Lily Hsieh, doctoral student in Health Policy concentrating in Decision Science, successfully defended her thesis, The Synergy of Real-World Evidence, Transportability, and Decision-Analytic Modelling in the Evaluation of Disease Screening Programs. Her dissertation committee was chaired by CHDS faculty Nicolas Menzies and included Issa Dahabreh and CHDS’ Jane Kim.
Hsieh’s thesis used decision-analytic modelling to evaluate the cost-effectiveness of an emerging class of tuberculosis (TB) infection diagnostics, leverages real-world data to improve understanding of long-term mortality following TB diagnosis, and develops a novel method to quantify and correct for bias in value of information analyses, enhancing the translation of clinical trial findings into real-world healthcare decisions. Her aim was to assess model-based evaluation of the effectiveness and cost-effectiveness of disease screening programs to inform clinical and policy decisions.
In her first chapter, Hsieh evaluated the cost-effectiveness of using a newer method of TB screening, host-response-based transcriptional signatures (HrTS), to screen for incipient tuberculosis (TB) among migrants arriving in the United States. Since the current standard of care has a low predictive value for identifying incipient TB, it is worthwhile to know if HrTS would be a cost-effective replacement. In this study, she created an individual-based discrete event simulation model to compare the projected health and economic impact of four post-arrival TB screening strategies. She found that HrTS may be cost-effective in specific migrant subgroups, however, results were sensitive to several assumptions, including progression risk trends post-entry.
For her second chapter, Hsieh estimated the elevated long-term mortality risks among individuals with pulmonary TB compared to matched individuals without TB using a retrospective longitudinal matched cohort study. She leveraged claims data, national TB and mortality registries, and electronic health record in Taiwan, and found that people with a history of pulmonary TB had a 17% lower 10-year survival probability, of which 35% was attributable to the post-TB period. Her findings highlight the importance of early TB detection and prevention. They also indicate that not accounting for long-term TB mortality risks may underestimate the value of TB intervention programs in policy modeling studies, affecting resource allocation decisions.
Finally, Hsieh proposed a novel framework to incorporate transportability methods from causal inference into value of information (VOI) estimation to enhance the translation of clinical trial findings into real-world healthcare decisions. Decisions on whether to adopt new healthcare technologies often rely on evidence from clinical trials but often the trial population differs from new intended recipients. VOI is often used to determine whether new evidence is needed before expanding the technology to the new groups. However, VOI analysis has historically focused on quantifying the benefit of reducing variance of the effect estimates, and not on the value of reducing uncertainty from applying study results from a trial population into a new target population. Hsieh’s framework defined metrics to quantify components of systematic errors that contribute to biases in the value of sample information when transportability issues are overlooked.
Related news: CHDS at INFORMS 2024
Related news: Hsieh Received Merit Fellowship