Rutgers Faculty Receives PCORI Contract to Develop Robust Methods for Longitudinal Causal Inference with Machine Learning


Newswise – Rutgers School of Public Health associate professor Liangyuan Hu has been awarded a three-year, $1,069,876 contract from the Patient-Centered Outcomes Research Institute (PCORI).

Hu will use the funding to develop robust and flexible causal inference methods and tools, leveraging machine learning, for longitudinal treatments. The methods developed will be used to analyze quasi-experimental or non-experimental data to generate concrete evidence for treatment decisions and intervention policies.

Real-world evidence is essential to answering clinical questions in comparative effectiveness research (CER) and patient-centered outcome research. In many circumstances, randomized controlled trials are not practical or ethical and their strict inclusion and exclusion criteria limit generalizability to vulnerable populations such as the frail and elderly, racially disadvantaged groups, and those at risk of serious morbidity and mortality. It is therefore essential to draw causal inference from large-scale data collected in real-life clinical settings to develop meaningful policy-related interventions with patient-centered outcomes.

There is a substantial set of causal inference methods with time-bound treatment. By comparison, causal inference methods, especially flexible ones using machine learning, for time-varying processing are relatively rare due to the additional complexities associated with time-varying confounding, selection bias, and longitudinal data structures. Additionally, existing approaches in this field no longer address the growing challenges posed by complex health data structures and processing models.

Hu and his team will propose new methods to address these critical methodological gaps in longitudinal research on causal inference and to answer important patient-centered outcome research questions.

“We will develop a suite of robust and flexible longitudinal causal inference tools that will provide a key analytical apparatus for researchers working with real-world clinical data,” Hu said. “Specifically, we will develop a new robust structural marginal quantile model to draw simultaneous causal inference about longitudinal treatments across the entire outcome distribution and further improve model flexibility using machine learning. For the censored survival outcomes, we will first develop a novel joint, continuous-time, marginal structural model for the restricted average survival times, and then develop a machine learning method based on Bayesian likelihood that can take into account time-varying covariates to estimate a set of weights to correct for time-varying confounding or selection bias due to informative censoring. To address the “no unmeasured longitudinal confounding” hypothesis, we will further develop a flexible and interpretable sensitivity analysis framework. Finally, we will develop open-source software within the R computing platform and implement our novel methods to address emerging questions about RECs in cardiovascular disease research and COVID-19 research.

PCORI is an independent, nonprofit organization authorized by Congress in 2010 with a mission to fund research that will provide patients, their caregivers, and clinicians with the evidence-based information needed to make more informed healthcare decisions.


About the Rutgers School of Public Health

The Rutgers School of Public Health is New Jersey’s only accredited public health school that seeks to improve health and prevent disease in diverse populations in New Jersey and around the world by educating students to be leaders, researchers, and qualified and effective public health researchers. practitioners; conduct research to advance science and public health policy; and deliver service programs that promote population and individual health. Visit us at and follow us on Facebook, Twitter, Instagram, and LinkedIn to learn about how we’re “keeping the ‘public’ in public health.”


Comments are closed.