Baylor University
Department of Statistical Science
College of Arts & Sciences

Baylor > Statistical Science > News > 2012 SPAIG Award

2012 SPAIG Award

Press release at Eli Lilly and Company


The American Statistical Association (ASA) presented its Statistical Partnership among Academia, Industry, and Government Award to Lilly and Baylor University at its annual meeting on July 31. The prestigious award, given only once each year, recognizes Lilly and Baylor for establishing a partnership to enhance pharmaceutical R&D.

Over the years, Lilly and Baylor have worked together to use Bayesian statistical and probability methods to solve preclinical and clinical problems. The partnership also has enabled Lilly to recruit top students from Baylor.

Lilly's initial collaboration with Baylor paved the way for the creation of the Lilly statistics network, which includes partnerships with other universities and consultants and companies from around the world. These collaborations have helped Lilly access statistical talent wherever it exists to help solve business and drug development problems.


Lilly/BU SPAIG Award



The American Statistical Association (ASA) awarded the 2012 SPAIG Award to Baylor University and Eli Lilly and Company. The citation read as follows:

The SPAIG committee is pleased to inform you that you have been selected to receive the 2012 SPAIG Award.  The committee was impressed by the strong and enduring relationship between Baylor University and Eli Lilly which has brought methodological enhancements to both clinical and non-clinical pharmaceutical applications.  We also noted that this successful collaboration has paved the way for partnerships between Eli Lilly and three other universities.  It is clear that this partnership has been beneficial to both parties.  The impressive number of graduate students and faculty supported by Eli Lilly as well as the hiring of Baylor graduate students indicates that the partnership is highly valued by both entities.


The formal award was presented at the Joint Statistical Meetings in San Diego on July 31, 2012.

The SPAIG award is as follows

The American Statistical Association, through the Statistical Partnerships among Academe, Industry and Government (SPAIG) Committee, is given in recognition of outstanding partnerships established between academe and business, industry, & government organizations–and to promote new partnerships. This award differs from other ASA awards in that it recognizes organizations instead of individuals.

Recent recipients are;

  • 2002 - Iowa State University, General Motors Corporation, and the Mayo Clinic
  • 2003 - Harvard University and Schering-Plough Research Institute
  • 2004 - Temple University and Merck & Co.
  • 2005 - NISS Affiliates (collaborations among 27 academic institutions, 9 government agencies, and 15 corporations)
  • 2006 - Worcester Polytechnic Institute and National Center for Health Statistics
  • 2007 - University of Michigan, University of Pennsylvania, and National Cancer Institute, National Center for Health Statistics, National Center for Chronic Disease Prevention and Health Promotion, and Information Management System
  • 2008 - GlaxoSmithKline and North Carolina State University
  • 2009 - Iowa State University and Los Alamos National Laboratory
  • 2010 - Pfizer and Rutgers University

Description of the SPAIG Award for the Collaborative research effort between the Department of Statistical Science at Baylor University and the Eli Lilly and Company.

The Award citation is

            “In recognition of the shared vision and mutual support between Eli Lilly & Company and Baylor University in the creation of a partnership for the advancement of biopharmaceutical statistical research and the training of statistical scientists.”

 

Brief history of the collaboration 

Collaboration between Baylor and Eli Lilly and Company began in 1996 when Stacy Lindborg (then Stacy David) joined Lilly after completing her Ph.D. at Baylor.  The initial project involved application of the methods Dr. Lindborg developed in her dissertation to Lilly data from an open-label comparative study involving approximately 1000 patients with Type I diabetes in a crossover design.  This work was conducted with Dr. John Seaman, her advisor at Baylor, and involved another Lilly statistician, Rocco Brunelle.  The work resulted in multiple presentations at professional meetings and furthered the understanding of Lilly data.

Subsequent collaborative work resulted from Dr. Lindborg’s research on missing data problems associated with her work on a drug for schizophrenia. This problem was addressed by a Baylor doctoral student under the direction of Dr. Seaman. This initial work was later expanded upon by several other Baylor doctoral students.  Dr. Aarti Shah, Vice President of Global Statistical Science and Advanced Analytics at Eli Lilly and company has described this initial effort as

The genesis of the Baylor/Lilly relationship came from the standard relationship between professor and former student. As Dr. Stacy Lindborg pursued extensions of her dissertation work with her advisor, Dr. John Seaman, the benefits to both parties became very clear. However, realization of these benefits would take many years of cultivation to achieve the level of outstanding statistical partnership consistent with the SPAIG Award.

