Statistical Considerations in Pilot Studies

Statistical Considerations in Pilot Studies

Matthew J. Gurka, PhD Professor, Dept. of Health Outcomes & Policy Associate Director, Institute for Child Health

Policy

Goals for Today

1. Statistical analysis plan for pilot studies

2. The basics of a power analysis (for planning future studies; why pilot studies are so important)

3. Pilot studies ? sample size considerations for feasibility

Pilot Study: DON'T DO THIS

? Analysis plan: Statistical procedures as appropriate

? Sample size: No sample size calculations are provided due to the pilot nature of this study.

Pilot Studies: Is it Feasible?

? Definition: "Preparatory studies designed to test the performance characteristics and capabilities of study designs, measures, procedures, recruitment criteria, and operational strategies that are under consideration for use in a subsequent, often larger, study."

(Very Good) Reference: Moore, Carter, Nietert, Stewart. Recommendations for planning pilot studies in clinical and translational sciences. Clinical and Translational Sciences 2011;4:332-337

Staging Research

Pilot study

Move forward

???

?????

Choose a New Research Topic

Abandon the Research Topic

Test the Research Topic in a Pilot

Study Framework

Find Nonsignificant (Negative) Results

(Figure) Moore, Carter, Nietert, Stewart. Recommendations for planning pilot studies in clinical and translational sciences. Clinical and Translational Sciences 2011;4:332-337

Pilot Studies

? Overarching Goal: Keep the Next Study in Mind

? Pilot study aims and methods should align with the goals of the subsequent study

? Aims of a pilot can range from evaluating feasibility of the protocol to investigating potential mechanisms of efficacy for a new intervention

6

Pilot Studies

? Pilot studies contribute to the development and design of future (larger) studies by:

? Refining the research hypotheses ? Identifying barriers to successful study

completion ? Evaluating acceptability of methods and

instruments to participants ? Estimating the time required for study

participation ? Providing estimates of missing data and dropout

7

Pilot Studies

? Pilot studies contribute to the development and design of future (larger) studies by:

? Estimating rates and variability in outcomes ? Testing mechanistic efficacy/ `proof of

concept'

8

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download