Clinical research is a delicate industry that only a few skilled specialists can handle head-on. Dr. Chen Yang, a prominent figure in biostatistics at the Icahn School of Medicine at Mount Sinai, spearheads innovative statistical analysis methods. With a PhD in statistics and extensive experience in Actuarial Science, Dr. Yang’s work is making significant strides in reshaping clinical research.
Methodological Innovation: Expanding Applications
The new tool enables researchers to investigate treatment effect disparities within the framework of stepped-wedge cluster randomized trials (SW-CRTs) and facilitates dynamic monitoring of trial progress across multiple periods, enhancing precision in clinical trial outcomes, particularly in studies focusing on health disparities.
A stepped-wedge cluster randomized trial (SW-CRT) is a unique research structure that assesses service delivery interventions. Within this design, clusters usually hospitals or particular communities are initially given no intervention during the baseline period and randomized to receive the intervention at subsequent periods.
Specifically, Dr. Yang’s paper proposes three approaches for determining the optimal sample size to detect interaction effects in cross-sectional SW-CRTs. Among these approaches, the generalized estimating equation (GEE) method demonstrated the largest simulated power, while the GEE-MD (Mancl and DeRouen correction) method had the smallest. With a cluster size of 120, the GEE approach achieved over 80% statistical power.
Furthermore, when the binary covariate was balanced (50%), the simulated power increased compared to an unbalanced binary covariate (30%). The paper also found that with an intermediate effect size of heterogeneity in treatment effect (HTE), only cluster sizes of 100 and 120 provided more than 80% power using GEE for both correlation structures examined.
When comparing the impact of increasing cluster size versus increasing the number of clusters, the latter showed a slightly higher gain in power. For example, when the cluster size changed from 20 to 40 with 20 clusters, power increased from 53.1% to 82.1% for GEE, 50.6% to 79.7% for GEE-KC (Kauermann and Carroll correction), and 48.1% to 77.1% for GEE-MD. In contrast, when the number of clusters changed from 20 to 40 with a fixed cluster size of 20, power increased from 53.1% to 82.1% for GEE, 50.6% to 81% for GEE-KC, and 48.1% to 79.8% for GEE-MD.
An SW-CRT is most useful when resources are scarce or when irreversible changes caused by the intervention prevent the use of traditional crossover-group randomized trials, for example, the intervention is a training program for physicians or redevelopment of communities.
Beyond the Clinic
Beyond clinical trials, Dr. Yang’s methodological expertise extends to the realm of public health research. In his 2024 paper, “Longitudinal Assessment of Association Between Tobacco Use and Tobacco Dependence Among Adults: Latent Class Analysis of the Population Assessment of Tobacco and Health Study Waves 1–4,” Dr. Yang and his colleagues explore the development of tobacco dependence among users of various tobacco products. Given that tobacco users often engage with multiple products simultaneously, traditional categorical distinctions among different user types are inadequate.
Leveraging Dr. Yang’s expertise in methodological research, the team successfully identified and quantified distinct tobacco use profiles. They identified three consistent profiles across four survey waves: dominant cigarette users (62%-68%), poly users with high propensity of using traditional cigarettes, e-cigarettes, and cigars (24%-31%), and dominant smokeless product users (7%-9%).
Furthermore, they examined the longitudinal association between tobacco use and dependence, finding that tobacco dependence was significantly lower among poly users and dominant smokeless users compared to dominant cigarette users. This finding is crucial for tailoring tobacco use reduction and cessation programs to the specific needs of different user profiles.
Dr. Yang’s application of latent class analysis to the high-quality longitudinal data from the PATH study has measurably advanced the understanding of the dynamic relationship between tobacco use and dependence over time. By quantifying the relative risks of tobacco dependence among different user profiles, his research provides actionable insights for public health interventions.
For example, the identification of cigarette-dominant users as the group with the highest risk of tobacco dependence can guide the allocation of resources and the design of targeted cessation programs. This data-driven approach to understanding and addressing tobacco dependence exemplifies how his methodological contributions can directly inform and improve public health strategies.
Dr. Yang’s contributions to public health also include the investigation of the association between social vulnerability and breast cancer incidence rates. In his recent paper, “Medicaid Expansion in California and Breast Cancer Incidence Across Neighborhoods with Varying Social Vulnerabilities,” Dr. Yang and his colleagues found that the 2014 Medicaid expansion significantly reduced disparities in localized-stage breast cancer incidence rates between individuals residing in communities with high and low levels of social vulnerability in California. These findings are instrumental in supporting the ongoing evaluation of healthcare reforms and informing targeted cancer prevention efforts.
In addition to public health issues, Dr. Yang has contributed his skills to health services research. In his recent paper, “Prognostic Understanding, Goals of Care, and Quality of Life in Hospitalized Patients with Leukemia or Multiple Myeloma,” Dr. Yang and his colleagues identified significant discordance between patients with blood cancers and their hematologists regarding prognosis and goals of care.
These results highlight a gap in patient-hematologist communication, as such discordance may hinder patients from making informed treatment decisions that reflect their values and preferences. The findings suggest the necessity for communication training programs for practicing hematologists, trainees, advanced practice nurses, and other members of the medical team, as well as the importance of robust psychosocial support programs for patients.
Emerging Trends and Future Directions
In the multifaceted landscape of research and academia in the United States, Dr. Yang’s work is poised to address the growing demand for sophisticated statistical methodologies. Federal funding for research is on the rise, and the integration of advanced technologies, such as AI and machine learning, is gaining traction faster than ever.
Dr. Yang’s contributions are well-positioned to meet the evolving needs of the medical field, which needs to continuously adapt and improve to thrive constantly.
Dr. Yang’s work exemplifies the crucial intersection of innovation and practical application in the field of biostatistics. Through life-changing methods in clinical research, risk management, and financial modeling, he has advanced the precision and applicability of statistical analysis.
As the scope of research and academia changes with increased funding and technological integration, Dr. Yang’s innovative approaches are well-suited to meet emerging challenges. Moreover, emphasizing rigorous validation, Dr. Yang and his peers underscore the importance of accuracy and reliability in statistical innovation.
Dr. Chen Yang offers a thoughtful perspective on his work: “The ultimate goal of our research is to enhance the precision and applicability of statistical methods across various domains. We aim to provide researchers effective tools to address complex, real-world problems by continuously refining our techniques.”
“The future of data science holds immense potential, and I am committed to contributing innovations that advance knowledge and drive positive outcomes in clinical research and beyond,” Dr. That concludes.
Comments are closed.