EVIDENCE BASED MEDICINE

Sef Lucrari Dr. Anna Marie Herghelegiu

Institutul National de Gerontologie si Geriatrie
“Ana Aslan”
 Clinical Patients’
Preferences
Circumstance

Research
Evidence

Clinical
Scientific Decision
Expertise Making

HEALTH
Evidence-based medicine = the process of systematically finding, appraising, and using
contemporaneous scientific research as the basis for clinical decisions.

This approach places less value on physiologic principles and clinical experience, instead
shifting the emphasis to
● systematic obser vation,
● understanding rules of evidence,
● systematic research and
● critical interpretation of literature.

Kyriacou DN. Evidence-based medical decision making: deductive versus inductive logical thinking. Acad Emerg Med. 2004 Jun;11(6):670-1.
 OBM EBM EVIDENCE BASED MEDICINE

● David M. Eddy - 1980, American Cancer Society - developed recommendations for

screening for cancer

- 1982, New England Journal of Medicine - the role clinical policies

and guidelines play in medical decision-making; explicit analysis of
evidence and estimation of outcomes

● Gordon Guyatt - 1992, JAMA - a new approach to teaching the practice of medicine
EBM coined in: Guyatt GH. Evidence-based medicine. ACP J Club. 1991;114:A–16.
● David Sackett - 1996 - “Evidence based medicine is the conscientious, explicit, and
judicious use of current best evidence in making decisions about the care of individual
patients. The practice of evidence based medicine means integrating individual clinical
expertise with the best available external clinical evidence from systematic research.”
(From BMJ 1996;312:71-72)

● Trisha Greenhalgh - 2010 - "the use of mathematical estimates of the risk of benefit
and harm, derived from high-quality research on population samples, to inform clinical
decision-making in the diagnosis, investigation or management of individual patients.“
(Greenhalgh, Trisha (2010). How to Read a Paper: The Basics of Evidence-Based
Medicine (4th ed.))
■ By integrating specialized medical education with clinical practice, evidence-based
medicine is promoted as a more efficient and effective method of clinical decision
making.

■ In the last decades evidence based medicine has been accepted as a method
logically superior to traditional medicine, and is a part of the specialized educational
systems.
 PATIENT CENTERED MEDICINE

Hudon C, Fortin M, Haggerty JL, Lambert M, Poitras ME. Measuring

patients' perceptions of patient-centered care: a systematic review
of tools for family medicine. Ann Fam Med. 2011 Mar-
Apr;9(2):155-64
PATIENT CENTERED MEDICINE

“A patient consults an orthopedist because of knee pain. The surgeon determines that no
operation is indicated and refers her to a rheumatologist, who finds no systemic inflammatory
disease and refers her to a physiatrist, who sends her to a physical therapist, who administers the
actual treatment. Each clinician has executed his or her craft with impeccable authority and skill,
but the patient has become a shuttlecock. Probably a hassled, frustrated, and maybe bankrupt
shuttlecock.”

Charles L. Bardes, M.D. Defining “Patient-Centered Medicine” N Engl J Med 2012; 366:782-783
 EBM

■ current best Evidence

■ clinical Expertize
■ patient’s values,
preferences and
circumstances
 5 As of EBM
Assess Ask
Incorporating best evidence into clinical
care requires a systematic approach in
order to be manageable.

A clear series of steps known as the

Evidence Cycle can provide an excellent
paradigm to guide you through this
process.
Apply Acquire
The foundation of evidence-based care
remains an excellent clinical evaluation.

Appraise
■ ASSESS the patient and the problem to determine the pertinent issues, which may include a
differential diagnosis, treatment decisions, or prognosis. A comprehensive understanding of
pathophysiology and the thorough history and physical remain a critical starting point for the process.

■ ASK a clear, answerable question to be pursued. The first critical step is to clarify one or two key issues
that come up in the course of caring for your patient and to develop a focused clinical question.

■ ACQUIRE the evidence from an appropriate source. Potential sources include original research studies,
systematic reviews, evidence-based journal abstracts, textbooks and computerized decision support
systems.

■ APPRAISE the evidence to further examine its worth and reliability.

■ Finally, the process must conclude by returning to the individual patient, as the clinician has to decide
whether it is appropriate to APPLY the evidence to the particular patient and their unique values and
circumstances. Evidence alone is never sufficient to direct decision making. Rather, it must be put into
context with a patient’s values.
We need to consider how to generalize the results from clinical trials to our individual patient. We must
consider whether the patient populations and treatments or interventions are comparable to our
setting.
ASK
■ The first stage of EBM is to ask a clearly focused question, because it makes it much easier to
find a reliable answer. Asking focused questions is directly relevant to patients’ problems and
phrased to direct your search to relevant and precise answers.
1 2 3 4
Intervention Comparison
Patient or Problem (a cause, prognostic Intervention Outcomes
factor, treatment, etc.) (if necessary)

Starting with your patient, Ask “What can I hope

ask “How would I describe a
group of patients similar to
Tips for Building
mine?”
Balance precision with
concision.
Ask “What is the main
Ask “Which main to accomplish?” or
alternative to compare
intervention am I “What could this
with the
considering?”Be exposure really
intervention?”Again, be
specific. affect?”Again, be
specific.
specific.

“… lead to lower
“… would adding
mortality or morbidity
“In patients with heart failure anticoagulation with “… when compared
from thromboembolism.
Example from dilated cardiomyopathy warfarin to standard with standard therapy
Is this enough to be
who are in sinus rhythm …” heart failure therapy alone …”
worth the increased risk
…”
of bleeding?”
ASK
Once you have formed the question using the PICO structure, you can think about what type
of question it is you are asking, and therefore what type of research would provide the best
answer.
 ASK (PICO) In older patients (>= 65 years of age) with knee joint pain does
physiotherapy reduce level of pain compared to NSAI medication ?

■ Population/Patient/Problem P: older patients

I: physiotherapy
■ Intervention C: NASI medication
■ Control/Comparison O: pain level

■ Outcome
Which is the best method of reducing knee joint pain in
older patients ?

P: older patients - define population

I: ?
C: ?
O: pain - measure

Ask: Ask an answerable clinical question based on the health needs of a specific patient
(may use the PICO framework: Patient, Intervention, Comparison, Outcome).
ACQUIRE Finding the Evidence

■ Finding the evidence is the second step in the five steps of evidence-based practice.
It is important when searching for evidence that search terms are referred back to your
original PICO question. The process of finding evidence therefore follow three key steps:
► Identify terms to fit your PICO question
► Look for secondary sources
► Search for primary sources
ACQUIRE Finding the Evidence

■ Look for secondary sources

● Guidelines: UK National Library for Health, NICE, SIGN; US National Guidelines Clearinghouse;
Canadian Medical Association; New Zealand Guidelines Group.
● Critically Appraised Topics (CATs): CAT Crawler
● Evidence-Based Summaries: Bandolier, Clinical Evidence
● Structured Abstracts: EBM Online, ACP Journal Club
● Systematic Reviews: Cochrane Library
● To search several of the databases simultaneously you can use: www.tripdatabase.com
ACQUIRE Finding the Evidence

■ Search for Primary Sources

Use methodological filters to target the right type of study. For instance, PubMED filters for:
● therapy
● diagnosis
● prognosis
● aetiology
APPRAISE

■ Critical appraisal is the systematic evaluation of clinical research papers in order to

establish:

► Does this study address a clearly focused question?

► Did the study use valid methods to address this question?
► Are the valid results of this study important?
► Are these valid, important results applicable to my patient or population?
 APPLY / MAKING A DECISION
■ When making decisions about whether to treat patients with therapeutic interventions, we must take into
account relevant information from randomised controlled trials (RCTs) or systematic reviews (SRs).

■ The integration of relevant evidence with clinical experience forms the cornerstone of evidence based-
practice.

■ To asses evidence of effectiveness of a therapeutic intervention there are two important concepts that need
to be addressed and understood. These are the internal and external validity of a given study.

■ External validity refers to the extent to which we can generalize the results of a trial to the population
of interest. Internal validity refers to the extent a study properly measures what it is meant to.
APPLY / MAKING A DECISION

■ Define the population and intervention

■ Search for and understand the biases
■ What is the summary statistic chosen to report the outcome?
■ Interpret the findings and apply them to your patient
CLINICAL GUIDELINES
PATIENTS
CLINICAL GUIDELINES
PUBLIC HEALTH
 GUIDELINES
■ based on a systematic review of existing evidence
■ developed by a knowledgeable, multidisciplinary panel of experts and
representatives from affected groups
■ consider patients subgroups and patient preferences
■ based on an explicit, transparent process that minimizes distortions, biases and
conflicts of interest
■ provides clear explanation of the logical relationship between alternative care
options and health outcomes and provide ratings of both quality of evidence and
the strength of recommendations
■ revised as appropriate when important new evidence warrants modification of
recommendations
Expert opinion: Classical textbooks
recommendations
why grade recommendations ?

■ strong recommendations
● strong methods
● large precise effect
● few downsides of therapy
■ weak recommendations
● weak methods
● imprecise estimates
● small effect
● substantial downsides
What is GRADE?
■ GRADE is a systematic and explicit approach to making judgements about quality of evidence
and strength of recommendations.
■ It was developed by the Grading of Recommendations, Assessment, Development and
Evaluations (GRADE) Working Group, and it is now widely seen as the most effective method
of linking evidence-quality evaluations to clinical recommendations.

Oxford Center for Evidence based Medicine
Scottish Intercollegiate Guidelines (SIGN)
US Preventative Services Task Force
American Professional Organizations
- AHA/ACC, ACCP, AAP, Endocrine Society etc
etc
common international grading system
Australian NMRC, SIGN, USPSTF, NICE, WHO,
Oxford CEBM, CDC, CC
high concentration of methodologists
frequent annual meetings
Code Quality of Evidence Definition

Further research is very unlikely to change our confidence in the estimate of effect.
A High •Several high-quality studies with consistent results
•In special cases: one large, high-quality multi-centre trial

Further research is likely to have an important impact on our confidence in the estimate of
effect and may change the estimate.
B Moderate
•One high-quality study
•Several studies with some limitations

Further research is very likely to have an important impact on our confidence in the estimate of
C Low effect and is likely to change the estimate.
•One or more studies with severe limitations

Any estimate of effect is very uncertain.

•Expert opinion
D Very Low
•No direct research evidence
•One or more studies with very severe limitations
Sources of Knowledge

■ Authority
■ Custom and tradition
■ Personal experience
■ Deductive reasoning
■ Scientific enquiry
Acquiring knowledge

■ RATIONALISM
■ EMPIRICISM
■ SKEPTICISM
■ TRANSCENDENTAL IDEALISM
■ REALISM
■ SCIENCE
■ science

a systematic enterprise that builds and organizes knowledge in the form of testable
explanations and predictions about the universe

"... modern science is a discovery as well as an invention. It was a discovery that nature
generally acts regularly enough to be described by laws and even by mathematics; and
required invention to devise the techniques, abstractions, apparatus, and organization for
exhibiting the regularities and securing their law-like descriptions.“
Heilbron, J. L. (editor-in-chief) (2003). The Oxford Companion to the History of Modern Science. New York: Oxford University Press
Roger Bacon (c. 1219/20 – c. 1292)

credited (mainly since the 19th century) as one of the earliest European advocates of the
modern scientific method inspired by Aristotle and by later scholars such as the Arab
scientist Alhazen

Galileo Galilei (1564 - 1642)

the “father” of science and scientific method

Research is either discovery of new facts, enunciation of new principles,
or fresh interpretation of the known facts or principles.

a systematic investigation towards increasing the sum of knowledge

Research is an organized and systematic way of finding answers to questions.

Scientific method

When conducting research, scientists use the scientific method to collect

measurable, empirical evidence in an experiment related to a hypothesis
(often in the form of an if/then statement), the results aiming to support
or contradict a theory.

draw conclusion about the target population from results of a sample

repeat the experiment (reproducibility)

Statistics

Statistics deals with all aspects of data including the planning of data
collection in terms of the design of surveys and experiments
Dodge, Y. (2006) The Oxford Dictionary of Statistical Terms, OUP.

Statistics is a mathematical body of science that pertains to the collection,

analysis, interpretation or explanation, and presentation of data
Moses, Lincoln E. (1986) Think and Explain with Statistics, Addison-Wesley, ISBN 978-0-201-15619-5 . pp. 1–3

Statistics is concerned with the use of data in the context of uncertainty and
decision making in the face of uncertainty
Moore, David (1992). "Teaching Statistics as a Respectable Subject". In F. Gordon and S. Gordon. Statistics for the Twenty-First Century.
Washington, DC: The Mathematical Association of America. pp. 14–25.
Statistical inference
Statistical inference is the process of deducing properties of an underlying distribution
by analysis of data.
Upton, G., Cook, I. (2008) Oxford Dictionary of Statistics, OUP.

probability distribution is a mathematical function that, stated in simple terms, can be

thought of as providing the probability of occurrence of different possible outcomes in
an experiment (e.g tossing a coin or a dice)
Inferential statistical analysis infers properties about a population: this includes testing
hypotheses (confirmatory data analysis) and deriving estimates. The population is
assumed to be larger than the observed data set; in other words, the observed data is
assumed to be sampled from a larger population.

Descriptive statistics (exploratory data analysis) is solely concerned with properties of the
observed data, and does not assume that the data came from a larger population.
 Theory

Hypothesis

Deduction
Observation

Confirmation

Theory

Tentative
Induction
Hypothesis

Pattern

Observation
Research Methodology

Planning Conducting Publishing

Research Planning

■ assemble a research team

■ ask a good research question
■ do a literature search
■ formulate the hypothesis and objectives
■ chose the appropriate study design
■ determine the sampling design
■ obtain ethical approval
Steps - Scientific Method

■ Make an observation or observations.

■ Ask questions about the observations and gather information.
■ Form a hypothesis — a tentative description of what’s been observed, and make
predictions based on that hypothesis.
■ Test the hypothesis and predictions in an experiment that can be reproduced.
■ Analyze the data
■ Interpret de data and draw conclusions; accept or reject the hypothesis or modify the
hypothesis if necessary.
■ Reproduce the experiment until there are no discrepancies between observations
and theory.
Garland, Jr., Theodore. "The Scientific Method as an Ongoing Process". U C Riverside.
PREINVESTIGATION STEPS

● IDENTIFY THE PROBLEM

For selection, match the research area to

(i) relevance and applicability for improving health in one way or the other,
(ii) interest and expertise of you and your collaborators,
(iii) the feasibility of completing the work with available resources, time, subjects, tools, etc.
● COLLECT AND EVALUATE EXISTING INFORMATION

Make sure that a reasonable answer to the proposed question is not already available.

The objective of all this exercise should be to identify the specific information gaps, and to
examine how the problem fits into the medical jigsaw puzzle.

Assess if the problem is really worth pursuing. If none or very little baseline information is
available, consider carrying out an exploratory study as a first step.
 What makes a good research question?

■ “Make sense”: In other words, you must clearly define your terms using known definitions outlined
in the literature.
For example, a poor research question would be: How do people’s lives improve after surgery? Not
only does this research question fail to specify the study population, it contains the vague term
“improve”. The researcher must specify what he/she means by this term—does it involve a physical
improvement or rather an improvement in mental state? The more specific your research question, the
better.

■ Address an important and relevant issue: A good research question will also always have relevance
to the time, place, and population of the study.
For example, a study of Vitamin A deficiency in Southern India would be a poor choice as this is not a
particularly significant problem in the area.
What makes a good research question?

■ Not already have been done: A good research study will be novel. When replicating a
pervious study, it is best to add or change one or two things to increase the novelty of the
research.

■ Be “operationalizable”: It should be assessed methodologically with research instruments.

■ Be within a reasonable scope: In general, the more focused the research question the more
likely it will be a successful project.

