DEVELOPING AND WRITING YOUR TRIOLOGICAL THESIS

1
DEVELOPING AND WRITING YOUR TRIOLOGICAL THESIS

• Maureen Hannley, PhD

2
(No Transcript)
3
THE ROAD TO TRIO

• Board-certified otolaryngologist
• In practice 3 years
• Published 2 papers attended 3 national
  meetings
• Be proposed by 2 active members approved by
  Council
• Write a thesis for review approval

4
(No Transcript)
5
THE MYTHOLOGY OF THE TRIOLOGICAL THESIS(or more
Urban Legends Debunked)
6
DONT DISS MY THESIS

• Claim The thesis has to be a production
  equivalent to a doctoral dissertation
• Status False
• Origins The Triological thesis guidelines
  describe a candidates thesis as a carefully
  written, comprehensive dissertation.. and in
  past years they were often vast productions
  several hundred pages long. But emphasis now
  is on timely publication
• Reality According to the guidelines, the thesis
  should be no more than 40 pages (including text
  and figures) double-spaced, with no more than 50
  references, in a format suitable for publication
  in The Laryngoscope. This is NOT equivalent to
  a doctoral dissertation!

7
Dissertation (n) a treatise advancing a new
point of view resulting from research document
that presents the authors research and findings
and is submitted in support of candidature for a
degree or professional qualifications. American
Heritage Dictionary of the American Language
8
HOW CAN I WRITE THEE? LET ME COUNT THE WAYS

• Claim The thesis work cannot have been used for
  any other purpose
• Status False
• Reality Candidates may use work begun in their
  residency that, in part, has been previously
  published or presented. It is permitted to use
  work that has been submitted as a Masters
  thesis, as long as it meets the format,
  publication, and originality requirements.

9
(SOMETIMES) IT TAKES A VILLAGE

• Claim The thesis cannot have multiple authors
• Status False
• Reality Although the candidate may be assisted
  in his/her work, the candidate must state in
  writing at the time of submission that he/she is
  personally responsible for 70 or more of the
  text and 70 or more of the research.

10
SUCCESSFUL THESESTopic Areas2001 2008 (n
168)
11
SUCCESSFUL THESESProject Approach2001-2008
12
STEPS TO BUILDING A THESIS

• Preliminary steps
• Read guidelines
• Develop the thesis plan
• Topic
• Idea
• Review literature
• Goals and objectives
• Statistical consult
• Strategy
• Variables
• Sample
• Hypothesis
• Submit IRB, IACUC application
• Conduct the thesis project
• Analyze the data
• Write the paper

13
THE MOST IMPORTANT RULE

• Read the guidelines for thesis format and
  submission
• Read them again
• Follow them to the letter

14
SOME EXCELLENT REFERENCES

• Troidl, H et al., Surgical Research Basic
  Principles and Clinical Practice (3rd Ed.). New
  York Springer, 1998
• Stone, J., Conducting Clinical Research A
  Practical Guide. Cumberland, MD Mountainside MD
  Press, 2006
• Belsher, J, Jacobsen, J, From Idea to Funded
  Project. Phoenix Oryx Press, 1992
• Spilker, B., Guide to Clinical Trials. New York
  Raven Press, 1991
• Norman, G, Streiner, D, PDQ Statistics (2d Ed.).
  St. Louis B.C. Decker, Inc., 1999
• Hulley et al., Designing Clinical Research (3rd
  Ed). Baltimore Lippincott, Williams Wilkins,
  2007

15
SELECT A TOPIC

• Relevance
• Avoidance of duplication originality
• Feasibility
• Ethical acceptability
• Applicability of possible results and
  recommendations
• Urgency of data needed
• The Big So-What

16
FOUR TESTS OF THESIS TOPIC

• Is it new?
• Is it true?
• So what?
• Who cares?

Fontanarosa, 2008
17
INTERESTS, TOPICS, QUESTIONS, PROBLEMS

• Find an interest in a broad subject area
• What am I interested/expert in?
• Pediatric otolaryngology with impaired children
• Narrow the interest to a plausible topic
• What are unsolved gaps or inconsistencies in
  this area?
• Candidacy of neurologically impaired children for
  tonsillectomy
• Develop questions within the topic
• What do I want to know?
• Any difference between normal impaired children
  in complications, recovery rate, outcomes?
• Develop a rationale for the project
• Why is it important to know this?
• Demonstrate candidacy, secondary benefits

18
DEVELOP AN IDEA

• Read the most authoritative sources until you
  come to a point where the sources disagree. This
  is where unsolved questions may reside.
• Talk with the leading figures in the area.
  Attend their lectures and be alert to problems
  they identify.
• Seek out and read strategic research plans of
  medical and research organizations.
• Contemplate your own experience. What are the
  problems or questions that frustrate you? Have
  you found a solution you think will benefit your
  peers and patients?

