Meta-analyses Are Only As Good As The Studies They Combine: A Lesson Learned From Meta-analyses On Ivermectin Against COVID-19 .

Meta-analyses can be powerful tools, but their conclusions only hold if the studies being combined are appropriate. Using ivermectin research as an example, two meta-analyses published around the same time reached opposite conclusions about whether the drug reduces mortality in COVID-19. One analysis included 15 trials and found a benefit, while the other included only 2 and found no clear effect. The difference came down to what each team allowed into the analysis—some studies used inactive comparators, mixed ivermectin with other unproven drugs, or included patients without confirmed infection. When those questionable trials are removed, the apparent benefit disappears, underscoring how meta-analysis results hinge entirely on what goes in.