Transcript Document

Reza Assadi
MD, HSCR, MPH, MPhil, PhD
candidate
DrAssadi.com
Sit in groups
with
one computer (laptop), clinicians with
different majors and with some basic
science specialists
Myth-ing Out?
 Myth #1: Creative individuals are born, not made
 Myth #2: Innovation is the sole purview of the lone creative geniuseative
individuals are born, not made
 Myth #3: Intellectual work is not creative work
 Myth #4: Creative work is totally original
 Myth #5: Creativity is open and free and cannot thrive in contexts of
constraint
 Myth #6: Creative work is not practical
How to Conjure up your Creative
Community
 Discussed your research focus with a member of your family whose education
ended with high school?
 Tried to explain what you're engaged in at school with a five-year old?
 Went to the magazine section of a mainstream bookstore and challenged
yourself to find a connection between your research and some of the latest
articles?
 Played devil's advocate with yourself and tried to come up with ten questions
that others might ask but you haven't addressed in your studies to date?
 Actively sought out members of your cohort who think very differently than you
(perhaps being detail oriented rather than a "big picture" person) and invited
them to participate in regular, general discussions about your discipline, or a
focused study group?
Generating Good Ideas and Executing Them
 A good idea is one that will make a difference and can actually be done.
The essence of generating good ideas involves listening and learninglistening to the situations that matter to you, to the people in those
situations, and for the inherent possibilities that are always there”
 LEARN
 CREATE
 DO
 ADAPT
LEARN
 Who are the people involved and what motivates them?
 • What are the most salient elements of the situation?
 • How do the different people and elements relate to and
 influence each other?
 • How is the situation unfolding over time?
 • Why is the situation the way it is?
CREATE
 Based on what you learned from understanding the situation and the
 people involved in it? What would it "look" like for that problem to be
 solved or the opportunity to be fully engaged?
 What are some possible ways to achieve the ideal situation? Of the
 ways available, which seems to be the best?
DO
 Create a specific plan for achieving those goals:
 • Identify the people you need to help you, either as part of your
 team or as part of your network, and how you will get them
 involved.
 • Identify the resources you need, materially and financially, and
 how you will secure them.
 • Develop a plan for letting people know what you are doing.
 • Develop a timeline with milestones
ADOPT
 What worked and did not work in the prototype OR in the plan thus far?
 • What needs to be changed or enhanced?
 • What are the best next steps to implement necessary adjustments?
Let’s TRY IT
 Given everything you've learned, articulate a major
challenge/problem/gap/opportunity that is presenting itself right now.
There may be more than one, so you may have to prioritize. Pick one to
focus on
 What would it "look" like for that problem to be solved or the opportunity to
be fully engaged? What are some possible ways to achieve the ideal
situation? Of the ways available, which seems to be the best?
 Translate your best solution into a set of definable goals that can be
measured. In other words, how will you know that you are achieving the
solution you are working towards?
 Use a prototype to test the plan. By definition, a prototype is not finished. It
has enough of the key functions that allows someone to test your product,
your process, or idea, etc.
THE IMPORTANCE OF "DOING
NOTHING"
 The point is to get out of
 your own way by calming your mind or giving it something simple to do so
you can focus. Whatever method of generating ideas you use, it is essential
to take time to reflect, think about your work, or just rest. This allows what
creativity theorists call the "incubation" phase, that phase in which ideas
seem to simply emerge while you are doing something "mindless" (walking,
showering, cooking, etc.).
Seven Things You Can Do RIGHT NOW
To Generate New Ideas
 ONE: Reflect on what attracted you to your discipline in the first place.
 TWO: Go talk with your adviser or a faculty member whose research
interests are in a similar area (or not!)...tell them what you're interested in
and listen to what they have to say about that
 THREE: Commit to skimming through the latest issue of three to five leading
journals in your field. Focus on the articles that capture your attention and
look at the discussion section for suggestions for further research.
 FOUR: Volunteer for as many on- or off-campus presentations and talks as
possible - even if you don't believe your ideas are well formed. The process
of putting together a presentation and speaking it out loud is a great way
to clarify your thinking and promote new directions for you
 FIVE: Socialize! talk briefly about your research interests, ask about theirs,
and look for connections that might be mutually beneficial.
 SIX: Register for courses offered by another department next semester. You
can gain invaluable knowledge about ethics, consulting, writing, teaching,
communication and many other topics, while meeting students from other
disciplines. Listen to how they respond to your explanation of your research
 SEVEN: Take a break. Creativity isn't about forcing yourself to do something,
it's about being receptive.
Cycles of Research
Question Development
O'Leary, Z. (2004) The Essential Guide to Doing Research. London: Sage.
Chapter Three
16
1) Flash of Brilliance Model
 You wake up one day with a new insight/idea
 New approach to solve an important open problem

