How to write an abstract

1
How to write an abstract

• Joost PH Drenth

Department of Gastroenterology and Hepatology
Radboud University Nijmegen Medical Center The
Netherlands
Berlin October 2006
2
What is an Abstract?

• A mini paper
• One paragraph (200-300 words)
• Describes what you are going to present
• Provides overview of the facts
• Gets reader interested (teaser)

3
What is an abstract? II

• Snapshot of your research efforts
• Sells your work
• Provides an "executive summary"
• A fully self-contained capsule
• Only piece of research your peers are likely to
  read from you..
• Adequate information enough to make a judgment of
  the results
• Balance data-rich vs. data-poor

4
An abstract is not

• Philosophical statement
• Prose
• Proposal description
• Case report
• Project conclusion without data
• Results will be presented at the congress..

5
Before you start

• Think ahead
• Think in deadlines
• The subject should be appropriate
• Real science
• Provide meat
• Avoid flimsy data
• No salami
• Do not overdo

6
Before Writing

• Know your audience
• Cardiologists are not into urological science
• What forum
• Scientific conference
• Journal article

7
Anatomy of an abstract

• Each section a single sentence
• Introduction / Background / Aim
• Hypothesis / Goal / Problem
• Experiment / Patients Methods
• Results (may not be fully available at time of
  writing)
• Conclusion
• Might include some discussion

8
How to write it

• Sketch out your findings
• Single out main findings
• Do an elevator pitch
• Construct the main sentence
• Hypothesis / goal / problem
• Include supporting evidence
• Methods / experiment / observations
• Tie everything together in logical order

9
Nuts Bolts first lines

• Authors
• Follow the ICMJE guidelines
• Contribution to conception, design, analysis,
  interpretation
• Wrote draft / revision
• Make your colleagues happy
• Affliations
• Brown1, Black2, Red 3
• Univ Miami1, Pittsburgh2, Dresden3

10
Introduction / background

• Motivation
• Why do we care about the problem and the results?
  
• Importance of your work
• Issues at stake
• Impact if successful
• Problem statement
• What problem will be solved?
• What is the scope of your work?

11
Methods

• Methods
• How did you do it?
• Did you use simulation, analytic models, or
  analysis of experimental data
• What important variables did you control, ignore,
  or measure?

12
Results

• Results
• What's the answer?
• Treatment better or worse?
• Be specific, mention the data

13
Conclusion

• Conclusion
• What are the implications?
• New road?
• Road sign (do turn here?)
• Well traveled road (confirmatory)

14
Abstract issues

• Never enough space
• Website restrictions
• Use abbreviations
• Sparsely
• Explain at the onset
• Include additional information in brackets
• Include table (only if appropriate)

15
Use of words

• Of the 29 patients with no baseline biopsy but
  only 1 biopsy, treatment was continued in 20
  patients (5 in another hospital), in 8 patients
  treatment was stopped (in 3 because of
  improvement or worsening of the situation of the
  skin, in 5 because of abnormal liver enzyme
  tests) and to 1 patient something else happened

16
Or table

• Clinical follow-up of 29 patients who had one
  single biopsy

Table
17
Before submitting it

• Has each and every author seen it?
• Is everybody relevant to the research included in
  the byline?
• Discuss drafts with your mentor
• You need a go-ahead
• Your (girl)friend should see it errors typos?
• Closely follow guidelines of the society
• Respect the deadline

18
Abstract example

• Title
• Liver injury in long-term Methotrexate treatment
  in psoriasis is relatively infrequent
• Authors
• M.A.M.Berends1, J.Snoek2, P.C.M. van de Kerkhof1,
  E.M.G.J. de Jong1, M.G.H. van Oijen2,
  J.P.H.Drenth2
• Affiliations
• Radboud University Nijmegen Medical Center,
  Departments of 1Dermatology and 2Gastroenterology
  Hepatology, Nijmegen, The Netherlands.

19
Abstract example

• Background
• Methotrexate (MTX)-induced liver injury in
  psoriasis patients has led to dermatologic
  guidelines that stipulate monitoring of liver
  injury by means of serial liver biopsies. Recent
  literature suggests that MTX is less hepatotoxic
  than previously assumed, which calls for
  reconsideration of the routine use of liver
  biopsies
• Aim
• An evaluation of prevalence and development of
  liver injury in patients with psoriasis on
  long-term MTX

20
Abstract example

• Methods
• Retrospective chart review (1976-2005) was
  performed to obtain information about
  demographics, details of MTX treatment and
  histology (Roenigk classification), presence of
  risk factors and liver enzyme test results.
  Information was collected using a structured
  database.

21
Abstract example

• Results
• 125 patients (F58/M67 mean age 45.0, SD 12.7
  yrs) received a median cumulative MTX dosage of
  2113 mg (range 180-20235) over a period of 228
  weeks (range 16-1763).
• We analyzed 279 liver biopsies, and 71 were
  classified as Roenigk grade I, 14 as Roenigk
  grade II, 12 grade IIIa, 2 grade IIIB and 2
  grade IV.
• - Elevated gamma glutamyl transferase levels
  predicted the presence of liver injury (Odds
  ratio 1.80 95 CI 1.30-2.49).
• - 68 patients had multiple biopsies, 3 improved,
  72 did not change and in 25 histology
  deteriorated. The majority of cases (67/80 84)
  who progressed to Roenigk 2 received a cumulative
  dose of less than 6000 mg, while there was no
  further liver injury in those who continued MTX

22
Abstract example

• Conclusion
• MTX related liver injury is less frequent than
  previously thought. Liver injury was unrelated to
  MTX dose and duration, but in cases with liver
  injury it occurred at cumulative doses of less
  than 6000 mg.