For example, a study that seeks to identify the prevalence of eye disease in a specific village is
more likely to succeed than a comparable study that seeks to identify eye disease prevalence in
the world population.
 ASK (PICO) In older patients (>= 65 years of age) with knee joint pain does
physiotherapy reduce level of pain compared to NSAI medication ?

■ Population/Patient/Problem P: older patients

I: physiotherapy
■ Intervention C: NASI medication
■ Control/Comparison O: pain level

■ Outcome
Which is the best method of reducing knee joint pain in
older patients ?

P: older patients - define population

I: ?
C: ?
O: pain - measure

Ask: Ask an answerable clinical question based on the health needs of a specific patient
(may use the PICO framework: Patient, Intervention, Comparison, Outcome).
FINER criteria for a good research question

F Feasible •Adequate number of subjects

•Adequate technical expertise
•Affordable in time and money
•Manageable in scope
I Interesting •Getting the answer intrigues investigator, peers and
community

N Novel •Confirms, refutes or extends previous findings

E Ethical •Amenable to a study that institutional review board will

approve

R Relevant •To scientific knowledge

•To clinical and health policy
•To future research
■ Hypothesis

scientific hypothesis = something that can be supported or refuted through carefully

crafted experimentation or observation (falsifiable and testable)

a prediction of the final outcome of the research

a concrete specific statement about the relationship between phenomena

the primary trait of a hypothesis is that something can be tested and that those tests can be
replicated - deriving predictions from the hypotheses about the results of future experiments, and
then performing those experiments to see whether they support the predictions
 HYPOTHESIS TESTING
RESEARCH STUDY CHARACTERISTICS
 Statistical inference

■ Bayesian inference

- probability of an event based on prior knowledge of conditions that might be related to the event
- probability measures a “degree of belief” - before and after accounting for evidence

■ Fiducial inference - historical value

■ Frequentist inference
- draws conclusions from sample data by emphasizing the frequency or proportion of the data
- probability measures a “proportion of outcomes”

■ Decision theory (or the theory of choice)

Statistical hypothesis testing
Accuracy/Precisio
n
■ Making an initial assumption -
Reliability/Validity
1. Define the null and alternate hypotheses under study /Consistency
Random
■ Collecting evidence (data) - error/Bias

2. Collect relevant data from a sample of individuals

Data
characteristics
■ Compare sets of data -
3. Calculate the value of the test statistics specific to the null hypothesis

Type I Error/
■ Decide whether to reject the null hypothesis Type II Error
4. Decide! - interpretation of probability (frequentist, critical value, p -value)
■ most frequent technique in statistical inference is
null hypothesis testing/ test of significance

■ recommended technique is estimation statistics

Testable

a property of an empirical hypothesis:

● counterexamples to the hypothesis are logically possible

● observing a reproducible series of such counterexamples is practical and
feasible

Falsifiable

● the intrinsic possibility that the hypothesis can be proven false

Scientific Hypothesis

The Scientific Hypothesis has to be testable and falsifiable !

Karl Popper

originally proposed testability and falsifiability as important criteria that

distinguish a scientific hypothesis from a common hypothesis ("educated guess“)
no number of experiments can ever prove a hypothesis, but a single experiment can contradict one

falsificationism thus strives for questioning, for falsification, of hypotheses instead of proving them

“My proposal is based upon an asymmetry between verifiability and falsifiability; an

asymmetry which results from the logical form of universal statements. For these are never
derivable from singular statements, but can be contradicted by singular statements.”

— Karl Popper, The Logic of Scientific Discovery,1959

In his Logic of Scientific Discovery (1959) Popper used a "black swan" to show how scientific
ideas can never be proven true, regardless of how many observations appear to support it.
However, a single contrary result can prove a theory to be false.

"[Popper] begins by pointing to a logical asymmetry between verification and

falsification. To express it in terms of the logic of statements: although no number of
observation statements reporting observations of white swans allows us to derive the
universal statement 'All swans are white', reporting one single observation of a black
swan, allows us to logically derive the statement 'Not all swans are white.' In this
important logical sense empirical generalizations, though not verifiable, are falsifiable.
This means that scientific laws are testable in spite of being unprovable, they can be
tested by systematic attempts to refute them." (Magee,Popper.Routledge.1974.pp 22)
Are all swans
white?
Are all swans Null hypothesis:
white? Not all swans are white Always Assumed True!

Alternate hypothesis:
All swans are white Probable!

SAMPLE!
■ in an European sample all swans were found white
the null hypothesis is REJECTED
the alternate hypothesis is ACCEPTED WITH A CERTAIN PROBABILITY - DEGREE OF BELIEF

■ the null hypothesis is either TRUE (always assumed true) or FALSE (rejected by the experiment)
NOT PROBABLE

■ only the alternate hypothesis can be PROBABLE (remember we make a prediction!)

NULL HYPOTHESIS

the term "null hypothesis" usually refers to a general statement or default position that
there is no relationship between two measured phenomena, or no association among
groups

rejecting or disproving the null hypothesis—and thus concluding that there are grounds
for believing that there is a relationship between two phenomena (e.g. that a potential
treatment has a measurable effect)—is a central task in the modern practice of
science, and gives a precise criterion for rejecting a hypothesis.
Null / Alternate hypothesis - mutually exclusive !!

■ "null hypothesis" - default position - there is no relationship between two measured

phenomena, or no association among groups

describes the present state of knowledge

the null hypothesis is always true or always false!

■ the null hypothesis is generally assumed to be true until evidence indicates otherwise
In statistics, it is often denoted H0 (read “H-naught”, "H-null", or "H-zero").
■ alternate hypothesis - prediction - that there is a relationship between two phenomena
(e.g. that a new treatment has a measurable effect)

describes what we wish to investigate

the alternate hypothesis is probably true or probably false!

■ the alternate hypothesis holds if the null hypothesis is rejected (“false”)

In statistics, it is often denoted H1

present state of
knowledge → formulate
null hypothesis
true? or not true? (test)
predict the future state of
knowledge → formulate
alternate hypothesis
how probable? (measure)
population → sample
collect data infer conclusions based
on the probability of the
alternate hypothesis to
be true given the initial
assumption that the null
hypothesis is true
analyse data/statistics
measure the probability
of alternate hypothesis
reject or fail to reject the
null hypothesis
is this probability
statistically significant?
■ two tailed test H0 A=B; prediction: reject H0 therefore H1 A ≠ B

■ one tailed test H0 A=B; prediction: reject H0 therefore H1 A > B or A < B

Hypothesis to determine whether a population mean, μ, is equal to some target value μ0 :

•Ho: μ = μ0

•H1: μ < μ0 (a lower-tailed test) or

•H1: μ > μ0 (a upper-tailed test) or

•H1: μ ≠ μ0 (a two-tailed test)

Hypothesis to determine whether one population mean, μ1, is equal to a different population
mean μ2 :

•H0: μ1 = μ2
•H1: μ1< μ2 or
•H1: μ1> μ2 or
•H1: μ1≠ μ2
 Truth

Decision Null Hypothesis Alternate Hypothesis

Do not reject null OK Type II ERROR

Reject null Type I ERROR OK

Type I error: The null hypothesis is rejected when it is true.

Type II error: The null hypothesis is not rejected when it is false.

■ The process of distinguishing between the null hypothesis and the alternate hypothesis is aided by
identifying two conceptual types of errors (type 1 & type 2), and by specifying parametric limits on e.g.
how much type 1 error will be permitted.

Type I error :
reject the null hypothesis when it is true - conclude there is an effect when in reality there is none
“false positive”
the probability of making a Type I error is denoted α = the significance level of test
α = 0,05 (by convention) (5%)
the significance level = the probability of rejecting the null hypothesis when it is true
Type II error :

do not reject the null hypothesis when it is false - conclude there is no effect when one really exists

“false negative”

the probability of making a Type II error is denoted β;

its complement (1 - β) is the power of the test

the probability of NOT making a Type II error is the power of test

the power of test = the probability of rejecting the null hypothesis when it is false
(ex.: the probability (chance) of detecting, as statistically significant, a treatment effect of a given size)
 Decision! - Test statistics

■ a standardized value that is calculated (mathematical formulas) from sample data

during a hypothesis test

■ compares your data with what is expected under the null hypothesis
measures the degree of agreement between a sample of data and the null hypothesis

■ reflects the amount of evidence in the data against the null hypothesis - the larger the
value, the greater the evidence
 Decision! - probability interpretations

■ P - value approach

■ frequentist approach
an event's probability as the limit of its relative frequency in a ratio from an infinite series of trials

■ critical value approach

region of acceptance - a range of values defined so that the chance of making a Type I error is
equal to the significance level
if the test statistics falls within the region of acceptance, the null hypothesis is not rejected
region of rejection - set of values outside the region of acceptance
if the test statistic falls within the region of rejection, the null hypothesis is rejected
 Decision! p value

■ p value = the probability of obtaining an effect at least as extreme as the one in your sample
data, assuming the truth of the null hypothesis

determine whether or not the average human body temperature is 37.0 °C

significance level α = 0,05
input sample data - calculate test statistics - calculate p- value

p-value is basically the probability of obtaining your sample data IF the null hypothesis was true

p = 0.85 - you have little reason to doubt the null hypothesis

p = 0.02 - there’s only a very small chance you would have obtained that data if the null hypothesis was
in fact true

■ p value is in the 0-1 range since p is also a probability

Decision! p value

■ p value = the strength of evidence in support of a null hypothesis

= the probability that alternate hypothesis is true when null hypothesis is true

■ relate the value of the test statistics to the known probability distribution → p value

● the smaller the p value, the greater the evidence against H0

● the greater the power of test and test statistics the greater the evidence against H0
“p-value is the probability that the null hypothesis is false”

“p-value is the probability that the null hypothesis is true”

“p-value is the probability that the alternate hypothesis is true”

“p-value is the probability that the alternate hypothesis is true under the null hypothesis”
■ The typical convention in most fields of
science allows for a 5% chance of
erroneously rejecting the null
hypothesis (of making a Type I error)

■ In other words, a researcher will

conclude that there is a significant
difference between the groups being
studied (will reject the null hypothesis)
only if the chance of being incorrect is
less than 5%
 Can We Accept the Null Hypothesis?

■ Some researchers say that a hypothesis test can have one of two outcomes: you accept
the null hypothesis or you reject the null hypothesis. Many statisticians, however, take
issue with the notion of "accepting the null hypothesis." Instead, they say: you reject the
null hypothesis or you fail to reject the null hypothesis.

■ Why the distinction between "acceptance" and "failure to reject?" Acceptance implies
that the null hypothesis is true. Failure to reject implies that the data are not sufficiently
persuasive for us to prefer the alternative hypothesis over the null hypothesis.
■ If the p-value is less than the required significance level (convention value), then we say
the null hypothesis is rejected at the given level of significance.

Rejection of the null hypothesis is a conclusion. This is like a "guilty" verdict in a criminal
trial: the evidence is sufficient to reject innocence, thus proving guilt. We might accept the
alternative hypothesis (and the research hypothesis) BUT we can discuss about the quality
and magnitude of the evidence (think of evidence in a trial).
■ If the p-value is not less than the required significance level then we say we failed to
reject the null hypothesis. The test has no result.

The evidence is insufficient to support a conclusion. This is like a jury that fails to reach a
verdict.

Whether rejection of the null hypothesis truly justifies acceptance of the research
hypothesis depends on the structure of the hypotheses.
---- Rejecting the hypothesis that a large paw print originated from a bear does not
immediately prove the existence of Bigfoot.

Hypothesis testing emphasizes the rejection, which is based on a probability, rather than
the acceptance, which requires extra steps of logic.
■ The successful hypothesis test is associated with a probability and a Type-I error rate.
The conclusion might be wrong.

■ The conclusion of the test is only as solid as the sample upon which it is based. The
design of the experiment is critical.

example of unexpected effect due to Type I error:

placebo effect (false positive effects)

■ hypothesis testing acts as a filter of statistical conclusions; only those results
meeting a probability threshold are publishable. The impact of filtering on
publication is termed publication bias.

■ multiple testing, in which a variety of tests for a variety of possible effects are
applied to a single data set and only those yielding a significant result are reported
■ Hypothesis testing - YES or NO - PROBABILITY
what about the magnitude of evidence?

■ p-value arises through the interaction of the effect size, the sample size (all things
being equal a larger sample size produces a smaller p-value) and sampling error

■ effect size is a quantitative measure of the strength of a phenomenon

 Variation
■ variability -
variabilitate (dex) / diversitate

= insușirea unui lucru, a unui fenomen etc. de a varia, de a lua forme și aspecte diferite.

= proprietate a ființelor vii de a-și schimba, sub influența mediului și a eredității, însușirile
morfologice, fiziologice, biochimice etc.

= proprietate pe care o are o cantitate, o mărime sau o funcție algebrică de a lua

succesiv o infinitate de valori diferite.
■ variation -
variație (dex) / modificare

= schimbare, transformare; stare a unui lucru care se prezintă sub diferite forme, în mod
variat; trecere de la o formă la alta; aspect variabil, schimbător
= schimbare a unei însușiri morfologice, fiziologice, biochimice etc. a organismului

■ variance → a measure of variability

varianța (dex)
= media aritmetică a pătratelor abaterilor valorilor individuale ale unui șir statistic de
experiențe, de la media aritmetică a șirului respectiv
Measures of variability (C4)

■ range,
■ interquartile range,
■ variance,
■ standard deviation
Variation

■ fluctuation among clinical measurements reflects the combined effects of several phenomena

■ the interpretation of clinical observations depends on the physician's ability:

– to recognize these sources of variation and
– to account for them in the diagnostic and therapeutic processes
■ True biologic variation

■ Systematic variation = Variation associated with making observations under

different conditions

■ Measurement variation
■ True biologic variation

Definition -true biologic variation in clinical measurements is the sum of many unknown
factors, each of which contributes a small random effect

random effects, based on the laws of probability are as likely to be positive (causing the
measurement to exceed the true value) as they are to be negative (causing the
measurement to be less than the true value)
-random effect is governed by the laws of probability = laws of chance
True biologic variation

a series of consecutive systolic blood pressure measurements on the same patient

under theoretically identical conditions will not be exactly equal because of true biologic
(random) variability inherent in these measurements within a given patient
■ Systematic variation = Variation associated with making observations under different
conditions

Definition
variation in clinical measurements occurs when the conditions under which the
measurements are made are known to affect the values obtained

this type of variation is more systematic rather than random, because its effect is
predictable and not based on the laws of chance
Systematic variation

-a patient's systolic blood pressure varies according to the time of day (temporal
variation) and the position of the person (postural variation) when the measurement is
taken

-the pattern associated with these fluctuations is somewhat regular and predictable
within a given person
■ Measurement variation

Definition – measurement variation (sometimes called measurement error) is the

variation among clinical observations that is attributed to the measurement process

measurement error may have both a:

– -random component

– -systematic component
random measurement error
– is governed by the laws of chance and
– results in a measurement that is either above or below the true value with equal
probability
– a series of measurements affected only by random variation will center on the true
value of variable being measured

systematic measurement error

– occurs when, as a result of a flaw in the measurement process, the measurements
no longer center around the true value but around a value that is systematically
higher or lower than the true value
Measurement variation

● for a given sample of urine, a series of measurements of pH made with the same meter
by the same analyst under theoretically identical conditions will not be exactly equal

- random variations in these measurements is the sum of many components,

including instrument precision

- systematic variation results if the instrument is out of calibration

Variation
■ Within–patient variation
Variation in the value of a clinical variable within a given patient may result from:
-biologic variations that result from:
▪ true biologic (random) variation
▪ systematic variation ("bias")
-varying conditions under which the measurement is made (systematic variation)
-measurement error
-a pathologic change in the biologic state of the patient
■ Variation among patients

Variation from patient to patient in the value of a clinical variable is attributable to:
- inherent biologic differences among patients – these biologic differences (variations)
can be:
▪ "true" – due to random changes
▪ "bias" – due to sampling error (a systematic change or variation)
- systematic variations (differences in conditions of measurement)
- measurement error
- a pathologic change
■ Error = The result of a measurement minus the “true”(conventional) value

● Random error (of a result) = A component of the error which, in the course of a
number of test results for the same characteristic,
varies in an unpredictable way.
– It is not possible to correct for random error

● Systematic error ( of a result) = A component of the error which, in the course of a

number of test results for the same characteristic,
remains constant or varies in a predictable way.
– Systematic errors and their causes may be known or unknown.
■ Bias = The difference between the expectation of the test results and an accepted
reference value.

