EVIDENCE-BASED
MEDICINE
Research is important because it is the foundation for knowledge,
problem-solving, innovation and development, evidence-based decision
making, and improving health and well-being.
Empirical research is a methodology in research in which data is collected
based on observations, measurements, or experiments to gather
knowledge and test hypotheses. It relies on observations of reality rather
than on theories or speculations. Can be quantitative or qualitative. It is important because it ensures that
scientific research is systematic, structured and an evidence-based approach to generate and validate
knowledge. It is essential to produce high-quality and reliable research. The empirical cycle starts with
observation, followed by induction, deduction, testing and evaluation.
Observation: An idea for proposing a hypothesis. The data is gathered using the observation.
Induction: The conclusion from the observation. It allows for generation of broader generalizations or
hypotheses.
Deduction: Means to draw a specific conclusion based on general principles or known facts. It is the
process of taking a general hypothesis and make specific predictions that can be tested (e.g., lower
morbidity, faster recovery, instead of ‘effective’)
Testing: Empirical methods to test the hypotheses through experiments, studies, or further
observations.
Evaluation: Analyse results and conclusions are drawn. The theory may be refined, accepted or
rejected. The cycle can begin again if new observations are made.
Advantages empirical research Disadvantages empirical research
Evidence-based: observable and measurable data, Time-consuming: Collecting data can take a
providing solid evidence for conclusions significant amount of time
Objective: Minimizes personal bias by focusing on Costly: requires significant resources, including
real-world data rather than subjective opinions funding for equipment, participants, and personnel
Reproducible: Other researchers can replicate Limited scope: Results are often specific to the
studies, which helps validate and strengthen studied population, setting, or time period, limiting
findings generalizability
Dynamic process: Allows for continuous refinement Ethical concerns: Involving human participants or
of hypotheses and theories as more data is gathered animal testing can raise ethical questions
Practical application: Provides actionable insights Risk of uncontrolled variables: uncontrolled factors
that can be directly applied to solve real-world may influence results, reducing reliability
problems
EVIDENCE-BASED MEDICINE
Evidence-based medicine is medicine practiced based on the basis of high-quality evidence. The fundamental
basis are clinical experience, patient values and best research evidence. 5 steps for applying EBM are:
1. Ask: Define a clinically relevant question
2. Acquire: Search for the best evidence
3. Appraise: Assess quality of the evidence
4. Apply: Act on the evidence to form a clinical decision
5. Assess: Evaluate the process
Achievements of EBM are the Cochrane library, clinical guidelines, full disclosure of clinical trial data, increased
awareness for overdiagnosis, and medical ethics (Tuskegee study). The Tuskegee study was conducted
between 1932 and 1972 amongst 400 African American men with syphilis. The purpose of the study was to
observe the effects of the disease when untreated. The men were not informed of the nature of the
experiment, and more than 100 died as a result. By the end of the study, medical advancement meant that
syphilis was entirely treatable.
Page | 1
,However, there were also troubling aspects of EBM. EBM movements argued that medical practice was too
subjective and sought greater reliance on published research. In doing so, it proposed that clinical judgement
and mechanistic reasoning are less reliable form of evidence in medicine. This view is abandoned, but
continuous to exert influence. Criticism and pitfalls of EBM are:
The curse of p-values
The absence of N-of-1 trials
Internal and external validity within systematic reviews
Clinical expertise and patient values are increasing being sidelines
Reimbursements and incentives tied to guidelines
Physician’s ignorance of statistics
The PICO-model is used in EBM to formulate the research question:
P: Patient or Population of interest
I: Intervention or Interest
C: Comparison intervention / group
O: Outcome
T: Time of observation
T: Type of question / study
In the research question, there is no hypothesis. So, what is the effect on recovery time, not whether it is
faster or not.
CLINICAL STUDY DESIGN
The basic components of a research study are:
1. Idea: what is the research problem? Make a research proposal
2. What is already known: review the literature
3. Making research questions, objectives, and hypotheses
4. Choose the study design
5. How many participants? Calculating on the sample size
6. Collecting data
7. Process and analyse the data
8. Write the research report –Disseminate and utilise results
Requirements needed for research are money, personnel, time, suitable participants, and patience.
Empirical research can be quantitative, qualitative, or a mix of both:
Quantitative: focuses on numerical data, statistical analyses, aims to quantify relationships, test
hypotheses, make generalisations.
o Experimental studies, cohort studies, surveys (closed questions)
Qualitative: Focuses on non-numerical data often semi- or unstructured methods(text, images,
narratives), aims to explore, understand and interpret complex phenomena surveys, interviews, etc.)
o In-depth (individual) interviews, focus groups, open text surveys
Page | 2
, Mixed methods: combines quantitative and qualitative approaches, aims to provide a more
comprehensive understanding
o Sequential (one after another), convergent (used concurrently)
In empirical research, there are lots of words to say the same thing, which can be confusing. For example,
quantitative is experimental and interventional, whereas analytical is non-interventional and observational.
