Characterizing the Risk Factors for Incidental Durotomy During Spine Surgery: A Matched Cohort Analysis of a Large, Multicenter Dataset

Introduction: Incidental durotomy occurs in up to 17% of spine surgeries and can significantly impact postoperative morbidity, patient-reported outcomes, and resource utilization. Identifying patients at high risk for durotomy is of interest to patients, payors, and physicians. Here we present a matched cohort analyses to identify critical patient- and procedure-specific risk factors associated with incidental dural tears during spine surgery.
We aimed to identify predictors of incidental durotomies during spine surgery using a matched cohort analysis from a multicenter academic institution.

Methods: All patient undergoing neurosurgical spine surgey between January 2018 and July 2023 at a large, multicenter academic institution were identified. Data on baseline demographics and procedural details were gathered. An optimal matching protocol was performed at a 4:1 ratio on 44 baseline variables to create two homogenous cohorts of patients that either did or did not experience intraoperative durotomy. Univariate and multivariate regression analyses were performed to identify independent risk factors of durotomy.

Results: Optimal matching yielded 3,535 patients, of whom 707 experienced a durotomy. On adjusted multivariate logistic regression analysis, patient characteristics independently predictive of durotomy were underweight BMI (OR: 2.58 per kg/m², p< 0.01), private [versus public] insurance (OR: 0.78, p< 0.01), hypertension (OR: 1.22, p=0.03), history of paralysis (OR: 2.95, p< 0.01), and total Elixhauser index >20 (OR: 2.14, p< 0.01). Procedure-based factors predictive of durotomy included lumbosacral (OR: 2.32, p< 0.01) or thoracic (OR: 1.64, p< 0.01) versus cervical procedural level, deformity surgery [versus non-deformity cases] (OR: 1.33, p=0.04), fusion [versus decompression alone] (OR: 0.61, p=0.01), revision [versus index] surgery (OR: 1.78, p< 0.01), employing a posterior [versus anterior] approach (OR: 1.41, p< 0.01), and placement of >8 implants (OR: 1.77, p< 0.01).

Conclusion : Unintentional durotomies appear more common in patients with underweight BMI and higher comorbidity burden. Additionally, revision surgery, placement of implants, and increased surgical complexity appear to be associated with increased risk. These results can assist surgeons in preoperatively counseling patients about their risk of intraoperative durotomy based upon the planned procedure and baseline patient demographics.