– Bias is the total systematic error as contrasted to random error.

There may be one or more systematic error components contributing to the bias.

errors and biases can occur at any stage of an

experiment: design, recruitment, measurements, data
collection or analysis
 Bias
■ a systematic component of both biologic variation and measurement variation

-unlike random variation, bias results in measurements that are systematically higher or
lower than the true underlying value of a diagnostic variable

■ may result from:

- a flaw in the measurement process or
- from sampling error – that is why bias is a systematic component of the biologic
variation (although the true biologic variation is random!)
■ Bias due to sampling error

– Sampling error arises when statements about reality (for an individual or group)
are made on the basis of incomplete information obtained from a sample

Example
-a biopsy of the right side of the liver will give biased results, or results that are
not representative of the truth (i.e., the presence of the tumor), if a tumor is
located on the left side of the liver
■ Bias due to a flaw in the measurement process

– A faulty measuring instrument may yield values that are consistently higher
or lower than the true value of the variable being measured

Example
-a physician with a hearing impairment may obtain diastolic blood pressure
measurements that are consistently above the true diastolic pressure
■ Relationship between bias and chance

– using a series of diastolic blood pressure measurements obtained from a patient by

two methods:
- an intra-arterial cannula (assumed to be exact and unbiased method) and
- a sphygmomanometer

– the sphygmomanometer values are systematically shifted (i.e., biased) to the right of
the true value, possibly owing to improper cuff size or hearing deficit of the person
making the observation

– the sphygmomanometer values also vary according to chance

these random variations are distributed equally to the left and to the right of the
prevailing value
Study/experiment measurement characteristics

conduct an experiment - evaluate / investigate / measure a phenomenon

● Accuracy Trueness ↔ Precision

acuratete adevarat ↔ precis

● Validity ● Reliability
validitate fiabilitate

● Consistency
consecventa/coerenta
● Accuracy the closeness of a measurement to the true value

When throwing darts at a dart board “accuracy” refers to whether the darts are hitting
the bull’s eye (an accurate dart thrower will throw darts that hit the bull’s eye)
Accuracy consists of Trueness (proximity of measurement results to the “true” value)
and Precision (repeatability or reproducibility of the measurement)
■ Accuracy - The closeness of agreement between a test result and the accepted reference value
Accuracy is essentially absence of error

■ Trueness - Closeness of agreement between the average value obtained from a large series
of test results and an accepted reference value
Trueness is equivalent to absence of bias

■ Precision - The closeness of agreement between independent test results obtained under
stipulated conditions

Repeatability conditions - the same method on identical test items in the same laboratory by the same
operator using the same equipment within short intervals of time

Reproducibility conditions - the same method on identical test items in different laboratories with different
operators using different equipment
● Validity – the extent to which a test measures what it is supposed to measure

- the extent to which a concept, conclusion or measurement is well-founded

and corresponds accurately to the real world

e.g Screening tests should be able to identify the presence or absence of a specific
disease (validity)
 ● Reliability - the degree to which an assessment tool produces stable and consistent results.

■ Inter-rater reliability - the degree of agreement between two or more raters in their appraisals.
■ Test-retest reliability - the degree to which test scores are consistent from one test administration
to the next. Measurements are gathered from a single rater who uses the same methods or
instruments and the same testing conditions. This includes intra-rater reliability.
■ Inter-method reliability - the degree to which test scores are consistent when there is a variation
in the methods or instruments used. This allows inter-rater reliability to be ruled out.
■ Internal consistency reliability - the consistency of results across items within a test.

The goal of estimating reliability is to determine how much

of the variability in test scores is due to errors in
measurement and how much is due to variability in true
scores
● Consistency

■ Consistency refers to logical and numerical coherence.

■ An estimator is called consistent if it converges in probability to its estimand as
sample increases
■ A consistent estimator is one for which, when the estimate is considered as a
random variable indexed by the number n of items in the data set, as n increases,
the estimates converge to the value that the estimator is designed to estimate.
■ A consistent test is one for which the power of the test for a fixed untrue
hypothesis increases to one as the number of data items increases.
■ An estimator that has Fisher consistency is one for which, if the estimator were
applied to the entire population rather than a sample, the true value of the
estimated parameter would be obtained.

■ Internal consistency refers to the general agreement between multiple items that
make-up a composite score of a survey measurement of a given construct. This
agreement is generally measured by the correlation between items.
Sensitivity and specificity
■ sensitivity
(true positive rate, the recall, or probability of detection) measures the proportion of positives that
are correctly identified as such (e.g., the percentage of sick people who are correctly identified as
having the condition).
avoid false negatives - Type II errors

■ specificity

(true negative rate) measures the proportion of negatives that are correctly identified as such (e.g.,
the percentage of healthy people who are correctly identified as not having the condition).
avoid false positives - Type I errors
■ sensitivity or true positive rate (TPR) ■ precision or positive predictive value (PPV)
TPR = TP/P = TP / (TP+FN) PPV = TP / (TP + FP)
■ specificity (SPC) or true negative rate ■ negative predictive value (NPV)
SPC = TN/N = TN / (TN+FP) NPV = TN / (TN + FN)

■ fall-out or false positive rate (FPR) ■ false discovery rate (FDR)

FPR = FP/N = FP / (FP + TN) = 1 - SPC FDR = FP / (TP + FP) = 1 - PPV
■ false negative rate (FNR)
FNR = FN / (TP + FN) = 1 - TPR

■ accuracy (ACC)
ACC = (TP + TN) / (TP + FP + TN + FN) = 1 / (FP + FN)
 Predicted condition

Predicted Condition Predicted Condition Prevalence =

Total population Σ Condition positive/
positive negative Σ Total population

True positive rate

False negative rate
condition False Negative Sensitivity
positive
True positive (TPR) = Σ True positive/
(FNR) = Σ False negative/
(Type II error) Σ Condition positive
Σ Condition positive
True
condition
True negative rate
False positive rate
condition False Positive Specificity
negative
True negative (FPR) = Σ False positive/
(TNR) = Σ True negative/
(Type I error) Σ Condition negative
Σ Condition negative

Positive predictive False omission

value (PPV) = rate (FOR) = Positive likelihood ratio
Accuracy (ACC) = Σ True positive / Σ False negative / (LR+)= TPR/FPR
Σ Test outcome positive Σ Test outcome negative
Diagnostic odds ratio
Σ True positive +
(DOR) = LR+/LR−
Σ True negative False discovery Negative predictive
/ Σ Total population rate (FDR) = value (NPV) = Negative likelihood ratio
Σ False positive / Σ True negative / (LR−)= FNR/TNR
Σ Test outcome positive Σ Test outcome negative
 Patients with bowel cancer 2030 subjects
(as confirmed on endoscopy)

Condition positive Condition negative

Positive predictive value= TP / (TP + FP)

Test
True positive False positive = 20 / (20 + 180)
outcome
Fecal (TP) = 20 (FP) = 180
positive
occult = 10%
blood
screen
test Negative predictive value= TN / (FN +
outcome Test TN)
False negative True negative
outcome = 1820 / (10 + 1820)
(FN) = 10 (TN) = 1820
negative
≈ 99.5%

Sensitivity= TP / (TP + FN) Specificity= TN / (FP + TN) 1 - FPR = TNR

1 - FNR = TPR confidence level
= 20 / (20 + 10) = 1820 / (180 + 1820)
power of test significance level = 1 - TNR = 0,09
≈ 67% = 91%
■ positive predictive value ■ sensitivity or true positive rate (TPR)

PPV = TP / (TP + FP) TPR = TP/P = TP / (TP+FN)

If a test subject has an abnormal screening The percentage of sick people who are
test (i.e., it's positive), what is the probability correctly identified as having the
that the subject really has the disease? condition

■ negative predictive value ■ specificity (SPC) or true negative rate

NPV = TN / (TN + FN) SPC = TN/N = TN / (TN+FP)

If a test subject has a negative screening The percentage of healthy people who
test, what is the probability that the subject are correctly identified as not having the
really does not have the disease? condition
 One more time!
■ Type I error - reject H0 when it is true - false positive (diagnosis or medical test)

probability of Type I error / “false positive” rate = significance level α

1 - α = specificity - quantifies the avoiding of false positive = true negative rate
1 - α = confidence level - the probability of NOT rejecting H0 when it is true

■ Type II error - do not reject H0 when it is false - false negative (diagnosis or medical test)

1 - probability of Type II error / “false negative” rate

= power of test = sensitivity - quantifies the avoiding of false negative = true positive rate
 Type I Type II
Error Error

false negative
false positive rate rate

significance level 1-β power of test

α

1 - FPR = 1 - FNR =

specificity sensitivity

confidence level power of test

■ Type I error rate
associated with the a-priori setting of the significance level by the researcher: the significance level
represents an acceptable error rate considering that all null hypotheses are true (the "global null"
hypothesis); the choice of a significance level may thus be somewhat arbitrary (i.e. setting 10% (0.1),
5% (0.05), 1% (0.01) etc.)

■ false positive rate

associated with a post-prior result, which is the expected number of false positives divided by the
total number of hypotheses under the real combination of true and non-true null hypotheses
(disregarding the "global null" hypothesis). Since the false positive rate is a parameter that is not
controlled by the researcher, it cannot be identified with the significance level.
■ Sensitivity to change

Definition: sensitivity of a measurement is the extent to which it changes to changing

subject (patient, variable being measured, etc.) abilities or condition or situation or
characteristics (sensitivity to change)
Uncertainty analysis investigates the uncertainty in a measurement.
An experiment designed to determine an effect, demonstrate a law, or estimate the numerical
value of a physical variable will be affected by errors due to instrumentation, methodology,
presence of confounding effects and so on

Measurement uncertainty is an estimation of the potential error in a measurement result that is

caused by variability in the equipment, the processes, the environment, and other sources. Every
element within a measurement process contributes errors to the measurement result, including
characteristics of the item being tested. Evaluation of the measurement uncertainty
characterizes what is reasonable to believe about a measurement result based on knowledge of
the measurement process.
Sensitivity analysis - a method to determine the robustness of an assessment by examining the
extent to which results are affected by changes in methods, models, values of unmeasured
variables, or assumptions, with the aim of identifying results that are most dependent on
questionable or unsupported assumptions

a series of analyses of a data set to assess whether altering any of the assumptions made
leads to different final interpretations or conclusions

is the study of how the uncertainty in the output of a mathematical model or system (numerical or
otherwise) can be apportioned to different sources of uncertainty in its inputs
• How confident can I be about the results?
• Will the results change if I change the definition of the outcome (e.g., using different cut-off
points)?
• Will the results change if I change the method of analysis?
• Will the results change if we take missing data into account? Will the method of handling
missing data lead to different conclusions?
• How much influence will minor protocol deviations have on the conclusions?
• What if the data were assumed to have a non-Normal distribution or there were outliers?
• Will the results change if one looks at subgroups of patients?
• Will the results change if the full intervention is received (i.e. degree of compliance)?
TYPES OF STUDY
SAMPLING
 Types of research

Type of information
Application Purpose
sought

Descriptive
Pure/Basic
Quantitative
Explanatory

Exploratory
Applied Qualitative
discovery of new knowledge, theoretical in nature
takes many years for the results of fundamental research to find a practical utility

→ solves specific, practical questions

→ policy formulation, administration, understanding of a phenomenon
→ improves products, processes
→ infers beyond the studied group or situation
→ usually exploratory or descriptive
→ on the basis of basic research
 Explanatory research Exploratory Research Descriptive research

Amount of uncertainty
characterizing decision Clearly defined Highly ambiguous Partially defined
situation

Key research statement Research hypotheses Research question Research question

Later stages of decision Early stage of decision Later stages of decision
When conducted?
making making making
Usual research
Highly structured Unstructured Structured
approach
‘Our sales are ‘What kind of people
‘Will consumers buy
declining for no patronize our stores
more products in a blue
apparent reason’ compared to our primary
package?’
competitor?’
Examples
‘What kinds of new
‘Which of two
products are fast-food ‘What product features are
advertising campaigns
consumers interested the most important to our
will be more effective?’
in?’ customers?’
Types of study

■ descriptive case reports

case series
cross sectional

■ analytical - observational: cohort studies, case control

- experimental: randomized trials
Descriptive statistics
■ solely concerned with properties
of the observed data,
■ does not assume that the data
came from a larger population.
to infer = make a well informed guess; conclude from evidence or by reasoning
Analytical/Statistical inference
■ infers conclusions about populations or scientific
truths from data: this includes testing hypotheses
and deriving estimates
■ the population is assumed to be larger than the
observed data set; in other words, the observed
data is assumed to be sampled from a larger
population.
■ the act of generalizing from the data (“sample”) to
a larger phenomenon (“population”) with
calculated degree of certainty
Study designs

■ descriptive studies
■ cross sectional studies
■ case control
■ cohort studies
■ randomized trials
Types of study design

■ there is no “best type of study design”

■ context, assumptions, paradigms and perspectives

Research steps

■ define the problem / research question / research goal

■ specify the hypothesis / objectives
■ select type of study / study design
■ select study sample
■ collect data
■ analyze data
■ formulate conclusions
Types of Clinical Questions

Clinical questions may be categorized as either background or foreground. Why is this important?
Determining the type of question will help you to select the best resource to consult for your answer.

Background questions ask for general knowledge about an illness, disease, condition, process or thing.

For example
•How overweight is a woman to be considered slightly obese?
•What are the clinical manifestations of menopause?
•What causes migraines?

Background questions are best answered by medical textbooks and narrative reviews.
Foreground questions ask for specific knowledge to inform clinical decisions.

These questions typically concern a specific patient or particular population.

Foreground questions tend to be more specific and complex compared to background questions. Quite often,
foreground questions investigate comparisons, such as two drugs, two treatments, two diagnostic tests, etc.

■ Foreground questions are best answered by consulting medical databases such as MEDLINE (via
PubMed or Ovid), Embase, Cochrane Database of Systematic Reviews and ACP Journal Club.
Foreground questions may be further categorized into one of 4 major types:

► treatment/therapy,
► diagnosis,
► prognosis,
► etiology/harm.

For example
• Is Chondroitin sulfate effective when compared with placebo in slowing the rate of functional
impairment in a 72 year old male patient with osteoarthritis?
• In pediatric patients with Allergic Rhinitis, are Intranasal steroids more effective than
antihistamines in the management of Allergic Rhinitis symptoms?
■ Therapy: Questions of treatment/intervention in order to achieve some outcome. May
include drugs, surgical intervention, change in diet, counseling, prevention, etc.

■ Diagnosis: Questions of identification of a disorder in a patient presenting with specific

symptoms.

■ Prognosis: Questions of progression of a disease or likelihood of a disease occurring.

■ Etiology/Harm: Questions of negative impact from an intervention or other exposure.

 Types of evidence to answer
Type or Domain Explanation
the question
Therapy (Treatment) Questions about the effectiveness of interventions in improving Randomised Controlled Trial
outcomes in sick patients / patients suffering from some (RCT)
condition. These are the most frequently asked. Among the many
treatments offered by clinicians are medications, surgical
procedures, excercise, and counseling about lifestyles changes.

Prevention Questions about the effectiveness of an intervention or exposure in RCT or Prospective Study
preventing morbidity and mortality. Similar to treatment questions.
When assessing preventive measures, it is particularly important to
evaluate potential harms as well as benefits.

Diagnosis Questions about the ability of a test or procedure to differentiate RCT or Cohort Study
between those with and without a condition or disease.

