Heterogeneity in Outcomes Reporting in Endoscopic Endonasal Skull Base Reconstruction: A Systematic Review

Background: There are currently no reporting standards for intraoperative or perioperative outcomes following endoscopic endonasal skull base surgery. This is problematic as it makes performing systematic reviews and meta-analysis impossible, impairing our ability to practice evidence-based medicine.

Objective: The goal of this study was to describe patterns of reporting outcomes in endoscopic endonasal skull base reconstruction.

Study Design: This study is a systematic review.

Methods: PubMed, EMBASE, CINAHL, Cochrane Review Databases, and Web of Science (2005 to June 21, 2019) were systematically searched for all publications on skull base reconstruction. Articles were limited to those with ≥25 patients and a focus on skull base reconstruction via an endoscopic endonasal approach. Articles with >20% of cases on primary CSF leaks (spontaneous, traumatic, or congenital) were excluded. Initial screening was based on review of title and abstract and those studies not meeting the above criteria were excluded. Full-text review was then performed to select the final studies. Subsequently, reported outcomes data was extracted from all included studies.

Results: A total of 111 studies were included in the review, including 20,548 patients undergoing skull base surgery. There was heterogeneity for pathologies treated, with 107 (96.4%) studies reporting the indication for reconstruction, among which sellar pathology was the most common. All studies reported some demographic data, but 17 (15.3%) did not report the gender distribution of patients and 16 (14.4%) did not report the age of patients. A total of 105 studies (94.6%) presented grouped or both grouped and individual patient data, while 6 (5.4%) studies presented solely individual patient data. Also, 48 studies (43.2%) reported patients’ history of prior surgery or radiotherapy and fewer studies (27, 24.3%) reported mean tumor size. Reporting of cerebrospinal fluid (CSF) findings had greater heterogeneity. Only 77 (69.4%) studies reported the presence or absence of intraoperative CSF leaks, with 42 of these studies (54.5%) providing a scaling system for these leaks. An even smaller number of studies (26, 23.4%) reported lumbar drain status. With regards to postoperative outcomes, few studies reported mean or median length of hospital stay (19, 17.1%) or of follow-up (51, 45.9%). The most commonly reported outcome was rate of postoperative CSF leak (108, 97.3%). Reporting of other surgical complications was less consistent, including meningitis (53, 47.7%), epistaxis (23, 20.7%), pneumocephalus (20, 18.0%), septal perforation (15, 13.5%), and flap necrosis (13, 11.7%). The rates of other general complications were rarely reported within the studies, including stroke (8, 7.2%), pulmonary embolism or deep vein thrombosis (14, 12.6%), and mortality (21, 18.9%).

Conclusion: There is significant heterogeneity in the outcomes reported in studies relating to endoscopic endonasal skull base surgery reconstruction. This highlights the need for the development of standard reporting outcomes to minimize bias and improve the execution of systematic reviews on the subject.