Lecture No.

8
Research Methodology and Biostatistics Course

Research Method Part 1:

Study Design
Aryandhito Widhi Nugroho, M.D., Ph.D.
Neurosurgeon, Dr. H. Chasan Boesoirie General Hospital, North Maluku, Indonesia
Lecturer, Dept. of Surgery, Faculty of Medicine, Khairun University
Introduction
• The set of methods & procedures to collect & analyze variables
• Each types has its advantages & limitations
• Determined by:
• The nature of research question
• The goal of research
• The availability of resources
 Important Vocabularies (i)
Bias:
A systematic error that
introduces uncertainty
in estimates of
effect/association
Blind/double blind:
The participant or/and
researchers are
unaware of intervention
status
Confounding:
Complex relationships
among variables that
may distort
association/relationship

Cohort: Cross - Intention-to-treat:

A group of participants with sectional: Participants are analyzed in the
common characteristics At a single original assigned group regardless
followed over time point in time of procedures completion
 Important Vocabularies (ii)

Per protocol: Prospective: Protocol:

Only participants that A study in which A step-by-step plan that
adhered to the study information is collected details aspects of study
protocol are analyzed looking forward in time design & data collection

Randomization: Retrospective: Variables:

Assignment of different A study in which A measurable attribute:
th/to participants information is collected - Exposure/intervention
based on probability looking backward in time - Outcome
Research Study Designs
• Primary
• Descriptive
• Analytical
• Secondary
• Systematic reviews
• Meta-analysis
Descriptive Study
• Designed to describe the distribution of one or more variables,
without regard to any causal or other hypothesis
• Case report and series
• Description of a patient w/unrecognized, novel condition
• A case series ~ multiple (often only a few) similar cases
• Cross sectional studies
• Information regarding one variable of interest/more (exposure
or outcome), in a defined population at one particular time
• May be a descriptive or analytical study
Descriptive Study
• Ecological/correlational studies
• Looking for association between an exposure and an outcome
across populations rather than in individuals
• Convenient since the data have often already been collected
and are available from a reliable source
• Limitation:
• Association may not be true at individual level (fallacy)
• Association may be related to a 3rd factor (‘confounding’)
• Migration of people between regions with different
exposure levels may also introduce an error
Analytical Observational Study
• Association ~ exposure(s) & outcome(s) is to be established
• The direction of enquiry can be:
• Forward (cohort)
• Backward (case-control studies)
AOS: Cohort Studies
• “A group of people with a shared characteristic”
• Different groups of people with varying levels of exposure are
followed over time to evaluate the occurrence of an outcome
• The presence/absence of exposure in each subject is recorded
• Typically prospective studies
AOS: Cohort Studies Advantages
• Temporality (and therefore probable causality) can be established
• For a given exposure, >1 outcome can be studied
• Often several exposures can be studied simultaneously
AOS: Cohort Studies Limitations
• Often require a long duration of follow‑up to determine whether
outcome will occur or not
• It is not uncommon for >1 unknown confounding factors to
affect the occurrence of outcome
AOS: Cohort Studies Uses & Variations
Uses
• Greater validity compared to other observational study designs
• Outcome risk determination among the exposed & unexposed
• Alternative for experimental study

Variations
• Prospective cohort study
• Retrospective cohort study
AOS: Case-Control Study
• Starts with cases (w/outcome) and controls (w/out outcome)
• Then, tries to elicit a history of exposure in each group
• Identification of appropriate controls is a key element
• Controls = cases in all aspects, except for absence of disease
• Sometimes, controls are individually matched with cases for
factors (except for the exposure of interest), e.g., age, sex
• To minimize assessment bias, the person assessing the history of
exposure is blinded to whether the participant interviewed is a
case or a control
AOS: Case-Control Study Advantages
• Often cheaper and less time-consuming than cohort study
• It is often easy to study the relationship of outcome with not one
but several exposures
AOS: Case-Control Study Limitations
• Temporality (whether the outcome/exposure occurred first) is
often difficult to establish
• Bias in selecting cases or controls is possible
• Confounding factors are prevalent
• Determination of exposure relies existing records/history taking
• Determination of “risk” (RR/risk ratio) is impossible
• Only “odds” (OR) can be estimated
AOS: Case-Control Study Uses &
Variation
Uses
• Ideal for rare diseases
• Because of their simplicity, often becomes the initial design used
to assess the relationship of a exposure and an outcome

Variation
• Nested case-control study
• Only “odds” (OR) can be estimated
Interventional Studies (“Trials”)
• The researcher actively interferes with nature
• By performing an intervention in some or all study
participants
• To determine the effect of intervention exposure on the
natural course of events
• By nature, is a prospective study
• Differentiate carefully with prospective cohort study
• Two main types exist:
• Controlled clinical trials  individual assignment
• Community/field trials  groups assignment
Randomized Controlled Trials (RCTs)
• Participants fulfilling inclusion & exclusion criteria is “randomly”
assigned to 2 separate groups, each with different intervention
• Randomization  equal chance of being allocated to 2 groups
• Active/intervention/treatment group
• Placebo/control groups
• Controlled  presence of a concurrent comparator group
• Some additional methodological features to improve validity:
• Blinding
• Intention-to-treat
• Minimizing drop-outs
Non-Randomized Controlled Trials (RCTs)
• Participants fulfilling inclusion & exclusion criteria is “not-
randomly” assigned to 2 separate groups
• May be based on the investigator’s convenience or whether
the participant can afford a particular drug or not
• Susceptible to bias  results validity is lower than RCT
Interventional Studies Without Concurrent
Control
• Assignment of a group of persons to new intervention and
comparing their outcome with similar group followed up in the
past without the intervention (“historical controls”)
• High risk of bias
Factorial Study Design
• If > 2 interventions are available for a particular disease:
• Whether each drug is efficacious
• Whether a combination of the two is more efficacious than
either of them alone
• The study of two interventions in the same trial is possible
without increasing the required number of participants
• The study of interaction between the two treatments is possible
Cluster Randomized Trials
• Sometimes, an intervention cannot be easily administered to
individuals but can be applied to groups
• In such cases, a trial can be done by assigning “clusters” – some
logical groups of participants
• Necessitates larger sample size than individual‑randomized
studies and special statistical tools
Narrative Review VS Systematic Review
Narrative Review
• Used informal methods to identify the key studies on the topic
• The final review thus is a summary of these “selected” studies
• Potential for subjectivity or selection bias
Systematic Review
• Involving a formal prespecified protocol with explicit, transparent
criteria for the inclusion and exclusion of studies
• Ensuring completeness of coverage of the available evidence
• Providing a more objective and comprehensive overview
Blinding
Meta-Analysis
• The follow-up of systematic reviews
• Using a mathematical/statistical technique to pool the data from
individual studies included in the review
• To obtain a single summary measure of effect
The Preferred Reporting Items for Systematic
Reviews and Meta-Analyses (PRISMA)
Summary (i)
1. Does the study describe the characteristics of a sample OR
does it attempt to analyze the relationship between two
variables?
• If no, then it is a descriptive study
• if yes, it is an analytical

2. If analytical, did the investigator determine the exposure?

• If no, it is an observational study
• If yes, it is an experimental study
Summary (ii)
3. If observational, when was the outcome determined?
• At the start of the study (case–control study)
• At the end of a period of follow‑up (cohort study)
• Simultaneously (cross sectional)