Prognosis (Forecast) Questions about the probable cause of a patient's disease or the Cohort Study and/or Case-
likelihood that he or she will develop an illness. Control Series

Etiology (Causation) Questions about the harmful effect of an intervention or exposure on Cohort Study
a patient.

Meaning Questions about patients' experiences and concerns. Qualitative Study

How to choose a study design

■ does it adequately test the hypothesis ?

■ does it identify and control extraneous factors ?

■ are results generalizable ?

■ can the hypothesis be rejected or supported via statistical methods ?

■ is the design efficient in using the available resources ?

How to select a study design
■ level of knowledge and expertise
■ nature of research phenomenon
■ nature of research purpose
■ ethical considerations
■ feasibility
■ availability and validity of data
■ precision
■ cost
 Descriptive studies (observational)

● attempt to systematically describe a situation/problem/phenomenon/service/etc

● provides an accurate portrayal of the characteristics of a particular individual/ group/
situation

■ less expensive and time consuming than analytical studies

■ can collect a large amount of data
■ helpful for developing hypotheses

■ does not identify a possible cause for a phenomenon

■ Case report

a descriptive study of one single patient

detailed profile of a rare presentation or treatment
helpful in developing hypotheses to be tested using analytical study

■ Case series

a descriptive study of multiple patients

rare phenomenon occurring multiple times
detailed profile of patients’ presentation and/or outcome
helpful in developing hypotheses
 Case report
Case series

■ can provide clues in identification of a new disease or adverse effect of

exposures/treatment/drugs

■ no appropriate comparison group

■ cannot be used to test for presence of a valid statistical association

■ presence of a risk factor may be purely coincidental

■ not a true epidemiologic design

■ Cross sectional

→ observation of a defined population at a single point in time or time interval

→ exposure and outcome are determined simultaneously

→ prevalence at a point in time; can be repeated at another point in time

→ pre-test/post-test studies measure the change - a variation of the cross-sectional design - two
sets of cross-sectional data collection on the same population to determine if a change has
occurred

→ often in the form of a survey

→ the sample frame used to select a sample and the response rate determine how well results
can be generalized to the population as a whole
 ■ Cross sectional

Advantages:
▪ short term ▪ design less complex

▪ fewer resources ▪ less statistical analysis

▪ more easily controlled ▪ useful for planning health services and medical programs

▪ relationship between attributes of disease and characteristics of various groups (e.g. elderly)

▪ many outcomes and risk factors can be assessed

 Disadvantages:

● represent only those who are evaluated

● prevalence-incidence bias (also called Neyman bias). Especially in the case of longer-lasting
diseases, any risk factor that results in death will be under-represented among those with the
disease

● do not determine temporal relationship between exposure and outcome (lacks time element)

● only a snapshot: the situation may provide differing results if another timeframe had been chosen
Explanatory studies (analytical)

■ Explain Why and How a phenomenon is going on

■ Driven by theory
■ Rely on experiments or quasi-experiments to demonstrate causality
■ Can be used for hypothesis testing
■ Allow for inferences to be drawn about associations and causality

■ identify the extent and nature of cause-and-effect relationships

■ focus on an analysis of a situation or a specific problem to explain the
patterns of relationships between variables
 Explanatory studies (analytical)
■ Advantages

1. may play an instrumental role in terms of identifying reasons behind a wide range of processes, as well
as, assessing the impacts of changes on existing norms, processes etc.
2. offer the advantages of replication if necessity arises
3. are associated with greater levels of internal validity due to systematic selection of subjects

■ Disadvantages

1. coincidences in events may be perceived as cause-and-effect relationships.

2. difficult to reach appropriate conclusions due to the impact of a wide range of factors and variables
while casualty can be inferred, it cannot be proved with a high level of certainty.

3. while correlation between two variables can be effectively established, identifying which variable is a
cause and which one is the impact can be a difficult task to accomplish.
■ Case control

→ examine multiple exposures in relation to an outcome

→ subjects are defined as cases and controls, and exposure histories are compared

→ begin with the outcomes and do not follow subjects over time

→ proceed from effect to the cause

■ Case control study: a study design that examines a group
of people who have experienced an event (usually an
adverse event) and a group of people who have not
experienced the same event, and looks at how exposure to
suspect (usually noxious) agents differed between the two
groups. This type of study design is most useful for trying to
ascertain the cause of rare events/dieases.
 Case control

Because participants are selected on the basis of disease,

exposures for ALL PARTICIPANTS are obtained RETROSPECTIVELY

PAST PRESENT

Exposure recall Selected Cases & Controls

Example: lung cancer cases and non-cancerous controls recall past exposure to cigarette smoke
 Case control

■ Advantages
Relatively inexpensive
Less time-consuming than cohort studies
Can evaluate effects of multiple exposures
Efficient for rare outcomes or outcomes with long induction or latency periods

■ Disadvantages
Limited to one outcome variable
Subject to recall bias (based on subjects’ memory and reports)
Difficult to establish clear chronology of exposure and outcome
Inefficient for rare exposures
 Cohort Study

■ when good evidence suggests an association of a disease with a certain exposure or exposures
■ examines multiple health effects of an exposure;
■ subjects are defined according to their exposure levels ;
■ subjects are followed over time for outcome occurrence;
■ subjects who presently have a certain condition or receive a particular treatment are followed
over time and compared with another group of subjects who are not affected by the condition;
■ the exposure of interest is determined for each member of the cohort and the group is followed
to document incidence of disease in the exposed and non-exposed members;
■ changes and variation in the disease or health status of a study population as the study group
moves through time
■ Cohort Study: a non-experimental study design that follows a group
of people (a cohort), and then looks at how events differ among
people within the group. A study that examines a cohort, which
differs in respect to exposure to some suspected risk factor (e.g.
smoking), is useful for trying to ascertain whether exposure is likely
to cause specified events (e.g. lung cancer). Prospective cohort
studies (which track participants forward in time) are more reliable
than retrospective cohort studies.
Cohort Study

■ Prospective – cohort characterized by determination of exposure levels (exposed vs. not

exposed) at baseline (present) and followed for occurrence of disease in future

■ Retrospective - makes use of historical data to determine exposure level at some baseline in
the past and then determine subsequent disease status in the present.

■ Restricted - limited exposure, narrow behavior (e.g. rehabilitation care inpatient)

Cohort Study

■ Advantages
Can evaluate multiple effects of a single exposure
More efficient for rare exposures and outcomes with long induction and latency periods
Can directly measure incidence
Clear chronological relationship between exposure and outcome

■ Disadvantages
Expensive
Time-consuming
Inefficient for rare outcomes with long induction or latency periods
 Experimental
■ research designs which use manipulation and controlled testing to understand
causal processes

■ one or more variables are manipulated to determine their effect on a dependent variable

■ often used where:

There is time priority in a causal relationship (cause precedes effect)
There is consistency in a causal relationship (a cause will always lead to the same effect)
The magnitude of the correlation is great.
■ An experiment is typically carried out by manipulating a variable, called the
independent variable, affecting the experimental group.

■ The effect that the researcher is interested in, the dependent variable(s), is
measured.

■ Identifying and controlling non-experimental factors which the researcher does not
want to influence the effects, is crucial to drawing a valid conclusion. This is often
done by controlling variables, if possible, or randomizing variables to minimize
effects that can be traced back to third variables.
■ aim:
investigate the possible cause and effect relationship by manipulating one variable to
influence the other variable in the experimental/study/intervention group and by
controlling the other relevant variables and measuring the effects of the manipulation
by statistical means
■ a true experiment - the researcher manipulates one variable, and control/randomizes
the rest of the variables
■ it has a control group, the subjects have been randomly assigned between the groups,
and the researcher only tests one effect at a time; know what variable(s) you want to test
and measure

■ a quasi experiment - the scientist actively influences something to observe the

consequences

■ Experiments are conducted to be able to predict phenomenons

■ Experiments are constructed to be able to explain some kind of causation
 is random assignment used?

yes no

randomized/ is there a control group or

true experiment multiple measurements?

yes no

quasi-experiment non-experiment
■ An ad hoc analysis is a hypothesis invented after testing is done, to try to explain why
the contrary evidence. A poor ad hoc analysis may be seen as the researcher's inability
to accept that his/her hypothesis is wrong, while a great ad hoc analysis may lead to
more testing and possibly a significant discovery.

■ First conduct a pilot-study or two before you do the real experiment. This ensures that
the experiment measures what it should, and that everything is set up right.
■ The key to designing any experiment is to look at what research variables could affect the
outcome.

■ There are many types of variable but the most important, for the vast majority of research
methods, are the independent and dependent variables.

■ The independent variable is the core of the experiment and is isolated and manipulated
by the researcher.
■ The dependent variable is the measurable outcome of this manipulation, the results of
the experimental design.

If you designed an experiment to determine how quickly a cup of coffee cools, the manipulated
independent variable is time and the dependent measured variable is temperature.
 Experimental
■ similar groups of individuals, from the same source population, are allocated at random to
receive and not to receive an intervention, then observed for occurrence of outcome(s)

random allocation
■ the “gold standard” of research design;
■ provide the most convincing evidence of the relationship between exposure and
outcome/effect - causal
■ expensive
■ ethics
■ dropout rates, lost to follow-up - intention to treat analysis (ITT)/efficacy subset analysis
Systematic Review

■ A type of literature review that collects and critically analyzes multiple research studies
or papers.
■ A systematic review is a critical assessment and evaluation of all research studies that
address a particular clinical issue.
■ The researchers use an organized method of locating, assembling, and evaluating a
body of literature on a particular topic using a set of specific criteria.
■ A systematic review typically includes a description of the findings of the collection of
research studies.
■ The systematic review may also include a quantitative pooling of data, called a meta-
analysis.
Systematic Review

■ Defining the review question(s) and developing criteria for including studies
■ Searching for studies
■ Selecting studies and collecting data
■ Assessing risk of bias in included studies
■ Analysing data and undertaking meta-analyses
■ Addressing reporting biases
■ Presenting results and "summary of findings" tables
■ Interpreting results and drawing conclusions
Systematic Review
The main stages of a systematic review are:

■ Defining a question and agreeing an objective method.

■ A search for relevant data from research that matches certain criteria.
■ 'Extraction' of relevant data. This can include how the research was done (often called the method
or 'intervention'), who participated in the research (including how many people), how it was paid
for (for example funding sources) and what happened (the outcomes).
■ Assess the quality of the data by judging it against criteria identified at the first stage.
■ Analyse and combine the data (using complex statistical methods) which give an overall result
from all of the data. This combination of data can be visualised using a blobbogram (also called a
forest plot).
■ Once these stages are complete, the review may be published, disseminated and translated into
practice after being adopted as evidence.
Systematic Review
■ Intervention reviews - assess the benefits and harms of interventions used in healthcare and
health policy.
■ Diagnostic test accuracy reviews - assess how well a diagnostic test performs in diagnosing
and detecting a particular disease.
■ Methodology reviews - address issues relevant to how systematic reviews and clinical trials are
conducted and reported.
■ Qualitative reviews - synthesize qualitative and quantitative evidence to address questions on
aspects other than effectiveness.
■ Prognosis reviews - address the probable course or future outcome(s) of people with a health
problem.
■ Overviews of Systematic Reviews (OoRs) are a new type of study in order to compile multiple
evidence from systematic reviews into a single document
Meta-Analysis
■ combines results from many studies dealing with the same topic

■ statistically combines results of existing research to estimate overall size of

relation between variables

■ helps in: - developing theory

- identifying research needs
- establishing validity

■ can replace large scale research studies

■ better evidence than literature review
Systematic Review

■ In 1996, to address the suboptimal reporting of meta-analyses, an international group developed

a guidance called the QUOROM Statement (QUality Of Reporting Of Meta-analyses), which
focused on the reporting of meta-analyses of randomized controlled trials.

■ A revision of these guidelines, renamed PRISMA (Preferred Reporting Items for Systematic
reviews and Meta-Analyses), which have been updated to address several conceptual and
practical advances in the science of systematic reviews has been published.
Section/topic # Checklist item
PRISMA 2009 Checklist
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both.
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and
interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review
registration number.

INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known.
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and
study design (PICOS).
METHODS
Protocol and 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information
registration including registration number.
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication
status) used as criteria for eligibility, giving rationale.
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the
search and date last searched.
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.

Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-
analysis).
Data collection 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and
process confirming data from investigators.
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

Risk of bias in 12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or
individual studies outcome level), and how this information is to be used in any data synthesis.
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means).
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I 2) for each
meta-analysis.
Section/topic # Checklist item PRISMA 2009 Checklist
Risk of bias across 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within
studies studies).
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-
specified.
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage,
ideally with a flow diagram.
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the
citations.
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).
Results of individual 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b)
studies effect estimates and confidence intervals, ideally with a forest plot.
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency.
Risk of bias across 22 Present results of any assessment of risk of bias across studies (see Item 15).
studies
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).

DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g.,
healthcare providers, users, and policy makers).
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research,
reporting bias).
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.

FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic
review.
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097
PRISMA 2009 Flow Diagram
Example Table: Summary of included studies evaluating the efficacy of antiemetic agents in acute gastroenteritis.
Example Figure: Overall failure (defined as failure of assigned regimen or relapse) with T-R versus T-S.
 Example Table: Quality measures of the randomized controlled trials that failed to fulfill any one of six markers of
validity.
Qualitative Research - not everything can be quantified
■ investigation in which the researcher attempts to understand some larger reality by
examining it in a holistic way or by examining components of that reality within their
contextual setting

“Qualitative Research…involves finding out what people think, and how they feel - or at
any rate, what they say they think and how they say they feel. This kind of information is
subjective. It involves feelings and impressions rather than numbers”
- Bellenger, Bernhardt and Goldstucker, Qualitative Research in Marketing, American Marketing Association
■ purpose is understanding
■ oriented toward discovery
■ uses subjective data
■ extracts meaning from data
■ interprets results in context
■ focus is holistic
■ Advantages
→ in depth examination of phenomena
→ not limited to rigidly definable variables
→ examines complex questions (impossible with numbers)
→ explore new areas of research (unanticipated responses)
→ build new theories

■ Disadvantages
→ subjectivity leads to procedural problems
→ replicability is difficult
→ researcher bias is built in and unavoidable
→ labor intensive, expensive
→ not understood well by “classical” researchers
■ Participant observation is appropriate for collecting data on naturally occurring
behaviors in their usual contexts.

■ In-depth interviews are optimal for collecting data on individuals’ personal histories,
perspectives, and experiences, particularly when sensitive topics are being
explored.

■ Focus groups are effective in eliciting data on the cultural norms of a group and in
generating broad overviews of issues of concern to the cultural groups or subgroups
represented.
→ Designing a research hypothesis is supported by a good research question and will
influence the type of research design for the study.

Acting on the principles of appropriate hypothesis development, the study can then
confidently proceed to the development of the research objective

→ The study objective is an active statement about how the study is going to answer
the specific research question.
Study: Warden SJ, Metcalf BR, Kiss ZS, et al. Low-intensity pulsed ultrasound for chronic
patellar tendinopathy: a randomized, double-blind, placebo-controlled trial. Rheumatology
2008;47:467–71.

Research question: How does low-intensity pulsed ultrasound (LIPUS) compare with a
placebo device in managing the symptoms of skeletally mature patients with patellar
tendinopathy?

Research hypothesis: Pain levels are reduced in patients who receive daily active-LIPUS
(treatment) for 12 weeks compared with individuals who receive inactive-LIPUS (placebo).

Objective: To investigate the clinical efficacy of LIPUS in the management of patellar

tendinopathy symptoms.
● IDENTIFY THE STUDY SUBJECTS

1. Identify the population of interest - a population is the group of people that you want to
make assumptions about - as you cannot study every individual from a population go to
next step
2. Specify a sampling frame - a sampling frame is the group of people from which you will
draw your sample - groups you have access to
3. Specify a sampling method - randomly (can use strata) or non-randomly
4. Determine the sample size
5. Implement the plan
■ Inferential statistics tries to infer information about a population by using information
gathered by sampling.

