Faculty in the Division of Biostatistics serve as principal investigators on several funded research projects, a sample of which are listed here.

Causal mediation analysis with machine learning to understand comparative treatment effects

Current methods for mediation analysis are limited and don’t work well with certain study designs and outcomes including those from observational studies or those with large amounts of data from electronic health records. In this study, the research team is using machine learning to create new methods for conducting mediation analyses, with application to treatments for opioid use disorder.

Principal Investigator: Ivan L. Diaz, PhD , Kara Rudolph, PhD

Funding: PCORI

Comparing treatment approaches to promote inpatient rehabilitation effectiveness for traumatic brain injury

The current proposal seeks to compare the effectiveness of specific inpatient rehabilitation approaches to optimize functional outcomes and community participation for persons with traumatic brain injury. Revealing the most effective approaches associated with the best outcomes will be investigated through a prospective observational cohort based on clinician-driven design of standardized electronic medical records documentation, analyzed with advanced causal inference methods. The results will have the potential to advance standards of practice in rehabilitation and to arm clinicians and healthcare administrators with evidence-based results to advocate for the use of the most effective practices in TBI rehabilitation.

Principal Investigators: Erinn M. Hade, PhD, Jennifer Bogner, PhD, Cynthia Beaulieu, PhD

Funding: NIH/NINDS

Complex WTC exposures impacting persistent large and small airflow limitation and vulnerable subgroups in the WTC survivor population

We propose to employ innovative statistical and machine learning methods to assess the effects of WTC exposures and comorbidities on the full spectrum of distributions of the post-bronchodilator spirometry and oscillometry measurements, both cross-sectionally and longitudinally. Another aim of this project is to identify and evaluate vulnerable subgroups that show heterogeneous susceptibilities to WTC exposures to reduce health disparity. The findings from this study will improve our knowledge of the health effects on severe respiratory diseases related to the 9/11 terrorist attacks in the survivor population as a whole and within specific vulnerable sub-population, and will inform future risk management, clinical treatment, and address health disparities of patients in the WTC-EHC cohort.

Principal Investigator: Mengling Liu, PhD and Joan Reibman, MD

Funding: CDC/NIOSH

Semiparametric methods for modeling of time-dependent environmental exposures

We seek to develop and implement a family of innovative statistical models to assess the effects of time- dependent environmental exposures, pinpoint critical windows of vulnerability, and identify susceptible subpopulations. We will apply these methods to ongoing collaborations to address critical scientific questions in environmental health research and to improve our understanding of how environmental exposures affect health. The statistical tools developed from this project will be disseminated to the general research community and will expand the analytic toolkits for environmental health research.

Principal Investigator: Mengling Liu, PhD

Funding: R01 NIH/NIEHS

Cognitive decline among WTC survivors with chronic mental and physical disorders

Twenty years after 9/11 terrorist attack on the World Trade Center (WTC) towers, there is a high prevalence of post-traumatic stress disorder (PTSD) and chronic lower respiratory symptoms among WTC-affected community members and a lack of information about cognitive status among the aging cohort of WTC-affected community members and about the association between WTC exposures, chronic PTSD, severe lung disorders, cognitive status, and cognitive decline over time. To improve diagnosis and treatment and to identify the biological underpinning of the WTC-related health disorders, this project proposes to investigate the biological mechanisms linking WTC exposures, comorbid PTSD and severe lung disorders, with cognitive function or decline rate borrowing strength from an existing CDC/NIOSH designated treatment/surveillance program for WTC Survivors and use a combined case-control and longitudinal study design. Latent class longitudinal regression and related causal inference are used for assessing the effect of inflammation and neurodegeneration biomarkers as mediators of acute and chronic WTC exposures and risk of cognitive abnormality as well as investigating these blood-based biomarkers as possible modifiable targets for novel interventions.

Principal Investigators: Yongzhao Shao, PhD , Joan Reibman, MD , Thomas Wisniewski, MD

Funding: U01 CDC/NIOSH

Data Coordinating Center: Pulmonary embolism: Thrombus removal with catheter-directed therapy (PE-TRACT Trial)

Submassive pulmonary embolism (PE) is a common dangerous disease that often causes cardiopulmonary disability, reduced functional capacity, and decreased quality of life; catheter-directed therapy (CDT), by acutely reducing clot burden and improving right ventricular function, has the potential to prevent these long- term complications. This study will determine if CDT, rather than anticoagulant therapy alone, should be used to treat patients with submassive PE. The Data Coordinating Center, led by experts with deep experience in adaptive and pragmatic clinical trial design and implementation, will oversee all aspects of study design, implementation, and analysis, and will provide state-of-the-art data collection and quality control infrastructure, reporting to the independent Data and Safety Monitoring Board, and provision of data and resources to the public.

Principal Investigators: Andrea B. Troxel, ScD, Thaddeus Tarpey, PhD

Funding: U24 NIH/NHLBI