Writing and Presenting a Research Abstract Marian Limacher, MD Director, CTSI Training and Professional Development Program Senior Associate Dean for Faculty.

Writing and Presenting a Research Abstract

Marian Limacher, MDDirector, CTSI Training and Professional Development Program

Senior Associate Dean for Faculty Affairs and Professional Development

Professor, Division of Cardiovascular Medicine

The UF CTSI is supported in part by NIH awards UL1 RR029890, KL2 RR029888 and TL1 RR029889

Abstracts for Scientific Meetings• Important for transmitting new information,

“cutting edge”• Important for career development• Important to become known in the field• Important for you to get to know others in the

field, enhance networking opportunities

Abstract Basics

• One abstract per research question• No duplicate abstracts at different meetings• Goal: 1 manuscript per abstract

Why are abstracts rejected

• Dull topic, nothing new• No context provided• Small number of subjects• All talk, no data (“results will be discussed…”• All data, no talk• Poor readability• Sloppy

Abstract Format

• Read the instructions• Format the “box” or other limitations• Include:

– Introduction/background/purpose– Methods– Results– Conclusions

Abstract format

• Introduction– 1-2 sentences– Include the research question (hypothesis) or

purpose at the end– Avoid jargon, excessive abbreviations– Assume reviewer(s) are not the most

knowledgeable in the field

Abstract format

• Methods– Include study design– Study population (#, age, M,F, inclusion criteria)– What was measured– How data was analyzed (if space), otherwise

include during the talk/presentation

Abstract format

• Results– Main finding of the study – in words– Give real numbers as well as significance– Tables can be helpful

Abstract format

• Conclusions– What do your findings mean– Do not repeat results– If you could have written your conclusion before

you knew your results – you don’t really have a conclusion

Abstract Presentations - oral• “The 10 minute talk”

– Garson, et al, Am Heart J 1986; 111: 193-203• Write out every word: better to read than to

stumble or go over time• Usual rule: 1 minute per slide (maximum 30

sec per slide for non-complex statements)• Absolute maximum = 14 slides per 10 min.

Abstract presentation

• Preparing slides• 20 point minimum (this is 32 point)• This is 20 point

• Black on white/light• Yellow/white on blue• Leave 10% margin• Avoid too much “glitz”

Abstract presentation

• Preparing slides• 20 point minimum (this is 32 point)• This is 20 point

• Black on white/light• Yellow/white on blue• Leave 10% margin• Avoid too much “glitz”

Abstract presentation

• Preparing slides• 20 point minimum (this is 32 point)• This is 20 point

• Black on white/light• Yellow/white on blue• Leave 10% margin• Avoid too much “glitz”

Abstract presentation

• Preparing slides• 20 point minimum (this is 32 point)• This is 20 point

• Black on white/light• Yellow/white on blue• Leave 10% margin• Avoid too much “glitz”

Abstract presentation

• Preparing slides• 20 point minimum (this is 32 point)• This is 20 point

• Black on white/light• Yellow/white on blue• Leave 10% margin• Avoid too much “glitz”

Abstract Presentations

• Avoid red lettering on blue backgrounds– Hard to read

AFRICAN - AMERICANS !! Are you between the ages of 18 and 65? Have you ever been told that your blood pressure is high? Are you on no medicine or only one medicine for your blood pressure?

You may qualify for a study that is looking to improve treatment for high blood pressure in African-Americans.

If you qualify, you can receive up to $750 forcompletion of the study. For details call the study staff at the Division of Nephrology at 352-273-5661.

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Bloo

d pr

essu

re s

tudy

352-

273-

5661

Abstract presentations

• Use horizontal format• Center all material• No more than 7 lines per slides• No more than 7 words per line

How to increase data per slide without reducing readability

• Use title/heading area or draw a line under title

• Use bullets, especially if the statement uses more than one line

• Underline the points• Use color

Slide tricks

1. Number the points2. Use underlining above subheadings

Slide tricks

• Purpose– Devise a new technique to keep your audience

awake– Increase odds of publishing

Slide tricks

• Use color– Methods

• 300 patients with CHF• Admitted for syncope• 100 M; 200 F• All underwent Holter monitoring, EP testing, tilt

table

Inexcusable!!!

• “I know you probably can’t make out what’s on this slide, but…”

• “I apologize for the complexity of this slide,…”• “Ignore the details on this slide, I just want to

focus on …”

Slide construction

• Title for every slide• Use phrases but should be readable• Common abbreviations ok if appropriate

– Pts, M, F, CHF, LV, MI– Be careful of language differences– Try to avoid ** and footnotes (distracting)

Slide construction

• Tables– Use if the numbers are important, rather than the

relationship– Avoid more than 4 rows (for 2 columns)– If 3 columns, use 3 or less rows– For a “p” value column, list “NS” for p>.05– For others use 0 before . “p<0.01”

Meta Analysis of Genetic risks for MI

OR CI

Factor V Leiden for MI 1.26 .94-1.67

Factor V Leiden for MI<55

1.34 .94-1.91

Prothrombin Mutation and MI

.89 .59-1.35

Prothrombin Mutation and MI < 55

1.86 .99-3.31

Boekholdt et al, Circulation 2001;104:3063-3068

Bar Graphs

• Used to compare results of different treatment to different groups

• Use less than 6 bars• (no more than 8 maximum)• If showing paired, no more than 3 groups• Label each bar

HT, Prothrombotic Mutations and Risk of Venous Thromboembolism

1

4.43.5

11

0

2

4

6

8

10

12

Odds Ratio

HT-/ Mu- HT-/ Mu+ HT+/ Mu- HT+/ Mu+

[Rosendaal, 77 cases, 163 controls age 45-64]

Graphs

• Pie graphs show how an entire population is distributed

• If showing %’s, indicate absolute numbers somewhere on the slide

Examples

Examples

Graphs

• Line graph– Shows a relationship between continuous

variables

Kaplan-Meier Estimates of Cumulative Hazards for CHD for WHI

0.0

0.01

0.02

0.03

0.04

0.05

0 1 2 3 4 5 6 7

Time (years)

E+PPlacebo

E+P 8506 8353 8248 8133 7004 4251 2085 814

Placebo 8102 7999 7899 7789 6639 3948 1756 523

HR 1.29

nCI (11.02, 1.63) aCI (0.85, 1.97)

Too Small!

