Teaching

Methodology &
Community
Medicine
Dr. Ghazal Hussain
Epidemiological Methods
Experimental studies and
Association
Community medicine 11/11/2017
 Review of last lecture
✓ Analytic epidemiology
✓ Cohort study
✓ Types of cohort study
✓ Method of cohort study
✓ Relative risk & Attributable risk
• Advantages & Disadvantages of cohort study

Community medicine 11/11/2017

 Lecture outline
❖ Experimental epidemiology
❖ Method of experimental study
❖ Blinding and its types
❖ Definition of Bias, Survey, Gold standard
❖ Association and its types
 Experimental studies

The study in which the researcher

exercises control over the independent
variables deliberately manipulating them is
called experimental studies.
• Experimental studies are similar in approach
to cohort studies excepting that the
conditions in which study is carried out are
under the direct control of the investigator.
• This may be conducted in animals or humans.
 Aims
• To provide scientific proof of etiological or
risk factors.
• To provide a method of measuring the
effectiveness and efficiency of health services.
 Types of experimental studies

1. Randomized control trials

2. Non-randomized control trials
 1.Randomized control trials

• Randomization is a process by which subjects

are allocated to one of two or more therapy
groups by chance, thus minimize selection bias.
• RCT is an epidemiological experiment.
• It is the number one method of evaluation for
new programs or new therapies.
 Types of RCTs

• Clinical trials
• Preventive trials
• Risk factors trials
• Cessation experiments
• Trials of etiological agents
• Evaluation of health services
 Method of conducting RCTs
Basics steps are:
1. Drawing up a protocol
2. Selecting reference population
3. Selecting experimental population
4. Randomization
5. Manipulation or intervention
6. Follow up
7. Assessment of outcome
 1.Drawing up a protocol

• It specifies aims, objectives, questions to be answered,

criteria or selection of study sample size.
• Protocol once formed, cannot be changed
• It reduces bias and sources of error
• Pilot study can be done in this step.
 2.Selecting reference population

• It is the population to which findings of the

trial, if found successful are expected to be
applicable.
• E.g. whole population or population of
school children
 3.Selecting experimental
population

• It is derived from reference population.

• It is actual population which participates in
study.
• They must be randomly selected from
reference population.
• Criteria for participants:

• Must give informed consent

• Should be representative of reference population

• Should be eligible for trial

 4. Randomization

• It is a statistical procedure by which participants are

allocated into groups usually called study and control
groups.
• Randomization is the heart of controlled trial.
• It will give the greatest confidence that the groups are
comparable so that like can be compared with like.
 5.Manipulation /intervention

• Manipulate the study group by the

deliberate application or withdrawal of
the suspected causal factor.
 6.Follow up

• It may be short or may require many years.

• It implies examinations of experimental and
control group at defined intervals of time
with same circumstances in a standard
manner.
 7.Assessment

• It is the final step of assessment of outcome.

• Positive results
• Benefits of experimental measure such as reduced incidence or severity of
disease
• Negative results
• No benefit but rather more side effects and complications
 Disadvantage of experimental
study

1. They are costly

2. Ethical problems exist
3. Non-feasibility
Non-Randomized Trials
 2.Non-Randomized Trials

• RCT is better one but it is not always possible to apply it.

So in following situations non-randomized trials are
applied.
▪ Uncontrolled trials
▪ Natural experiments
▪ Before and after comparison
 Comparison b/w observational and
experimental study

Observational Experimental
1. The study in which nature is allowed to 1. The study in which investigator
take its own course, investigation deliberately manipulation independent
measures but not intervene. variables.
2. Researcher simply observe different 2. Investigator intervene, giving the drug
patients who had taken or not taken the to one group of study but not other.
drug.
3. Ethical issues can be raised and
3. Ethical and practical impractical.
4. Not expensive 4. More expensive
 BIAS

• It is any factor which adversely

affects the outcome of study.
• It is a systematic error.
 Sources of Bias in observational
study

• Confounding
• Memory /recall bias
• Selection bias
• Interviewer's bias
• Design bias
• Sampling bias
 Sources of Bias in RCT

• Bias in evaluation
• Bias in observer
• Bias in subject
 Blinding
• Bias may be arise from errors of assessment
of the outcome due to human element.
• In order to reduce this problem a technique
known as blinding adopted.
 Types of blinding

