Research Methods: Trial Design, Statistics, Industry and NIH
Designing Rigorous, Investigator-Led Trials
Suresh Vedantham urged investigators to test falsifiable hypotheses rather than seek proof, guard against confirmation bias from narrow subspecialty perspectives, and collaborate across fields, as the ATTRACT trial did in reshaping aligned hematology and vascular surgery guidelines.
Sample Size and Statistics in RCTs
Sameer Parpia explained that sample size depends on alpha, power, outcome variance, and target treatment difference, illustrated by the C-TRACK trial’s calculation, and cautioned that a single p value of 0.05 offers supportive but not definitive evidence of efficacy.
Industry’s Role in Clinical Trials
Mark Garcia argued industry funding provides essential infrastructure, regulatory expertise, and cost coverage for device trials as government funding has declined, and found little evidence that industry sponsored trials are more likely to report favorable results.
Understanding the NIH Grant Review Process
Andrei Kanchelskis described how the Center for Scientific Review assigns applications to study sections that score significance, innovation, and approach on a one to nine scale, discuss top ranked applications, and issue a summary statement guiding funding decisions.
