Transcript EVIDENCE-BASED MEDICINE METHODOLOGICAL PRINCIPLES

INTRODUCTION
Why Do I Need to Know About
Methodology ?
???????
 As
a graduate student...
– To be able to read and understand the
empirical literature in your field; to become a
critical consumer of information.
 As
a graduate student preparing for a
thesis or dissertation…
– To be able to both design and implement your
thesis or dissertation as well as future studies
that interest you.
????
 As
a future practitioner…
– To be able to intelligently participate in
research projects, evaluations, and studies
undertaken by your institution.
 As
an educated citizen ...
– To understand the difference between
scientifically acquired knowledge and other
kinds of information.
What’s the Difference Between
“Method” and “Methodology”?
Method:
 Techniques for gathering evidence
 The various ways of proceeding in
gathering information
Methodology:
 The underlying theory and analysis of how
research does or should proceed.
(Sandra Harding)
What It Takes?







Creativity
Open mind
Curiosity
Patience
Persistence
Positive Attitude
Discipline and focus
Research Sequence
Topic selection
 Research planning
 Literature survey
 Formulating the problem
 Creating new solutions
 Verification of analytical results
 Communication of results
 Commercialization of research outcomes

How Extraordinary Creative
Ideas Occur?
 Sudden
spontaneous visions
 Dreams
 Cross-pollination
from different fields
Stories of Extraordinary Inventors
Sudden Vision Discoveries
 Tesla's
idea of the rotating magnetic
field came to him instantly while he
was walking in a park.
 He
drew a picture of the rotating
magnetic field in the ground of the
park.
Stories of Extraordinary Inventors
Sudden Vision Discoveries
 The
great mathematician Gauss
proved in an instant a theorem on
which he had worked unsuccessfully
for four years. "As a sudden flash of
light, the enigma was solved. . . .“
 Similar accounts given by
extraordinary creative people such
as Mozart, Tchaikovsky, Poincare
etc.
Stories of Extraordinary Inventors
Dream Discoveries
 Frederick
Kekule fell asleep and
dreamed of the benzene molecule
as a snake biting its tail.
 Otto Loewi had a dream that led to
his discovery of the chemical
transmission of nerve impulses.
.
HOW TO SEARCH THE
LITERATURE
.


BASIC SEARCH
ADVANCED SEARCH
National Library of Medicine(PUBMED):
www.ncbi.nlm.nih.gov/pubmed
 MEDLINE
.
Medline is the National
Library of Medicine's (NLM) premier
bibliographic database covering the
fields of medicine, nursing, dentistry,
veterinary medicine, the health care
system, and the preclinical sciences
as far back as 1966.
.
MeSH a Medical Subject Heading.
A controlled vocabulary term used by all
the indexers in an organization to ensure
consistency in assigning terms to articles
(or records) on the same topic.
 They are also used in searching to avoid
having to think of all the possible words
that various authors could have used to
express the same concept.

For example:
The MeSH AGED is assigned instead of:
 aging,
 older,
 elderly,
 senior citizen,
 geriatric person.
BASIC SEARCH STRATEGY




Pick a search topic: for example, previous BMD
studies in the literature?
Divide topic into concepts
Find words (keywords) or subject headings
(MeSH) to represent these concepts
Combine concepts with AND, OR, or NOT
When should I Use Basic or
Advanced Search Mode?
Here are some general rules to go by !
Basic


Use the Basic Search when you have only
a single term to search.
Use the Basic Search when you are only
searching by the author.
Advanced



Use the Advanced Search when you have
more than one concept to search.
Use the Advanced Search when you want
to search in a specific field.
Use the Advanced Search when you want
to limit search by language or publication
type
BASIC SEARCH

Enter your search term(s) in the query box.
You may use "OR", "AND", or "NOT";
PubMed will automatically combine terms
with "AND" using automatic term mapping.

Terms are matched against a MeSH
translation table, a Journals translation table,
a phrase list, and an Author index -- in that
order.
.
 PubMed
will display your results as
brief citations in batches of 20 per
page, by default.
 To change the "Number of documents
to display per page:", select a higher
number from the pull-down menu
below the query box.
.
 The
Entrez Date is the date the citation
was added to the PubMed database,
which differs from the Date of
Publication.
 Use the "Entrez Date limit" pull-down
menu to select a range of days or years
Author names should be entered in the form
Smith JB, but initials are optional. PubMed
automatically truncates the author's last name
to allow for varying initials and designations
such as Jr. or 2nd. If only the author's last name
is entered (no initials), PubMed will search that
name in All Fields not just the Author field.
•Journal titles may be entered in full:
•as valid MEDLINE abbreviations:
•or as ISSN numbers:
Citation display format displays the source, title,
author and affiliation, abstract, MeSH and unique
identifier
Advanced Search
Searching in the different fields
In Advanced search, you now have the option to
search in the different Medline fields
Just click on the arrow in the “Search field” box to
see the list
Highlight the one you want and click once
How to combine search terms

Advanced search’s main feature is the ability
to do a search with 2 or more terms

The first step is to select the search terms
that apply to your question using MeSH
Browser

Example: are there any articles on the
diagnosis and chemoprevention of TB?