The collaborative effort entered into a new phase in the fall of 2005 when Baylor was awarded $100,000 by the Lilly Foundation for the purpose of expanding the existing informal relationship across Lilly (i.e., non BU alumni) and Baylor (i.e., engaging all faculty members).  In 2006 projects were solicited across the statistics group at Lilly and prioritized by Lilly based on business value.  This list was then shared with Baylor Faculty and six problem areas were selected for Baylor faculty and students to work on with Lilly statisticians through out the summer.  The projects spanned areas of development at Lilly touching early clinical work (PK/PD and other pre-Phase III), Phase III, commercialization (Phase IV) and manufacturing spanning a broad range of problems:  Semi-parametric Modeling of PK, Metabolic Syndrome, Adaptive Models, QC Time Series, Nonparametric Modeling and PK Modeling.

In May of 2006, a group of Lilly statisticians traveled to Baylor to introduce the statistical and clinical nature of each problem to Baylor’s faculty and students and to begin work on the research.  Lilly statisticians serving as the industry lead for these research projects included both Baylor alumni and alumni from a variety of other institutions.  Additionally, the entire faculty from Baylor were involved in the research in addition to all graduate students currently active in the graduate program.  At the conclusion of the summer, the students presented their results as posters for faculty, students (both undergraduate and graduate) and administration within Baylor University. From this group several students and faculty traveled to Indianapolis to share the results through a poster session and oral talks broadly with Lilly staff.  As the research matured, 6 separate dissertations emerged from the research involving 4 Baylor Faculty.  The collaborations also produced multiple presentations at JSM and a subsequent internship at Lilly for a Baylor doctoral student. 

In 2008 the collaborative relationship shifted course at Lilly’s request whereby Baylor faculty and doctoral students would be working on projects from Lilly’s portfolio driven by business need. This change was done in order to expand the current Baylor/Lilly relationship to include opportunities for the Baylor faculty and students to provide formal consultation on Lilly projects and receive appropriate funding for their work.  Baylor is one of four universities with Master Service Agreements (MSA) in place through the Global Statistics department at Lilly.  This pursuit of MSA with other universities grew out of the maturity and productivity between Lilly & Baylor.  The initial amount of the MSA with Baylor from the Statistics department was $50,000. To date this project has generated nearly $200,000 for the department.

 

Specific Benefits of the Partnership

Baylor University perspective

·       Students have benefited from having meaningful research experiences before beginning their dissertation.  Baylor Ph.D. students were eligible to work on these problems beginning in their second summer in the program, by which time they will have acquired their M.S. The partnership enabled students to have this research experience for two summers prior to beginning their dissertation research. It should be noted that it was not the original intent of the program to use these projects as a part of a student’s dissertation problem but several students have already selected dissertation areas that are natural extensions of their earlier work. 

·       Ph.D. students and faculty have benefited from interacting with statisticians and other scientists from the pharmaceutical industry.  In our experience with the five students mentioned above, we found that interaction with other Lilly scientists was greatly beneficial for the student, underscoring the need to communicate well and to understand the science involved, as well as the statistical methodology to be used.  A side-benefit is that students learned about the working world of biopharmaceutical statisticians.

·       The doctoral program has benefited by

-        Enhancing our ability to attract potential graduate students. Many bright students are attracted to graduate programs in the sciences and medicine because they want to work on problems that contribute to the betterment of society.  This goal is rarely realized before they begin their dissertation research in their third or fourth year.  Our joint work with Lilly engaged students in important research as early as the beginning of their second year.  This prospect is a valuable recruiting tool for our program.

-        Providing the faculty and students with a wealth of potential dissertation topics.  Not surprisingly, the teams' efforts uncovered more research problems than could be addressed quickly. Some students continued to work on these problems for their dissertation.