■ Statistics: The collection of methods used in planning an experiment and analyzing

data in order to draw accurate conclusions.

■ Population: The complete set of elements/units is termed the population.

■ Sample: A sample is a portion of a population selected for further analysis.

■ Parameter: A parameter is a characteristic of the whole population.

■ Statistic: A statistic/estimate is a characteristic of a sample, presumably measurable.

Sampling

■ the process of selecting a number of units (individuals, lab animals,

devices, etc.) for a study in such a way that these units represent the larger
group from which they were selected

■ to gather data about the population in order to make an inference that can
be generalized to the population
Population

■ sample - “a smaller and representative collection of units from a population used to

determine truths about that population” (Field, 2005)

■ study population - the segment from the target population you have access to

■ target population - a set of elements/units larger than the sampled population to which
the researcher would like to generalize the study findings
 population sample

■ what is your population of

interest ?
■ to whom do you want to
generalize your results ?
the sample should be
representative to the population
- have all the characteristics of
the target population

older people (≥ 65 years) with DM

eligibility criteria
middle aged men with STEMI
sampling error - systematic error - response rate - coverage error

The definition of the subject of study and the target population should be clearly spelt out.

Besides inclusion criteria, the exclusion criteria should also be clearly stated.

If there are two or more groups, define them.

 The Universe

Study Study
Sample Sample
Study
Population
REPRESENTATIVENESS

● sampling procedure
● sample size
● participation (response)
When might you sample the entire population ? statistic = parameter

■ when your population is very small

ex - toti bucurestenii internati prin camera de garda in luna decembrie si care au fost internati in
INGG in ultimele 6 luni

■ when you have extensive resources

ex. - recensamantul populatiei

■ when you don’t expect a very high response

ex. - chestionar privind frecventa activitatii sexuale in randul varstnicilor
The sampling process - stages:

■ Defining the population of concern (target and study)

■ Specifying a sampling frame, a set of items or events possible to measure
■ Specifying a sampling method for selecting items or events from the frame
■ Determining the sample size
■ Implementing the sampling plan
■ Sampling and data collecting
■ sampling frame - a list of all those within a population who can be sampled

■ unit or element - the subject on which information is obtained

■ sampling unit - the unit of selection

■ sample size - the number of units sampled for inclusion in the study

■ sampling technique - method of selecting sampling units from sampling frame

who are accessible to the researcher

who meet eligibility criteria

Sampling method

■ probability sampling - random - some representativeness

every unit in the population has a chance (greater than zero) of being selected in the sample, and
this probability can be accurately determined
- 'equal probability of selection' (EPS)

■ non probability sampling - non-random - no representativeness

some elements of the population have zero chance of selection, or the probability of selection can't
be accurately determined
non-probability sampling

■ usually not representative of the population

■ ability to generalize results is limited
■ largely used in qualitative research

■ methods: ● convenience sample (pilot study)

● snowball sampling
● quota sampling
● purposive/judgmental sampling
● expert sampling
● diversity sampling
probability sampling

■ more accurate ■ Every element has a known nonzero

probability of being sampled
■ more representative
■ Involves random selection at some
■ can calculate sampling error
point.
probability sampling

■ simple random sample (SRS)

■ systematic sampling (interval sampling)
■ stratified sampling
■ cluster sampling
■ multi-stage random sampling
Matched samples

■ a sample in which the individuals selected for analysis share all properties except
that under investigation

■ reduces bias due to confounding

■ Matching is a statistical technique which is used to evaluate the effect of an
intervention by comparing the “treated” and the “non-treated” units in an
observational study or quasi-experiment (i.e. when the treatment is not randomly
assigned)

■ The goal of matching is, for every treated unit, to find one (or more) “non-treated”
unit(s) with similar observable characteristics against whom the effect of the
intervention can be assessed.

■ By matching “treated” units to similar “non-treated” units, matching enables a

comparison of outcomes among “treated” and “non-treated” units to estimate the
effect of the treatment reducing bias due to confounding
Paired/dependent - independent samples

■ Paired samples (also called dependent samples) are samples in which

natural or matched couplings occur. This generates a data set in which
each data point in one sample is uniquely paired to a data point in the
second sample.
■ Examples of paired samples include:

• pre-test/post-test samples in which a factor is measured before and after an intervention,

• cross-over trials in which individuals are randomized to two treatments and then the
same individuals are crossed-over to the alternative treatment,

• matched samples, in which individuals are matched on personal characteristics such as

age and sex,

• duplicate measurements on the same biological samples,

• any circumstance in which each data point in one sample is uniquely matched to a data
point in the second sample.
■ In a paired sample design, we model the data as if there is a single bucket of balls,
and each draw from the bucket results in a pair of numbers (that we can distinguish
as first and second).

■ In a paired design, the method of analysis is as follows:

- Take individual differences for each pair
- Treat the differences as a sample from a single population

■ Notice that the mean of differences is equal to the difference of means.

■ However, the standard error for a paired design does not equal the standard error
computed from the same data assuming two independent samples
Sample Size

the objective of any medical study is to estimate from a

sample the corresponding population parameter or parameters

● you need to have enough research participants (patients) to show a difference

● perform a power analysis to determine the sample size
■ Acceptable level of significance
■ Power of the study
■ Expected effect size
■ Underlying event rate in the population estimates
(prevalence)
assumptions
■ Standard deviation in the population
■ Confidence interval
■ Confidence level
■ Population size
DATA MANAGEMENT
 Variables

■ The purpose of data analysis is to identify whether your assumptions

were correct or not, and to highlight possible new views on the problem
under study
■ The ultimate purpose of analysis is to answer the research questions
outlined in the objectives with your data
Variables

■ when reduced to most simple terms, all medical research is simply the
study of relationships among variables

■ most often, medical investigators are interested in studying either

associations or differences among variables
Variables

■ a variable is any:
■ quality
■ characteristic
■ constituent
of a person or thing
that can be measured
■ by definition a variable is subject to change - variability
 Variables

■ When you selected the variables for your study, you did so with the
assumption that they either:
■ would help to define your problem (dependent variables -
represents the output or effect, or is tested to see if it is the effect)
and its different components
or that
■ they were contributory factors to your problem (independent
variables – the inputs or causes, or are tested to see if they are the
cause)
Variables
■ Other way of expressing that is:
– dependent variable is the response
– independent variable is the predictor
■ Ex: – relationship between:
– cardiovascular risk (response) and body mass (predictor)
– species number (response) and island area (first predictor) and
island age (second predictor)
■ The key to designing any experiment is to look at what research variables could affect the
outcome.

■ There are many types of variable but the most important, for the vast majority of research
methods, are the independent and dependent variables.

■ The independent variable is the core of the experiment and is isolated and manipulated
by the researcher.
■ The dependent variable is the measurable outcome of this manipulation, the results of
the experimental design.

If you designed an experiment to determine how quickly a cup of coffee cools

the manipulated independent variable is time and the dependent measured variable is temperature.
Classification of variables

■ independent variables
■ controlled variables - confounding variables/third variable
■ dependent variables
 Classification of variables
■ Univariate data. When we conduct a study that looks at only one variable, we say that
we are working with univariate data.

Suppose, for example, that we conducted a survey to estimate the average weight of high school
students. Since we are only working with one variable (weight), we would be working with univariate
data.

■ Bivariate data. When we conduct a study that examines the relationship between two
variables, we are working with bivariate data.

Suppose we conducted a study to see if there were a relationship between the height and weight of
high school students. Since we are working with two variables (height and weight), we would be
working with bivariate data.
 Bias, confounding and effect modification

■ Bias: A systematic error in the design, recruitment, data collection or analysis that results in
a mistaken estimation of the true effect of the exposure and the outcome.

■ Confounding: A situation in which the effect or association between an exposure and

outcome is distorted by the presence of another variable. Positive confounding (when the
observed association is biased away from the null) and negative confounding (when the
observed association is biased toward the null) both occur.

■ Effect modification : a variable that differentially (positively and negatively) modifies the
observed effect of a risk factor on disease status. Different groups have different risk
estimates when effect modification is present.
■ If the method used to select subjects or collect data results in an incorrect
association,
THINK >> Bias!

■ If an observed association is not correct because a different (lurking) variable is

associated with both the potential risk factor and the outcome, but it is not a
causal factor itself,
THINK >> Confounding!

■ If an effect is real but the magnitude of the effect is different for different
groups of individuals (e.g., males vs females or blacks vs whites).
THINK >> Effect modification!
 Classification of variables
■ Between-subjects Variable Values of the variable are examined as they are observed to
change from one individual to another. When a variable is administered Between-
Subjects, subjects each receive ONE level of the Independent Variable

■ Within-subject Variable Values of the variable are examined as they are observed to
change from time to time within an individual. When a variable is administered
Within-Subject, subjects receive ALL levels of the Independent Variable (also called
Repeated Measures).
Data analysis

■ Analysis of data is a process of inspecting, cleansing, transforming, and

modeling data with the goal of discovering useful information, suggesting
conclusions, and supporting decision-making. Data analysis has multiple
facets and approaches, encompassing diverse techniques under a variety
of names, in different business, science, and social science domains.
 Data management
■ data requirements - the sample, the variables
■ data collection:
▪ CRF (case report form - clinical trials)
- the first step in translating the protocol-specific activities into data being generated
- data fields should be clearly defined and be consistent throughout
- filling instructions for error-free data acquisition
- annotation - coded terms used to indicate the variables in the study
▪ database design -
▪ data entry - double data entry - missing data
■ data processing - data validation in accordance with the protocol specifications
- data cleaning (incomplete, duplicates, errors, outliers)
- medical coding
- database locking
■ exploratory data analysis - analyze data sets to summarize their main characteristics
▪ initial data analysis
- data cleaning

■ modeling and algorithms, inferential statistics

Data management

■ descriptive statistics,
■ exploratory data analysis (EDA) - discovers new features in the data - patterns
■ confirmatory data analysis (CDA) - rejects or fails to reject existing hypotheses
p -value
CDA

EDA+IDA
 Variables

■ In analysing our data, it is important first of all to determine the type of

data that we are dealing with
■ This is crucial because the type of data used largely determines the type
of statistical techniques that should be used to test whether the results
of the study are significant
Variables
■ First, before we look at how variables may be affecting one another, we
need to summarise the information obtained on each variable in
simple tabular form or in a figure

■ Some of the variables may have produced numerical data, while other
variables produced categorical data
Data Levels of measurement

■ Qualitative/categorical data
● nominal
● dichotomous
● ordinal

■ Quantitative/numerical data
● ordinal
● interval
● ratio
Qualitative variables

– are variables that are measured at a nominal level

– a diagnostic test for pregnancy gives a result of either "positive" or "negative”

The diagnostic test variable is a qualitative variable

Quantitative variables

– are variables that are measured on an ordinal or interval/ratio scale

- a measurement of serum sodium concentration (e.g., 140 mEq/l) expresses the exact amount
of sodium in the serum
- serum cholesterol level, systolic blood pressure, and blood ureea nitrogen (BUN) level are other
quantitative clinical variables
Scales used to measure variables

■ 1.Nominal Scale
■ categories that cannot be ordered one above the other
■ limited number of categories
■ 2.Ordinal Scale
■ limited number of categories that are ranked in terms of a graded order
■ 3.Interval Scale
■ unlimited number of categories that are equally spaced
■ NO true zero point
■ 4.Ratio Scale
■ true zero point
 Levels of Measurement

■ Nominal: Nominal data have no order and thus only gives names or labels to various
categories.
■ Ordinal: Ordinal data have order, but the interval between measurements is not
meaningful.
■ Interval: Interval data have meaningful intervals between measurements, but there is
no true starting point (zero).
■ Ratio: Ratio data have the highest level of measurement. Ratios between
measurements as well as intervals are meaningful because there is a starting point
(zero).
 Types of data

■ Qualitative/categorical data are nonnumeric.

Composed of categories which are not comparable in terms of magnitude
Which?
■ Quantitative data are numeric.
Can be ordered with respect to magnitude on some dimension - numbers
● Discrete data - numeric data that have a finite number of possible values
- data can differ only by well-defined steps with no intermediate values possible
How many?
● Continuous data - can be measured with an arbitrary degree of precision
- any two points on a scale of a continuous data have an infinite number of values in between
How much?
■ nominal - name - label

dichotomous
■ ordinal - the order
■ interval - space in between - no absolute 0
■ ratio
Time

Nominal time of day - categories; no additional information

Ordinal time of day - indicates direction or order of occurrence; spacing between is uneven

Interval time of day - equal intervals; analog (12-hr.) clock, difference between 1 and 2 pm is same as
difference between 11 and 12 am

Ratio - 24-hr. time has an absolute 0 (midnight); 14 o'clock is twice as long from midnight as 7 o'clock
Descriptive statistics

■ quantitatively describe or summarize features of a collection of information

■ descriptive statistics, unlike inferential statistics, are not developed on the basis of
probability theory

■ descriptive statistics aim to summarize a sample, rather than use the data to learn
about the population that the sample of data is thought to represent

■ useful for two purposes: 1) to provide basic information about variables in a dataset
and
2) to highlight potential relationships between variables.
■ quantitative
■ pictorial
- frequency tables
- graphs

● distribution
● distribution shape
● location - central tendency - quantiles
data patterns
● dispersion (scatter, spread)
● association
■ Distribution

The distribution of a statistical data set is a listing or function showing all the possible
values (or intervals) of the data and how often they occur (frequency).

A summary of the frequency of individual values or ranges of values for a variable

Frequency tables and graphs

Frequency distributions are used for both qualitative and quantitative data.

- when a distribution of categorical data is organized, you see the number or percentage of
individuals in each group (category)
- the simplest distribution would list every value of a variable and the number of persons
who had each value
- for the distribution of numerical data we use groups or class intervals
■ relative frequency - represent the percentage of the number of observations falling in a
specific category/group

■ fraction frequency - represents the number of observations falling in a specific

category/group expressed as a proportion (fraction) = probability !

■ cumulative frequency - is the fraction (or percentage) of observations that are less than the
upper limit of each interval/category/group; is obtained by adding the relative frequencies for
the intervals in questions
one bag of M&M; there are 55 M&M's: 17 brown, 18 red, 7 yellow, 7 green, 2 blue, and 4 orange

Table I. Frequencies of colors in one bag of M&M's

Relative frequency
Color Frequency (%)
Brown 17 30.9
Red 18 32.72
Yellow 7 12.72
Green 7 12.72
Blue 2 3.63
Orange 4 7.27
Total 55 100
Figure 1. Distribution of 55 M&M's.
 what about the distribution of all M&Ms?

the manufacturer reports proportions

Since every M&M is one of the six

familiar colors, the six proportions
shown in the figure add to one.

This is a probability distribution

because if you choose an M&M at
random, the probability of getting, say,
a brown M&M is equal to the
proportion of M&M's that are brown
(0.30)

Figure 2. Distribution of all M&M's.

Some of the graphs that can be used with frequency distributions are:

- pie charts, nominal data

qualitative data
- bar charts, nominal and ordinal data

- histograms continuous data → ordinal

quantitative data
- scatterplots, interval and ratio data
 the offspring cohort of the Framingham Heart Study 3,539 subjects

Sex Frequency Relative Marital Status Frequency Relative Frequency,

Frequency, %
%
Single 203 5.8
Male 1,625 45.9 Married 2,580 73.1
Widowed 334 9.5
Female 1,914 54.1
Divorced 367 10.4

Total 3,539 100.0 Separated 46 1.3

Total 3,530 100.0

Table 1 - Frequency Distribution Table for Sex Table 2 - Frequency Distribution Table for Marital Status
 Fig. 2 - Frequency bar chart for marital status

Fig. 1 Relative frequency

distribution pie chart for sex

45.9

54.1

male female
 Smoking Frequency Relative
Status Frequency, %

Non-Smoker 1,330 37.6

Former 1,724 48.8

Current 482 13.6

Relative frequency bar chart for smoking status
60
Total 3,536 100.0
50

40

30

20

10

0
non-smoker former current
Numerical data – frequency distribution

Health centers of District X are submitting numbers of malaria cases and you wish to
summarize them
Compare the daily and weekly summaries of the same data as presented in next Table
Table: Daily and weekly summaries of malaria cases in health centers in District X
– Both daily and weekly data show an increasing amount of malaria, but the
improving situation shown in days 19, 20 and 21 is not reflected in the weekly
summary.