• Single blind trial

• Double blind trial
• Triple blind trial
• Single blind trial
• This trial is so planned that the participant is not aware whether
he belongs to the study group or control group.
• Double blind trial
• This trial is so planned that neither the doctor nor the participant
is aware of the group allocation and the treatment received.
• Triple blind trial
• This is one step further.
• The participant, the investigator, the person analyzing the data are all blind.
• Ideally triple blinding should be used.
• But the double blinding most frequently used method.
 Palcebo and Nocebo Effect

• Patient is given inert substance during study but he starts

feeling well and thinks that he is given actual drugs, it is called
Placebo effect.
• Patient is given inert substance during study but he starts
feeling adverse effects of the drug and thinks that he is given
actual drugs, it is called Nocebo effect.
 The 2*2 contingency table

• A table that consists of two columns

(vertical) that represent the presence or
absence of disease and two rows (horizontal)
that represent a positive or negative test
result is called a 2-by-2 table.
 Uses
✓ 2-by-2 table may be employed in risk factor studies
(where rows represent the presence or absence of
risk factor)
✓ It may be employed in hypothesis testing (where
the rows represent the acceptance or rejection of
null hypothesis)
 Survey
• A canvass of selected people or households in a population usually used to
infer demographic characteristics or trends for a larger segment or all of the
population.
• It can be defined as an investigation of opinions ,behavior for a particular
group of people which is usually done by asking them questions.
• Open ended questionnaire
• A question on a survey that does not restrict the respondent to specific
choices ,but allows for a free response.
• Closed ended questionnaire
• A question on a survey that offers a set of
specific response choices those are mutually
exclusive and exhaustive.
 Gold Standard

• Any diagnostic procedure believed to identify diseased

person with certainty e.g biopsy for prostate cancer.
• It is a term used to describe a method ,procedure or measurement that is
widely accepted as being the best available.

• It is often used to compare with new methods.

 Association

• Concurrence of two variables more often

than would be expected by chance is
called association.
 Co-efficient of Correlation

• It indicates the degree of association b/w two

variables.
• Its symbol is “r”
• Its value ranges from -1 to +1.
 Interpretation of association
• When r value is equal or more than +1, it indicates a strong association b/w
x and y variables. when one variable increases other also increases.
• When r value is equal or less than -1, it indicates a strong negative
association b/w x and y variables. when one variable increases other
decreases.
• When r=0, it indicates that there is no association.
• When association is very strong, it is called
causation.
• e.g Myobacterium tuberculosis is a definite
cause of TB but smoking is not a definite
cause of lung cancer.
 Classification of association

1. Spurious association
2. Indirect association
3. Direct association (one –to-one causal & Multifactor
causation)
 Spurious association

• It means an observed association b/w a disease and suspected factor may

not be real.
• For example:
• Apparent perinatal mortality was higher in the hospital births than in the
home births. It might be concluded that homes are safer place fro delivery
than hospitals.
• Such conclusion is spurious. In general hospitals attract women at high risk
of delivery becoz of their special equipment and expertise.
 Indirect association

• It is statistical association b/w a variable of interest and a disease

due to the presence of another factor known or unknown, that is
common to both the characteristics and the disease.
• The third common factor is called confounding factor.
• Confounder is a variable associated with disease and exposure
and is distributed unequally among study and control groups and
it adversely affects outcome of study.
• High altitude & endemic goiter
• Endemic goiter is generally found in high altitudes, showing thereby an
association b/w altitude & endemic goiter. It shows how a common factor i.
e. iodine deficiency has resulted in an apparent association b/w two
variables ,when no association exists.
• Sucrose and CHD
 Direct (causal) association
• Cause is there & there comes effect.
• One –to-one causal relationship
• Two variables (A,B) are said to be causually related,if a change in A is followed by a
change in B. This model suggest that when the factor A is present The disease B Must
result.
• E.g. Mycobacterium tuberculosis TB
Scarlet fever
• Hemolytic streptococci Streptococcal tonsillitis
Erypsipelas
 Multi factorial causation
Model of multifactorial causation
Factor 1
• :

Reaction
Factor 2 at Factor 3
cellular
level

Disease
• There may be alternative factors each acting
independently – e.g smoking, air pollution,
and exposure to asbestos can produce lung
cancer independently.
Community medicine 11/11/2017
Community medicine 11/11/2017