Related Experiment Videos
Interpretation and power of a pooled index
C H Goldsmith1, H A Smythe, A Helewa
1Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada.
The Journal of Rheumatology
|March 1, 1993
Summary
The pooled index, combining multiple measures, offers a more powerful way to detect clinical effects in rheumatoid arthritis research than single measures alone. This statistical technique enhances outcome detection in clinical trials.
Area of Science:
- Rheumatology
- Clinical Trials
- Biostatistics
Background:
- Individual clinical outcome measures in rheumatoid arthritis (RA) studies often lack the statistical power to detect significant effects.
- There is a need for more sensitive and robust methods to assess treatment efficacy in RA.
Purpose of the Study:
- To describe and illustrate the concept of a pooled index for clinical outcome assessment.
- To provide statistical rationale and clinical examples of the pooled index technique in RA research.
Main Methods:
- Utilized data from randomized control trials of patients with rheumatoid arthritis.
- Compared interventions including inpatient/outpatient care, occupational therapy, and physiotherapy.
- Analyzed variables such as joint count, stiffness, grip strength, erythrocyte sedimentation rate, and a functional change score to compute a pooled index.
Main Results:
- The pooled index successfully detected clinically important effects in all studies.
- Individual outcome measures frequently failed to achieve statistical significance, highlighting the pooled index's superior sensitivity.
Conclusions:
- The pooled index serves as a more powerful clinical outcome measure than individual metrics.
- Its effectiveness is maximized when derived from at least two relevant, low-correlation clinical outcome measures.