.

3. The main objective of cluster sampling is to reduce costs by increasing sampling
efficiency
Cluster Vs Stratified Sampling
Cluster Sampling
When natural groupings are evident in a statistical population, cluster sampling
technique is used.
Cluster sampling can be opted if the group consists of homogeneous members.
The advantages of cluster sampling over other sampling methods is, it is cheaper
as compared to the other methods.
The main disadvantage of cluster sampling is, it introduces higher sampling error.
This sampling error can be represented as design effect.
Stratified Sampling
Stratified sampling is a method where in, the member of a group are grouped into
relatively homogeneous groups.
For heterogeneous members in the groups, stratified sampling is a good option.
The advantages of stratified sampling are, this method ignores the irrelevant ones
and focuses on the crucial sub populations. Another advantage is, with this
method, for different sub populations, you can opt for different sampling
techniques. This sampling method also helps in improving the efficiency and
accuracy of the estimation. This sampling method allows greater balancing of
statistical power of tests.
The disadvantages of this sampling method are, it requires choice of relevant
stratification variables which can be tough at times. When there are homogeneous
subgroups, it is not much useful. Its implementation is expensive. If not provided
with accurate information about the population, then an error may be introduced
47
Figure: Cluster sampling
Population
Cluster one
Cluster two
Cluster three
Randomly sample within each cluster
Sample one
Stage 1:
Randomly
selected
districts
within the
country
Sample two
Stage 2:
Randomly
selected
upzillas
within those
districts
Stage 3:
Randomly
selected
village within
those upzilla
48
Sample three
Stage 4:
Randomly
selected
Tuberculosis
patients with
each village for
study

Figure: Cluster sampling
49

Statistics
Definition: Statistics may be defined as the science and art of
collection, organization, analysis, interpretation and presentation of
numerical data from which a definite conclusion can be made.
C
O
A
I
P
collection
organization
analysis
interpretation
presentation
Parameter: Character of population
Statistic: Character of sample
Biostatistics:
It deals with data relating to biological aspects of human being.
Biostatistics is much broader term and includes the vital statistics.
E. g. Marriage, divorce birth, death etc.
Data: Data is a set of values recorded on one or more observational
units.
Variable:
Definition:
A variable is a characteristic of a person, object or phenomenon that
can take on a different value.
50

Types of variable
A variable can be classified in a number of ways.
A. The causal relationship
B. The design of the study
C. The unit of measurement
From the viewpoint of causation:
In studies that attempt to investigate a causal relationship or
association, four sets of variables may operate:
1. Changes variables, which are responsible for bringing about
change in a phenomenon
2. Outcome variables, which are the effects of a change variable;
3. Variables which affect the link between cause-and-effect
variables;
4. Connecting or linking variables, which in certain situations are
necessary to complete the relationship between cause-and-effect
variables.
Connecting or linking
variables (4)
Cause
Change variables (1)
Effect
Outcome variable (2)
Variables that affect the relationships (3)
In research terminology:
Change variables are called: independent variables
Outcome/effect variables are called: dependent variables
The unmeasured variable affecting the cause-and-effect relationship
are called extraneous variables
The variables that link a cause-and-effect relationship are called
intervening variables.
Hence:
1. Independent variable:
The cause supposed to be responsible for brining about
change(s) in a phenomenon or situation
2. Dependent variable: The outcome of the change(s) brought
about by introduction of an independent variable
51

3. Extraneous variable: These factors, not measured in the study,

may increase or decrease the magnitude or strength of the
relationship between independent and dependent variables
4. Intervening variable:
SYN. Intermediate variable
Sometimes called the confounding variable
It links the independent and dependent variables
Example: Smoking causes lung cancer
Cancer(Assumed
effect)Dependent
variable
Smoking(Assum
ed
cause)Independ
ent variable
Change
variables (1)
Affect the relationship
The age of the person

The extent of his/her smoking
The duration of smoking
The extent of daily
exerciseExtraneous variables
Figure: Independent, dependent and extraneous variables in a

causal relationship
52

MortalityIndepe
ndent variable
Use of contraceptive
FertilityDepende
nt variable
Intervening variables
Attitudes towards CP use among the population

