Research is the question: It’s the
CLINICAL STUDY "what" and "why" you want to
DESIGNS IN learn or investigate.
DRUG
INFORMATION
Research design is the plan: It’s the map or
blueprint you follow to get your answers,
find people, and gather your facts.
Systematically: A study design is a
structured methodological framework that
guides how research is conducted,
ensuring validity, reliability and
reproducibility.
Exposure versus Intervention
Exposure: This is looking at the Intervention: It focuses on what
past or present. It focuses on the researcher will give people
what people are already doing specifically for the study to see
on their own or exposed to what happens next in the
currently or in the past (take a future (e.g., giving a medicine
prescribed medicine or have to investigate efficacy and/or
been using). safety)
Backward (Retrospective):
Capturing an outcome (result
or effect) in the present (e.g.,
a side effect) and then
looking back to find the cause
(exposure
Forward (Prospective):
Exposure Capturing an exposure (Use
of a medicine). Then
timeline following people to see what
happens next (to observe the
outcome).
Snapshot (Simultaneous):
Capturing exposure and
outcome at same time in the
present.
Observational study
design
• This design is to capture
exposure and outcome(s)
• The researcher observes an
outcome(s) without actively
applying an intervention to
participants.
Observational study design
B-Case series (descriptive mode): To
describe what a group of patients
A-Case report (descriptive mode): To
(e.g., with similar clinical
describe what single patient clinically
characteristics like type 2 diabetes)
experienced or exposed to
clinically experienced or what they
(symptoms, diagnosis, treatment,
were exposed to (e.g., use of the
and outcome).
same drug) without a comparison
control group.
Observational studies based on
exposure timeline
Design What capturing Direction
(observing)?
Case-control (analytical) Outcome (Effect) Backwards to find the
exposure (from now toward
the past)
Cohort (Analytical) Exposure (Cause) Forwards to find the
outcome (from now toward
future)
Cross-sectional (descriptive Exposure and outcome Currently (at the same time)
mostly, and/or analytical)
Observational studies based on exposure
timeline
C: Cross-sectional study design (The Snapshot)
Observat
• It measures exposures (e.g., taking Paracetamol) and
outcomes (e.g., liver hepatotoxicity) at the same time in
ionalone group of people.
• It could be descriptive (Common): To find the prevalence
studies
(How many people have a cough right now due to the
based onasking why.
use of lisinopril?). Thus, it describes what is happening
without
exposure
• Analytical (Sometimes): Used to compare groups within
the same sample (commonly no control group) at the
timeline
same time (e.g., comparing cough rates in people on
lisinopril vs. those on propranolol).
• Mixed mode (descriptive and analytical): Provides both
the prevalence and the associations [The link between
the medicine (exposure) and the side effect (outcome)].
Observational studies based on
exposure timeline
• D: Cohort study design (Analytical):
• The Concept: It enrolls a large group
(cohort) to find associations between an
exposure and an outcome over time (long
period: years, decades).
• It compares exposed individuals (e.g.,
patients taking Metformin) vs. non-exposed
individuals to observe who develops a
specific health issue (outcome) (e.g., lactic
acidosis).
• Based on the exposure timeline, the design
can be:
• I. Prospective (Looking Forward): Identify the
exposure now, wait to see the outcome in the
future.
• (e.g., Start tracking new Metformin users
today to see if they develop lactic acidosis by
2030).
• II. Retrospective (Looking back through
records): Identify the exposure in past records,
follow those records forward to the present to
see the outcome. (e.g., Using medical records
from 2010 to identify who started Statins, then
following their files up to 2026 to see if they
developed Diabetes).
Observational studies based on
exposure timeline
• E-Case-control study design
(Analytical)
• it starts with the outcome (the effect)
and looks backward to see if a prior
exposure (the cause) occurred.
• The participants are split into cases
(those with the disease/outcome) and
controls (those without it). Then the
researcher compares the past exposure
(e.g., medication use) between these
two groups.
• Example (Osteoporosis): Cases are
patients who already have fractures,
while controls are Similar patients who
do not have fractures. Then the
researcher compare how many people in
each group used corticosteroids in the
past.
• Observational designs are essential
for public health research when it is
not ethically permissible to assign
Application of participants to potentially harmful
observational exposures.
studies • Instead, researchers observe and
analyze data from individuals who
have already been exposed to certain
factors in their natural environments
to understand health risks.
Application of observational studies
Study design Best used For. Key strength
Cross-Sectional Prevalence (Snapshot) Fast and inexpensive; good for identifying
the "burden" of a disease right now.
Cohort Rare Exposures & Incidence Best for seeing the sequence of
events (proving exposure came first).
Case-Control Rare Diseases Perfect for outcomes that take a long
time to happen or are very rare.
Interventional design
• They are designs where the investigator actively
changes something (provide an intervention (new drug) to
investigate toxicity, safety, dose escalation and efficacy.
• The objective is to prove cause and effect (does Drug
A cause Result B?).
• Unlike observational studies (where you just observe
exposure), the researcher is the driver of the trial.
• It is the gold standard of evidence because it provides the
highest level of scientific evidence for drug safety and efficacy.
Interventional design
They have a robust internal validity because the design minimise the bias and control
confounders through:
1- Assignment of participants into two arms (interventional and control) (For pairwise
comparison of the efficacy). The control could be independent (A separate group gets a
placebo or standard care) or within-group (Comparing the participants of the same arm
pre- and post intervention).
2- Assignment of the participants randomly (By chance) to balance hidden factors (e.g.,
genetics, age, gender), ensuing balanced interventional and control group)
3- Aplplying blinding: which could be single (Patient doesn’t know what the
intervention and the assignment), double-blind (Neither the participants nor the
researchers know the intervention of the allocation) or triple-blind (The participant, the
researchers and the data analyst do not know the assignment and the intervention).
Phases of interventional trials
Phase Main Goal Participants Key Focus
Phase I Safety & Dosage 20–100 Healthy Volunteers Is it safe? What is the best dose?
How does the body process it
(PK)?
Phase II (RCT) Efficacy 100–300 Patients with the Does it work for the sick? Which
disease dose is most effective?
Phase III (RCT) Confirmation 1,000–3,000+ Diverse Is it better than the current
Patients standard? (Required for FDA
approval).
Phase IV Surveillance General Population (Millions) Long-term safety and rare side
effects in the "real world."
Quasi-Experimental
Design
• Quasi-experimental designs are
interventional studies that lack
random assignment and blinding,
making them ideal for scenarios
where RCTs are impractical,
unethical, or too costly. Participants
are assigned to intervention or
control groups based on non-random
criteria, allowing for pre-post
comparisons or control group
analyses