-        Fostering more research interaction among the faculty and with Lilly statisticians.  One of our professors has enjoyed such joint research with Lilly since 1996.  As the collaboration has progressed other faculty members have also been allowed to identify interesting biopharmaceutical problems and work in cooperation with Lilly scientists.

-      Having Lilly statisticians and other scientists serve as outside members on dissertation committees. This did much to enhance the quality of the committee and the direction given to the doctoral student.  (Since 2002, four of our Ph.D. students have had Lilly statisticians serve on their dissertation committees with great success.)

Eli Lilly & Company perspective

·       Within the industry, we are almost always constrained by time (i.e., deadlines).  There are times when we, as scientists, might feel the methods we’re using are adequate but perhaps not optimal (e.g., historical use of LOCF).  However, time constraints might not allow us the opportunity to pursue methodological advances.  This collaboration with Baylor enables a number of benefits. 

-       Facilitate the furthering of science through collaborative research that has directly advanced our understanding of specific clinical trial data. 

-      Validation of methodological work and research.  We have over 350 statisticians at Lilly, many of whom are active in research.  Collaboration with academia has the potential to strengthen the research (validates it) and encourage publication/application outside Lilly.

-      Sharing of Resources:  At Lilly we have an endless source of very interesting and relevant research problems; including Data & Banking databases to go with them.  Under multiple circumstances we have been able to share data (with appropriate confidentiality agreements in place) with external researchers and together learn beyond what we could have done individually.

-      Lilly statisticians to contribute within Academia by helping expose students to the pharmaceutical industry.  Most students don’t know what a job in the industry entails, yet they will need to make crucial career decisions as they approach graduation.  Working closely with students allows arm students with knowledge so that they make the best career decision for them.

-      Recruiting:  while this has never been an explicit goal tied to the Baylor/Lilly collaboration, working closely with talented students has facilitated hiring a number of Baylor students for employment at Lilly upon their graduation.  The exposure the students get to industry data and methodological challenges through out their graduate training as a result of the program facilitates those hired by Lilly to get up to speed quickly upon employment.


Professor John Seaman's response when asked to answer some questions concerning the benefits of the collaborative effort with Lilly.

  1. Did you imagine your work with Dr. Lindborg leading to this collaboration between the Department of Statistical Science and Eli Lilly? What was your original goal (i.e. long term/short term?)?  My goal at the time was to simply work with a former PhD student, which was common for me.  That she was working in industry was different at the time (1996), and it prompted us both to wonder how we  could sustain joint scientific pursuits given the differences in our respective institutions.  Very early we naturally focused on problems that were of interest to us both, and of potential practical application at Lilly.  We also knew we could not pursue problems with hard deadlines—bottom-line issues for Lilly—if we were ever to involve PhD students.  That was a major consideration in our long term goals—we wanted to be able to involve PhD students.  We continue to work on such problems, but the collaboration has expanded to include joint research primarily between Baylor faculty and Lilly scientists.  
  2. How does it feel to know that your work with Dr. Lindborg led to such a beneficial collaboration?  Like any professor, I'm delighted to see work with a former student become so fruitful.  That it continues to enrich the research experience of current students is icing on the cake.  
  3. What is the biggest benefit, in your opinion, for the university with this collaboration?  Our continued involvement with rich, practical biomedical problems.  Faculty and student research benefit tremendously from this collaboration.  Many of our students work at biomedical research institutions—research hospitals, pharmaceutical firms, etc--upon graduation, so having one or more chapters in their dissertations with that focus is very important to them, both in landing such jobs, and in being successful in them.  
  4. What does winning the SPAIG award mean to you? What does it mean for the department? The university?  For me and the department, it is an acknowledgment of over 15 years of effort in establishing this partnership and in keeping so many of our former PhD students involved with us.  
  5. Where do you see the collaboration going in the future?   I think that, in addition to continuing to play a major role in the research we do, it affords involvement in other scientific efforts of importance to us.  For example, we are involved with an international effort to incorporate certain advances in statistical science into biopharmaceutical research.  This effort is being led by another former student, Dr. Karen Price.  She plays a critical role in the Baylor/Lilly collaboration and, because of that connection, thought of us when she began organizing this new effort.