The search terms for this query would be:
•
tuberculosis
•
diagnosis
•
prevention
.
Article
Types
ORIGINAL RESEARCH
 Original
Articles are scientific
reports of the results of original
clinical research.
The text is limited to 2700 words,
with an abstract, a maximum of 5
tables and figures (total), and up to
40 references.
ORIGINAL RESEARCH
 Special
Articles are scientific
reports of original research in such
areas as economic policy, ethics, law,
and health care delivery.
The text is limited to 2700 words,
with an abstract, a maximum of 5
tables and figures (total), and up to
40 references
CLINICAL CASES
 Brief
Reports usually describe one
to three patients or a single family.
The text is limited to 2000 words, a
maximum of 3 tables and figures
(total), and up to 25 references.
They begin with a brief summary of
no more than 100 words.
CLINICAL CASES
 Clinical
Problem-Solving
manuscripts consider the step-bystep process of clinical decision
making.
Information about a patient is
presented to an expert clinician.
CLINICAL CASES
 The
clinician responds (in regular
type) as new information is
presented, sharing his or her
reasoning with the reader.
The text should not exceed 2500
words, and there should be no more
than 20 references
REVIEW articles
 Review
articles are usually solicited
by the editors.
All review articles undergo the same
peer-review and editorial process as
original research reports.
Clinical Practice articles

Are evidence-based reviews of
topics relevant to practicing
physicians, both primary care
providers and specialists.
Clinical Practice articles

Articles in this series should include:
-the clinical problem
-strategies and evidence
-areas of uncertainty
-guidelines from professional
societies
-conclusions and recommendations.
Clinical Practice articles
The text is limited to 2500 words
and a small number of figures and
tables.
 These articles do not include an
abstract.

Clinical Therapeutics articles
 Are
evidence-based reviews of topics
relevant to practicing physicians.
 The series focuses on clinically
oriented information about specific
forms of therapy:
-drugs
-devices
-procedures
Clinical Therapeutics articles

Each article in the series begins with
a clinical vignette describing a
patient with a specified condition for
whom the treatment under
discussion has been recommended.
Clinical Therapeutics articles








definition of the clinical problem
patho- physiology and how the therapy works
clinical evidence
clinical use (including costs)
adverse effects
areas of uncertainty
guidelines
recommendations
The text is limited to 2500 words. These articles do
not include an abstract.
Current Concepts articles
 Focus
on clinical topics, including
those in specialty areas but of wide
interest.
The text is limited to 2400 words,
with a maximum of 4 figures and
tables (total), and up to 50
references. These articles do not
include an abstract.
Drug Therapy articles
 Detail
the pharmacology and use of
specific drugs or classes of drugs, or
the various drugs used to treat
particular diseases.
The text is limited to 4000 words,
with a maximum of 6 figures and
tables (total), and up to 120
references. These articles do not
include an abstract.
Mechanisms of Disease articles
 Discuss
the cellular and molecular
mechanisms of diseases or
categories of diseases.
The text is limited to 3500 words,
with a maximum of 6 figures and
tables (total), and up to 100
references. These articles do not
include an abstract.
Medical Progress articles
 Provide
comprehensive, scholarly
overviews of important clinical
subjects, with the principal (but not
exclusive) focus on developments
during the past five years.
Medical Progress articles

Each article details how the perception of
a disease, disease category, diagnostic
approach, or therapeutic intervention has
evolved in recent years.
The text is limited to 3500 words, with a
maximum of 6 tables and figures (total),
and up to 100 references. These articles
do not include an abstract.
OTHER SUBMISSIONS
.
Editorials
Usually provide commentary and analysis
concerning an article in the issue of the
Journal in which they appear.
 They may include 1 figure or table.
 They are nearly always solicited, although
unsolicited editorials may occasionally be
considered.
 Editorials are limited to 1000 words, with
up to 15 references

Perspective articles
 Cover
a wide variety of topics of
current interest in health care,
medicine, and the intersection
between medicine and society.
Perspective articles are limited to
1000 to 1200 words and usually
include one figure. There is a
maximum of 5 references
Sounding Board articles
 Are
opinion essays.
 They are similar to editorials but are
not tied to a particular article.
 They often present opinions on
health policy issues and are normally
unsolicited. The text is limited to
2000 words.
Clinical Implications of Basic
Research articles
 Discuss
single papers from preclinical
journals.
 The purpose is to explain the findings
and comment on their possible
clinical applications in fewer than 750
words. There may be 1 figure and up
to 4 references.
Special Reports
 Are
miscellaneous articles of special
interest to the medical community.
They are limited to 2700 words.
Letters to the Editor
 Provide
a forum for readers to
comment about articles recently
published in the Journal, and they
are a place to publish concise
articles, such as reports of novel
cases.
.
Reading medical
articles critically
Getting Ready
Choose a few good peer-reviewed
journal articles
 Relevant