Level of education of the population
Socioeconomic status
Provision and quality of health services
Motivation of the individual
Age
Religion
Extraneous variables
Figure: Independent, dependent, extraneous and intervening

variables
53

From the viewpoint of the study design:

In controlled experiments the independent (cause) variable may be
introduced or manipulated either by the researcher.
In these situations there are two types of variables:
1. Active variables
2. Attribute variables
Active variables: Those variables that can be manipulated, changed
or controlled
Attributed variables: Those variables that cannot be manipulated,
changed or controlled
Table: Active and attribute variables
Study intervention
Different treatment models
Experimental intervention
Program service
Study population
Age
Gender
Level of motivation
Attitudes
Religions
Active variables
A researcher can manipulate
Attribute variables
A researcher cannot manipulate
54

From the viewpoint of the unit of measurement

Variables:
Type:
1. Qualitative or categorical
2. Quantitative
Nature:
1. Discrete
2. Continuous
Scale:
1. Nominal
2. Ordinal
3. Interval
4. Ratio
Qualitative Variable:
Variables that take non-numeric narrative values are called Qualitative,
or Categorical variable
E.g. Gender which takes the non-numeric narrative value of Male or
Female.
Two types:
a) Dichotomous:
a. Only two categories
b. Yes/no, good/bad, rich/poor
b) Polytomous
a. More than two categories
b. Religion (Christian, Hindu, Muslim), Socioeconomic
condition (lower, mid and upper)
Quantitative Variable:
Variables that take numerical values are called Quantitative variable.
This variable has two characteristic namely discrete and continuous.
Discrete: A variable that takes only isolated or absolute values is
called discrete e.g. parity, family size, pulse rate etc.
Continuous: When a variable can assume any value within a
meaningful range or continuum it is called continuous, for example,
age, height, weight etc.
55

Variable
Qualitative
(Outcome: Narrative)
Quantitative
(Outcome: Numerical)
Discrete
Discrete
Nominal
Ordinal
Continuous
Interval
Ratio
Scale of measurement:
Proposed by S S Stevens in 1946
a) Nominal or classificatory scale
b) Ordinal or ranking scale
c) Interval scale
d) Ratio scale
a) Nominal variable (scale)
Definition: A nominal variable consists of named categories, with no
implied order among the categories.
Properties:
Observations of a qualitative variable can only be classified and
counted. There is no natural order (for example- Gender).
Data categories are represented by labels or names
Even when the labels are numerically coded, the data
categories have no logical order.
e.g.
Gender:
1. Male
2. Female
56

Marital Status:
1. Unmarried
2. Married
3. Divorced
4. Widow
b) Ordinal or ranking scale
Definition: An ordinal variable consists of ordered categories, where
the differences between cannot be considered to be equal.
Properties:
The categories are distinct, mutually exclusive and exhaustive
The categories are possible to be ranked or order
The distance or difference from one category to the other
category is not necessarily constant
Variables with an ordered series
e.g.
Mild
Moderate
Severe
What Does Mutually Exclusive Mean?
A statistical term used to describe a situation where the occurrence of one event is
not influenced or caused by another event. In addition, it is impossible for mutually
exclusive events to occur at the same time.
Mutually Exclusive Events
Non-Mutually Exclusive Events
Two events are mutually exclusive if
Two events are non-mutually exclusive if
they cannot occur at the same time
they have one or more outcomes in
(i.e., they have no outcomes in
common.
common).
In the Venn Diagram above, the

probabilities of events A and B are
represented by two disjoint sets (i.e.,
they have no elements in common).
In the Venn Diagram above, the

probabilities of events A and B are
represented by two intersecting sets (i.e.,
they have some elements in common).
Interval variables:
57

