Influence of unintended dural tears on postoperative outcomes in lumbar surgery patients: a multicenter observational study with propensity scoring

Background context: Unintended dural tears (DTs) are common in spinal surgeries. Some authors have reported that the outcomes in lumbar surgery patients with DTs are equivalent to those in patients without DTs, but this remains uncertain.

Purpose: To assess the effect of unintended DTs on postoperative patient-reported outcomes.

Study design/setting: A multicenter retrospective observational study.

Patient sample: We enrolled patients undergoing lumbar spine surgery at eight hospitals between April 2017 and November 2018.

Outcome measures: We collected data regarding patients' backgrounds, operative factors, occurrence of unplanned DTs during surgery, postoperative complications, patient-reported outcomes, such as pain or dysesthesia of the lower back, buttock, leg, or plantar area, EuroQol 5 Dimension (EQ-5D), Oswestry Disability Index (ODI) scores, and postoperative satisfaction.

Methods: We divided the patients into a DT- group (without DTs) and a DT+ group (with DTs). First, multivariate logistic regression analyses were conducted to reveal risk factors for occurrence of DTs. Then, we used propensity score matching to obtain a matched DT- group (mDT- group) and a matched DT+ (mDT+ group). Student's t test was used for comparing continuous variables and Pearson's chi-square test for comparing categorical variables between the two groups.

Results: We enrolled 2,146 patients in this study. The number of patients with unintended DTs was 166 (7.7%). Older age, body mass index, ossification of posterior longitudinal ligament / yellow ligament, spinal deformity, and revision surgery were significant risk factors for DTs. We used propensity score matching to compare 163 of the patients with DTs with 163 patients without DTs. No significant difference was found in postoperative pain or dysesthesia of the lower back, buttock, leg, and plantar area between the mDT- and mDT+ groups. When comparing preoperative with postoperative pain and dysesthesia, a statistically significant improvement was found in each group (p<.01 for all variables) except for sensory disorder of the plantar area, where a significant improvement was only observed in dysesthesia of the mDT- group (p<.01). Although some improvements were observed, they were not statistically significant in terms of pain in the mDT- (p=.06) and mDT+ (p=.13) groups and dysesthesia in the mDT+ (p=.13) group. No significant differences were found in postoperative outcomes, such as EQ-5D (p=.44) and ODI (p=.89) scores, and postoperative satisfaction (p=.73) between the two groups.

Conclusions: Although insufficient improvement of sensory disorder of the plantar area was observed, patients with DTs showed almost equivalent postoperative outcomes compared with patients without DTs.