to biomedical informatics
Interesting for you & audience
Important result or approach
Getting Ready
Send choices to instructor for feedback
(2 weeks before your presentation)
 Choose one article
 Read critically
 Take notes while reading
 Reread even more critically
 Skim related articles or web information
Creating the Content
 Introduce
yourself
 Why did you choose this paper?
 How does it relate to your interests?
 Summarize article (briefly!)
 Provide context
 Critique, question, react
 Conclude
Summarize Article
Assume audience read the paper
 Do not assume audience understood it
 Provide context
When?
Where?
Why?
 State authors’ take home messages
 Focus on interesting/controversial issues

Critique, Question, React
Objectively
􀃘 Did the author(s) support their point?
􀃘 Was their support valid?
 Subjectively
􀃘 How does it relate to your own
experiences?
􀃘 Why did you find this paper interesting or
important?
􀃘 What do you think the impact of this
paper is?

Conclude
 Restate
author’s take-home message
 State your own take-home message
 Provide a personal perspective
 Be provocative
Summary
 Choose
a good article
 Introduce yourself
 Summarize article
 Provide context
 Critique, question, react
 Conclude
.
Clinical Trials
What Are Clinical Trials?
 Research
 Try
studies involving people.
to answer scientific questions and
find better ways to prevent, diagnose, or
treat disease
Why Are Clinical Trials
Important?
 Clinical
trials translate results of
basic scientific research into better ways
to prevent, diagnose, or treat disease
 The more people take part, the faster we
can:
- Answer critical research questions
- Find better treatments and ways to
prevent disease
What Are the Different Types
of Clinical Trials?
 Treatment
 Prevention
 Early
detection/screening
 Diagnostic
 Quality of life/supportive care
Treatment Trials
 What
new treatments can help
people with a particular disease?
 What
is the most of effective
treatment for people with that
disease?
Prevention Trials
Evaluate the effectiveness of
ways to reduce the risk of a particular
disease
 Enroll healthy people at high risk
for developing that disease

Prevention Trials
 Action
studies (“doing something”)
studies (“taking something”)—also
called “chemoprevention studies”
 Agent
Chemoprevention Trials
 Phase
3 chemoprevention trials
compare a promising new agent with
either a:
--Standard agent
--Placebo
Fundamentals of Clinical Trials
A clinical trial is a prospective study
assessing the effect and value of
intervention(s) vs. control in human
subjects
Each clinical trial has a PRIMARY question
(outcome)
There may be multiple SECONDARY
outcomes

Clinical Trial Phases
Phase 1 trials
 How does the agent affect the human
body?
 What
dosage is safe?
Clinical Trial Phases
Phase 2 trials
 Does the agent or intervention have an
effect on the disease?
Clinical Trial Phases
Phase 3 trials
 Is the new agent or intervention (or new
use of a treatment) better than the
standard?
 Participants have an equal chance to be
assigned to one of two or more groups
Randomized Trials
Participants have an equal chance to be assigned
to one of two or more groups:
 One
gets the most widely accepted treatment
(standard treatment).
 The other gets the new treatment being tested,
which researchers hope and have reason to
believe will be better than the standard
treatment.
Why is Randomization
Important?
 So
all groups are as alike as possible.
 Provides
the best way to prove the
effectiveness of a new agent or
intervention.
Clinical Trial Protocol
 Strict
scientific guidelines:
-Purpose of study
-How many people will participate
-Who is eligible to participate
-How the study will be carried out
-What information will be gathered
about participants
-Endpoints
Benefits of Participation
Possible benefits:
 Patients will receive, at a minimum, the
best standard treatment (if one exists)
 If the new treatment or intervention is
proven to work, patients may be among
the first to benefit
 Patients have a chance to help others
and improve patient care
Risks of Participation
Possible risks:
 New
treatments or interventions under
study are not always better than, or
even as good as, standard care
 Even if a new treatment has benefits, it
may not work for every patient
Patient Protection
 There
have, unfortunately, been past
abuses in patient protection
 Federal regulations ensure that people
are told about the benefits, risks, and
purpose of research before they agree
to participate
How Are Patients’ Rights
Protected?
 Informed
consent
 Scientific review
 Institutional review boards (IRBs)
 Data safety and monitoring boards
(DSMBs)
.
.
Time (In Years) To Develop A New
Drug (Average) Pre-Clinical Testing
To Approval 1960s-1990s (US FDA)
Pre-clinical
Clinical
Approval
1960s
3.6
2.8
2.4
1970s
4.6
4.4
2.1
1980s
4.7
5.7
2.8
1990s
4.7
6.6
1.9
Phase
Phase
Phase
Pitfalls Of The Current Model Of Drug Discovery & Development.
- High and unaffordable costs of R&D and
consequently of new drugs.
- Too many products with identical
pharmacological profile and mechanism – the
milligram battle.
- Low therapeutic rationale and advance for new
drugs.
- Many products especially biotech products
developed through technology push rather than
medical demand pull.
Can We Make Regulatory Submissions More Crisp And Meaningful
No. of Words In Documents
Pythagoras
Theorem
24
Archimedes
Principle
67
The
Ten Commandments
American
Declaration of
Independence
European Legislation on when and
where one can smoke
Average
IND Submission to FDA
179
300
24,942
800,000
At The Same Time Regulations in New
Drugs Research Are Needed To:
- To ensure safety and efficacy of New Drugs by an
independent authority
- To ensure that uniform and well-laid out
standards apply to all products
- To ensure that products are continuously
monitored, post-marketing
- To review safety and efficacy standards based on
new knowledge
- To recommend appropriate amendments to
Drugs and Cosmetics Act
FDA REGULATIONS MOSTLY A REACTIVE RESPONSE
1906
History of FDA, USA, concerned only with
purity.
1936
Elixir Sulfanilamide 10% solution in 70%
Diethylene Glycol implicated in 105 deaths.
1938
Federal Food, Drug & Cosmetic Act passed
concerned with safety, not efficacy – NDA to
be approved.
1961
Thalidomide disaster in Europe.
1962
Kefauver-Harris Amendment passed regarding
extensive safety and efficacy studies – IND
mandatory.
1987
Format of IND changed.
.
Criticisms & Negative Perceptions On
Clinical Trials
 Trial objectives skewed in favor of
potential positive outcomes.
 Cutting out tests likely to end in negative
results.
 Manipulation of Subject inclusion &
exclusion criteria.
 Outright fraud in selection of investigators
with vested interests.
 Suppression of publication of negative
results.
To Eliminate Or Minimize These Negatives,
Clinical Trials Need To Be Conducted Under
Internationally Accepted Good Clinical
Practice (GCP) Guidelines.
 GCP protects patients/subjects.

