.jpg)
Lead Biostatistician
VA Puget Sound Health Care System
I do not have any relevant financial / non-financial relationships with any proprietary interests.
OMB No. 0925-0001 and 0925-0002 (Rev. 11/16 Approved Through 10/31/2018)
BIOGRAPHICAL SKETCH
Provide the following information for the Senior/key personnel and other significant contributors.
Follow this format for each person. DO NOT EXCEED FIVE PAGES.
NAME: Anna Korpak
eRA COMMONS USER NAME (credential, e.g., agency login):
POSITION TITLE: Biostatistician
EDUCATION/TRAINING (Begin with baccalaureate or other initial professional education, such as nursing, include postdoctoral training and residency training if applicable. Add/delete rows as necessary.)
INSTITUTION AND LOCATION DEGREE
(if applicable)
Completion Date
MM/YYYY
FIELD OF STUDY
Cornell University (School of Arts & Sciences), Ithaca NY BA 05/2000 Biology & Society, Linguistics
University of Washington (School of Public Health), Seattle WA PhD 08/2016 Biostatistics
A. Personal Statement
I am the Lead Biostatistician with the Seattle Epidemiologic Research & Information Center (ERIC), at VA Puget Sound. Throughout my career, I have enjoyed working at various levels with experts within and outside of statistics in planning, conducting, and reporting scientific research. I have over 10 years of analytic experience in biomedical and public health research, including observational research and clinical trials in ICU and office settings. My earlier work coordinating multiple clinical trials with the ARDSNet in Boston provided broad experience and insight into how a multi-center national project is run, and inspired me to pursue advanced training in statistics. I completed my PhD in Biostatistics at the University of Washington, whose rigorous program offered me the chance to conduct my own methods research and to serve as an instructor and mentor to others. Throughout my career, I have worked to expand and hone my applied statistical skill set. I have excelled at working both independently and collaboratively with interdisciplinary teams as a statistician on projects ranging from clinical trials in dentistry (Northwest PRECEDENT) to geriatric research in the dementia care setting (UW Nursing). My extensive training has provided me with a versatile skill set, which I am eager to apply to the SPiRE study of chronic back pain in the veteran population.
B. Positions and Honors
Positions:
2000-2006 Data Manager / Analyst, ARDSNet Coordinating Center,
Massachusetts General Hospital Biostatistics Center, Boston, MA
2006-2007 Research Assistant / Data Manager, Casa Pia DCC, Seattle, WA
2007-2014 Research Assistant / Statistician, Northwest PRECEDENT, Seattle, WA
2011-2012 Biostatistics Teaching Assistant, University of Washington, Seattle, WA
2014-2017 Research Analyst / Statistician, University of Washington Nursing School (Psychosocial &
Community Health), Seattle, WA
2017-- Research Health Science Specialist / Lead Biostatistician, Seattle ERIC | VA Cooperative
Studies Program (CSP) | VA Puget Sound Health Care System
Other Experience and Professional Memberships:
2006-2013 UW Dept. of Biostatistics Computing Policy Committee, student representative
2006-- American Statistical Association, member
2006-- Institute of Mathematical Statistics, member
2006-2011 Kirschstein Training grant (NIDCR)
2010-2012 University of Washington IRB Committee A, student member
C. Contributions to Science
1. As I am early in my career, I will focus on my methodology research from my graduate work. My dissertation, "Methods for Hypothesis Testing in Animal Carcinogenicity Experiments" evaluates the underlying hazard assumptions and performance characteristics (type I error and power) of existing methods for analyzing experiments with time to event data where the event is not directly observed; such data frequently occurs in 2-year rat and mouse studies where tumor onset is the event of interest. A very popular statistical test used in this setting is the poly-k test, which can be biased when its parametric assumptions are not met and treatment toxicity induces differential intercurrent mortality among treatment groups. My work expands on what is understood about the underlying assumptions of the poly-k test in particular, and introduces a new method to extend the poly-k by using a maximum likelihood estimator (MLE) to supply its k parameter. The new MLE-based test performs as well as the standard poly-k with correctly-specified k, and outperforms the poly-k with incorrectly-specified k in the presence of treatment toxicity. Crucially, the MLE approach is found to perform better than an earlier competitor estimator, with the added appeal that it does not require that the experiment implement interim sacrifices.
