Professional Documents
Culture Documents
Volume 2, Issue 1
Winter 2007 biostatisphere
Conference Kicks Off Breslow Lectureship
The UW department of Biostatis- on absolute risk relied on many of those tools. I am
tics established the Norman E. confident that Norm’s wonderful record of accom-
Breslow Distinguished Lecture- plishment and the warmth and spirit of Norm and his
ship, to be awarded annually to colleagues will enrich a long succession of Norman E.
a biostatistical scholar who has Breslow Distinguished Lectures.” The 2007 Breslow
significantly contributed to meth- Lecture will be given by Norm’s long-term collaborator
odology and applications in the Nick Day, dates TBA.
health sciences.
120 alumni, faculty, and colleagues attended the con-
The Lectureship kicked off this past ference, which featured an all-star line-up of experts,
summer at the Conference on Sta- providing an overview of past, current and prospec-
tistical Methods in Epidemiology and tive research in the field. “I was extremely impressed
Norm Breslow, Mitch Gail and Bruce Weir Observational Studies in Seattle, an by the enormous effort put into the Breslow Confer-
at Breslow Conference event that celebrated the scientific ence by Tom Fleming and by organizers Xihong Lin,
achievements of Professor Norm Thomas Lumley and David Yanez,” said Breslow. “It
Breslow, former department Chair. The conference, was wonderful to hear from many conference partici-
held in Seattle, showcased the initial lecture, delivered pants during the subsequent Joint Statistical Meet-
by Dr. Mitchell Gail, Chief of the Biostatistics Branch ings how much they had enjoyed the event and, in
“ I hope the Depart- of the National Cancer Institute. particular, the outstanding inaugural Breslow Lecture
ment will enjoy the given by my old friend Mitch Gail. I hope the Depart-
“It was a great pleasure for me to be asked to give a ment will enjoy the benefits of visits by other world
benefits of visits by lecture in Norm Breslow’s honor,” said Gail, pictured class lecturers in future years.”
other world class lec- here between Breslow and department chair Bruce
turers in future years.” Weir. “He and his colleagues have provided the epi- If you wish to make a donation to the Breslow Lec-
demiology and biostatistics community with insights tureship fund, visit https://secure.gifts.washington.
and tools we use every day—and indeed my lecture edu/sphcm/gift.asp?code=breslo&page=make.
Student-Faculty Collaboration:
Air Pollution Methods for Epidemiologists
Featured Publication: Janes H, Sheppard L, Lumley to work interpreting issues that sprung from a publica-
T. Assessing the Health Effects of Air Pollution Us- tion by another Biostat professor, Thomas Lumley. Their
ing the Case-Crossover Design. Epidemiology, 2005, resulting collaboration was published in Epidemiology.
16:717-26.
The Lumley Environmetrics article (citation above) dealt
Janes and Sheppard
Also: Lumley T and Levy D (2000). Bias in the case- with case-crossover design, one of several study de-
crossover design: implications for studies of air pollu- signs that are used to assess the health effects of air
tion. Environmetrics, 11: 689-704, 2000. pollution. He and epidemiology graduate student Levy
discovered that the commonly used reference selec-
“ …this method has When Holly Janes (Ph.D., 2005) started at UW Biostat, tion strategies, or implementations of the design, in-
some tricky issues in- she soon connected with Professor Lianne Sheppard, an corporated certain types of bias. They recommended
volved.” expert in the environmental statistics. Sheppard set her alternative strategies that circumvented the bias.
Continued on pg. 4.
CAST had a startling result, which was that the new The other is my involvement with Medic One Foun-
generation anti-arrhythmic drugs, so-called miracle dation. Early on, I worked as their statistician. And
drugs, were killing people—a totally unsuspected re- because of spending so much time with the data, 25
sult. So that study is quoted often in statistical litera- years ago I formulated the hypothesis that compres- “We’ve had many
ture because it’s an example of how far off surrogate sions alone, without ventilations, were the saving
trials where results
endpoints can be. factor in CPR.
were contrary to what
What is the Center’s greatest accomplishment? 18 years ago, I finally talked the city into a random- everyone had previ-
We’ve had many trials where results were contrary ized telephone CPR trial. When people called 911 ously believed.”
to what everyone had previously believed. AVID (An- to report a cardiac arrest and didn’t know CPR, the
tiarrhythmics Versus Implantable Defibrillator trial) dispatcher would give them one of two sets of di-
verified that the devices increase survival, although rections—one with alternating ventilations and com-
the magnitude was infinitely less than what had pressions, one with just compressions.
been assumed.
In the end, the ones who just did chest compres-
When implantable defibrillators went from a simple sions had 14% survival and the ones doing ventila-
shock box to computer driven devices that paced
Continued on pg. 4.
“Thomas’ paper was about this particular method that lined strategies, and advising that the newer recom-
has some tricky issues involved,” said Janes. “We mendations be universally adopted.
worked to thoroughly understand these issues in order
to write up a paper that summarized in plain language Janes continues her work with environmental issues.
the tricky statistical issues in a way that would be help- She is currently at the Department of Biostatistics at
ful to epidemiologists.” Johns Hopkins University, working with Francesca
Dominici and Scott Zeger in environmental epidemiol-
One of the original pioneers of case-crossover design, ogy and developing a new method for diagnosing con-
Murray Mittleman, wrote a commentary to accompany founding bias.
the article in the journal, basically sanctioning the out-
tions and compressions only had 10% survival. And What do you like to do in your spare time?
now the standard is to do 100 compressions when I love to build things. Over the years, I’ve built sev-
you first get there, before you do anything else. Then eral of the houses my family has lived in.
maybe do ventilations.
We have a house now near Lake Union, one I didn’t
So I feel very proud about that because I was fight- build, and I’ve finally convinced my wife that since
ing tradition at the very beginning just to bring it up. we’ll soon be doing a lot of world travel, we should
move downstairs and rent the upstairs. So now I get
to do the downstairs kitchen.
UPCOMING EVENTS
March 1 UW Biostat and EH Professor Gerald June 25-28 2007 WNAR/IMS Western Regional
van Belle will give the SPHCM Dis- Meeting in Irvine, CA.
tinguished Faculty Lecture, “Biosta-
tistics as a Distillation of Everyday July 29-Aug 2 Joint Statistical Meetings at Salt
Experience,” 2:00-3:20 p.m., T639 Palace Convention Center in Salt Lake
Health Sciences Center. City, UT.
May 31 Betz Halloran will deliver the 2007 July 30 UW Biostat Alumni Reception at JSM.
Ross Prentice Lecture.
Sept. 20-21 Department Retreat at Pack Forest.
June 11-29 The 12th Annual Summer Institute
in Statistical Genetics, organized by For more information on UW Biostat events, please
Institute director Bruce Weir, at UW visit the UW Biostatistics Web site at www.biostat.
South Campus Center. washington.edu.