The Limited Medium Survival (RMST) analysis technique was introduced in research on health care about 25 years ago and has since been widely used in financial, mechanical, businesses and other occupations.
In clinical arrangements, RMST is useful because it is a simple way to understand the average survival time-the time that patients live after diagnosis or treatment and the factors affecting this time.
In addition, unlike COX regression models and other popular models, estimates and comparisons made with the use of RMST are not based on the proportional risk of risk that the chance of an event will be stable over time.
But there is one catch: RMST can test the differences on the effect of treatment between groups from the original value at a time point-the threshold — but the recognition of the ideal threshold in clinical and epidemiological studies is difficult. This leads to results that are less statistically strong than they could be. ”
Gang Han, PhD, Professor of Biostatism, Texas A&M University School of Public Health
To tackle this challenge, Han and his colleagues in academia and industry have developed a new method that uses an existing mathematical tool-the reduced partial exponential model to determine the ideal or optimal threshold time in the limited average survival time.
“This is especially important in medical studies, as the likelihood of a specific event that happens can change in relation to the various stages of treatment,” said Matthew Lee Smith, PhD, Professor of Health Behavioral Health at the Texas A&M School of Public Health, who participated in this research.
To determine the best limit, the team calculated one year of the important points of change in risk rates and compared what they found with the longest possible threshold time.
Their research document, published in American newspaper of EpidemiologyIt showed the benefits of the proposed method in multiple simulation studies and two real examples, a clinical study and a study of epidemiology.
They used the new method to measure type 1 error rates and statistical power in simulations at which the risk rate was stable for one group and changed for another group. The groups were compared using the standard Logrank and their new model.
“Our model performed the best,” said Marcia G. Ory, PhD, Regents and a distinguished professor at the School of Public Health, which research methods of prevention based on evidence. “This happened when we applied it to two real world scenarios.”
For both scenarios, traditional statistical analysis methods did not reveal remarkable differences between two treatments. When the new model was applied, however, the results for each scenario found that a treatment was clearly superior.
The first scenario compared two treatments within seven months for patients with lung cancer non -small cell cells who had lower levels of a basic biomarker. The second used a standard evaluation to measure the time to reduce people with mild dementia who lived with caregivers compared to those who did not live with caregivers.
“These results are very promising and more research is needed that compares more than two groups and uses multiple converts, such as the age, nationality and socio -economic status of participants,” Han said. “Still, based on these first results, we believe that this method could be more powerful than all existing comparisons for two groups in analysis of the results of time for examination.”
Others participating in the study were the Department of Epidemiology and Biostatic Doctoral Students Laura Hopkins, Raymond Carroll, PhD, a distinguished professor in the Department of Statistics and Foreign Affiliates of Texas A&M by Eli Lilly and Company and H. Lee Mofitt Center Center & Research Institute.
Source:
Magazine report:
Han, G., et al. (2025). Determination of threshold time in limited average survival time analysis for two comparisons of groups with applications in clinical and epidemiological studies. American newspaper of Epidemiology. https://doi.org/10.1093/aje/kwaf034.
