Clinical researchers often encounter situations where one treatment arm involves clustering, such as group therapy, while the control arm does not. This scenario requires a specific partially nested trial design to ensure statistical validity. Consequently, determining the correct sample size becomes a challenge, especially when researchers lack prior knowledge of variance parameters. A new maximin approach now offers a robust solution for these complex research scenarios.

The Role of Maximin Approaches in Partially nested trial design

Standard trial designs usually require precise estimates of intraclass correlation before the study begins. However, these parameters are often unknown during the planning stage. The maximin approach addresses this uncertainty by ensuring a required power level across a range of plausible values. Furthermore, this method identifies the best treatment-to-control allocation ratio. It also optimizes the balance between the number of therapy groups and the number of participants per group. Specifically, it maximizes power for the worst-case parameter scenarios, protecting the study from being underpowered.

Practical Impact on Research Budgets

One of the primary advantages of this approach is its ability to minimize the total research budget. Moreover, empirical evidence shows that maximin designs significantly reduce the required subjects compared to traditional equal-allocation designs. Because of this efficiency, researchers can conduct high-quality trials even with limited funding. Additionally, a user-friendly R Shiny app now allows clinicians to perform these sophisticated calculations without needing advanced biostatistical expertise. Therefore, this tool is becoming essential for researchers aiming to produce reliable evidence in fields like psychiatry and community medicine.

Frequently Asked Questions

What is a partially nested randomized trial?

It is a clinical trial where participants in one arm are grouped in clusters, such as therapy sessions or specific clinics, while the other arm consists of independent individuals.

How does the maximin approach help with unknown data?

The maximin approach calculates the necessary sample size based on a range of possible values for unknown variables, ensuring the trial maintains power even in the most difficult statistical scenario.

Disclaimer: This content is for informational and educational purposes only. It does not constitute professional medical advice or a substitute for consultation with a certified biostatistician. Refer to the latest local and national guidelines for clinical practice.

References

Candel MJ et al. Efficient design of partially nested randomized trials: A maximin approach. Stat Methods Med Res. 2026 Mar 13. doi: 10.1177/09622802251409388. PMID: 41823059.

van Breukelen GJP, Candel MJJM. Efficient design of cluster randomized and multicentre trials with unknown intraclass correlation. Stat Methods Med Res. 2015;24(5):540-556.

Candlish J, et al. Appropriate statistical methods for analysing partially nested randomised controlled trials with continuous outcomes: a simulation study. BMC Med Res Methodol. 2018;18(1):105.