Evaluating pituitary adenomas using national research databases: systematic review of the quality of reporting based...

The use of national research databases has become more prevalent for studying various neurosurgical diseases. Despite the advantages of using large databases to glean clinical insight, variation remains in the methodology and reporting among studies. Using STROBE and RECORD guidelines, we evaluated the quality of reporting of the database literature investigating surgical management of benign pituitary adenomas. In this systematic review of the PubMed/MEDLINE database, we identified studies employing large national research databases of patients who underwent surgery for benign pituitary adenoma. We evaluated each of these studies using the STROBE-RECORD reporting guideline criteria to assess their quality. A total of 42 studies from 2003 to 2020 were identified for inclusion. The two raters demonstrated a κ = 0.228 with 84% overall agreement. Commonly underreported criteria included bias (discussed in 56% of studies), main result reporting (70%), subgroup analysis (69%), generalizability (68%), and funding (57%). These factors, in addition to the data sources/measurement criteria, also had the largest discrepancies between reviewers. About 20% of administrative database reviews did not accurately address bias or control for confounding variables. We found frequent underreporting of crucial information and criteria that can be challenging to identify may limit large database studies of pituitary adenomas. Improved reporting of certain criteria is critical to optimize reader understanding of large database studies. This would allow better dissemination and implementation of study findings, especially as the use of these research tools increases.