– It would therefore be better to use the daily data if you want to indicate when
exactly the numbers of reported malaria cases started going down.
Class intervals

Definition – class intervals are a set of non-overlapping, contiguous intervals

encompassing the range of values assumed by the clinical variable being described

-the intervals are chosen such that a given data value can be placed in only one
interval
the number of class intervals used to group data is arbitrary, although 5–12 intervals
usually suffice
– too few, obscure detail and lead to loss of information
– too many, defeat the purpose of grouping

-the endpoints of the intervals can be specified in terms of true class limits, which are
specified to one decimal place more than the recorded data and reflect the precision
of the measuring process
■ Histogram

- distributions for continuous variables are called continuous distributions. They also
carry the fancier name probability density.

- numerical data are often presented in histograms, which are very similar to the bar
charts which are used for categorical data.

- a histogram is a graphic representation of a frequency distribution consisting of a

series of rectangles whose widths represent class intervals and whose areas are
proportional to the corresponding frequencies of particular data values within each
interval
- In a histogram the width of the bars represents a class interval ( a numeric value)
on the x axis ( numeric values) ; in bar chart the width of the bar is not important as
it represents a category on the x axis of labels

- In a histogram as well as in a bar chart, the height of the bars represent the
number of items (frequency) which fall into a class interval or a category

- An important difference is that in a histogram the ‘bars’ are connected (as long as
there is no gap between the data), whereas in a bar chart the bars are not
connected, as the different categories are distinct entitles.
■ bar chart ■ histogram

- categorical (nominal, ordinal data) - numeric data

- on the x axis we have categories - on the x axis we have values (numbers) which
(no low end and high end) are ordered (low end and high end)

- each column represents a group - each column represents a group defined by a

defined by a categorical variable quantitative variable (class interval)
(label)

- there is no space between columns

- there is a space between columns

The height of the column indicates the size of the group defined by the column
a histogram reveals the shape of a frequency distribution
■ Distribution shape

The shape of a distribution is described by the following characteristics.

• Symmetry. When it is graphed, a symmetric distribution can be divided at the center so that each half is
a mirror image of the other.

• Number of peaks. Distributions can have few or many peaks. Distributions with one clear peak are
called unimodal, and distributions with two clear peaks are called bimodal. When a symmetric
distribution has a single peak at the center, it is referred to as bell-shaped (normal distribution/Gaussian
distribution).
• Skewness. When they are displayed graphically, some distributions have many more observations on
one side of the graph than the other. Distributions with fewer observations on the right (toward higher
values) are said to be skewed right and have a left tail; and distributions with fewer observations on
the left (toward lower values) are said to be skewed left and have a right tail.

• Uniform. When the observations in a set of data are equally spread across the range of the
distribution, the distribution is called a uniform distribution. A uniform distribution has no clear peaks.
remember! frequency distribution is a probability distribution! sample → population; probability of population falling
into sample! area under the curve!
a histogram reveals the shape of a frequency distribution
Unusual Features

Sometimes, statisticians refer to unusual features in a set of data. The two most
common unusual features are gaps and outliers.

• Gaps. Gaps refer to areas of a distribution where there are no observations.

• Outliers. Sometimes, distributions are characterized by extreme values that differ

greatly from the other observations. These extreme values are called outliers.
As a "rule of thumb", an extreme value is often considered to be an outlier if it is at least 1.5 interquartile
ranges below the first quartile (Q1), or at least 1.5 interquartile ranges above the third quartile (Q3)
■ Central tendency

what’s typical for a set of data

mean (media) - the arithmetic mean is the most common measure of central
tendency. It is simply the sum of the numbers divided by the number of numbers.
median (mediana) - the midpoint of a distribution: the same number of scores is
above the median as below it
the median of 2, 4, and 7 is 4
the median of the numbers 2, 4, 7, 12 is (4+7)/2 = 5.5
mode (modul) - the most frequently occurring value - for continuous data - grouped
frequency distribution = histogram
■ Dispersion/spread

how wide the set of data is

The range (amplitudine) is a basic statistic that tells you the range of values.
For example, if your minimum value is $10 and the maximum value is $100 then the range is
$90 ($100 – $10). A similar statistic is the interquartile range, which tells you the range in the
middle fifty percent of a set of data; in other words, it’s where the bulk of data tends to lie.
The standard deviation (SD) is a measure of how spread out data is around center of the
distribution (the mean)

the standard deviation is the square root of variance

The variance is a very simple statistic that gives you an extremely rough idea of how spread
out a data set is.

The average of the squared differences from the mean

Quartiles divide your data set into quarters according to where those numbers falls on the
number line. Like the variance, the quartile isn’t very useful on its own. Instead, it’s used to find
more useful values like the interquartile range.
divide your data into four segments according to where the numbers fall on the number line. The four
quarters that divide a data set into quartiles are:

The lowest 25% of numbers.

The next lowest 25% of numbers (up to the median).
The second highest 25% of numbers (above the median).
The highest 25% of numbers.
percentiles - a measure used in statistics indicating the value below which a given percentage
of observations in a group of observations fall. For example, the 20th percentile is the value (or
score) below which 20% of the observations may be found.

- two largely unequal sets of data - both large

boxplot

■ a method for graphically depicting groups of numerical data through their quartiles
■ central tendency and shape of the distribution
■ lines extending vertically from the boxes (whiskers) indicate variability outside the upper and
lower quartiles

■ the bottom and top of the box are always the first and third quartiles, and the band inside the
box is always the second quartile (the median)

■ any data not included between the whiskers should be plotted as an outlier with a dot, small
circle, or star

■ the ends of the whiskers can represent several possible alternative values:
- the minimum and maximum of all of the data
- the lowest datum still within 1.5 IQR of the lower quartile, and the highest datum still within 1.5
IQR of the upper quartile (often called the Tukey boxplot)
- one standard deviation above and below the mean of the data
■ “In a skewed distribution, the mean is farther out in the long tail than is the
median.” (Moore and McCabe 2003, p. 43)

■ “For skewed distributions, the mean lies toward the direction of skew (the longer
tail) relative to the median.” (Agresti and Finlay 1997, p. 50)

for continuous variables in a normal distribution

■ Association

relative risk
odds ratio
chi square
student t test
■ A t-test requires two variables; one must be categorical and have exactly two levels, and the
other must be quantitative and be estimable by a mean. For example, the two groups could
be Republicans and Democrats, and the quantitative variable could be age.

■ A chi-square test requires categorical variables, usually only two, but each may have any
number of levels. For example, the variables could be ethnic group — White, Black, Asian,
American Indian/Alaskan native, Native Hawaiian/Pacific Islander, other, multiracial; and
presidential choice in 2008 — (Obama, McCain, other, did not vote.
■ The t-test allows you to say either "we can reject the null hypothesis of equal
means at the 0.05 level" or "we have insufficient evidence to reject the null of
equal means at the 0.05 level."

■ A chi-square test allows you to say either "we can reject the null hypothesis of no
relationship at the 0.05 level" or "we have insufficient evidence to reject the null at
the 0.05 level."
Binary or categorical outcomes (proportions)
Are the observations correlated?
Outcome
Variable independent
Binary or Chi-square test:
categorical compares proportions
(e.g. between more than two
fracture, groups
yes/no)
Relative risks: odds
ratios or risk ratios
Logistic regression:
multivariate technique
used when outcome is
binary; gives multivariate-
adjusted odds ratios
correlated
McNemar’s chi-square test:
compares binary outcome
between correlated groups
(e.g., before and after)
Conditional logistic
regression: multivariate
regression technique for a
binary outcome when groups
are correlated (e.g., matched
data)
GEE modeling: multivariate
regression technique for a
binary outcome when groups
are correlated (e.g., repeated
measures)
Alternative to the chi-square
test if sparse cells:
Fisher’s exact test:
compares proportions between
independent groups when
there are sparse data (some
cells <5).
McNemar’s exact test:
compares proportions between
correlated groups when there
are sparse data (some cells
<5).
Chi-square test
From an RCT of probiotic supplementation during pregnancy to prevent eczema in the infant:

Table 3. Cumulative incidence of eczema at 12 months of age

Probiotics group Placebo group p-value Adjusted OR(95% CI) p-value

Cumulative incidence at 12/33 (36.4%) 22/35 (62.9%) 0.029* 0.243(0.075–0.792) 0.019†

12 months

*Significant difference between the groups as determined by Pearson's chi-square test.

†p value was calculated by multivariable logistic regression analysis adjusted for the antibiotics use, total duration of
breastfeeding, and delivery by cesarean section.

Kim et al. Effect of probiotic mix (Bifidobacterium bifidum, Bifidobacterium lactis, Lactobacillus acidophilus) in the primary prevention of eczema: a double-blind,
randomized, placebo-controlled trial. Pediatric Allergy and Immunology. Published online October 2009.
Statistical question: Does the proportion of infants with eczema differ in the
treatment and control groups?

■ What is the outcome variable? Eczema in the first year of life (yes/no)
■ What type of variable is it? Binary
■ Are the observations correlated? No
■ Are groups being compared and, if so, how many? Yes, two groups
■ Are any of the counts smaller than 5? No, smallest is 12 (probiotics group
with eczema)

 chi-square test or relative risks, or both

Chi-square test of Independence

Chi-square test allows you to compare proportions between 2 or

more groups (ANOVA for means; chi-square for proportions).
 Review Question 1
I divide my study population into smokers, ex-smokers, and never-smokers; I
want to compare years of schooling (a normally distributed variable)
between the three groups. What test should I use?

a. Repeated-measures ANOVA.
b. One-way ANOVA.
c. Difference in proportions test.
d. Paired ttest.
e. Chi-square test.
 Review Question 1
I divide my study population into smokers, ex-smokers, and never-smokers; I
want to compare years of schooling (a normally distributed variable) between
the three groups. What test should I use?

a. Repeated-measures ANOVA.
b. One-way ANOVA.
c. Difference in proportions test.
d. Paired ttest.
e. Chi-square test.
 Review Question 2
I divide my study population into smokers, ex-smokers, and never-smokers; I
want to compare the proportions of each group that went to graduate school.
What test should I use?

a. Repeated-measures ANOVA.
b. One-way ANOVA.
c. Difference in proportions test.
d. Paired ttest.
e. Chi-square test.
 Review Question 2
I divide my study population into smokers, ex-smokers, and never-smokers; I
want to compare the proportions of each group that went to graduate school.
What test should I use?

a. Repeated-measures ANOVA.
b. One-way ANOVA.
c. Difference in proportions test.
d. Paired ttest.
e. Chi-square test.
 Risk ratios and odds ratios
From an RCT of probiotic supplementation during pregnancy to prevent eczema in the infant:

Table 3. Cumulative incidence of eczema at 12 months of age

Placebo
Probiotics group group p-value Adjusted OR(95% CI) p-value

Cumulative incidence at 12/33 (36.4%) 22/35 (62.9%) 0.029* 0.243(0.075–0.792) 0.019†

12 months

*Significant difference between the groups as determined by Pearson's chi-square test.

†p value was calculated by multivariable logistic regression analysis adjusted for the antibiotics use, total
duration of breastfeeding, and delivery by cesarean section.

Kim et al. Effect of probiotic mix (Bifidobacterium bifidum, Bifidobacterium lactis, Lactobacillus acidophilus) in the primary prevention of eczema: a double-blind,
randomized, placebo-controlled trial. Pediatric Allergy and Immunology. Published online October 2009.
Corresponding 2x2 table
Treatment Group

Treatment Placebo
Eczema

+ 12 22

- 21 13
 Risk ratios and odds ratios
Statistical question: Does the proportion of infants with eczema differ in the
treatment and control groups?

■ What is the outcome variable? Eczema in the first year of life (yes/no)
■ What type of variable is it? Binary
■ Are the observations correlated? No
■ Are groups being compared and, if so, how many? Yes, binary
■ Are any of the counts smaller than 5? No, smallest is 12 (probiotics group
with eczema)

 chi-square test or relative risks, or both

Odds vs. Risk (=probability)

If the risk is… Then the odds

are…
½ (50%) 1:1
¾ (75%) 3:1
1/10 (10%) 1:9
1/100 (1%) 1:99
Note: An odds is always higher than its corresponding probability,
unless the probability is 100%.
 Risk ratios and odds ratios
■ Absolute risk difference in eczema between treatment and placebo:
36.4%-62.9%=-26.5% (p=.029, chi-square test).

36.4%
■ Risk ratio:  0.58 There is a 26.5%
62.9% decrease in absolute
risk, a 42% decrease in
■ Corresponding odds ratio: relative risk, and a 66%
decrease in relative
odds.
36.4% /(1  36.4%)
 0.34
62.9% /(1  62.9%)
Why do we ever use an odds ratio??

■ We cannot calculate a risk ratio from a case-control study

(since we cannot calculate the risk of developing the
disease in either exposure group).
■ The multivariate regression model for binary outcomes
(logistic regression) gives odds ratios, not risk ratios.
■ The odds ratio is a good approximation of the risk ratio
when the disease/outcome is rare (~<10% of the control
group)
 Interpretation of the odds ratio:

 The odds ratio will always be bigger than the corresponding

risk ratio if RR >1 and smaller if RR <1 (the harmful or
protective effect always appears larger)

■ The magnitude of the inflation depends on the prevalence

of the disease.
The odds ratio vs. the risk ratio

Rare Outcome
Odds ratio Odds ratio

Risk ratio 1.0 (null) Risk ratio

Common Outcome
Odds ratio Odds ratio

Risk ratio 1.0 (null) Risk ratio

 When is the OR is a good
approximation of the RR?

General Rule of
Thumb:
“OR is a good
approximation as
long as the
probability of the
outcome in the
unexposed is less
than 10%”
In a study that is designed and conducted as a case-control study, you cannot
calculate incidence. Therefore, you cannot calculate risk ratio or risk difference.
You can only calculate an odds ratio. However, in certain situations a case-
control study is the only feasible study design.
 Interpreting ORs when the outcome is
common…
If data are from a cross-sectional or cohort study, then you can convert ORs
(from logistic regression) back to RRs with a simple formula:

OR
RR 
(1  Po )  ( Po  OR)
Where:
OR = odds ratio from logistic regression (e.g., 3.92)
P0 = P(D/~E) = probability/prevalence of the outcome in the unexposed/reference group (e.g. ~45%)

Formula from: Zhang J. What's the Relative Risk? A Method of Correcting the Odds Ratio in Cohort
Studies of Common Outcomes JAMA. 1998;280:1690-1691.
MEDICAL RESEARCH
ETHICS
World Medical Association (WMA)

■ The World Medical Association (WMA) is an international organization representing

physicians. It was founded on 17 September 1947, when physicians from 27
different countries met at the First General Assembly of the WMA in Paris. The
organization was created to ensure the independence of physicians, and to work
for the highest possible standards of ethical behaviour and care by physicians, at
all times. This was particularly important to physicians after the Second World War,
and therefore the WMA has always been an independent confederation of free
professional associations. Funding has been by the annual contributions of its
members, which has now grown to 111 National Medical Associations.

https://www.wma.net/who-we-are/
■ As an organization promoting the highest possible standards of medical ethics, the
WMA provides ethical guidance to physicians through its Declarations, Resolutions
and Statements. These also help to guide National Medical Associations,
governments and international organizations throughout the world. The
Declarations, Resolutions and Statements cover a wide range of subjects,
including an International Code of Medical Ethics, the rights of patients, research
on human subjects, care of the sick and wounded in times of armed conflict,
torture of prisoners, the use and abuse of drugs, family planning and pollution.

