The split-plot design was useful for evaluating complex, multilevel interventions, but there is need for improvement...

Area of Science:

• Health Services Research
• Clinical Trials Methodology
• Biostatistics

Background:

• Split-plot (S-P) designs involve two randomization levels, posing unique challenges in healthcare trials.
• Effective implementation requires careful consideration of sample size calculation and participant flow.

Purpose of the Study:

• To describe sample size calculation, analysis, and reporting practices in healthcare split-plot randomized controlled trials.
• To identify areas for improvement in the methodology and reporting of S-P trials.

Main Methods:

• Comprehensive search of the EMBASE database (1946-2016) for healthcare trials using S-P designs.
• Screening, data extraction, and assessment of 18 S-P studies based on CONSORT reporting standards.

Main Results:

• Nine different designations were used for S-P trials, with randomization units unclear in nine abstracts.
• Ten studies reported sample size calculations accounting for clustering; however, participant flow diagrams were incomplete in 14 articles.
• Lack of explicit rationale for design choice was noted in several studies.

Conclusions:

• Split-plot designs offer utility but require clear reporting of rationale, sample size, and participant flow.
• A suggested CONSORT-style participant flow diagram is provided to enhance reporting.
• Further research is needed on sample size calculation methods for S-P trials.