Sample size and power considerations for cluster randomized trials with count outcomes subject to right truncation

Area of Science:

• Epidemiology and Public Health
• Biostatistics
• Clinical Trial Design

Background:

• Cluster randomized trials (CRTs) are essential for evaluating population-level interventions in public health.
• A key challenge in CRTs, particularly for vector-borne disease control (e.g., malaria), is analyzing right-truncated count outcomes.
• Existing sample size formulas for CRTs are inadequate for designs with truncated count data.

Purpose of the Study:

• To develop and validate statistical methods for analyzing CRTs with right-truncated count data.
• To derive closed-form sample size formulas specifically for CRTs with truncated count outcomes.
• To assess the impact of right truncation on statistical power in CRT design.

Main Methods:

• Utilized two marginal modeling approaches for analyzing CRTs with truncated counts.
• Developed two corresponding closed-form sample size formulas.
• Validated the proposed formulas through extensive simulations and applied them to a malaria control CRT example.

Main Results:

• The proposed marginal modeling approaches effectively handle right-truncated count data in CRTs.
• The derived sample size formulas provide a computationally efficient tool for trial design, avoiding complex simulations.
• The study quantifies the influence of right truncation on statistical power and demonstrates practical application in malaria control.

Conclusions:

• New sample size formulas are presented for CRTs with right-truncated count outcomes, facilitating robust trial planning.
• These methods address a critical gap in the statistical design of public health and epidemiological studies.
• The findings are directly applicable to optimizing the design and power calculations for vector-borne disease control interventions.