GCP ensures that clinical trials produce accurate,
credible data by:
- defining standards
- defining responsibilities
Good Clinical Practice
1962
1961
Drug Amendments Act
Thalidomide
1963
IND Procedure
ORIGINS
1977
1964
Proposed FDA Regulations covering
obligations of Sponsors, Monitors
and Clinical Investigators
Declaration of Helsinki
The Emergence Of ICH
Guidelines
ICH was evolved to negotiate common
standards for the regulation of pharmaceutical
products in Europe, Japan and U.S.A.
According to its Mission Statement –
ICH exists “to provide a forum for a
constructive dialogue between the regulatory
authorities and the pharma industry on the real
and perceived differences in the technical
requirements for product registration”.
ICH -International Conference on
Harmonization of Technical Requirements for
Registration of Pharmaceuticals for Human
Use
(Objectives)
- Eliminate redundant & duplicative technical
development.
- Expedite global development.
- Expedite availability of new medicines.
- Maintaining safeguards on quality, safety &
efficacy.
ICH
Some key areas of interest :
 Safety
reporting/Adverse Events
(definitions and timings)
 New standards/templates (e.g.,
protocol, investigator’s brochure)
 Essential documents
Clinical Research An important Component Of
The Regulatory System Has The Following
Components
-Candidate Drug
-Trial Sponsor
- CRO/Monitor
- Investigators & Trial Centers
- Trial Subjects : Healthy Volunteers &
Patients
- Biostatisticians
The Investigator’s Obligations
are spread over three phases.
1) Prior to Initiation of the trial
2) During The Trial
3) Post Trial
Prior To The Trials
- The investigator should be excited about the
study for its scientific merit rather than other
considerations.
- Should ensure that confidentiality should be
maintained.
- Should have sufficient time allocated to involve in
the trials
- Should be familiar with the product and well
acquainted all pre-clinical data & IND dossier.
- Ensure that there are adequate resources
available & allocated.
- Should discuss the Protocol & details of trials with
Sponsor/Monitor.
- Finalize local clearances, IEC, IRB etc as per the
ICH guidelines, Informed Consent Forms .
During The Clinical Studies
- Investigator should prepare a File containing all documents
related to the Trial which should be kept in a secure area
accessible to only him & his staff.
- Patient identification codes & details should be preserved at
least for 15 years and should be destroyed only with prior
permission from the sponsor.
- General medical Care should be provided to all subjects by the
investigator or family physician.
- Ensure strict adherence to randomization, unblinding etc as per
the protocol.
-- Safety & adverse reaction reports to be regularly obtained and
action taken as per protocol to cause no or least damage to
subjects. (ICH 4.11)
-- Ensure proper completion/validation of the Case record Forms..
- Assess causality in terms of ‘not related, unlikely, possible,
probable and most probable’, send reports and take corrective
action.
INVESTIGATOR SOPs
Objectives
Ensuring
that the investigator understands the
responsibilities and obligations of the study.
Planning & conducting the study as per the approved
protocols and complying with ICH/GCP guidelines.
Ensuring that the safety & welfare of the subjects are
always the prime concern.
Proper and accurate collection , Documentation &
analysis of Data from the trials.
Cooperate in inspections, monitoring and auditing of
the study by third parties.
Investigators To Be Familiar With Documents Used By
Ethics Committee.
-GCP Guideline same
-Investigator Brochure and safety Information.
-Trial Protocol.
-Consent Forms & Trial Information Sheets
-Subject Recruitment Procedures.
-Information on payment & remuneration to subjects.
-Any Amendments To The Protocols Or SOPs..
-Any other document required by IEC/IRB.
See ICH Guidelines 3.1.2.
Nature Of Regulatory Inspections
1) Study related Inspections
2) Investigator Related Inspections
-
-
Based on the pivotal nature of the study
Sponsor’s Difficulties in getting some Reports.
-
Violation of trial protocol.
-
Work involved turns out to be outside
competence of the investigator .
-
Results at variance with those of other
investigators.
Inspection Reporting Systems
1) No lapse.
2) Requires corrective action for remediable
lapses.
3) Warning letter if corrective action not taken
or delayed, with copies to sponsor
&reviewing IRB.
4) Inviting for possible hearings.
5) Disqualification when the Investigator has
deliberately violated the Agency’s regulatory
standards or submitted false information.
Clinical Trials Deficiencies* (1999)
Apart From Deliberate Fraud, Clinical Trials fail due to following
categories of deficiencies
Protocols
-
28%
Records & Documentation - 20%
Adverse Drug Reaction Reporting - 15%
Informed Consent
-
10%
Drug Accounting
-
10%
*Data From U.S.FDA.
FIDDES CASE- FRAUD AT ITS
WORST
From early 1990s Dr. Fiddes, President
of a California based CRO had
conducted over 200 Clinical Trials for
47 companies. Engaged in extensive
fraudulent and falsified data, he was
sentenced and jailed in Federal Prison
for 15 months, a penalty of
$ 800000 imposed in 1998 and was
disqualified as a clinical investigator in
1999.
Glaxo Paxil Case-Emergence Of Transparency In Clinical Trials
In the lawsuit against GSK, New York Attorney General Eliot Spitzer
asserted that a novel fraud of suppression of information was committed in

