INTRODUCTION
Unilateral biportal endoscopic (UBE) spine surgery has emerged as a promising minimally invasive alternative for lumbar decompression, offering superior visualization, reduced muscle trauma, and faster postoperative recovery compared to conventional open or tubular techniques [1]. With the increasing adoption of UBE worldwide, attention has shifted toward understanding its unique complication profile. Among these, postoperative spinal epidural hematoma (PSEH) represents a potentially devastating but under-recognized event that may cause acute neurological deterioration if not promptly diagnosed and managed [2].
PSEH can be defined as accumulation of coagulated blood within the operated spinal canal following surgery [3]. The incidence of PSEH after biportal endoscopic spine surgery has been reported to be as high as 23.6%, whereas the rate of symptomatic PSEH is approximately 1.1% [3,4]. UBE itself has been suggested as a risk factor for PSEH [5], others include continuous saline irrigation masking epidural venous bleeding, the confined working corridor and absence of compressive wound closure limiting natural tamponade, and the reduced effectiveness of hemostatic agents such as bone wax in a persistently wet environment.
However, no study to date has specifically examined the relationship between postoperative drain output and PSEH in the setting of UBE spine surgery. In this study, we aim to determine the incidence of PSEH following single-level UBE decompression with or without discectomy, identify perioperative factors associated with its development, and evaluate the association between postoperative drain output, hematoma grade, and corresponding magnetic resonance imaging (MRI) findings. By clarifying these relationships, we seek to provide simple, clinically applicable indicators that may facilitate earlier detection and safer postoperative management in patients undergoing UBE spine surgery.
MATERIALS AND METHODS
1. Study Design and Ethics
This retrospective cohort study included patients who underwent single-level UBE decompression, with or without discectomy, between January 2022 and June 2024. The study protocol was reviewed and granted an exemption from IRB review by the Public Institutional Bioethics Committee (IRB Exemption No. P01-202511-01-003; November 4, 2025), and all procedures were performed in accordance with the Declaration of Helsinki. The requirement for informed consent for study participation was waived due to the retrospective design and use of deidentified data. However, written informed consent for surgical treatment and permission to use anonymized clinical data were obtained from all patients prior to surgery. All operations were performed by surgeons experienced in UBE procedures using a standardized operative technique and irrigation protocol.
2. Patient Selection
3. Surgical Technique
All procedures were performed under spinal anesthesia and standard UBE approach. Antithrombotic agents were withheld preoperatively according to institutional protocol (typically 5 days, depending on agent and indication). A 3-L saline irrigation bag suspended at 1.7 m for gravity-assisted flow was used for cases. Target blood pressure was 100/60 mmHg. Hemostasis was achieved with bipolar radiofrequency ablation and hemostatic sealants like CollaStat (Dalim Tissen Co., Ltd., Korea) and TachoSil (Takeda Pharma A/S, Denmark) as required. Systolic blood pressure was maintained around 90–100 mmHg during final hemostasis. Floseal hemostatic matrix (Baxter, USA) was routinely applied at the end of surgery in a standardized manner equally for all patients. Tranexamic acid was not used in any case. At the end of the procedure, a closed-suction Jackson-Pratt drain (Barovac 1.6 mm × 3.2 mm) was inserted and connected to a low-pressure vacuum system. Patients typically began mobilization 4–6 hours postoperatively. Drain output was recorded daily, and the drain was removed when the 24-hour output fell below 50 mL.
4. Data Collection
Demographic variables (age, sex, body mass index [BMI]), comorbidities (hypertension, anticoagulant use), and preoperative coagulation parameters (prothrombin time [PT], international normalized ratio [INR]) were recorded. Operative duration, level of surgery, and total drain output were also documented.
5. Postoperative Imaging and Hematoma Grading
All patients underwent routine postoperative MRI within 12 hours of surgery to evaluate decompression adequacy and detect epidural hematoma. Early imaging was chosen to identify both asymptomatic and potential early compressive collections before clinical deterioration.
