Conference Management:
Your Talk, Their Questions,
Your Questions
Theodore J. Iwashyna, MD, PhD
University of Michigan
Ann Arbor VA Center for Clinical Management Research
on sabbatical at ANZIC-RC at Monash University
7 September 2015 – SPHPM HDR Student Professional Development
Before your talk
Giving your talk
Managing their questions
Making connections
Asking good, appropriate
questions
Mythbusters. http://www.discovery.com/tv-shows/mythbusters/videos/polishing-a-turd-minimyth/
http://www.wonder-tonic.com/geocitiesizer/index.php
We shall go on to the end, we shall fight in France, we shall fight
on the seas and oceans, we shall fight with growing confidence
and growing strength in the air, we shall defend our Island,
whatever the cost may be, we shall fight on the beaches, we shall
fight on the landing grounds, we shall fight in the fields and in the
streets, we shall fight in the hills; we shall never surrender… we
would carry on the struggle, until, in God's good time, the New
World, with all its power and might, steps forth to the rescue and
the liberation of the old.
http://www.winstonchurchill.org/resources/speeches/1940-the-finest-hour/we-shall-fight-on-the-beaches
GEOGRAPHIC CONTEXT
• Locations of fight
• to end
• France
• seas
• oceans
• with growing:
• confidence
• strength in the air
• beaches
• landing grounds,
• fields
• streets
• hills
http://www.winstonchurchill.org/resources/speeches/1940-the-finest-hour/we-shall-fight-on-the-beaches
Peter Norvig. http://norvig.com/Gettysburg/index.htm
Peter Norvig. http://norvig.com/Gettysburg/index.htm
METHODS
•
•
•
•
•
•
•
Statistical Analyses
Our primary outcome variable was in-hospital mortality. Our primary exposure variables were newly developed and validated
risk-of death scores, which had two parts: a simple antecedent component (including demographics, chronic health scores,
admission from home or chronic care facility, presence of treatment limitations, with adjustment for hospital site and temporal
trends9) and an acute illness component (including characteristics of the patient’s acute reason for being in the ICU, such as
ICU admission diagnosis, the components of the APACHE III acute physiology score, time of ICU admission and the source of
that admission, ICU care type (ICU versus high dependency unit), and pre-ICU hospitalization length of stay). Details on the
split sample development and validation are presented in Appendix 1.
Our primary multivariable analyses were conducted using logistic regression. We examined all patients in ICU for ICU days 1
through 21; that is, for the analysis on ICU day 4, we included all patients in the regression whose ICU stay was 4 days or
longer . In these analyses we adjusted our standard errors for clustering of patients within ICUs via //however Michael did it//.
We evaluated the contribution of types of variables to mortality by examining differences in the Areas Under the Receiver
Operating Characteristics (AUROC) curve in the validation sample, using XXX test. //will need to update from Bailey//
For diagnosis-specific analyses in the 15 most common diagnoses, predicted probabilities were used to summarize the output.
Probabilities were predicted for representative patients without applying APACHE III chronic health evaluation points ; the
exact characteristics of the representative patients on each variable are shown in the Appendix. Age, acute physiology score,
and diagnoses were varied systematically. We used representative patients, rather than average adjusted predictions to
prevent the effects of differential patient composition across different days of ICU stay that would have occurred with average
adjusted predictions. These predicted probabilities were graphed on a linear probability scale to facilitate interpretation;
graphical presentations for each diagnosis are truncated once the confidence intervals on the predicted probabilities exceed
20 percentage points. Measured levels are presented at time points chosen post-hoc to help explain the visualized patterns.
Analyses were conducted in Stata 13.1 (College Station, TX) and SAS 9.4 (Research Triangle, NC). A p-value of 0.05 was
considered statistically significant, and two-sided tests are presented.
The Alfred Hospital Human Research Ethics Committee, Melbourne, Australia, approved the study with a waiver of informed
consent.
Van den Berge (1998) Verhandelingen - Koninklijke Academie voor Geneeskunde van Belgie 60:487.
http://innovation.cms.gov/Files/reports/ChronicallyCriticallyIllPopulation-Report.pdf;
Kahn et al (2010) JAMA 303:2253.; Kahn et al (2015) Crit Care Med 43:282.
It is our hypothesis that
there exists a substantial and growing group of patients
who are ICU-dependent, in the sense that they are are unable to
live for more than a few days outside of intensive-care-like services
whose current problems are driven by their ongoing cascading
critical illnesses rather than their original ICU admitting diagnosis
who account for a substantial portion of our bed-days
for whom we have little specific expertise in promoting their
recovery (as opposed to continuing their resuscitation)
but who are not immutably fated to such limbo, but rather whose
care we could improve both via improved ICU patient selection but
also by changing care & communication practices in the ICU
Iwashyna, Hodgson, Pilcher, Orford, Santamaria, Bailey, Bellomo (2015) Crit Care & Resusc forthcoming.
Before your talk
Giving your talk
Managing their questions
Making connections
Asking good, appropriate
questions
Before your talk
Giving your talk
Managing their questions
Making connections
Asking good, appropriate
questions
Steve Maccone (2010) The New Yorker.
What was your worst?
Handling GOOD questions
• Stalling to think
• When you do not have the answer, or had not thought of that
Handling HORRIBLE questions
• “What the American people really want to talk about, though, is…”
• “Complex issue, and I think you and I should discuss it when I can
more fully understand your views”
Jon Krakeuer https://www.youtube.com/watch?v=LNmtDsb43mQ
Before your talk
Giving your talk
Managing their questions
Making connections
Asking good, appropriate
questions
Repeat as necessary,
but in general
not more often than annually with any individual
Before your talk
Giving your talk
Managing their questions
Making connections
Asking good, appropriate
questions
https://sliscon.files.wordpress.com/2011/03/conference-asking-questions-resource.pdf
Jack’s Guidelines to Asking Questions
Something you think other people will want to hear the
answer to?
When you ask, be long enough to be clear, short enough to
be brief. (~3-5 sentences)
Something you think the presenter is going to be able to
speak about cogently?
Will not make the session run late (help the facilitator)
Be constructive, even when being devastating. (If you can’t
find the constructive, almost never good to ask in public.)
Never, ever punch down.
Really, never, ever punch down.
Before your talk
• do good science
• prepare and practice
Give your talk
• with your audience in mind
Manage their questions
• with your whole audience in mind
Make connections
• outside and around your talk
Ask good questions
• that make their talk better
I would be happy to continue the conversation. Email me at
jack.iwashyna.on.sabbatical @ gmail.com for copies of my
slides or to talk. I also tweet @iwashyna.