Naumann in Belcher Jacobsen, 1992
19
Practical Problem
motivates
helps to solve
?
Research Answer
Research Question
finds
defines
Research Problem
Booth, Columb, Williams, 1995
20
SIX STEPS TO DEVELOP A COMPELLING NOVEL IDEA

• Identify the niche area
• Collect and critically analyze background
  information related to the problem
• Develop a preliminary idea (dont force it)
• Assess the ideas potential for success and
  modify it, if necessary
• Seek constructive criticism from colleagues
• Refine the idea to maximize its potential for
  impact on your field

21
A PLACE IN HISTORY

• Determine current status of topic area
• If controversial, identify both sides of the
  argument
• Use recent citations
• Do not cite references that you havent read
• Do not duplicate published projects

22
SOME INVESTIGATIONAL CATEGORIES

• Descriptions of problems required for planning
  interventions
• A Standardized Regimen of Antibiotics Prevents
  Infectious Complications in Skull Base Surgery
• Transtympanic vs. Sustained Release
  Administration of Gentamicin Kinetics,
  Morphology, and Function
• Information required to evaluate ongoing
  interventions, assess progress
• Therapeutic Selective Neck Dissection 25 Year
  Review
• Long-term Follow-up for Children Treated with
  Surgical Intervention for Chronic Sinusitis
• Information required to define problems
• Direct Evidence of Bacterial Biofilms in Otitis
  Media
• Analyze possible causes
• Direct Nasopharyngeal Reflux of Gastric Acid is
  a Contributing Factor in Refractory Chronic
  Rhinosinusitis
• or offer solutions
• Intratympanic Dexamethasone for Sudden
  Sensorineural Hearing Loss Following Failure of
  Systemic Therapy

23
ASKING THE RIGHT QUESTION

• Ask a question with clinical relevance, not just
  clinical application
• Ask a question that interests you and will
  sustain your interest
• Ask the question in such a way that either a
  positive or negative answer will be interesting
• Be specific
• Make sure you have time, resources
• Subjects
• Database/access, technical assistance
• Collaborators if appropriate

24
YOUR NEW BEST FRIEND

• Consult a statistician UP FRONT!
• Question design statistical
  treatment
• Sample size estimations
• Bias issues
• Involve sponsor/mentor in planning process
• Careful, comprehensive literature review

25
THE RESEARCH QUESTION

• Begin by asking the question as a general
  statement
• In patients with recurrent acute sinusitis by
  accepted criteria is ESS the best treatment
  option to improve symptoms and disease-specific
  QOL?
• Use PICO to help structure the question

26
PICO

• P Population, Patient or Problem
• I Intervention
• C Control or Comparison
• O Outcome

27
THE RESEARCH QUESTION

• P (Pediatric, adult) patients with recurrent
  acute sinusitis
• I Endoscopic sinus surgery
• C Medical treatment
• O Symptoms and disease-specific
  QOL
• In (age) patients with recurrent acute sinusitis
  by accepted criteria does ESS compared to medical
  treatment improve symptoms and disease-specific
  QOL?

Patients
Intervention
Comparison
Outcome
28
REFINE YOUR RESEARCH QUESTION

• Define the population to be studied
• Define the period of time for the study
• Select the variables to be measured
• Change non-specific variables into variables that
  can be measured.

29
WRITE THE HYPOTHESIS

• Write what you expect to find from your study.
• What are the general relationships implied by
  your hypothesis?
• Are there any specific alternative relationships
  or explanations that would serve as competing or
  rival hypothesis?
• State your hypothesis in a clear, concise
  sentence.
• Should be directional and quantifiable
• Should be simple, specific, and stated in advance

30
TYPES OF HYPOTHESES

• Null hypothesis (H0)
• There is no difference in symptom resolution or
  disease-specific QOL in children with recurrent
  acute sinusitis treated with ESS and those
  treated medically.
• Formal basis for testing statistical significance
• Alternative hypothesis
• Children with recurrent acute sinusitis treated
  with ESS will have significantly better symptom
  resolution and better disease-specific QOL than
  those treated medically.

Definable amount that is clinically significant
31
NOTE (NB)

• An objective is NOT synonymous with an
  hypothesis.
• Objective what you want to accomplish with your
  project where you are going
• One step along continuum of progress toward a
  long-term goal
• Hypothesis best guess as to how the project will
  turn out, given the way you do it (subjects,
  methods, analysis)
• Both MUST be stated.