Furthermore, sex and gender are not the same thing. Sex is based on biology and gender is based on norms
and behaviours. This is important because 70% of research is performed on white males or male animals.
Protocols, guidelines and therapies are based on this research.
DESCRIPTIVE STUDY DESIGNS
An observational ecological study design is a type of research in which scientists analyse the relationship
between environmental factors and health outcomes within entire populations or groups, rather than focusing
on individual participants. It uses group level or aggregate level data, to find a potential association between 2
variables. An observational individual study design is a type of research in which the focus is on individual
participants rather than groups or populations. In observational studies, there is a risk of the Hawthorne effect,
which is that people know they are being watched and change their behaviour.
Observational ecological Observational individual
Easy to conduct, inexpensive Individual analysis
Study large populations Establish temporal relationships
Pro Examine associations at population level (exposure outcome)
Control confounding factors at individual
level
Limited control over confounding factors Smaller sample sizes
Ecological fallacy (incorrectly inferring Potential for selection bias
Con individual-level relationships from group May be more expensive and time-
data) consuming than ecological
No temporality, can’t establish causality
A cross-sectional study design is an observational study where researchers collect data on both exposures and
outcomes at a single point in time across a specific population or sample. It provides a "snapshot" of the
population, showing the prevalence of particular characteristics, conditions, or behaviour.
A case report is a detailed description of a single patient’s medical history, symptoms, diagnosis, treatment,
and follow-up. It’s typically written to highlight a unique or noteworthy case that may provide insights for
further research or clinical practice. A case series is a collection of multiple case reports that document the
characteristics of a group of patients with a similar condition, treatment, or exposure. Unlike case reports, case
series allow for a broader look at patterns among cases and can help identify trends or common features
within the group. Case reports/series usually concern rare diseases or things that do not happen often. They
are very low in the pyramid of evidence.
Cross-sectional Case report(s)
Relatively quick and inexpensive Detailed information on rare or novel
Study multiple outcomes simultaneously diseases
Pro Can determine prevalence at a certain Useful for hypothesis generation
time Relatively easy to conduct
Hypothesis generation possible
Not for rare diseases/ rapidly changing Can’t establish causality
Potential for selection bias and Findings not generalisable
Con
confounding Potential publication bias
No temporality, can’t establish causality
ANALYTICAL NON-INTERVENTIONAL REAL-WORLD STUDY DESIGNS
A cross-sectional study can also be analytical, by which they take 2 snapshots at different timepoints and
compare them.
Page | 3
MEDICINE
Research is important because it is the foundation for knowledge,
problem-solving, innovation and development, evidence-based decision
making, and improving health and well-being.
Empirical research is a methodology in research in which data is collected
based on observations, measurements, or experiments to gather
knowledge and test hypotheses. It relies on observations of reality rather
than on theories or speculations. Can be quantitative or qualitative. It is important because it ensures that
scientific research is systematic, structured and an evidence-based approach to generate and validate
knowledge. It is essential to produce high-quality and reliable research. The empirical cycle starts with
observation, followed by induction, deduction, testing and evaluation.
Observation: An idea for proposing a hypothesis. The data is gathered using the observation.
Induction: The conclusion from the observation. It allows for generation of broader generalizations or
hypotheses.
Deduction: Means to draw a specific conclusion based on general principles or known facts. It is the
process of taking a general hypothesis and make specific predictions that can be tested (e.g., lower
morbidity, faster recovery, instead of ‘effective’)
Testing: Empirical methods to test the hypotheses through experiments, studies, or further
observations.
Evaluation: Analyse results and conclusions are drawn. The theory may be refined, accepted or
rejected. The cycle can begin again if new observations are made.
Advantages empirical research Disadvantages empirical research
Evidence-based: observable and measurable data, Time-consuming: Collecting data can take a
providing solid evidence for conclusions significant amount of time
Objective: Minimizes personal bias by focusing on Costly: requires significant resources, including
real-world data rather than subjective opinions funding for equipment, participants, and personnel
Reproducible: Other researchers can replicate Limited scope: Results are often specific to the
studies, which helps validate and strengthen studied population, setting, or time period, limiting
findings generalizability
Dynamic process: Allows for continuous refinement Ethical concerns: Involving human participants or
of hypotheses and theories as more data is gathered animal testing can raise ethical questions
Practical application: Provides actionable insights Risk of uncontrolled variables: uncontrolled factors
that can be directly applied to solve real-world may influence results, reducing reliability
problems
EVIDENCE-BASED MEDICINE
Evidence-based medicine is medicine practiced based on the basis of high-quality evidence. The fundamental
basis are clinical experience, patient values and best research evidence. 5 steps for applying EBM are:
1. Ask: Define a clinically relevant question
2. Acquire: Search for the best evidence
3. Appraise: Assess quality of the evidence
4. Apply: Act on the evidence to form a clinical decision
5. Assess: Evaluate the process
Achievements of EBM are the Cochrane library, clinical guidelines, full disclosure of clinical trial data, increased
awareness for overdiagnosis, and medical ethics (Tuskegee study). The Tuskegee study was conducted
between 1932 and 1972 amongst 400 African American men with syphilis. The purpose of the study was to
observe the effects of the disease when untreated. The men were not informed of the nature of the
experiment, and more than 100 died as a result. By the end of the study, medical advancement meant that
syphilis was entirely treatable.