Warnings:


This rarely happens
Even if it does, you may not be able to
find an advisor who agrees
2) The Apprentice Model
 Your Colleague has a list of topics
 Suggests one (or more!) that you can work on
 Can save you a lot of time/anxiety

Warnings:


Don’t work on something you find
boring, fruitless, badly-motivated,…
Several students may be working on the
same/related problem
3) The Phoenix Model
 You work on some projects and think very hard about what
you’ve done looking for insights
 Re-implement in a common framework
 Identify an algorithm/proof problem inside
 The topic emerges from your work
 Especially common in systems

Warnings:

You may be working without “a topic” for
a long time
4) The Stapler Model
 You work on a number of small topics that turn into a series of
conference papers
 E.g., you figure out how to apply a technique (e.g., ILP) to a number
of key problems in an area
 You figure out somehow how to tie it all together, create a
chapter from each paper, and put a big staple through it

Warnings:

May be hard/impossible to find the tie
5) The Synthesis Model
 You read some papers from other subfields in computer
science/engineering or a related field (e.g., biology)
 And look for places to apply insight from another (sub)field to
your own
 E.g., databases to compilers

Warnings:


You can spend a career reading papers!
You may not find any useful connections
6) The Expanded Term Project Model
 You take a project course that gives you a new perspective
 E.g., theory for systems and vice versa
 The project/paper combines your research project with the
course project
 One (and ½) project does double duty

Warnings:

This can distract from your research if you
can’t find a related project/paper
Attributes of a good research topic
 practical considerations - requirements of your uni
 your capability – have/develop the skills and
competencies to manage the research
 excitement - your genuine interest
 resources – money, equipment, expert knowledge, time
 accessibility of data – especially when sensitive topics
 linkage to the theory - previous theories give a context to
your topic; avoid you to “discover” well-known things
 clarity of research questions and objectives – will there be
new insights or not
 practical considerations AGAIN - Congruence with your
career goals – in the end, you will be, want it or not, kind
of an expert in the specific area you studied
Generating and refining research ideas
 Generating research ideas: range of tips and techniques
rational AND creative techniques
 Using these techniques usually may have two outcomes:
one or more possible project idea VERSUS absolute panic
Rational techniques
examine your own strengths and interests – look
back and look into the future
look at past project titles – let your imagination
and attention to guide you; note down
discussion – colleagues; friends; university
supervisors, tutors; practitioners; proffesionals
searching the literature – academic and
professional journals; review articles
Creative techniques
 keep a notebook of ideas – short-time
memory is a ... (you fill the gap)
 explore your personal preferences using past
projects –
1) what appeals me about the project?
2) what is good about this project?
3) Why is the project good?
Find out aspects in projects you consider as
excellent and important to you
 relevance trees or mind mapping
 Brainstorming – not only for solving, but also
for generating problems!
‫ایده جدید‬
‫‪ ‬شناسایی وضعیت موجود‬
‫‪ ‬مقایسه دو وضعیت‬
‫‪ ‬اصالح وضعیت موجود‬
‫‪ ‬کشف حوزه های جدید دانش‬
Start with an important big question
Focus to solvable question
Observe
Analyze data
Reach conclusions
Generalize back to big problem
Clinical Study:
Design and
Methods
SOURCE: HAIL M. AL-ABDELY, MD, CONSULTANT, INFECTIOUS DISEASES, KING FAISAL SPECIALIST
HOSPITAL & RESEARCH CENTRE
Where to Start?

A good clinical study starts with


a good question based on good hypothesis that is based on good and
comprehensive review of the available evidence from pre-clinical and
clinical data
Type of design depends on the question to be answered
31
Formulating a Research Question

Focused and specific
What is the prevalence of Hepatitis B surface Antigen in Saudi
Arabia? Cross-sectional study
 What are the risk factors for hepatitis B infection? Prospective cohort or

case-control


Is vancomycin better than ceftazidime against gram negative
organisms?
Not a replication of already established evidence



clinical trial
Supported by available data


Is interferon a useful therapy for hepatitis B infection? Therapeutic
Is smoking associated with lung cancer?
Ethical
Answerable

Methods, resources ….etc
32
Objectives

Specific aims


Clear and detailed
End point(s)

