Claims of Safety in Neurosurgical Literature: We Shouldn’t Call Something Safe Without Strong Justification

Abstract

BACKGROUND: Many articles published in the neurosurgical literature make claims of safety regarding interventions. The strength of evidence supporting these claims has not been systematically evaluated. OBJECTIVE: To evaluate the quality of evidence supporting claims of safety in the neurosurgery literature. METHODS: The 14 neurosurgical journals listed in the Web of Science database with the highest impact factors for 2018 were included in the study. The titles and abstracts of all the articles published in these journals in 2018 were searched for the word "safety" and reviewed by 2 independent observers to identify those making a claim of safety and assess their evidence. RESULTS: The search identified 323 articles, representing 5% of the articles published in the 14 searched journals in 2018. Of these, 114 (1.8% of all articles and 35% of the identified articles) were judged to make a claim of safety of a neurosurgical intervention. A total of 98 (86%) of the 114 articles making a safety claim were retrospective cohort studies. Patient cohort sizes ranged from 1 to 1565 (median of 34), leading to wide variations in the CIs for estimated complication rates. No article made a statement of the assumptions underlying its claim of safety. CONCLUSION: Almost no articles that make claims of safety for neurosurgical interventions explicitly state the (1) patient population, (2) comparison intervention, nor (3) specific outcomes. To assure that claims of safety are supported by high-quality evidence, authors should address these elements early and systematically when designing studies.