How to Evaluate a Clinical Study
Red Flags (Signs to be Skeptical)
No study pre-registration in a clinical study database
Study is company-run without independent oversight
Unclear methodology that leaves questions about how a study was conducted
No control or placebo group
Vague descriptions of the treatment
Emphasis on statistical trends instead of clear between-group differences
No conference presentation of results, peer-reviewed publication, or clear plan to share results
Methods — Study Registration
Is the study pre-registered in a database of clinical research studies (e.g., ClinicalTrials.gov with an NCT number)?
Does the published study appear to follow the original plan?
Methods — Who Participated
Was there clear inclusion criteria (who was allowed in)?
Was there clear exclusion criteria (who was left out)?
Can you tell who the results apply to?
Methods — Study Design
Are there two or more groups (e.g., treatment and control/placebo)?
Were participants randomly assigned to groups?
Were participants blinded (did they not know which treatment they received)?
Were researchers blinded, if possible?
Methods — Treatment Details
Is the treatment clearly described (what it was, how much, how often, how long)?
Is the control or placebo clearly described?
Did all groups receive similar care except for the treatment being tested?
Methods — Study Size
Do the authors report a power analysis?
Does the study include enough participants to detect meaningful differences?
Results — Between-Group Comparisons
Did the authors compare outcomes between the treatment and control groups?
Are improvements greater in the treatment group than in the control group?
Results — Statistical Significance
Are results statistically significant (e.g., p < 0.05 or p < 0.01)?
Are confidence intervals reported?
Results — What Not to Over-Interpret
Are results more than just baseline or within-group changes?
Do the authors avoid overstating results based on placebo effects?