32
LINKAGE OF THREE KEY COMPONENTS

• Long-Term Goal Broadest (the Forest)
• Projects your continuum of research
• Overall Objective Narrower
• Step along the continuum
• Must be achieved regardless of how the hypothesis
  tests
• Central Hypothesis Narrowest (the Tree)
• Best bet, but could be invalid (alternative
  presented later, therefore)

33
DETERMINE WHAT YOU ARE GOING TO MEASURE

• Variables that you are going to need to measure
• Be parsimonious if it wont add to your answer,
  dont do it
• ButRecognize the value of independent converging
  measures
• Each additional variable complicates your
  statistics increases your sample size
  requirement
• Reliability and validity of instruments
  relevance to question/topic

34
VARIABLES
Types of Variables
Dependent
Independent
Qualitative
Quantitative

(Reflux)
(Proton pump inhibitor vs pepsin antagonist)
Nominal
Interval
Confounding
(Sex, gender)
(36o 38o)
(Gastric ulcer)
Ratio
Ordinal
(BMI, vital capacity)
(TNM stage)
Norman Streiner, 2003
35
SELECT VARIABLES

• Numerical, e.g.
• Age
• Number of documented recurrences
• Categorical, e.g.
• Involved site
• Disease outcome
• Tissue type
• Operational e.g.
• Dependent
• Independent
• Confounding
• Background

36
SELECT A STUDY STRATEGY

• Basic vs. clinical
• Prospective vs. retrospective
• Efficacy vs. effectiveness
• Duration of study
• Cross-sectional vs. longitudinal
• Medical vs. patient outcome

37
QUALITY vs BIAS IN CLINICAL RESEARCH

• Case report
• Case series
• Case-control
• Cross-sectional studies
• Cohort studies
• Clinical trial
• Randomized clinical trial
• Blinded randomized clinical trial

BIAS
BEST EVIDENCE
38
TYPES OF STUDIES
39
TYPES OF STUDIES
40
FORMULATE PROJECT OBJECTIVES, HYPOTHESES

• Formulation of objectives will help
• Focus the study, narrowing it to essentials
• Avoid collection of data not strictly necessary
  for solving defined problem
• Organize the study in clearly defined parts or
  phases
• Each objective should have a hypothesis
• Formulation of hypotheses will provide a basis to
  test achievement of objectives

41
SELECT YOUR SAMPLE(S)

• Describe the characteristics of the subjects who
  will be eligible for participation in your study
• Describe the characteristics of the subjects who
  will be excluded from your study
• Describe the population (beyond your sample) to
  which you wish to generalize your conclusions
• Determine sample size necessary for desired
  statistical power

42
ELIMINATE OR CONTROL THE BIASES

• Effects of historical events
• Effects of maturation, gender, ethnicity
• Effects of repeated measurement
• Instrument decay
• Effects of statistical regression
• Subject selection
• Loss of subjects
• Patient recall
• Investigator bias

43
TO ENHANCE CREDIBILITY

• Appropriate controls
• Appropriate operational definitions
• Appropriate measurement tools
• Appropriate design and analysis
• Balanced perspective
• Cite others work
• If there are 2 camps, make sure you cite both
  sides

44
CONDUCTING THE THESIS PROJECT
45
SCIENTIFIC INTEGRITY AND ETHICS

• Ethics in scientific design and conduct
• Ethical justifiability
• Clinical equipoise
• Informed, comprehending, voluntary consent
• Confidentiality
• Conscientious, ethical use of animals
• Scientific adequacy
• Conflict of interest

Riis, in Troidl, et al., 1998
46
MANAGE THE DATA

• Collect the data
• Establish and follow protocol
• Know elements of informed consent
• Maintain quality control over the data
• Enter the data to a database
• Store the data
• Analyze the data

47
DESIGN AND METHODOLOGY

• Who?
• What?
• Where?
• When?
• How much?