Definition: An interval variables has equal distances between values,

but the zero point is arbitrary
Properties:
These are equally spaced variables, e.g. temperature
Zero has meaning
The data classifications are mutually exclusive and exhaustive
The data can be meaningfully ranked or ordered
Distance between attributes does have meaning. For example,
when we measure temperature (in Fahrenheit), the distance from
30-40 is same as distance from 70-80. (e.g. 30-40 =70-80)
The ratio between number in the scale is not, however,
necessarily the same as that between the amounts of the
attribute. A room at 200C is not twice as hot as one at 100C.
Example: Temperature (Celsius, Fahrenheit), IQ, SAT score, PH
Temperature, expressed in F or C, is not a ratio variable. A temperature of 0.0 on either of
those scales does not mean 'no temperature'. However, temperature in degrees Kelvin in a
ratio variable, as 0.0 degrees Kelvin really does mean 'no temperature'.
Ratio variables (scale):

Definition: A ratio variables has equal intervals between values and a
meaningful zero point
Properties:
Data classifications are ordered according to the amount of the
characteristics they possess.
Equal differences in the characteristics are represented by equal
differences in the numbers assigned to the classifications.
The zero means absence and
Ratio between the 2 numbers is meaningful.
Example: age, Temperature (Kelvin)
Difference:
Nomina Ordina Interva
Ratio
l
l
l
Frequency distribution
Yes
Yes
Yes
Yes
Median and percentiles
No
Yes
Yes
Yes
Add or subtract
No
No
Yes
Yes
Mean, standard deviation, standard error of the

mean
No
No
Yes
Yes
Ratio, or coefficient of variation
No
No
No
Yes
58

Flow chart regarding variables:

Do the numbers express a quantitative value or order?
If no
If yes,
ask
Nominal
Do the differences between the numbers

represent equal units of measurement (e.g.
3-2=4-3)?
If no
If yes,
ask
Ordinal
Does the measurement have an

absolute zero?
If no
If yes
Interval
Ratio
59

Variable based on threshold values:

A particular group of derived variables are those based on threshold
values of a measured variable.
Derived variable
Original variables
LBW
<2500 g
Yes
2500 g
No
Vitamin A status
<0.35mol/L
Severe deficiency
0.35-0.69 mol/L
Mild/moderate deficiency
0.70 mol/L
Normal
Commonly used alternatives for describing exposure and
outcome variables
Outcome variable
Exposure variable
Response variable
Dependent variable
y-variable
Case-control group
Explanatory variable
Independent variable
x- variable
Risk factor
Treatment group
60


Lecture X

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture X

Uploaded by

Copyright:

Available Formats

Department of Public Health and Informatics

The project title:

Dr. Md. Atiqul Haque, Assistant professor

Some important points:

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Components of the clinical question

Population - type of person /patient

Refining the clinical question:

Are anticoagulant agents useful in

The well-formulated question

Do anticoagulant agents improve outcomes in

patients with acute ischemic stroke

compared with no treatment?

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

One-and two-sided alternative hypothesis

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Sources of information are generally categorized as primary, secondary or

Primary sources are original materials on which other research is based

Some examples of primary sources:

Dr. Md. Atiqul Haque, Assistant professor

Secondary Sources (Secondary literature)

generally, secondary sources

Dr. Md. Atiqul Haque, Assistant professor

Tertiary Sources (Tertiary literature)

Works which list primary and secondary resources in a specific subject

Some examples of tertiary sources:

Dr. Md. Atiqul Haque, Assistant professor

ii. Analytical study:

Dr. Md. Atiqul Haque, Assistant professor

Table: Types of epidemiological study (2)

Dr. Md. Atiqul Haque, Assistant professor

Descriptive type of study:

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Assessing public opinion or individual characteristics by the use of

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Usually Hypothesis Generating

Difference between ecological study and cohort study:

Type of ecologic study:

Dr. Md. Atiqul Haque, Assistant professor

The ecologic trend study:

Cross sectional study

Dr. Md. Atiqul Haque, Assistant professor

outcomes depends on the cause-and-effect hypotheses of the investigator

Dr. Md. Atiqul Haque, Assistant professor

Case control study:

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Dr. Md. Atiqul Haque, Assistant professor

Large sample size

Exposure may be distorted

Not suitable for rare diseases

Not suitable for diseases with long-latency

Unexpected environmental changes may influence the association

Nonresponse, migration and loss-to-follow-up biases

Confounding variables are the major problem in analyzing cohort

Dr. Md. Atiqul Haque, Assistant professor

The Hawthorne effect

Dr. Md. Atiqul Haque, Assistant professor