■ The WMA is in official relations with the World Health Organization (WHO).
WMA

■ Declaration of Geneva - the modern Hippocratic Oath

■ Declaration of Helsinki - medical research involving human subjects
■ Declaration of Tokyo - guidelines for physicians to prevent torture
■ Declaration of Taipei - research on health databases, big data and biobanks
Adopted by the 2nd General Assembly of the World Medical Association, Geneva, Switzerland, September 1948
and amended by the 22nd World Medical Assembly, Sydney, Australia, August 1968
and the 35th World Medical Assembly, Venice, Italy, October 1983
and the 46th WMA General Assembly, Stockholm, Sweden, September 1994
and editorially revised by the 170th WMA Council Session, Divonne-les-Bains, France, May 2005
and the 173rd WMA Council Session, Divonne-les-Bains, France, May 2006
and amended by the 68th WMA General Assembly, Chicago, United States, October 2017

The Physician’s Pledge

AS A MEMBER OF THE MEDICAL PROFESSION:
I SOLEMNLY PLEDGE to dedicate my life to the service of humanity;
THE HEALTH AND WELL-BEING OF MY PATIENT will be my first consideration;
I WILL RESPECT the autonomy and dignity of my patient;
I WILL MAINTAIN the utmost respect for human life;
I WILL NOT PERMIT considerations of age, disease or disability, creed, ethnic origin, gender, nationality, political
affiliation, race, sexual orientation, social standing or any other factor to intervene between my duty and my patient;
I WILL RESPECT the secrets that are confided in me, even after the patient has died;
I WILL PRACTISE my profession with conscience and dignity and in accordance with good medical practice;
I WILL FOSTER the honour and noble traditions of the medical profession;
I WILL GIVE to my teachers, colleagues, and students the respect and gratitude that is their due;
I WILL SHARE my medical knowledge for the benefit of the patient and the advancement of healthcare;
I WILL ATTEND TO my own health, well-being, and abilities in order to provide care of the highest standard;
I WILL NOT USE my medical knowledge to violate human rights and civil liberties, even under threat;
I MAKE THESE PROMISES solemnly, freely, and upon my honor.
 Declaration of Helsinki
Adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964
and amended by the:
29th WMA General Assembly, Tokyo, Japan, October 1975
35th WMA General Assembly, Venice, Italy, October 1983
41st WMA General Assembly, Hong Kong, September 1989
48th WMA General Assembly, Somerset West, Republic of South Africa, October 1996
52nd WMA General Assembly, Edinburgh, Scotland, October 2000
53rd WMA General Assembly, Washington DC, USA, October 2002 (Note of Clarification added)
55th WMA General Assembly, Tokyo, Japan, October 2004 (Note of Clarification added)
59th WMA General Assembly, Seoul, Republic of Korea, October 2008
64th WMA General Assembly, Fortaleza, Brazil, October 2013

Preamble
1. The World Medical Association (WMA) has developed the Declaration of Helsinki as a
statement of ethical principles for medical research involving human subjects, including research on
identifiable human material and data.

The Declaration is intended to be read as a whole and each of its constituent paragraphs should be
applied with consideration of all other relevant paragraphs.

2. Consistent with the mandate of the WMA, the Declaration is addressed primarily to physicians.
The WMA encourages others who are involved in medical research involving human subjects to adopt
these principles.
General Principles

3. The Declaration of Geneva of the WMA binds the physician with the words, “The health of my patient
will be my first consideration,” and the International Code of Medical Ethics declares that, “A physician shall
act in the patient’s best interest when providing medical care.”

4. It is the duty of the physician to promote and safeguard the health, well-being and rights of patients,
including those who are involved in medical research. The physician’s knowledge and conscience are
dedicated to the fulfilment of this duty.

5. Medical progress is based on research that ultimately must include studies involving human subjects.

6. The primary purpose of medical research involving human subjects is to understand the causes,
development and effects of diseases and improve preventive, diagnostic and therapeutic interventions
(methods, procedures and treatments). Even the best proven interventions must be evaluated continually
through research for their safety, effectiveness, efficiency, accessibility and quality.

7. Medical research is subject to ethical standards that promote and ensure respect for all human
subjects and protect their health and rights.
8. While the primary purpose of medical research is to generate new knowledge, this goal can never take
precedence over the rights and interests of individual research subjects.
9. It is the duty of physicians who are involved in medical research to protect the life, health, dignity,
integrity, right to self-determination, privacy, and confidentiality of personal information of research subjects.
The responsibility for the protection of research subjects must always rest with the physician or other health
care professionals and never with the research subjects, even though they have given consent.
10. Physicians must consider the ethical, legal and regulatory norms and standards for research involving
human subjects in their own countries as well as applicable international norms and standards. No national or
international ethical, legal or regulatory requirement should reduce or eliminate any of the protections for
research subjects set forth in this Declaration.
11. Medical research should be conducted in a manner that minimises possible harm to the environment.
12. Medical research involving human subjects must be conducted only by individuals with the appropriate
ethics and scientific education, training and qualifications. Research on patients or healthy volunteers requires
the supervision of a competent and appropriately qualified physician or other health care professional.
13. Groups that are underrepresented in medical research should be provided appropriate access to
participation in research.
14. Physicians who combine medical research with medical care should involve their patients in research
only to the extent that this is justified by its potential preventive, diagnostic or therapeutic value and if the
physician has good reason to believe that participation in the research study will not adversely affect the health
of the patients who serve as research subjects.
15. Appropriate compensation and treatment for subjects who are harmed as a result of participating in
research must be ensured
Risks, Burdens and Benefits
16. In medical practice and in medical research, most interventions involve risks and burdens.
Medical research involving human subjects may only be conducted if the importance of the objective
outweighs the risks and burdens to the research subjects.
17. All medical research involving human subjects must be preceded by careful assessment of
predictable risks and burdens to the individuals and groups involved in the research in comparison with
foreseeable benefits to them and to other individuals or groups affected by the condition under investigation.
Measures to minimise the risks must be implemented. The risks must be continuously monitored, assessed
and documented by the researcher.
18. Physicians may not be involved in a research study involving human subjects unless they are
confident that the risks have been adequately assessed and can be satisfactorily managed.
When the risks are found to outweigh the potential benefits or when there is conclusive proof of definitive
outcomes, physicians must assess whether to continue, modify or immediately stop the study.
Vulnerable Groups and Individuals

19. Some groups and individuals are particularly vulnerable and may have an increased
likelihood of being wronged or of incurring additional harm.
All vulnerable groups and individuals should receive specifically considered protection.

20. Medical research with a vulnerable group is only justified if the research is responsive to the
health needs or priorities of this group and the research cannot be carried out in a non-vulnerable
group. In addition, this group should stand to benefit from the knowledge, practices or interventions
that result from the research.
Scientific Requirements and Research Protocols

21. Medical research involving human subjects must conform to generally accepted scientific principles,
be based on a thorough knowledge of the scientific literature, other relevant sources of information, and
adequate laboratory and, as appropriate, animal experimentation. The welfare of animals used for research
must be respected.

22. The design and performance of each research study involving human subjects must be clearly
described and justified in a research protocol.
The protocol should contain a statement of the ethical considerations involved and should indicate how the
principles in this Declaration have been addressed. The protocol should include information regarding
funding, sponsors, institutional affiliations, potential conflicts of interest, incentives for subjects and
information regarding provisions for treating and/or compensating subjects who are harmed as a
consequence of participation in the research study.
In clinical trials, the protocol must also describe appropriate arrangements for post-trial provisions.
Research Ethics Committees

23. The research protocol must be submitted for consideration, comment, guidance and approval to the
concerned research ethics committee before the study begins. This committee must be transparent in its
functioning, must be independent of the researcher, the sponsor and any other undue influence and must be
duly qualified. It must take into consideration the laws and regulations of the country or countries in which
the research is to be performed as well as applicable international norms and standards but these must not
be allowed to reduce or eliminate any of the protections for research subjects set forth in this Declaration.

The committee must have the right to monitor ongoing studies. The researcher must provide monitoring
information to the committee, especially information about any serious adverse events. No amendment to
the protocol may be made without consideration and approval by the committee. After the end of the study,
the researchers must submit a final report to the committee containing a summary of the study’s findings
and conclusions.
Privacy and Confidentiality
24. Every precaution must be taken to protect the privacy of research subjects and the confidentiality of
their personal information.

Informed Consent
25. Participation by individuals capable of giving informed consent as subjects in medical research must
be voluntary. Although it may be appropriate to consult family members or community leaders, no individual
capable of giving informed consent may be enrolled in a research study unless he or she freely agrees.
26. In medical research involving human subjects capable of giving informed consent, each potential
subject must be adequately informed of the aims, methods, sources of funding, any possible conflicts
of interest, institutional affiliations of the researcher, the anticipated benefits and potential risks of
the study and the discomfort it may entail, post-study provisions and any other relevant aspects of
the study. The potential subject must be informed of the right to refuse to participate in the study or to
withdraw consent to participate at any time without reprisal. Special attention should be given to the specific
information needs of individual potential subjects as well as to the methods used to deliver the information.
After ensuring that the potential subject has understood the information, the physician or another
appropriately qualified individual must then seek the potential subject’s freely-given informed consent,
preferably in writing. If the consent cannot be expressed in writing, the non-written consent must be formally
documented and witnessed.
All medical research subjects should be given the option of being informed about the general outcome and
results of the study.
27. When seeking informed consent for participation in a research study the physician must be
particularly cautious if the potential subject is in a dependent relationship with the physician or may consent
under duress. In such situations the informed consent must be sought by an appropriately qualified
individual who is completely independent of this relationship.

28. For a potential research subject who is incapable of giving informed consent, the physician must
seek informed consent from the legally authorised representative. These individuals must not be included in
a research study that has no likelihood of benefit for them unless it is intended to promote the health of the
group represented by the potential subject, the research cannot instead be performed with persons capable
of providing informed consent, and the research entails only minimal risk and minimal burden.

29. When a potential research subject who is deemed incapable of giving informed consent is able to
give assent to decisions about participation in research, the physician must seek that assent in addition to
the consent of the legally authorised representative. The potential subject’s dissent should be respected.
30. Research involving subjects who are physically or mentally incapable of giving consent, for example,
unconscious patients, may be done only if the physical or mental condition that prevents giving informed
consent is a necessary characteristic of the research group. In such circumstances the physician must seek
informed consent from the legally authorised representative. If no such representative is available and if the
research cannot be delayed, the study may proceed without informed consent provided that the specific
reasons for involving subjects with a condition that renders them unable to give informed consent have been
stated in the research protocol and the study has been approved by a research ethics committee. Consent to
remain in the research must be obtained as soon as possible from the subject or a legally authorised
representative.
31. The physician must fully inform the patient which aspects of their care are related to the research. The
refusal of a patient to participate in a study or the patient’s decision to withdraw from the study must never
adversely affect the patient-physician relationship.
32. For medical research using identifiable human material or data, such as research on material or data
contained in biobanks or similar repositories, physicians must seek informed consent for its collection,
storage and/or reuse. There may be exceptional situations where consent would be impossible or
impracticable to obtain for such research. In such situations the research may be done only after
consideration and approval of a research ethics committee.
Use of Placebo

33. The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of
the best proven intervention(s), except in the following circumstances:
Where no proven intervention exists, the use of placebo, or no intervention, is acceptable; or
Where for compelling and scientifically sound methodological reasons the use of any intervention less
effective than the best proven one, the use of placebo, or no intervention is necessary to determine the
efficacy or safety of an intervention
and the patients who receive any intervention less effective than the best proven one, placebo, or no
intervention will not be subject to additional risks of serious or irreversible harm as a result of not receiving
the best proven intervention.
Extreme care must be taken to avoid abuse of this option.
Post-Trial Provisions
34. In advance of a clinical trial, sponsors, researchers and host country governments should make
provisions for post-trial access for all participants who still need an intervention identified as beneficial in the
trial. This information must also be disclosed to participants during the informed consent process.
Research Registration and Publication and Dissemination of Results
35. Every research study involving human subjects must be registered in a publicly accessible database
before recruitment of the first subject.
36. Researchers, authors, sponsors, editors and publishers all have ethical obligations with regard to the
publication and dissemination of the results of research. Researchers have a duty to make publicly available
the results of their research on human subjects and are accountable for the completeness and accuracy of
their reports. All parties should adhere to accepted guidelines for ethical reporting. Negative and inconclusive
as well as positive results must be published or otherwise made publicly available. Sources of funding,
institutional affiliations and conflicts of interest must be declared in the publication.

Reports of research not in accordance with the principles of this Declaration should not be accepted for
publication.
Unproven Interventions in Clinical Practice
37. In the treatment of an individual patient, where proven interventions do not exist or other known
interventions have been ineffective, the physician, after seeking expert advice, with informed consent
from the patient or a legally authorised representative, may use an unproven intervention if in the
physician’s judgement it offers hope of saving life, re-establishing health or alleviating suffering. This
intervention should subsequently be made the object of research, designed to evaluate its safety and
efficacy. In all cases, new information must be recorded and, where appropriate, made publicly available.
Good Clinical Practice (GCP)
International Council of Harmonisation (ICH)

■ Good clinical practice (GCP) is an international ethical and scientific quality standard for
designing, recording and reporting trials that involve the participation of human subjects.
Compliance with this standard provides public assurance that the rights, safety and wellbeing
of trial subjects are protected and that clinical-trial data are credible.
International Council of Harmonisation of Technical Requirements for Registration
of Pharmaceuticals for Human Use (ICH) 1990, 2015

- brings together the regulatory authorities and pharmaceutical industry to discuss scientific and
technical aspects of drug registration
- to respond to the increasingly global face of drug development
- mission is to achieve greater harmonisation worldwide to ensure that safe, effective, and high
quality medicines are developed and registered in the most resource-efficient manner.
USA
Europe
Japan
■ European Commission
■ European Federation of Pharmaceutical Industries and Associations (EFPIA)
■ Ministry of Health, Labour and Welfare (Japan) (MHLW)
■ Japan Pharmaceutical Manufacturers Association (JPMA)
■ Food and Drug Administration (FDA)
■ Pharmaceutical Research and Manufacturers of America (PhRMA)
■ Medicine developers are responsible for ensuring that they and any parties working for them
comply with standards set out in European Union (EU) legislation and guidelines for good
clinical practice (GCP), good laboratory practice (GLP) and good manufacturing
practice (GMP) for investigational medicinal products.
■ The European Medicines Agency (EMA) is responsible for harmonising these standards at EU
level. It also coordinates inspections to verify that medicine developers comply with them.
■
E1 Clinical Safety for Drugs used in Long-Term Treatment
■ E2A - E2F Pharmacovigilance
■ E3 Clinical Study Reports
■ E4 Dose-Response Studies
■ E5 Ethnic Factors
■ E6 Good Clinical Practice
■ E7 Clinical Trials in Geriatric Population
■ E8 General Considerations for Clinical Trials
■ E9 Statistical Principles for Clinical Trials
■ E10 Choice of Control Group in Clinical Trials
■
E11 - E11A Clinical Trials in Pediatric Population
■ E12 Clinical Evaluation by Therapeutic Category
■ E14 Clinical Evaluation of QT
■ E15 Definitions in Pharmacogenetics / Pharmacogenomics
■ E16 Qualification of Genomic Biomarkers
■ E17 Multi-Regional Clinical Trials
■ E18 Genomic Sampling
■ E19 Safety Data Collection
■ Cross-cutting Topics
1 December 2016
EMA/CHMP/ICH/135/1995 Committee for
Human Medicinal Products