Hematomas were graded on T2-weighted axial images [3] according to the degree of canal compression using PACS (picture archiving and communication system) measurement as follows:
• Grade 0: Hematoma with no canal compression
• Grade I: <25% canal compromise
• Grade II: 25%–50% canal compromise
• Grade III: 50%–75% canal compromise
• Grade IV: >75% canal compromise
The critical ratio was defined on MRI as the postoperative spinal canal cross-sectional area divided by the preoperative spinal canal cross-sectional area. Symptomatic PSEH was defined as any hematoma causing new or progressive neurological deficit requiring surgical evacuation.
6. Statistical Analysis
Descriptive statistics summarized demographic, perioperative, laboratory, drain, and MRI variables, reporting continuous data as mean±standard deviation (or median [interquartile range] if nonnormal) and categorical data as counts (%). Comparisons between patients with and without PSEH used chi-square (or Fisher exact) tests for categorical variables and independent-samples t-tests (or Mann-Whitney U-for nonparametric data) for continuous variables. Across MRI-based hematoma grades, categorical variables were analyzed with chi-square tests and continuous variables with the Kruskal-Wallis test. All analyses were 2-sided (p<0.05) and performed using IBM SPSS Statistics ver. 25.0 (IBM Corp., USA).
RESULTS
Among the 238 patients included in the study, 115 (48.3%) were female and 123 (51.7%) were male. Most patients (n=194, 81.5%) underwent UBE decompression alone, whereas 44 (18.5%) underwent combined UBE decompression and discectomy. The operated levels were predominantly midlumbar: L3–4 in 88 patients (37.0%) and L4–5 in 89 patients (37.4%), with fewer cases at L1–2 (8 patients, 3.4%), L2–3 (36 patients, 15.1%), and L5–S1 (17 patients, 7.1%).
Postoperative drains were typically maintained for 2–3 days, with majority (62.6%) having the drain in situ for 2 days. Hypertension was present in 136 patients (57.1%), and 63 patients (26.5%) were on anticoagulant therapy. The mean age of the cohort was 68.79±10.81 years (range, 27–90 years), and the mean BMI was 24.44±3.09 kg/m2 (range, 16.0–33.4 kg/m2). The average duration of surgery was 63.59±20.20 minutes, with a range of 21–130 minutes. The mean PT was 11.70±1.24 seconds (range, 9.30–14.70 seconds), and the mean INR was 0.97±0.07 (range, 0.82–1.23).
Mean drain output on postoperative day (POD)-1 was 91.27±36.95 mL, decreasing to 40.37±20.38 mL on POD-2 (n=218), 23.62±11.75 mL on POD-3 (n=69), and 14.00±4.00 mL on POD-4 (n=3). The total cumulative drainage output over the duration of drain placement was 135.27±55.05 mL (range, 0–443 mL). The mean preoperative MRI canal area was 94.87±30.86 mm2 (range, 39.45–206.63 mm2), which increased postoperatively to 195.29±47.08 mm2 (range, 67.61–313.14 mm2).
Postoperative epidural hematoma was observed in 126 of 238 patients (52.9%), while 112 patients (47.1%) had no hematoma. Table 1 summarizes distribution of hematoma groups based on MRI findings. On univariate analysis, postoperative epidural hematoma was more frequent in males (73 of 123 [59.3%] vs. 53 of 115 [46.1%]; p=0.041) and was strongly associated with surgical level (p=0.001), highest at upper lumbar levels L1–3 (up to 87.5%) compared with L4–S1 (29.4%–40.4%). Age>60 years showed a nonsignificant trend toward higher rates (p=0.077). Procedure type, drain duration, hypertension, and anticoagulant use were not associated with hematoma occurrence. Table 2 summarizes the association of perioperative factors and PSEH.