the promotion of Paxil for use as an anti-depressant.
In August 2004, Glaxo started posting the full details of Clinical trials in
their Website, creating a Clinical register and started inclusion of Safety &
Efficacy data, off-label drug use issues etc in Medical Information Letters to
Physicians... GSK also paid a fine of $ 2.5 million.
Profile of Regulatory Agencies: More
applicable to Developing Countries
- Need to balance the interests of the consumers as well
as the industry.
- Under-staffed and over-burdened: working on shoestring
budgets.
- No independent database to arrive at timely and scientific
judgements.
-Vulnerable to pressures from politicians and consumer
activists.
--No control
Investigators.
on
recruitment
of
Clinical
Centers
&
- Dependency on ‘Experts’ who have little stake in the
impact of their judgements on consumers or producers.
Not equipped physically or technically to ensure
compliance with GCP/ICH Guidelines
Fundamentals of Clinical Trials
􀃘 The study population must be defined in
advance, with clear rationale and
eligibility criteria
􀃘 A control group must always be used
against which the new intervention can
be compared
􀃘 Randomization is the most reliable way
to assign participants to the treatment
groups
Fundamentals of Clinical Trials
A double blind design is the best way
to avoid bias during collection of data
If not possible, single blind and other
methods should be utilized
 Calculation of sample size should be
sufficient to provide adequate power
and levels of statistical significance

Fundamentals of Clinical Trials
Relevant BASELINE data should be
acquired before initiation of
intervention
 All efforts should be made to collect
accurate and consistent data
 Need plan for assessment, analysis
and reporting of adverse effects (AEs)
and serious adverse events (SAEs)

Evidence Based Medicine: Levels
of Evidence
Class I
1. Randomized double-blind, placebo-controlled trials
2. Meta-analyses of such RDBPC trials
 Class II
• Observational trials (case-control studies or concurrent
control studies)
 Class III
• All other controlled trials (including well-defined natural
history controls or patients serving as own controls) in a
representative population, where outcome assessment
is independent of patient treatment.
 Class IV:
• Evidence from uncontrolled studies, case series, case
reports, or expert opinion.