WHY?
48
ANIMAL USE AND CARE

• Projects using vertebrate animals must be
  reviewed by your home institution IACUC
• Justify choice of species
• Specify and justify analgesia, euthanasia methods

49
DATA ANALYSIS

• Descriptive (should always be included)
• Numbers, demographics n, age, gender, ethnicity
• Central tendency mean, median, mode
• Variation range, standard deviation, percentile
• Inferential (depends on design, question)
• Process of inferring features of population from
  observations of a sample
• Likelihood that observed effect/differences could
  be due to chance standard error of mean
• Many inferential tests depending on research
  design and whether data are normally distributed
  (parametric) or not (non-parametric)

50
A difference to be a difference should make a
difference. Gertrude Stein
51
WRITING THE THESIS
process, a uniform level of scientific rigor can
be attained to achieve three objectives 1)
provide support for the most meritorious research
in otolaryngology and head
52
APPROACH TO THESIS PREPARATION
Writing is easy. All you have to do is sit and
stare at the blank sheet of paper until the drops
of blood form on your forehead.
Gene Fowler
53
THE ABCs OF SCIENTIFIC WRITING

• Accuracy
• Brevity
• Clarity

54
AN INFREQUENT (but useful) PRACTICE

• Tell the story of your project in your own words
  technical language can come later
• What you did
• Why you did it
• What you found
• What it means
• Start with whatever section you feel most
  comfortable
• Sketch out figures/tables you anticipate will be
  most supportive

55
THE BAIT AND THE PUNCHLINE

• Introduction
• Opening quotation or fact
• Context of past research
• Condition of ignorance
• Cost of that ignorance
• Gist of solution

• Conclusion
• Gist of solution
• Larger significance or application
• What is still not known
• Call for further research
• Closing quotation or fact

Booth, Colomb, Williams, 1995
56
STATE OBJECTIVES

• Cover different aspects of problem and
  contributing factors in coherent way and logical
  sequence.
• Use specific action-oriented verbs (to verify
  to compare to establish etc.) that correspond
  with goals and methodology.
• Avoid weak verbs (to find out, to
  characterize, to demonstrate) that give the
  impression of a hypothesis seeking project,
  a.k.a. a fishing expedition
• Avoid non-specific descriptors (worse, more
  satisfied, less severe etc)
• Objectives should be demonstrably achievable
  through selected methodology and subjects.
• State assumptions underlying your project.

57
METHODS AND PROCEDURES

• Reveal how carefully you conducted your research
• Allow another researcher to replicate your study
• Sections
• Subjects or Participants
• Experimental design
• Equipment or measures
• Procedures
• Subject selection
• Measurement of dependent variables
• References to support choice of procedures,
  especially if options available

58
(No Transcript)
59
A USEFUL EXERCISE

• Convert pooled data to charts (bar, pie, scatter,
  etc.)
• Write 1-2 bullets/sentences about what chart
  tells you
• Write 1-2 bullets/sentences about any alternative
  explanations for the display
• This can serve as outline for Results and
  Discussion sections

60
RESULTS

• Report only results that bear on your question
  (not every analysis needs to be shown)
• Consider possible contributing factors to data
  outcomes
• A good Results section should tell a story
• Analyses that support the integrity of the study
  (internal consistency, variance, etc)
• Present analyses in logical sequence
• Use tables figures to relieve clutter of
  numbers

61
ILLUSTRATIONS

• To illustrate this
• Process
• Logical relationships
• Object
• Parts of complex object
• Action, step in process
• Results

• Use this
• Flow chart, decision tree
• Diagram, matrix
• Photo, drawing
• Microscopic view, drawing
• Schematic, photo, diagram
• Photo, graph, matrix

62
DISCUSSION/CONCLUSION

• Considers findings in light of other theories and
  past research
• Begin with brief overview of problem and your
  findings should not reiterate Results, but
  state whether Hypothesis accepted or rejected
  should be consistent with Results
• Relevance to clinical problems practice
• Identify limitations of your research why they
  arent fatal flaws and can represent
  opportunities for further research

63
THE BEST DEFENSE..

• Your Answers
• I claim that
• I limit it to
• I offer as evidence
• I offer this principle
• I can rebut that. First.
• My claim stands as long as...
• Well, I have to admit that

• The Questions
• What is your point?
• What is your claims scope?
• What evidence do you have?
• What links evidence to claim?
• But what about.?
• But what if.?
• No problems at all?

64
RED FLAG LANGUAGE

• Using the standard techniques
• I hope the reviewers know what I mean
• A representative population was studied
• The nurse recruited the subjects with no formal
  inclusion/exclusion criteria someone else did
  the chart review
• Visual inspection of the results by two
  specialists did not reveal any significant
  differences
• Statistical analysis was not conducted
• 15 surgeons, including 12 residents and 3
  attendings participated in collecting the data
• I havent read the guidelines
• The average patient was a middle aged male with
  a long history of.
• I didnt want to be in Trio anyway

65
JUST DO IT!
66
CONTACT

• Maureen Hannley, PhD
• 414-805-8308
• mhannley_at_mcw.edu
• auntiemo_at_aol.com