Researchers Assess the Scientific Integrity of Systematic Reviews

Systematic reviews, which aggregate data from multiple studies for authoritative conclusions, have gained popularity in research and policy, particularly regarding gender-affirming care. However, a recent study published in JAMA Network Open highlights concerns about the integrity of systematic reviews, revealing that 299 out of 200,000 reviews cited retracted articles from low-quality “paper mills.” Experts warn this poses a risk of integrating flawed studies into influential systematic reviews. Lisa Bero noted that not all studies termed “systematic reviews” follow critical scientific criteria. The static nature of these reviews often means they don’t account for newly retracted studies, which could lead to misleading conclusions. An alternative, “living evidence syntheses,” updates systematic reviews regularly but is time-intensive. To improve efficiency, researchers like Christian Cao are developing AI tools to automate the process, which have proven promising but raise concerns about the potential for generating subpar reviews. The dialogue on standards in systematic reviews is ongoing.

Source link