The Title

• Ideally: it should encourage the reader to be interested enough to want to see more

• It should not lead to any of the following reactions– “not again!”– “I don’t believe this”– “How boring”– “Huh?”

Title

Should say a lot about your researchWhen in doubt: choose a declarative statementAsk colleagues and coauthors (should have a

consensus)The key features should be in the title

(clinical trial, children with ear infections, new antibiotic)

Title – Giving the results?

• “Antibiotic X improves survival in neutropenic sepsis”

• Those who say no:– Interpreted as bragging?– Will get lower marks if not the final trial results

(e.g. preliminary findings)

Titles – stating results

• May highlight what’s new about your study, especially if the population or intervention have been studied before

• Can interest the reader • And the reviewer• Key point is to be objective – don’t overstate

Other considerations

• Pros and cons of stating the results: “B52 genotype reduces survival in Disease X”– Will clearly give the reader the content of your

study– But, may be “too assertive”

• Can seem like bragging• What if it conflicts with previous work, e.g. the

reviewers?• More objective (i.e. safer): “Effect of B52 genotype

on Survival in Disease X”

Highlight particular strengths or new methodology• Especially if this is one of many studies on the same

topic– “Association of protein Z with low cholesterol

levels and dementia”• Specify study population (e.g. rodents or humans)

when submitting to a journal/meeting where either might be presented– “Risk of renal failure in trauma patients with

hematuria”

Title Checklist

• Are the title and research question closely related?

• Is the title objective in tone?• Are special features of the study mentioned –

especially the study population (elderly, women, racial/ethnic group)

Examples

• What not to do (from the world of journalism)

At least we know it’s effective!

Giving the talk

• Have your 1st slide up – Usually, title of talk, authors, affiliation

• Begin thanking the moderators (learn their names)

• “Thank you Dr. Jones. Dr. Miller, ladies and gentlemen,…”

• Or, “Dr. Jones, Dr. Miller, ladies and gentlemen….”

Giving the Talk

• Be sure you use the same words in your talk as are on your slide

Example

• Slide: All sudden death outside hospital

• Say, “All sudden deaths occurred outside the hospital.”

• Do not say, “All patients who died suddenly did so outside the medical center.”

Slide transitions

• Pause slightly• You have control• Indicate in your written text when you will transition

– symbol, e.g. • For graphs, first orient the audience

– “This slide shows the occurrence of cardiac events over time. The horizontal or X axis displays time in years and the vertical axis shows the incidence of cardiac events in percent.”

Abstract presentations

• If you have more data in your presentation than you showed in the abstract, make a brief mention, then go on.– “As you can see, the numbers on this slide are

larger than those reported in the abstract as we included 50 additional cases.”

Talking about graphic results• Describe the overall findings after orientating the

audience.– “We found that heart attacks increased for every

quintile of BMI.”

• Point to only the key features– “More men than women developed cancer”– “Women had significantly more embolic events

than men.”

Giving the talk

• Talk about everything on the slide• If there’s too much, eliminate it from the slide

– You’ll have more room in the manuscript

• For conclusions:– Use complete sentences– Read words as in the slides. Some additional transition is

allowed.– Orient the audience, by “First,…” “Our second

conclusion…” “Finally, …”

At the end

• After your last conclusion recommendation, at the end of the last slide, – stop, – look up, – say “Thank you.”

• Wait for the moderator to take over.

Questions and answers

• If no microphone, repeat the questions• Have a pen to jot notes as long-winded questions get

asked• Be calm• Be professional• Do not disagree, if possible• Be brief• Be appreciative, thank person if someone makes a

good point.• OK to say, “We haven’t analyzed that yet.”

Giving the talk

• Practice• Time yourself• Give before an audience• Speak slowly

Precautions

• Bring back-ups– Flash drive, email to yourself– Be sure someone else has the slides

• Day of the talk– Be sure your slides are loaded– Check in the room of the talk to be sure

everything is ready

Pointer etiquette

• Do not point at every word• Do not “circle” or “wave” or “squiggle” with

your pointer• Only point at key aspects of a graph or picture.

If you are clear (in your slides and your talk) you won’t really need the pointer. Audience will follow with you.

Posters

• Similar in organization as oral presentations• Use format in the meeting instructions• Title Lettering: minimum 1 inch high• Other type: suitable for reading at 3-4 ft.• Group by Intro, Methods, Results, Conclusions.• Consider Powerpoint template through Med

Illustrations (BMS), VA Med media or a colleague.

Other Talks

• Don’t read from text– Unless you’re acknowledging the Nobel Prize

• Develop an outline• Tell what you will be covering• Highlight your objectives• Look at the audience

General tips for giving other talks• Plan how to cover your content• Break up sections• Stop and synthesize• Never go over time• Encourage questions• Humor can be good – in good taste