Fundamentals of Clinical Trials
FDA requires evidence of SAFETY and
EFFICACY to consider approving medications
for prescription use.
 The process by which medications come to
market goes through:
􀃘 preclinical development
􀃘 dose finding
􀃘 then safety trials
􀃘 exploratory efficacy/safety trials
􀃘 larger scale trials to determine the size and
consistency of clinical effects, and develop more
safety data

Drug Development Process
Discovery
(2 - 10 years)

Pre-clinical Testing
Laboratory and animal testing
 Phase 1- 20-80 healthy volunteers used to
determine safety and dosage
 Phase 2 - 100-300 patient volunteers used to
look for efficacy (POC) and side effects
 Phase 3 - 3,000-5,000 patient volunteers used
to monitor adverse reactions to long-term use
 FDA Review/Approval

Diagnostic Test Studies
 How
well does a test
identify people with the disease?
exclude people without the disease?
 Compare test results on people with
the disease with test results on
people without the disease.
 Need to know who has the disease.
Diagnostic Test Studies
Two designs
 Prospective or cohort design, or crosssectional design: take a sample of subject
eligible for the test, test them all and get
true diagnosis on them all.
 Retrospective or case-control design: take
a sample with true diagnosis established
as positive and another sample of
controls. We may have negative diagnosis
established on controls and we may not.

Who has the disease?
True diagnosis.
 We can never be absolutely sure that the
‘true’ diagnosis is correct.
 We decide to accept one method as ‘true’:
call this the gold standard or reference
standard.
 Often more invasive than the test, e.g.
histopathology compared to ultrasound
image.
 It is always possible that the reference
standard is wrong for some subjects.
Statistics of diagnostic test
studies
Sensitivity
 Specificity
 Receiver operating characteristic curve
(ROC curve)
 Likelihood ratio (LR) for positive test
 Likelihood ratio (LR) for negative test
 Odds ratio (OR)
 Positive predictive value (PPV)
 Negative predictive value (NPV)

Statistics of diagnostic test
studies
 Example:
diabetic eye tests (cross-
sectional)
test = direct opthalmoscopy
reference standard = slit lamp
stereoscopic biomicroscopy
 Single sample of subjects all received
reference standard test.
Statistics of diagnostic test
studies
 Sensitivity
= proportion of reference
positive cases who are positive on
the test = proportion of true cases
that the test finds.
 Specificity = proportion of reference
negative cases who are negative on
the test = proportion of true noncases that the test finds.
Example: eye disease in
diabetics
 45
reference standard positive cases
of whom 40 were positive on the
test, 275 reference standard
negative noncases of whom 237
were negative on the test.
 Sensitivity = 40/45 = 0.89 = 89%.
 Specificity = 237/275 = 0.86 = 86%.
Statistics of diagnostic test
studies
 Positive
predictive value (PPV) =
proportion of test positives who are
reference positive.
 Negative predictive value (NPV) =
proportion of test negatives who are
reference negative.
Example: eye disease in
diabetics
 78
test positives of whom 40 were
positive on the reference standard,
242 test negatives of whom 237
were negative on the reference
standard.
 PPV = 40/78 = 51%.
 NPV = 237/242 = 98%.
.
Types of medical
studies.
Classification of types of clinical research
Lancet 2002; 359: 57-61
The evidence pyramid
.
.
.
.
Observational Studies

Cohort
– Question answered: “what will happen?”
– Prospective/forward
– e.g. Framingham study of cardiovascular
disease: started in 1948, 6000 citizens
participated, followed for 20 years (study in
1970 by Gordon and Kannel)
– Possible uses:
 Typical cohort study
 Outcome assessment (patient outcomes:
economic, functional, satisfaction, QOL,
..)
Cohort Studies
Cohort Studies
 Advantages:
– Design of choice for studying cause of a
disease, course, risk factors
 Disadvantages:
– Cannot be used to prove causation
– Long studies can be costly
– Vulnerable to patient attrition, migration
.
Observational Studies

Case-Control
– Retrospective
– Question answered: “What
happened?”
– Matching needed for controls
Case Control Studies
Case-control Studies: research in reverse
Example: association between smoking and lung cancer.
People with lung cancer are enrolled to form the case group, and people without lung cancer
are identified as controls.
Researchers then look back in time to ascertain each person's exposure status (smoking history),
(retrospective design). Investigators compare the frequency of smoking exposure in the case group
with that in the control group, and calculate a measure of association.
Case-Control Studies

Advantages:
– Quickest
– Least expensive
– Good for rare diseases and diseases that take long
time
– Good for investigation of a preliminary hypothesis
– Time factor research

Disadvantages:
– Large biases
– Difficult to find matching controls
Both Cohort and Case-Control studies
are called
“Longitudinal Studies”
Notion of time
.
Cross Sectional Studies
 Advantages:
– Best for:
Determining status quo of a disease
Prevalence of disease
Evaluation of diagnostic procedures
– Relatively quick and inexpensive
 Disadvantages:
– Provide only a snapshot in time
.
Observational Studies
Case Series
Case-Control
Cohort
Longitudinal Studies
Cross Sectional
Experimental Studies



AKA Clinical Trials (Involve humans)
Easier to identify (usually explicitly stated
in the abstract)
Two main categories of clinical trials:
1.
2.
Controlled trials
Uncontrolled trials
.
Experimental Studies
Controlled Trials