Korpak A, McKnight B. Maximum Likelihood Estimation of the k Parameter in the Poly-k Trend Test.
(in submission)
2. My collaborative applied work has included contributions to publications in the clinical trials literature. Two examples are described below.
The ARDSNet conducted several multicenter pulmonary ICU trials, assessing drug treatments, ventilator settings, and fluid management strategies for patients with acute lung injury (ALI) and acute respiratory distress syndrome (ARDS). There was interest in determining whether usual-care practices were modified following results from one of the studies, which compared ventilator tidal volumes of 6ml/kg to 12ml/kg, and concluded that lower tidal volume was preferable for patient management. Using data from the first 6 ARDSNet studies, which spanned nearly a decade, we were able to examine how practice changed over time, and also examined how some pre-enrollment characteristics were associated with outcome.
Checkley W, Brower R, Korpak A, Thompson BT. Effects of a Clinical Trial on Mechanical Ventilation Practices in Patients with Acute Lung Injury. Am. J. Respir. Crit. Care Med. 177, no. 11 (2008) 1215-1222. doi: 10.1164/rccm.200709-1424OC.
The Northwest PRECEDENT was a dental practice-based research network which conducted several randomized and observational studies in general dentistry, craniofacial, and orthodontic care. One study was concerned with white spot lesions following removal of orthodontic appliances. White spot lesions are common in this population, and beyond posing an aesthetic problem, are an early indication of tooth decay. Subjects were randomized to either usual home hygiene, or one of two commonly used agents for reduction of white spot lesions: MI Paste Plus or PreviDent fluoride varnish. Our analysis did not find evidence that either of the treatment groups differed significantly from usual care.
Huang GJ, Roloff-Chiang B, Mills BE, Shalchi S, Spiekerman C, Korpak AM, Starett JL, Greenlee GM, Drangsholt RJ, Matunas JC. The effectiveness of MI Paste PlusTM and PreviDent® fluoride varnish for treatment of white spot lesions: a randomized controlled trial. American Journal of Orthodontics and Dentofacial Orthopedics 143 (2013): 31-41.
3. In another collaboration, I have had the opportunity to contribute to development and validation of clinical screening tools for use with family/friend caregivers of patients with dementia (PWDs). One of the challenges in working with caregiver/care-recipient dyads is in identifying caregivers who are most in need of additional support, information, or training resources. Existing tools that measure stress only address one facet of the difficulties that caregivers may be facing. We evaluated a new survey instrument to assess caregiver activation – a construct that includes self-assessed skills, confidence, and level of engagement – which has the potential to improve and guide interactions between clinicians and caregivers of PWDs.
Sadak T, Korpak A, Borson S. Measuring Caregiver Activation for Health Care: Validation of PBH-LCI:D. Geriatric Nursing 36, no. 4 (July 2015): 284–92. doi:10.1016/j.gerinurse.2015.03.003.
D. Additional Information: Research Support and/or Scholastic Performance
Current Support
Seattle ERIC staff (VHA Cooperative Studies Program) 04/2017--
Director: Nicholas Smith
As a member of the research / statistical staff at the Seattle ERIC, my position is VA-funded through the Cooperative Studies Program.
Past Research Support
Robert Wood Johnson Foundation grant, NIH R21AG054144-01 03/2014 – 04/2017
PI: Tatiana Sadak
Dr. Sadak contributed to a pool of funding used to fund research by the dementia caregivers research group at the UW School of Nursing, for which I served as primary statistician.
Predoctoral Research Position with UW Dept of Biostatistics Data Coordinating Center 07 09/2007 – 03/2014
(PRECEDENT practice-based research network, funded by NIDCR U01 grant DE016752)
PI: Brian Leroux
The National Institute of Dental and Craniofacial Research (NIDCR) awarded a grant to fund a practice-based research network (PBRN) in dentistry to the University of Washington, which operated the Northwest PRECEDENT PBRN jointly with OHSU in Portland. My projects under this grant focused on statistical methods and applications to dental research, including human subjects (RCT and observational) studies on third molar extraction outcomes, white spot lesions, and other topics of clinical importance.
