Contrasting clinical and statistical significance within the research setting

Area of Science:

• Clinical Trials
• Biostatistics
• Pulmonary Medicine

Background:

• Effective clinical trial design requires careful consideration of clinical significance, statistical significance, power, and sample size.
• Overemphasis on statistical significance can lead to overlooking crucial clinical implications.

Purpose of the Study:

• To evaluate the frequency with which key design parameters are addressed in published pulmonary function test research.
• To assess the reporting of clinical significance, statistical significance, sample size, and statistical power in scientific literature.

Main Methods:

• A systematic review of articles published in 1990 concerning pulmonary function tests (specifically forced vital capacity).
• Articles were identified using Minnesota MEDLINE and screened for the discussion of specific design parameters.
• A total of 121 articles met the inclusion criteria for review.

Main Results:

• Statistical significance (P values) was reported in 92.6% of the articles.
• Clinical significance was discussed in 13.2% of studies.
• Sample size was addressed in 21.5% of studies.
• Statistical power was reported in only 5.0% of studies.
• No articles discussed all four critical design parameters.

Conclusions:

• The reporting of clinical significance, sample size, and statistical power in clinical research, particularly in pulmonary function tests, is inadequate.
• Future research should prioritize the comprehensive reporting of all essential design parameters for robust study interpretation.
• Establishing clear methods for determining clinical significance, such as using effect sizes based on population distributions, is recommended.