Primary


The main answer to the research question
Secondary

Answer other related questions
33
Study Design

Your question
Describe
 Analyze


Your resources
Retrospective
 Prospective


Community

Acceptance of research
Observational
 Interventional

34
Clinical Study Types

Observational Studies






Cohort (Incidence, Longitudinal)
Case-Control
Cross-Sectional (Prevalence)
Case Series
Case Report
Experimental Studies
Uncontrolled Trials
 Controlled Trials

35
Levels of Evidence
36
Hierarchy of Strength of Evidence for Treatment Decisions
Level I:
Level I (A):
Level I (B):
Level II (A):
Level II (B):
Level III:
Level IV:
N of 1 randomized trial (double-blinded, cross-over)
Systematic reviews of randomized trials
Single randomized trial
Systematic review of observational studies addressing
patient-important outcome
Single observational study addressing important outcome
Physiologic studies
Unsystematic clinical observations (case-reports, anecdotal)
JAMA 2000; 284(10):1290-96
Observational study  Clinical trial
describe as
occurring in nature
observational
study
exposed
outcome
allocate
randomly
non exposed
Clinical
Trial
Ethics!
37
Important issues in Study Design

Validity: Truth

External Validity:


Internal Validity:


Can the study be generalized to the population
Results will not be due to chance, bias or confounding factors
Symmetry Principle: Groups are similar
38
Important issues in Study Design


Confounding: distortion of the effect of one risk
factor by the presence of another
Bias: Any effect from design, execution, &
interpretation that shifts or influences results

Confounding bias: failure to account for the effect
of one or more variables that are not distributed
equally

Measurement bias: measurement methods differ
between groups

Sampling (selection) bias: design and execution
errors in sampling
39
Introduction
Why this study is needed ?
What is the purpose of this study?
 Was purpose known before the study?
 What has been done before and how does this
study differ?


inadequacies of earlier work or next step in an
overall research project
Does the location of the study have relevance?
40
Why doing a study?

Alternative:
 census: test every individual in the population
 use available data, e.g. hospitals
But:
- data availability
- data quality
- cost
- questions require specific type of data and
circumstances
41
Types of observational studies

CROSS - SECTIONAL STUDY

COHORT STUDY

CASE CONTROL STUDY

CASE SERIES/CASE REPORTS
42
Characteristics of observational studies

No control over study units




need to clearly describe study individuals
Can study risk factors that have serious consequences
Study individuals in their natural environment (>>
extrapolation)
Possibility of confounding
43
Aims of observational studies

Evaluate the effect of a suspected
risk factor (exposure) on an outcome
(e.g. disease)
 define ‘exposure’ and ‘disease’

Describe the impact of the risk factor
on the frequency of disease in a
population
44
Cross - Sectional
Study
45
Cross - Sectional Study (1)

46
Exposure and disease measured once, i.e.
at the same point in time
n
exposed ?
diseased ?
past
present
future
Cross - Sectional Study (2)

Random sample from population
 i.e.
results reflect reference population

Estimates the frequencies of both
exposure and outcome in the population

Measuring both exposure and outcome
at one point in time

Typically a survey
47
Cross - Sectional Study (3)






Can study several exposure factors and
outcomes simultaneously
Determines disease prevalence
Helpful in public health administration &
planning
Quick
Low cost (e.g. mail survey)
Limitation:
Does not determine causal relationship
 Not appropriate if either exposure or outcome is rare

48
Pediatricians per
1000 Children
Cross-Sectional: Pediatrician-to-Child Ratio
40
35
30
25
20
15
10
5
0
1981
1986
1991
1996
Rural
Urban
Greg et al. (2001) Pediatrics.107(2):e18
49
Cross-Sectional: Risk Factors for Smoking
Variable
No. friends who smoke:
- all vs. none of them
- most vs. none of them
- about half vs. none of them
- a few vs. none of them
Any siblings who smoke: Y vs. N
OR
95% CI
36.5
18.4
7.5
2.1
2.8
9.3 – 142.8
5.5 – 61.8
2.2 – 26.0
0.6 – 7.9
1.8 – 4.3
Mother smokes:
Yes vs. No
Have no mother vs No
1.9
3.5
1.3 – 2.9
0.8 – 15.0
50
Cohort Studies
51
Cohort studies

Follow-up studies; subjects selected on
presence or absence of exposure & absence
of disease at one point in time. Disease is then
assessed for all subjects at another point in
time.