Trials with independent concurrent
controls
– Double or single blind
– Best is randomized
assignment
– Same point in time
– These include:
 RCT:
– The epitome of all
research designs
– Provides the
strongest evidence of
concluding causation
– Best insurance that
results are due to the
intervention
 Nonrandomized trials:
– Assignment not
randomized
– Opened to biases


Trials with self controls
– Subject to bias (Hawthorne
effect)
– Can do crossover study
(with washout period in
between)
Trials with external control
– Uses the results of another
investigator’s research as a
comparison
– Historical controls can also
be used: for disease with
no cures yet
.
Randomized Controlled Studies
The Double Blind Method
Experimental Studies


Uncontrolled trials
Investigator’s experience with the new
drug or procedure is described but not
formally compared with another one
More likely to be used for interventions
that are procedures rather than drug
Clinical Trials
 Advantages:
– RCT is the gold standard or reference
 Disadvantages:
– Expensive
– Long duration
Experimental Studies
Controlled trials
Uncontrolled trials
Independent
concurrent
controls
Self controls
RCT
External controls
Non-randomized
Systematic Reviews &Metaanalyses
.
.
Study Pyramid
Best
Worst
EXPERIMENT 1
A
researcher assembles two groups of
study participants with Lyme Disease. She
administers the antibiotic doxycycline to
one group and amoxicillin to the other. The
researcher then measures which has more
of a beneficial effect.
EXPERIMENT 2
A
researcher identifies two groups of
elderly immobilized patients in a nursing
home. One group of patients is
repositioned every two hours. The other
group is repositioned every 1.5 hours. The
researcher measures whether the
incidence of bed sores is lower in the
group repositioned every 1.5 hours than
the group repositioned every 2 hours .
EXPERIMENT 3
A
researcher develops a counseling
program designed to increase medication
compliance in AIDS patients. The
researcher delivers the program to one
group of patients but not another, then
studies whether the group that underwent
the counseling program adhered to their
medication better than the group that did
not .
What do all three experiments
have in common?
?
.
 In
each case the researcher actively
interfered with one group of study subjects
and then compared the outcome with a
similar group of subject that did not
receive the same intervention.
 Experiments in which researchers interfere
with their study subjects are known as
clinical investigations.
 All examples above pertain to humans,
but many clinical investigations are done
using animal subjects, particularly for new
or risky interventions
EXPERIMENT 1
A
researcher identifies a group of patients
who were incorrectly diagnosed with Lyme
Disease, then records whether the
patients experienced adverse effects from
the antibiotics they were prescribed for
treatment of Lyme
EXPERIMENT 2
A
researcher identifies a group of elderly
immobilized patients in a nursing home.
The researcher follows the patients for 6
months and records the administration of
any topical (skin) agent. At the end of the
study period, the researcher conducts an
analysis to determine whether certain
topical regimens related to reduced
incidence of bed sores
EXPERIMENT 3
A
researcher identifies a group of
individuals with AIDS who have low
medication compliance. The researcher
administers a questionnaire to the patients
to determine what factors relate to their
poor compliance
What do all three experiments have
in common?
?
.
.
 In
each case the researcher does not
actively manipulate or intervene with
patients; rather, the researcher
merely observes the effect of one or
more factors on an outcome. It is
easy to see why observational
studies received their moniker
.
Medical Writing
The Basic Structure of Article
TITLE
(S)Summary
(Structured Abstract)
(I) Introduction (What Question was asked?)
(M)Methods
(How was it Studied?)
(R)Results
(What was Found?)
(A)Analysis
(How data was analysed?)
(D)Discussion
(What Do the Findings Mean?)
Acknowledgements
References
Title
 Should
be specific but
comprehensive
 Short but sufficiently
descriptive
 No abbreviations
 Should be easy to catalogue
Structured abstract
•
•
•
•
•
•
•
Objective
Design
Setting
Patients and methods
Results
Conclusions
Key words
Abstract
most important part of your paper
 Introduction/background:
what was
the purpose (10%)
 Material and methods: what was the
study design, techniques, and
statistical methods (30-35%)
 Results: what are the most important
findings (35-45%)
 Conclusions: why are the results
important (20-25%)
Cornett, 2001
Goals of abstract
maximum info into minimum space
Structured
 Use headings to
identify
 Follows IMRAD
format
 Typical in North
American journals
 250 words
Unstructured
 Arranged in 1
paragraph
 Follows IMRAD
format
 Typical in European
journals
 125-150 words
6 steps for writing your abstract
1.
2.
3.
4.
5.
6.
Identify guidelines, meeting or
journal
Highlight key features
Insert sentences into abstract
format
Write, revise, and condense
Edit sentences and words
Check final
Introduction
Existing state of knowledge
Gaps in knowledge which research will fill.
State what you Intend to do
Give pertinent references
It does not
Review the history of the subject
Does not identify all the other gaps in knowledge
Don not include methods, results and discussion
Introduction
begin to tell a story
Part 1
2
3
Begin with what is known
State what is unknown
End with what your study will
answer
Methods
 What Subjects/patients/animals/specimens

