Gina Pugliano McKernan, PhD

Director, Biostatistics Core, Department of Physical Medicine and Rehabilitation
Assistant Director for Data Science, Human Engineering Research Laboratories

I am a data scientist and biostatistician, with over 15 years of industry and academic experience. I have a multidisciplinary background in statistics, measurement, and research, working within Integrated Delivery and Finance Systems (IDFS), the University of Pittsburgh, and the VA Pittsburgh Healthcare System. I obtained my doctorate in Research Methodology from the University of Pittsburgh, concentrating on advanced statistical methods. I have led the healthcare marketing industry in using statistical methods and Machine Learning (ML) algorithms to design service utilization, cost-savings, and capacity planning models in both the payor and provider space. I mined both patient electronic health record (EHR) and medical claims data to create Custer Lifetime Value (CLV) metrics for a multi-hospital/multi-plan IDFS. The first two years of my academic career has resulted in an immersion of primary and secondary research projects focused on the evaluation of technology, protocols, interventions, and data associated with improving the quality of life in persons with disabilities, Spina Bifida (SB), traumatic brain injury (TBI), spinal cord injury (SCI), chronic lower back pain (cLBP) and stroke. I have collaborated with investigators across multiple disciplines on federally-funded grant applications and currently serve as co-investigator (CO-I) on 11 active projects, totaling over $20 million in research funds. I was recently awarded a VA pilot grant to predict post-TBI recovery trajectories, MH outcomes, and healthcare costs from individual clinical, social determinants of health (SDOH), geographic and treatment level characteristics. In addition, my current research interest and funded work includes: using ML techniques to analyze behavioral and psychosocial outcomes of people with disabilities, creating enhanced data structures by appending socio-geographic and demographic data to existing clinical data, and analysis of randomized controlled trials involving neuro-stimulation and pharmacologics for individuals who have experienced stroke, TBI, and other conditions. I currently serve as the Co-director for the data center and Informatics Core of the NIH HEAL initiative at the University of Pittsburgh, overseeing data governance, monitoring, sharing, and statistical analysis.

Representative Publications

McKernan, GP. Identification of latent relationships between disability rates and socio-geographic variables, such as: neighborhood characteristics, living conditions, access to care, and health/preventative behaviors in veterans. Abstract presented at the American Statistical Association’s Data Science Symposium. 2021 June
Vijapur SM, Vaughan LE, Awan N, DiSanto D, McKernan GP, Wagner AK. Treelet Transform Analysis to Identify Clusters of Systemic Inflammatory Variance in a Population with Moderate-to-Severe Traumatic Brain Injury. Brain Behav Immun. 2021 Jan 29:S0889-1591(21)00030-1. doi: 10.1016/j.bbi.2021.01.026. Epub ahead of print. PMID: 33524553.
Rigot, S. K., Boninger, M. L., Ding, D., McKernan, G., Field-Fote, E. C., Hoffman, J., ... & Worobey, L. A. (2021). Towards Improving the Prediction of Functional Ambulation after Spinal Cord Injury Though the Inclusion of Limb Accelerations During Sleep and Personal Factors. Archives of Physical Medicine and Rehabilitation.
McKernan, G, Izzo S, Crytzer TM, Houtrow AJ, Dicianno BE. The Relationship between Motor Level and Wheelchair Transfer Ability in Spina Bifida: A Study from the National Spina Bifida Patient Registry [published online ahead of print, 2020 Jul 16]. Arch Phys Med Rehabil. 2020;S0003-9993(20)30430-5. doi:10.1016/j.apmr.2020.06.016
Chandrasekaran S, Nanivadekar AC, McKernan GP, Helm ER, Boninger ML, Collinger JL, Gaunt RA, Fisher LE. Sensory restoration by epidural stimulation of dorsal spinal cord in upper-limb amputees. eLife 2020;9:e54349 DOI: 10.7554/eLife.54349

Research Interests

Social determinants of health
Machine Learning
NLP of clinical notes
EMR/claims data
Image analysis