Typically prospective but can be retrospective,
depending on temporal relationship between
study initiation & occurrence of disease.
52
Cohort Study (1)

53
Individuals selected by exposure status and
future occurrence of disease measured
n
n
Exposed yes
no
past
disease
?
yes
Exposed
disease ?
no
present
disease ?
disease ?
future
Cohort studies (2)

More clearly established temporal sequence between
exposure & disease

Allows direct measurement of incidence

Examines multiple effects of a single exposure (nurses’ health
study, OC and breast, ovarioan cancers)
54
Cohort studies (3)

Limitations:
time
consuming and expensive
loss
to follow-up & unavailability of
data
potential
confounding factors
inefficient
for rare diseases
55
Prospective Cohort Study
with outcome
Exposed
without
outcome
Cohort
with outcome
Unexposed
Onset
of study
without
outcome
Time
Direction of inquiry
Q: What will happen?
56
Prospective Cohort Study

Appropriate for frequent disease

Can examine only few risk factors

Usually expensive

RR = ‘relative risk’ = incidence rate ratio

AR = incidence difference
57
Case-Control
Studies
58
Case-Control Study (1)

Retrospective

Can use hospital or health register data

First identify cases

Then identify suitable controls


Hardest part: who is suitable ??
Then inquire or retrieve previous exposure

By interview

By databases (e.g. hospital, health insurance)
59
Case-Control Study (2)
60
Diseased and non-diseased individuals
are selected first
 Then past exposure status is retrieved

n
exposed ?
exposed ?
past
yes disease
no
present
future
Case-Control Study (3)

Good for rare disease (e.g. cancer)

Can study many risk factors at the same time

Usually low cost

Confounding likely

OR (not RR !!)
61
Case-Control Study Design
62
Exposed
Cases
Unexposed
Exposed
Controls
Unexposed
Data
collection
Direction of inquiry
Q: What happened?
Time
Case-Control study (4)
•
Study subjects selected on basis of whether they have
(case) or do not have (control) a disease
•
Useful for disease with long latency period
•
Efficient in terms of time & costs
•
Particularly suited for rare diseases
•
Examines multiple exposures to a single disease
63
Case-control study (5)
Limitations:
(1) susceptible to bias (particularly selection & recall)
(2) difficulties in selection of controls
(3) ascertainment of disease & exposure status
(4) inefficient for rare exposures unless attributable risk is high
64
Case Selection
• Define source population
• Cases
– incident/prevalent
– diagnostic criteria (sensitivity + specificity)
• Controls
– selected from same population as cases
– select independent of exposure status
65
Control Selection
• Random selection from source population
• Hospital based controls:
– convenient selection
– controls from variety of diagnostic groups other
than case diagnosis
– avoid selection of diagnoses related to
particular risk factors
– limit number of diagnoses in individuals
66
Summary of Observational Studies
Characteristic
Sampling
Time
Causality
Frequency
measure
Risk
parameter
Cross Sectional
Case Control
Cohort
Random sample: Purposive sample:
population
diseased/nondiseased
One point
Retrospective
Statistical
Screening for
association
many risk factors
Prevalence
None
Purposive sample:
Exposed/nonexposed
Prospective
Testing one (or
few) risk factors
Incidence
Prevalence (risk) Odds ratio
ratio, odds ratio
Relative risk, odds
ratio
67
Clinical Trials
68
Types of trials
Trial
Controlled
Randomised
Blinded
Not controlled
Not randomised
Not blinded
Clinical Trials – Drug Development
Basic
Research
In-Vitro
Screening
In-Vivo
Screening
Isolated cells
& tissues
Drug
Licensing
& Release
Clinical
Trials I - III
In Humans
Safety
Testing
In Animals
Novel
Compounds
70
Clinical trials in drug development
(Any alternatives)
In-Vitro Tests Can Show Whether:
• A compound has the desired effect on isolated
cells or tissues
• There are adverse effects on those tissues
• In-Vitro Tests Cannot Show Whether:
• The desired effect will occur in a complete living
system
• There will be any adverse effects in a complete living
system
71
Clinical trials in drug development
(Any alternatives)


Animal Tests Can:

Suggest which drugs are likely to be effective in humans

Indicate which drugs may not be harmful in humans
Animal Tests Cannot:

Predict with absolute certainty what will happen in humans
72
Clinical trial vs. Cross-sectional

Clinical trial:

Cross Sectional Study:

Individuals selected by
entry condition

Individuals selected
randomly

Control over exposure


Exposure groups fully
comparable
Exposure observed as
occurring in nature
(groups not ‘identical’)

Exposure AND outcome
measured at one point
in time

No causal
interpretation


Outcome measured
after allocating
individuals to exposure
Therefore: causal
association likely
73
Clinical Trials-Phases

Phase I - Does it hurt the Patient?


Phase II - Does it help the Patient?


On patients to confirm the effectiveness of the drug
Phase III - Is it any better?


Usually in normal volunteers, small groups for safety testing
Large groups of patients for statistical confirmation of effect
and incidence of side-effects
Phase IV - Does it work in the community?

