Neurosurgical Evidence and Randomized Trials: The Fragility Index

Abstract

Background: Neurosurgical randomized controlled trials (RCTs) are difficult to carry out due to the low incidence of certain diseases, heterogeneous disease phenotypes, and ethical issues. This results in a weak evidence base in terms of both the number of trials and their robustness. The fragility index (FI) measures the robustness of an RCT and is the minimum number of patients in a trial whose status would have to change from a nonevent to an event to change a statistically significant result to a nonsignificant result. The smaller the FI, the more fragile the trial's outcome.

Methods: RCTs that have influenced neurosurgical practice were included in this analysis. Simulations were run to calculate the FI. To determine associations with a high or low FI, multivariable logistic regression was used to calculate adjusted odds ratios and 95% confidence intervals adjusting for baseline confounders.

Results: Of 2975 papers screened, 74 were included. The median FI was 4.5 (interquartile range: 1.5-10). RCTs included a median of 165 patients (interquartile range: 75-330), with a maximum of 10,008. A total of 38 trials had lost to follow-up patients that might have impacted the robustness of the results (51%).

Conclusion: Results of neurosurgical RCTs on which we base our clinical decision-making and treatment guidelines are often fragile. Improved methodologies, international collaboration, and cooperation between specialties might improve the evidence base in the future.

Keywords: Effect size measures; Evidence; Fragility index; Neurosurgical trials; Randomized clinical trials; Study design.