Drain versus no drain after intradural spine surgery: a 10-year multicenter study.

objectivesA postoperative drain is commonly used after intradural spine surgery to prevent hematoma formation, but it may lead to other complications. We aimed to investigate the efficacy and safety of postoperative drain use after intradural spine surgery.

methodsPatients who underwent intradural spine surgery at three tertiary neurosurgical centers from January 2015 to December 2024 were retrospectively analyzed. Data collected included preoperative (demographics, comorbidities, symptoms), intraoperative (pathology, details of surgery, use of drain, duration and type of drain [active or passive], duration of surgery) and postoperative variables. Primary outcome measures included cerebrospinal fluid (CSF) leak requiring repair, postoperative hematoma, and infection. Secondary outcome measures included length of stay and 30-day mortality.

results229 patients comprising 93 males (40.6%) and 136 females (59.4%) were included. 79 patients (34.5%) had a postoperative drain (Drain group) while 150 patients (65.5%) did not (No Drain group). Median age was 61 years (interquartile range [IQR] 50-69). The Drain group had more 3- and 4-level surgeries and longer operative times at baseline, but these factors were not shown to be significantly associated with CSF leak requiring repair. The Drain group had a higher observed rate of postoperative CSF leak requiring repair than the No Drain group (4, 5.1% versus 1, 0.7% respectively, OR = 7.95, 95% CI: 0.87-72.35, p = 0.049). Active drains had a higher, though not statistically significant, rate of CSF leak requiring repair compared to passive drains (3, 10.3% versus 1, 2%, OR = 5.65, 95% CI: 0.56-56.9, p = 0.137). No significant differences were found between the Drain and No Drain groups in the rates of postoperative hematoma (0 versus 1, 0.7% respectively, p = 1.000), infection (2, 2.5% versus 4, 2.7% respectively, p = 1), 30-day mortality (1, 1.3% versus 0 respectively, p = 0.345), and length of stay (median 9 days, IQR 6-18 versus median 9 days, IQR 5-15 respectively, p = 0.134). Median duration of follow-up was 17 months (IQR 6-42). On multivariable analysis, with the caveat of low event numbers and wide confidence intervals, younger age (OR = 0.87, 95% CI: 0.79-0.97 for every increase in age of 1 year, p = 0.008) and diabetes mellitus (OR = 181.73, 95% CI: 4.41-7496, p = 0.006) were found to be significant predictors of CSF leak requiring repair.

conclusionsThe Drain group had a higher observed rate of postoperative CSF leak requiring repair than the No Drain group, with similar hematoma and infection rates. However, drains were preferentially used in cases with more surgery levels and longer operative times. Therefore, a postoperative drain may not be universally necessary for all cases of intradural spine surgery, but should be individualized depending on surgical complexity.