Post marketing studies. Fine tuning and new rare findings from a
very large population
74
Clinical Trial: Study Design


Uncontrolled
Controlled
Before/after (cross-over)
 Historical
 Concurrent, not randomized
 Randomized

75
Non-randomized Trials
May Be Appropriate
• Early studies of new and untried therapies
• Uncontrolled early phase studies where the
standard is relatively ineffective
• Investigations which cannot be done within the
current climate of controversy
• Truly dramatic response
76
Advantages of Randomized
Control Clinical Trial
1.
Randomization "tends" to produce
comparable groups
2.
Assure causal relationship
3.
Randomization produces valid statistical
tests
77
Disadvantages of
Randomized Control Clinical
Trial
1. Generalizable Results?

Participants studied may not represent
general study population.
2. Recruitment

Hard
3. Acceptability of Randomization Process

Some physicians will refuse

Some participants will refuse
4. Administrative Complexity
78
Clinical Protocol (1)

Background/Justification
--Where we are in the field
--What the study will add that is important

Objectives
--Primary hypothesis
--Secondary hypotheses
--Other
79
Study Population
Subset of the general population determined
by the eligibility criteria
General population
Eligibility criteria
Study population
Enrollment
Study sample
Observed
80
Clinical Protocol (2)

Study Design and Methods

Type of study, comparison

Inclusion and exclusion criteria

Description of intervention (what, how)

Concomitant therapy

Examination procedures (baseline, follow-up, outcome
assessment)

Intervention assignment procedure

Data collection sheet

Informed consent
81
Eligibility Criteria
(inclusion & exclusion)

State in advance

Consider

Potential for effect of intervention

Ability to detect that effect

Safety

Ability for informed consent
82
Method Outlines (1)


The independent (predictor) and dependent (outcome)
variables in the study should be clearly identified, defined,
and Measured?
How to choose subjects?




Random or not
Are they going to be representative of the population?
Random selection is not random assignment
Types of Blinding (Masking) Single, Double, Triple.
 Control group? How is it chosen?
 How are patients followed up? Who are the dropouts?
 How is the data quality insured? Reliability?
 Consider independent review of data? Compliance?
83
Methods outlines (2)

Reference any unusual methods?

Statistical methods specified in sufficient details

Is there a statement about sample size issues or statistical power?

? multicenter study. Quality assurance measures should be employed to obtain
consistency across sites?
84
Comparing Treatments
• Fundamental principle
• Groups must be alike in all important aspects and only differ in the
intervention each group receives
• In practical terms, “comparable treatment groups” means
“alike on the average”
• Randomization
• Each participant has the same chance of receiving any of the
interventions under study
• Allocation is carried out using a chance mechanism so that neither the
participant nor the investigator will know in advance which will be
assigned
• Blinding
• Avoidance of conscious or subconscious influence
• Fair evaluation of outcomes
85
Patients and Clinicians Kept
Blind To Treatment?
 Investigator
 Care
taker
86
Methods outlines (3)

Monitoring and Management
--Data and safety monitoring
--Adverse event assessment, reporting
--Contingency procedures
--Withdrawal criteria
87
Regular Follow-up

Routine Procedures (report forms)

Interviews

Examinations

Laboratory Tests

Adverse Event Detection/Reporting

Quality Assurance
88
Compliance/adherence

Pill counts and computers

Diaries

Biological tests
89
Lipid lowering drugs after myocardial
infarction
Mortality
clofibrate
18.2%
placebo
19.4%
Overall
Clofibrate
18.2%
Clofibrate Adherence
 80
< 80%
15.0%
24.6%
90
Methods outlines (4)

Statistics
--Sample size
--Stopping guidelines
--Analysis plans

Participant protection issues
91
Sample Size

The study is an experiment in people

Need enough participants to answer the question

Should not enroll more than needed to answer the question

Sample size is an estimate, using guidelines and assumptions
92
Contingency Plans

Patient management

Evaluation and reporting to all relevant persons and groups

Data monitoring plans

Protocol amendment or study termination
93
Human Subjects Protection
•
Institutional Review Board
•
Informed consent
•
Different levels of risk
•
Confidentiality as well as risk of new tx
•
Patient can refuse to participate w/o effect
•
Path to exit study known
•
Compensation
94
Summary
Selection of design should be made on the basis
of the particular hypothesis to be tested with
consideration of current state of knowledge
 Consider available resources when deciding on
a study design
 A clear and organized study design leads to
successful results
 Observational studies are especially valuable in
epidemiology
 Clinical trials carry the highest level of evidence
and should be pursued whenever feasible

95