techniques were used?
Reason for selecting the experimental design
of the study

Statistical methods used for analysis

The section should be called "Materiel and
Methods" only if inanimate specimens have
been used.
Material and Methods
reads like a cookbook
 Write
this section first!
 How was the problem studied
 Usually arranged chronologically
 Subdivided into subsections
according to type of information
– Material: chemicals, experimental
materials, animals or humans
– Methods: preparation, protocol,
purposes of protocol and methods,
methods, data analysis
Techniques
Give enough details for readers to
assess the validity of the results,
and repeat the study
 If standard techniques is used, give
appropriate reference,any
modifications should be clearly
explained
 If drug trial- clear description of
trial

Statistics
Clearly mention the statistical methods used
for appropriate verification of reported
results.
(consult a statistician before starting the
study)
Example Material & Methods
Clinical
 Patient population

Protocol
(independent,
dependent,
control)

Methods for each

Analysis of data
Anatomical
 Materials



Sample
preparation for
each
Procedures for the
dependent
variables
Analysis of data
Results
 Communication of facts, measurements,
and observations gathered by the author



Start with the results that are easier to
interpret
Results should be set out in tables and
figures
Do not duplicate illustrations
Results
what were the findings
 Functions:
state results of
experiments
and present data that supports
results
 Content: results and data
 Consistency: check this in results,
discussion, and abstract
 Organization: chronological or from
most to least important
Results
uses the same order as Methods
Material & Methods
 Study subjects
 Study protocol
 Calculations
 Statistical analysis
Results
 Clinical
characteristics
 Systemic effects
 Specific effects
 Adverse effects
Results
see the forest through the trees
Be brief and uncluttered
 Mention species and material again
 Use past tense
 Give specific comparisons
 Present detailed data in figures and tables
to keep written data to a minimum
 Express results, give data
 Provide appropriate statistical details
 Begin each paragraph with a result

Visual data
Type of visual
Flow charts
(algorithms)
Tables
Line graphs
Bar graphs
Pie charts
Diagrammatic
illustrations
What it summarizes
Protocols
Complex data
Response to treat.
Complex data to
compare categories
Percentage of the
whole
Simplify or enhance
figures
Discussion
what do your findings mean
 Answers
the question posed in
Introduction
 Explain the significance of your
results
 Explain the findings, relationships,
and generalizations of your results
 Explains how results support answers
and how answers fit with existing
knowledge on the topic
 Has a beginning, middle, and end
Discussion
What gaps in knowledge remain to be
filled?

Main results should be summarised at
the beginning of discussion

Only mention previous results or
comments which illuminate or which are
illuminated by the present results.

Discussion
• Final paragraph in which the message
of the article is firmly stated.
• Point out where further gaps in
knowledge could usefully be filled instead
of "further research is needed".
• Intention of author to explore the
"Gaps" further.
Discussion
Part 1, beginning
The question posed in Introduction…
“… to test whether abnormal
vasospasm before angioplasty
increases the likelihood of
restenosis.”
Is answered in your Discussion…
“…the presence of abnormal
vasospasm before angioplasty was
associated with an increased
likelihood…”
Part 1, Beginning
present strongest evidence first
Begin with significance of your results
– Never begin with background
information
– Never repeat information from
Introduction
– Never begin with historical overviews
This is what everyone is waiting for!
Part 2, Middle
interpret your results
–Show how your results fit into the
literature and how they support
your answer
–Give in descending order of
importance
–Compare your results with other
studies, your work or others
–Use one idea per paragraph
Part 3, Ending
be strong
 Restate
answer to question, and
signal this ending
 Mention possible applications,
implications, or speculations
 Pull out as a separate Conclusions
section
 Suggest future work if needed
References
1.
Standard journal article
Vega K., Pina I., Krevisky B. Heart
transplantation is associated with an increased
risk for pancreaticobiliary disease. Ann Intern
Med 1996 Jun 1: 124(11): 980-3
2.Books and other monographs
Phillips SJ, Whisnant JP. Hypertension and
stroke. In: Laragh JH, Bremier BM, editors.
Hypertension: pathophysiology, diagnosis, and
management. 2nd ed. New York: Raven Press;
1995. P.465-78.
References
3.
Disertation
Kaplan SJ. Post-hospital home health care: the
elderly's access and untilizaiion (dissertation].
St. Louis (MO): Washington Univ: 1995.
4. Unpublished material
Leshner AL. Molecular mechanisms of cocaine.
N Engl J Med. In press 1996.
5. Electronic material
Morse SS. Factors in the emergence of infectious
disease. Emerg Inftect Dis [serial online] 1995
Jan-Mar (cited 1996 Jun 5]; 1(1): [24 screens].
Available from URL:
http:/www.cdc.gov/ncdod/EID/eid.htm