Incidental dural tears under continuous irrigation in endoscopic lumbar spine surgery: A systematic review and meta-analysis

To estimate the incidence of recognised incidental durotomy in irrigated uniportal and biportal endoscopic lumbar surgery, and to describe conversion, delayed cerebrospinal fluid sequelae and repair.

The protocol was registered (PROSPERO CRD420261507619) and the review follows PRISMA 2020. MEDLINE, CENTRAL and Europe PMC were searched directly to 16 September 2026. Embase, Scopus and CINAHL were not directly accessible and were approached only indirectly, a limitation of the search. Consecutive series reporting recognised intraoperative tears with a denominator were eligible. Proportions were pooled with a binomial generalised linear mixed model (GLMM), with the prespecified Freeman-Tukey model as a sensitivity analysis.

Nineteen series (6 823 operations, 177 recognised tears) were included. The pooled rate was 2.74 % (95 % CI 1.72 to 4.16; Freeman-Tukey 3.00 %), with I² 86 % and a prediction interval of 0.44 % to 15.3 %. Disc herniation strata pooled at 1.07 % and stenosis strata at 5.47 % (odds ratio 4.53, p = 0.003). After adjustment for indication, biportal and uniportal series did not differ significantly (odds ratio 1.58, 95 % CI 0.63 to 3.94; very low certainty). Conversion occurred in 6 of 113 documented tears (5.3 %) and delayed fistula or pseudomeningocele in 3 of 80 (3.8 %).

Rates vary widely between series, and no single pooled figure describes the risk of an individual operation. About one tear per hundred discectomies and one per twenty stenosis decompressions is a reasonable basis for consent, adjusted for revision and experience. A difference between techniques is neither shown nor excluded. Occult tears after transforaminal discectomy should be quoted separately.