New SMART design improves efficiency in adaptive clinical trials

Photo by CDC / Unsplash

medRxiv Published April 30, 2026 Medically reviewed April 30, 2026 Study authors: Chen, Z.; Hartman, H. DOI ↗ By Dr. Julia Lee, PhD · Oncology, Genomics & Drug Development

Key Takeaway

Consider RA-TF-SMART as a methodological concept for adaptive trials, but evidence is lacking for clinical application.

This study describes a new adaptive clinical trial design called the response-adaptive tailoring function SMART (RA-TF-SMART). The design was compared with balanced randomized SMARTs (BR-SMARTs), tailoring function SMARTs (TF-SMARTs), and generalized outcome-adaptive SMARTs (GO-SMARTs). The primary outcome and sample size were not reported, and no specific population or setting was described.

As a methods paper, no patient-level results, safety data, or adverse events were reported. The study did not provide numerical comparisons or statistical analyses. The design is intended to improve efficiency in sequential multiple assignment randomized trials by adapting randomization based on participant response.

Key limitations include the absence of empirical data, lack of reported funding or conflicts of interest, and no information on follow-up duration. The study does not provide evidence of clinical benefit or harm.

For clinicians, this is a methodological concept that may inform future trial designs. It should not be interpreted as a proven strategy for patient care. Further research with real-world data is needed to assess its practical utility.

Study Details

Study typeRct

EvidenceLevel 2

PublishedApr 2026

View Original Abstract ↓

We present a novel sequential multiple assignment randomized trial (SMART) design that integrates response-adaptive randomization with tailoring functions (RA-TF-SMART). We develop percentile-based and Z-score RA-TFs that incorporate both within-patient and between-patient adaptation to map continuous outcomes to randomization probabilities. We apply Q-learning, tree-based reinforcement learning, and G-estimation to estimate dynamic treatment regimens (DTRs). We compare our RA-TF-SMART designs to balanced randomized SMARTs (BR-SMARTs), tailoring function SMARTs (TF-SMARTs), and generalized outcome-adaptive SMARTs (GO-SMARTs). This study addresses limitations in SMART methodology by presenting designs where randomization probability does not require dichotomization of continuous outcomes and utilizes both individual patient outcomes and accumulated treatment efficacy data from prior participants. RA-TF-SMARTs offer a flexible framework that maximizes benefit for trial participants while maintaining statistical validity for post-trial DTR estimation.