Research Grants

U19AR076725 (SOWA/VO) 9/26/19-5/31/24 2.40 CM

National Institutes of Health $16,756,172

Role: Co-Investigator

HEALing LB3P: Profiling Biomechanical, Biological and Behavioral phenotypes

The purpose of the University of Pittsburgh Low Back Pain: Biological, Biomechanical, Behavioral Phenotypes (LB3P) Mechanistic Research Center (MRC) is to perform in-depth statistical phenotyping of patients with chronic low back pain (CLBP), using a multi-modal approach, to characterize patients and provide insight into the phenotypes associated with the experience of CLBP to direct targeted and improved treatments. The proposed center will integrate novel characterizations of all of the critical contributors to chronic low back pain, including clinical, biological, behavioral, and biomechanical factors, to establish unique phenotypes associated with response to treatment. Through improved patient phenotyping that maintains the complex interaction of the various contributors to low back pain, the work of the proposed center will support the BACPAC research community, and lead to improvements in outcomes and costs for this common problem.

69A3552047140 (COOPER) 8/1/20-1/31/22 0.60 CM

US Department of Transportation $999,654

Role: Co-Investigator

ASPIRE CENTER – AUTOMATED VEHICLE SERVICE FOR PEOPLE WITH DISABILITITES – INVOLVED RESPONSE ENGINEERING

We propose to create the Automated vehicle Service for People with disabilities – Involved Response Engineering (ASPIRE) Center. This Center will investigate the implications of accessible automated vehicles and mobility services for people with disabilities and their caregivers by conducting three related projects (see Figure 1). The three projects are: 1) a systematic literature review to describe the current state of the science, 2) a Voice of the Consumer and Provider survey, journey mapping, and focus groups which elicit input from all key stakeholders on their experiences with, barriers to, and future needs and capabilities for accessible automated transportation, and 3) a project that presents summary findings, extrapolates findings to the greater population of potential automated vehicle users, combines our data with publicly available datasets to understand factors that influence travel, displays clusters of users based on their characteristics and needs, and develops solid model drawings that illustrate key features and parameters for implementing automated vehicles and mobility services.

Not Assigned (MCKERNAN) 1/1/20-5/31/2021 0.6 CM

Department of Veterans Affairs $32,310

Role: PI

HERL COMPETITIVE PILOT: IDENTIFICATION OF LATENT RELATIONSHIPS BETWEEN DIABILITY RATES AND SOCIO-GEOGRAPHIC VARIABLES IN VETERANS

The goal of this project is to use Machine Learning (ML) to identify socio-geographic and demographic factors related to disability and to inform and impact resource allocation, health care planning, access, and affordability initiatives, and care management and caregiver support programs at the community level.

Not Assigned (MCKERNAN) 1/1/2021 -12/31/2021 1.2 CM

Department of Veterans Affairs $24,228

Role: PI

HERL COMPETITIVE PILOT: MACHINE LEARNING FOR PREDICTING MENTAL HEALTH RISK FACTORS AND OUTCOMES FOR INDIVIDUALS WITH TRAUMATIC BRAIN INJURY

The goals for this study include the creation of an analytic data set of individuals who have experienced moderate-to-severe TBI derived from the Optum® de-identified Integrated Claims-EHR data, and predict post-TBI recovery trajectories, MH outcomes, and healthcare costs from person, SDOH, and treatment level characteristics. Models will examine multi-symptom recovery trajectories, comorbidity/ cumulative burden, and medical cost.

CTTP (MCKERNAN) 1/1/21-12/31/21 0.24 CM

Department of Transportation $20,000

Role: PI

MACHINE LEARNING APPROACHES TO CLASSIFY COMMUTING BEHAVIOR USING INDIVIDAUL- AND NEIGHBORHOOD-LEVEL SOCIODEMOGRAPHIC, GEOGRAPHIC, AND HEALTH FACTORS

The objective of this project is to expand the body of research on Commuting in America (CIA) by developing a single brief describing how variations in geography and demographics are related to choice of transportation mode and commuting behavior. We will create a single database containing unique socio-geographic and health related information at the geo-level and use this database to create variable structures and individual predictors of commuting behavior utilizing traditional machine learning (ML) methodologies.