■ A standard for the design, conduct, performance, monitoring, auditing, recording, analyses,
and reporting of clinical trials that provides assurance that the data and reported results are
credible and accurate, and that the rights, integrity, and confidentiality of trial subjects are
protected.
Implementation
■ :EC, Europe - Adopted by CHMP, 15 December 2016, issued as EMA/CHMP/ICH/135/1995
■ :MHLW/PMDA, Japan - To be notified
■ :FDA, US - To be notified
■ :Health Canada, Canada - To be notified
■ :Swissmedic, Switzerland - Refer to the press release on Swissmedic, Switzerland's website
 The principles of ICH GCP

■ 2.1. Clinical trials should be conducted in accordance with the ethical principles that have
their origin in the Declaration of Helsinki, and that are consistent with GCP and the applicable
regulatory requirement(s).
■ 2.2. Before a trial is initiated, foreseeable risks and inconveniences should be weighed against
the anticipated benefit for the individual trial subject and society. A trial should be initiated
and continued only if the anticipated benefits justify the risks.
■ 2.3. The rights, safety, and well-being of the trial subjects are the most important
considerations and should prevail over interests of science and society.
■ 2.4. The available nonclinical and clinical information on an investigational product should be
adequate to support the proposed clinical trial.
■ 2.5. Clinical trials should be scientifically sound, and described in a clear, detailed protocol.
■ 2.6. A trial should be conducted in compliance with the protocol that has received prior
institutional review board (IRB)/independent ethics committee (IEC) approval/favourable
opinion.
■ 2.7. The medical care given to, and medical decisions made on behalf of, subjects should always
be the responsibility of a qualified physician or, when appropriate, of a qualified dentist.
■ 2.8. Each individual involved in conducting a trial should be qualified by education, training, and
experience to perform his or her respective task(s).
■ 2.9. Freely given informed consent should be obtained from every subject prior to clinical trial
participation.
■ 2.10. All clinical trial information should be recorded, handled, and stored in a way that allows its
accurate reporting, interpretation and verification.
■ 2.11. The confidentiality of records that could identify subjects should be protected, respecting
the privacy and confidentiality rules in accordance with the applicable regulatory requirement(s).
■ 2.12. Investigational products should be manufactured, handled, and stored in accordance with
applicable good manufacturing practice (GMP). They should be used in accordance with the
approved protocol.
■ 2.13. Systems with procedures that assure the quality of every aspect of the trial should be
implemented.
 Good Laboratory Practice (GLP)
The Organisation for Economic Co-operation and Development (OECD)

■ Good Laboratory Practice (GLP) is a quality system concerned with the organisational process
and the conditions under which non-clinical health and environmental safety studies are
planned, performed, monitored, recorded, archived and reported.

■ Principles of Good Laboratory Practice should be applied to the non-clinical safety testing of test
items contained in pharmaceutical products, pesticide products, cosmetic products, veterinary
drugs as well as food additives, feed additives, and industrial chemicals. These test items are
frequently synthetic chemicals, but may be of natural or biological origin and, in some
circumstances, may be living organisms. The purpose of testing these test items is to obtain
data on their properties and/or their safety with respect to human health and/or the
environment.
■ The principles of Good Laboratory Practice (GLP) promote the quality and
validity of data generated in the testing of chemicals and prevent fraudulent
practices.

■ The principles have been developed in accordance with the Organisation for
Economic Cooperation and Development (OECD) and the EU has adopted
these principles and the revised OECD Guides for Compliance Monitoring
Procedures for GLP as annexes to its two GLP Directives.
The Organisation for Economic Co-operation and Development (OECD)

The Organisation for Economic Co-operation and Development (OECD) is an intergovernmental

economic organisation with 35 member countries, founded in 1961 to stimulate economic
progress and world trade.

- a forum of countries committed to democracy and the market economy, providing a setting to
compare policy experiences, seek answers to common problems, identify good practices, and co-
ordinate domestic and international policies

- its mandate covers economic, environmental, and social issues. The organisation works through
consensus to develop policy recommendations and other "soft law" instruments to encourage
policy reform in member countries.
 Good manufacturing practice (GMP)

■ Good manufacturing practice (GMP) describes the minimum standard that a medicines
manufacturer must meet in their production processes. The European Medicines Agency (EMA)
coordinates inspections to verify compliance with these standards and plays a key role in
harmonising GMP activities at European Union (EU) level.

■ Any manufacturer of medicines intended for the EU market, no matter where in the world it is
located, must comply with GMP.
■ GMP requires that pharmaceutical products, dietary supplements, and medical devices:
- are of consistent high quality;
- are appropriate for their intended use;
- meet the requirements of the marketing authorisation or clinical trial authorisation.
Good epidemiological practice (GEP)

International Epidemiological Association IEA-European Federation

Research should be an activity devoted to the exploration of the laws of nature driven only by a
desire to know the truth. In the real world, other factors often interfere with this ideal aspiration
and can result in conflicts of interest. Research has to be funded, conducted and published, and
researchers like to promote their reputations. Research is, therefore, often carried out amid many
competing elements.
■ Research findings in public health should serve the public good but may not be welcomed by all.
Findings of serious side-effects of a given drug, for example, may be welcomed by those taking the
drug and their physicians, but may seriously reduce sales and the expected profits of the
manufacturer. Alternatively, a promising hypothesis may not be supported by new data.

■ We must also be sure to obtain a reasonable balance between ethical constraints and the
opportunities for legitimate research of importance to public health.
■ background to epidemiological research
■ the role of ethics committees
■ general ethical principles for research
■ informed consent
■ rules for good research behaviour under the headings of working with personal
data, data documentation, publication, and exercise of judgment
■ scientific misconduct
ETHICAL PRINCIPLES OF RESEARCH

There are four general ethical principles for research:

■ ·Autonomy (Respect for individual rights)

■ ·Beneficence (Do good)
■ ·Non-maleficence (Do no harm)
■ ·Justice
Informed consent

■ A process by which a subject voluntarily confirms his or her willingness to

participate in a particular trial, after having been informed of all aspects of the trial
that are relevant to the subject's decision to participate. Informed consent is
documented by means of a written, signed and dated informed consent form.
 Council for International Organisations of Medical Sciences (CIOMS)

Informed consent is a decision to participate in research, taken by a competent

individual who has received the necessary information; who has adequately
understood the information; and who, after considering the information, has arrived at
a decision without having been subjected to coercion, undue influence or inducement,
or intimidation.

The Council for International Organizations of Medical Sciences (CIOMS) is an international, non-
governmental, non-profit organization established jointly by WHO and UNESCO in 1949. CIOMS
represents a substantial proportion of the biomedical scientific community through its member
organizations, which include many of the biomedical disciplines, national academies of sciences
and medical research councils. CIOMS mission is to advance public health through guidance on
health research including ethics, medical product development and safety.
■ information,
■ understanding,
■ consent
 information

■ adequate disclosure of information to enable the potential participant in research to make

an informed choice
■ the language used in the oral and written information about the trial, including the written
informed consent form, should be as non-technical as practical and should be
understandable to the subject or the subject's legally acceptable representative and the
impartial witness, where applicable
■ the subject or the subject's legally acceptable representative should be provided ample
time and opportunity to inquire about details of the trial and to decide whether or not to
participate in the trial. All questions about the trial should be answered to the satisfaction
of the subject or the subject's legally acceptable representative.
Both the informed consent discussion and the written informed consent form and any other written
information to be provided to subjects should include explanations of the following:

■ That the trial involves research.

■ The purpose of the trial.
■ The trial treatment(s) and the probability for random assignment to each treatment. (case-
control; genetic, etc)
■ The trial procedures to be followed, including all invasive procedures.
■ The subject's responsibilities.
■ Those aspects of the trial that are experimental.
■ The reasonably foreseeable risks or inconveniences to the subject and, when applicable, to
an embryo, fetus, or nursing infant.
■ The reasonably expected benefits. When there is no intended clinical benefit to the subject,
the subject should be made aware of this.
■ The alternative procedure(s) or course(s) of treatment that may be available to the subject,
and their important potential benefits and risks.
■ The compensation and/or treatment available to the subject in the event of trial-related
injury.
■ The anticipated prorated payment, if any, to the subject for participating in the trial.
■ The anticipated expenses, if any, to the subject for participating in the trial.
■ That the subject's participation in the trial is voluntary and that the subject may refuse to
participate or withdraw from the trial, at any time, without penalty or loss of benefits to which
the subject is otherwise entitled.
■ That the monitor(s), the auditor(s), the IRB/IEC, and the regulatory authority(ies) will be
granted direct access to the subject's original medical records for verification of clinical trial
procedures and/or data, without violating the confidentiality of the subject, to the extent
permitted by the applicable laws and regulations and that, by signing a written informed
consent form, the subject or the subject's legally acceptable representative is authorizing
such access.
■ That records identifying the subject will be kept confidential and, to the extent permitted by
the applicable laws and/or regulations, will not be made publicly available. If the results of
the trial are published, the subject’s identity will remain confidential.
■ That the subject or the subject's legally acceptable representative will be informed in a timely
manner if information becomes available that may be relevant to the subject's willingness to
continue participation in the trial.
■ The person(s) to contact for further information regarding the trial and the rights of trial subjects,
and whom to contact in the event of trial-related injury.
■ The foreseeable circumstances and/or reasons under which the subject's participation in the trial
may be terminated.
■ The expected duration of the subject's participation in the trial. The approximate number of
subjects involved in the trial.
■ The written informed consent form and any other written information to be provided to subjects
should be revised whenever important new information becomes available that may be relevant
to the subject’s consent. Any revised written informed consent form, and written information
should receive the IRB/IEC's approval/favourable opinion in advance of use. The subject or the
subject’s legally acceptable representative should be informed in a timely manner if new
information becomes available that may be relevant to the subject’s willingness to continue
participation in the trial. The communication of this information should be documented.
 understanding

■ Individuals should be able to understand what they are told and to make a reasoned choice
based on that information. Experience shows that this ideal is difficult to achieve. Too much
information may be given in excessive detail, mainly to protect the researcher and the
institution. An understanding of the main ideas and risks of the study may be more important
than being informed about all the specific scientific details.
■ When a clinical trial (therapeutic or non-therapeutic) includes subjects who can only be
enrolled in the trial with the consent of the subject’s legally acceptable representative
(e.g., minors, or patients with severe dementia), the subject should be informed about the
trial to the extent compatible with the subject’s understanding and, if capable, the subject
should sign and personally date the written informed consent.
 consent

■ Neither the investigator, nor the trial staff, should coerce or unduly influence a subject to
participate or to continue to participate in a trial.

■ None of the oral and written information concerning the trial, including the written informed
consent form, should contain any language that causes the subject or the subject's legally
acceptable representative to waive or to appear to waive any legal rights, or that releases or
appears to release the investigator, the institution, the sponsor, or their agents from liability for
negligence.
■ Prior to a subject’s participation in the trial, the written informed consent form should be
signed and personally dated by the subject or by the subject's legally acceptable
representative, and by the person who conducted the informed consent discussion.

■ If a subject is unable to read or if a legally acceptable representative is unable to read, an

impartial witness should be present during the entire informed consent discussion. The
witness should sign and personally date the consent form. By signing the consent form,
the witness attests that the information in the consent form and any other written
information was accurately explained to, and apparently understood by, the subject or the
subject's legally acceptable representative, and that informed consent was freely given by
the subject or the subject’s legally acceptable representative
■ In obtaining and documenting informed consent, the investigator should comply
with the applicable regulatory requirement(s), and should adhere to GCP and to the
ethical principles that have their origin in the Declaration of Helsinki. Prior to the
beginning of the trial, the investigator should have the IRB/IEC's written
approval/favourable opinion of the written informed consent form and any other
written information to be provided to subjects
 Research Misconduct
Plagiarism

European Federation of Academies of Sciences and Humanities (ALLEA)

- founded in 1994
- currently brings together 59 Academies in more than 40 countries from the Council of Europe
region

ALLEA Permanent Working Group on Science and Ethics (PWGSE)

Horizon 2020 funded ENERI project (European Network of Research Ethics and Research Integrity)
- aims to train experts in ethics related issues and to harmonise research integrity infrastructures
across Europe
"The European Code of Conduct for Research Integrity" developed in 2011 by All European
Academies (ALLEA) and the European Science Foundation (ESF)

Research misconduct is traditionally defined as fabrication, falsification, or plagiarism (the so-

called FFP categorisation) in proposing, performing, or reviewing research, or in reporting
research results:

• Fabrication is making up results and recording them as if they were real.

• Falsification is manipulating research materials, equipment or processes or changing,

omitting or suppressing data or results without justification.

• Plagiarism is using other people’s work and ideas without giving proper credit to the original
source, thus violating the rights of the original author(s) to their intellectual outputs.
 unacceptable practices

• Manipulating authorship or denigrating the role of other researchers in publications.

• Re-publishing substantive parts of one’s own earlier publications, including translations, without
duly acknowledging or citing the original (‘self-plagiarism’).
• Citing selectively to enhance own findings or to please editors, reviewers or colleagues.
• Withholding research results.
• Allowing funders/sponsors to jeopardise independence in the research process or reporting of
results so as to introduce or promulgate bias.
• Expanding unnecessarily the bibliography of a study
• Accusing a researcher of misconduct or other violations in a malicious way.
• Misrepresenting research achievements.
• Exaggerating the importance and practical applicability of findings.
• Delaying or inappropriately hampering the work of other researchers.
• Misusing seniority to encourage violations of research integrity.
• Ignoring putative violations of research integrity by others or covering up inappropriate responses
to misconduct or other violations by institutions.
• Establishing or supporting journals that undermine the quality control of research (‘predatory
journals’).
ENERI project (European Network of Research Ethics and Research Integrity)

The bar chart below shows a comparison of AIMM scores for the 27 countries surveyed, according to scores based on survey results for nine categories
as shown. The categories were scored in the range 0-4, the nine scores were summed (equal weighting) to give a maximum score of 36
Consiliul Național de Etică a Cercetării Științifice, Dezvoltării Tehnologic și Inovării
(CNECSDTI)

privind buna conduita in cercetarea

Lege nr. 206/2004
stiintifica, dezvoltarea tehnologica si inovare

Legea nr. 319/2003 Codul de etica

Legea nr. 398/2006 publicata in Monitorul Oficial, Partea I nr. 892 din 02/11/2006. -
Ordonanta nr. 28/2011 publicata in Monitorul Oficial, Partea I nr. 628 din
02/09/2011
■ plagiatul - expunerea intr-o opera scrisa sau o comunicare orala, inclusiv in format
electronic, a unor texte, expresii, idei, demonstratii, date, ipoteze, teorii, rezultate ori
metode stiintifice extrase din opere scrise, inclusiv in format electronic, ale altor autori, fara
a mentiona acest lucru si fara a face trimitere la sursele originale;

■ autoplagiatul - expunerea intr-o opera scrisa sau o comunicare orala, inclusiv in format
electronic, a unor texte, e) autoplagiatul - expunerea intr-o opera scrisa sau o comunicare
orala, inclusiv in format electronic, a unor texte, expresii, demonstratii, date, ipoteze, teorii,
rezultate ori metode stiintifice extrase din opere scrise, inclusiv in format electronic, ale
aceluiasi sau acelorasi autori, fara a mentiona acest lucru si fara a face trimitere la sursele
originale
 evitarea plagiatului

■ NU folositi copy/paste
■ mai mult de 3 cuvinte fara semnele citarii si bibliografie - plagiat; USA mai mult de 8 cuvinte -
federal offence
■ parafrazarea, rezumatul, reformularea frecventa SI fara citarea sursei in text - plagiat
■ citarile/bibliografia din text trebuie sa se regaseasca in lista bibliografica de la finalul lucrarii si sa
corespunda informatiei respective
■ orice material grafic refolosit NUMAI cu permisiunea autorilor
■ utilizarea propriilor cuvinte, idei, informatii publicate anterior fara citare - autoplagiat
■ cunostintele generale (“omul este un mamifer”) nu necesita citare
■ rezumat, rezultate, concluzii - fara citari