Page | 1
,However, there were also troubling aspects of EBM. EBM movements argued that medical practice was too
subjective and sought greater reliance on published research. In doing so, it proposed that clinical judgement
and mechanistic reasoning are less reliable form of evidence in medicine. This view is abandoned, but
continuous to exert influence. Criticism and pitfalls of EBM are:
The curse of p-values
The absence of N-of-1 trials
Internal and external validity within systematic reviews
Clinical expertise and patient values are increasing being sidelines
Reimbursements and incentives tied to guidelines
Physician’s ignorance of statistics
The PICO-model is used in EBM to formulate the research question:
P: Patient or Population of interest
I: Intervention or Interest
C: Comparison intervention / group
O: Outcome
T: Time of observation
T: Type of question / study
In the research question, there is no hypothesis. So, what is the effect on recovery time, not whether it is
faster or not.
CLINICAL STUDY DESIGN
The basic components of a research study are:
1. Idea: what is the research problem? Make a research proposal
2. What is already known: review the literature
3. Making research questions, objectives, and hypotheses
4. Choose the study design
5. How many participants? Calculating on the sample size
6. Collecting data
7. Process and analyse the data
8. Write the research report –Disseminate and utilise results
Requirements needed for research are money, personnel, time, suitable participants, and patience.
Empirical research can be quantitative, qualitative, or a mix of both:
Quantitative: focuses on numerical data, statistical analyses, aims to quantify relationships, test
hypotheses, make generalisations.
o Experimental studies, cohort studies, surveys (closed questions)
Qualitative: Focuses on non-numerical data often semi- or unstructured methods(text, images,
narratives), aims to explore, understand and interpret complex phenomena surveys, interviews, etc.)
o In-depth (individual) interviews, focus groups, open text surveys
Page | 2
, Mixed methods: combines quantitative and qualitative approaches, aims to provide a more
comprehensive understanding
o Sequential (one after another), convergent (used concurrently)
In empirical research, there are lots of words to say the same thing, which can be confusing. For example,
quantitative is experimental and interventional, whereas analytical is non-interventional and observational.
Furthermore, sex and gender are not the same thing. Sex is based on biology and gender is based on norms
and behaviours. This is important because 70% of research is performed on white males or male animals.
Protocols, guidelines and therapies are based on this research.
DESCRIPTIVE STUDY DESIGNS
An observational ecological study design is a type of research in which scientists analyse the relationship
between environmental factors and health outcomes within entire populations or groups, rather than focusing
on individual participants. It uses group level or aggregate level data, to find a potential association between 2
variables. An observational individual study design is a type of research in which the focus is on individual
participants rather than groups or populations. In observational studies, there is a risk of the Hawthorne effect,
which is that people know they are being watched and change their behaviour.
Observational ecological Observational individual
Easy to conduct, inexpensive Individual analysis
Study large populations Establish temporal relationships
Pro Examine associations at population level (exposure outcome)
Control confounding factors at individual
level
Limited control over confounding factors Smaller sample sizes
Ecological fallacy (incorrectly inferring Potential for selection bias
Con individual-level relationships from group May be more expensive and time-
data) consuming than ecological
No temporality, can’t establish causality
A cross-sectional study design is an observational study where researchers collect data on both exposures and
outcomes at a single point in time across a specific population or sample. It provides a "snapshot" of the
population, showing the prevalence of particular characteristics, conditions, or behaviour.
A case report is a detailed description of a single patient’s medical history, symptoms, diagnosis, treatment,
and follow-up. It’s typically written to highlight a unique or noteworthy case that may provide insights for
further research or clinical practice. A case series is a collection of multiple case reports that document the
characteristics of a group of patients with a similar condition, treatment, or exposure. Unlike case reports, case
series allow for a broader look at patterns among cases and can help identify trends or common features
within the group. Case reports/series usually concern rare diseases or things that do not happen often. They
are very low in the pyramid of evidence.
Cross-sectional Case report(s)
Relatively quick and inexpensive Detailed information on rare or novel
Study multiple outcomes simultaneously diseases
Pro Can determine prevalence at a certain Useful for hypothesis generation
time Relatively easy to conduct
Hypothesis generation possible
Not for rare diseases/ rapidly changing Can’t establish causality
Potential for selection bias and Findings not generalisable
Con
confounding Potential publication bias
No temporality, can’t establish causality
ANALYTICAL NON-INTERVENTIONAL REAL-WORLD STUDY DESIGNS
A cross-sectional study can also be analytical, by which they take 2 snapshots at different timepoints and
compare them.
Page | 3
