Written by Robert Hamama
Photo by CDC on Unsplash
Clinical trials serve as test runs that generate answers to questions about the safety and effectiveness of medical treatments. The data gathered from clinical trials may appear as a series of numbers, but they represent something much deeper than that: a cure.
“Statistics are not just numbers; they are the backbone of decision-making in healthcare,” says Dr. Chen Yang, a prominent biostatistician at the Icahn School of Medicine at Mount Sinai.
Today, data drives a good portion of people’s lives. Walking one foot ahead of the trend, Dr. Yang’s innovations breathe new life into clinical trials, offering new insights and methodologies vital for disparity research and beyond.
Navigating the Complexity of Risk Pricing: A Fresh Approach to Long-Term Insurance Sustainability
In his early work, Dr. Yang delved into the intricate challenges posed by pricing risks tied to long-term insurance products, such as annuities and long-term care expenses. By focusing on the stochastic processes underlying these payout patterns, he highlighted a critical risk: even without large, unexpected payouts, insurance companies could face insolvency if these risks are not priced fairly.
Dr. Yang’s research underscores the importance of developing more accurate pricing models that allow insurers to mitigate risks through reinsurance. Although simplifying stochastic assumptions yields elegant results, he acknowledges the limitations of these models in real-world applications, advocating for gradual refinement to better reflect market complexities.
Building on his previous work, Dr. Yang expanded his research into risk-based clustering analysis of time series, offering an innovative approach to financial data categorization. By clustering time series based on tail order and jump tail dependence, his method enhances portfolio resilience, minimizing the chances of simultaneous, catastrophic losses across assets during critical events like financial crises. This breakthrough provides a powerful tool for investment decision-making, yet the process of determining risk-based similarity measures proved challenging, often requiring extra assumptions.
To address this, Dr. Yang developed a nonparametric solution-the “average block-minima estimator”―which computes risk-based similarity directly from the data, further solidifying his role as a leader in risk management.
The Metamorphosis of Clinical Trials
Dr. Yang’s journey from theoretical frameworks to practical applications began long before his current role at Mount Sinai. With a PhD in statistics and years of experience as an assistant professor of Actuarial Science in Wuhan University, he has always been fueled by a passion for applying mathematical principles to real-world problems. Further solidifying his capabilities as an expert in his field, Dr. Yang embarked on a 10-year pursuit of post-secondary education in mathematics and statistics, a 10-year research project in actuarial sciences, and a 6-year work period in economics and management school.
Dr. Yang joined Mount Sinai as a biostatistician when researchers at the Mount Sinai Hospital System were planning a special multi-period clinical trial for advanced cancer patients at high risk of death. In this trial, physicians may or may not receive a combination of a training program on how to conduct Goal-of-Care discussions with patients and a clinical decision support system based as a reminder.
At the end of the trial, whether a Goal-of-Care discussion occurred with each patient would be recorded as a binary outcome to examine if racial disparity existed for the effect of the program combination. Once a physician receives the training, they cannot return to the untrained state. Since the training program is irreversible, researchers considered using a popular clinical trial design known as the Stepped-Wedge Cluster Randomized Trial (SW-CRT). Unfortunately, due to the complexity of this trial design, there were no existing tools to evaluate the statistical power of such a clinical trial, posing a challenge in determining the number of patients needed to achieve sufficient statistical power.
Unlike many modern analytical tasks, clinical trials lack data at the power calculation stage. Therefore, the entire analytical procedure must be performed within a theoretical framework, resulting in a numerical approximation of statistical power based on the chosen statistical model.
Additionally, the limited number of clinics involved in this trial prevents the direct application of large sample theory. As a result, the theoretical framework must incorporate bias-reduction techniques to accommodate the small number of clinics.
Although the power calculation for this trial initially appeared intractable, Dr. Yang applied his expertise in mathematics and probability theory to the project. Drawing on his deep theoretical understanding of statistical models, he successfully transformed the statistical power into a computable function, enabling it to be calculated using software.
Consequently, Dr. Yang developed a hybrid methodology that combines both mathematical and computational components to provide a relatively conservative estimate of statistical power. The effectiveness of his proposed power calculation method was rigorously assessed through a comprehensive simulation study led by him. This drive is evident in his groundbreaking 2024 paper, “Power Calculation for Detecting Interaction Effect in Cross-Sectional Stepped-Wedge Cluster Randomized Trials: an Important Tool for Disparity Research.”
Later, Dr. Yang developed a power calculation tool particularly for this type of design based on his proposed method. This new tool allows researchers to quickly calculate the statistical power when investigating treatment effect disparities using stepped-wedge cluster randomized trials.
“Clinical trials are required to have sufficient statistical power. However, the problem is that at the design stage of clinical trials, there is no actual data collected to help determine the statistical power. As a result, probabilistic methods are needed to provide insight into power calculations,” Dr. Yang shares. “When I joined Mount Sinai, the institute was conducting a research project which evaluated the effectiveness of a training program for physicians to improve communication skills related to the Goal-of-Care discussion with cancer patients.”
The dedication to bridging this gap is clearly demonstrated in Dr. Yang’s groundbreaking 2024 paper, “Power Calculation for Detecting Interaction Effects in Cross-Sectional Stepped-Wedge Cluster Randomized Trials: An Essential Tool for Disparity Research.” Building on his proposed method, he later developed a power calculation tool for this trial design. This innovative tool enables researchers to swiftly calculate statistical power when examining treatment effect disparities using stepped-wedge cluster randomized trials.
Furthermore, it offers the capability to dynamically monitor whether the trial’s statistical power can be maintained as the trial proceeds across multiple periods, representing a significant leap over traditional, less flexible formulas.
“The ability to accurately calculate power in clinical trials is crucial, especially when studying the heterogeneity of treatment effects,” Dr. Yang explains. “Our invention enhances precision while providing flexibility that was previously unattainable. Now this tool is employed by an upcoming NIH grant application regarding improvement of end-of-life outcomes among minorities with advanced cancer.”
Clinical Methods of the Future
As Dr. Yang reflects on the positive impact of his professional initiatives, he remains optimistic about the future of research and academia.
“The field of medical research is evolving rapidly, and it’s an exciting time to be in this industry,” he concludes. “Our work is powered by a desire to solve real-world problems and improve outcomes, whether in healthcare, finance, or beyond. It’s about turning theoretical knowledge into practical, impactful solutions.”
Dr. Yang’s leap from theoretical mathematician to practical problem-solver exemplifies the dynamic nature of modern research. His innovations address present-day challenges while paving the way for future advancements. He makes sure that the field of clinical trials and statistical methodologies continues to develop meaningfully.
The research aficionado has built a reputation as a distinguished figure in terms of advanced applied science. His forward-thinking work goes beyond refining research methodologies; it sparks a movement among emerging researchers to integrate theoretical insights with practical applications.
This convergence is driving significant, real-world developments and resetting the future of research across various disciplines. Through his leadership, Dr. Yang is fostering a new era of scientific inquiry, where the impact of academic research extends tangibly into everyday life.