A review of Neurosurgical RCTs in the Cochrane Database of Systematic Reviews: Key findings and implications for future research

World Neurosurg. 2024 Nov 20:S1878-8750(24)01908-9. doi: 10.1016/j.wneu.2024.11.054. Online ahead of print.

Abstract

Objectives: Evidence available to clinicians and patients to inform treatment decisions is ideally produced by randomised controlled trials (RCT). The objective of this study was to assess the extent to which Neurosurgical practice is supported by RCT-level evidence.

Methods: A search of the Cochrane Library was conducted to find reviews of the effectiveness of Neurosurgical operative interventions. Data were extracted on the intervention, patient population and outcome measures as well as the strength of evidence, as rated by the Cochrane authors. The extracted data were analysed to identify the gaps and areas of (in)consistency across the RCTs included within the Cochrane Reviews.

Results: A total of 52 Cochrane reviews met the inclusion criteria, which covered 8 Neurosurgical subspecialties. However, only 28 were published after 2015. There was limited coverage of multiple commonly performed Neurosurgical interventions and 9 reviews found no RCTs related to their selected topic. In 28 reviews, results were synthesized from 5 or fewer trials. Primary outcomes also varied among trials examining similar interventions. The overall quality rating of the evidence for the different subspecialties varied, with the majority of reviews rating the evidence as being from very low to low.

Conclusion: The RCT level evidence supporting Neurosurgical practice is varied and the outcomes tested remain predominantly heterogenous. There remain important Neurosurgical conditions where treatment strategies are not underpinned by high quality evidence. Pragmatic RCTs, well-designed observational studies as well as robust audit and registry processes may provide the real-world evidence for treatment decisions in Neurosurgical care.

Keywords: Evidence-based Medicine; Neurosurgery; Outcomes.