Among the 126 patients with MRI-graded postoperative hematoma, surgical level was strongly associated with hematoma severity (χ²=32.450, p=0.001): upper lumbar levels (L1–3) more often had moderate (25%–50%) hematomas, whereas L4–5 and L5–S1 predominantly showed mild (<25%) collections as shown in Table 3. Age, drain duration, hypertension, and anticoagulant use were not significantly associated with hematoma grade (all p>0.05), though hypertension showed a trend toward higher grades (χ²=6.889, p=0.076).
Across MRI-based hematoma grades, operative duration did not differ significantly (Kruskal-Wallis p=0.398), although grade 3 group had a numerically longer mean time (79.0±26.38 minutes). Drain output decreased significantly with increasing hematoma severity on POD-1 and POD-2 (p=0.024 and p=0.028) and for total drain output (p=0.020), with the 50%–75% group showing the lowest medians (POD-1: 45 mL; POD-2: 15.5 mL; total: 65.5 mL). Postoperative MRI canal measurements and critical ratio declined stepwise with higher hematoma grades (p=0.003 and p=0.034) as shown in Table 4. Two patients developed symptomatic PSEH who underwent revision surgery. Their clinical characteristics are summarized in Table 5.
1. Illustrative Cases
1) Example 1
A 79-year-old man with hypertension who was receiving anticoagulant therapy presented with progressive bilateral buttock and posterior thigh pain, which worsened with walking and lumbar extension. On examination, the straight-leg raise test was negative. He had weakness of the extensor hallucis longus (EHL) (power 4/5). Lumbar MRI demonstrated severe L4–5 canal stenosis (Figure 1A).
After temporary discontinuation of anticoagulant therapy, he underwent L4–5 UBE decompression. By POD-2, his buttock pain worsened (visual analogue scale [VAS] 6/10) and he developed progressive weakness of the left EHL (power 3/5). MRI done on POD-2 demonstrated a postoperative epidural hematoma (Figure 1B). By POD-6, he reported severe bilateral buttock and posterior thigh pain (VAS 10/10) with markedly limited ambulation. Repeat MRI confirmed a persistent postoperative epidural hematoma at L4–5 (Figure 1C).
On POD-7, he underwent UBE re-exploration with L4–5 decompression and hematoma evacuation. His pain gradually improved. At discharge, he was ambulatory and neurologically intact with MRI showing good decompression (Figure 1D). At 3-month follow-up, he reported complete resolution of buttock and posterior thigh pain (VAS 1/10) and was able to walk independently without claudication. Motor power, including the left EHL, was normal (5/5), with no sensory deficits. He had returned to baseline activities without recurrence of symptoms.
2) Example 2
A 66-year-old female with diabetes mellitus and hypertension presented with severe right L4–5 radiculopathy characterized by intense buttock pain (VAS 7/10) and new-onset motor weakness. She reported persistent symptoms despite prior conservative treatment. On neurological examination, right EHL strength was 3/5. Lumbar MRI demonstrated a right L4–5 herniated nucleus pulposus (Figure 2A). Given progressive weakness and intractable pain, she underwent right-sided L4–5 UBE decompression with discectomy. Her radicular pain initially improved after surgery (VAS 2/10).
On POD-6, she developed recurrent and worsening right buttock and leg pain (VAS 10/10). Repeat MRI showed a postoperative epidural hematoma at L4–5 (Figure 2B), and she underwent urgent surgical evacuation. Following hematoma removal, her symptoms improved by approximately 50%–60%. She had residual L5-distribution paresthesia and persistent EHL weakness, but subsequently demonstrated steady recovery after the second surgery with gradual improvement in motor strength (Figure 2C).
At 3 months postoperatively, she reported marked improvement in right buttock and leg pain (VAS 1/10), with only mild intermittent L5-distribution paresthesia. Motor function improved progressively, with ankle dorsiflexion and EHL recovering to near-normal strength. She was ambulatory without support, had resumed routine daily activities, and had no further symptom deterioration.
DISCUSSION
UBE offers enhanced visualization, less soft-tissue damage, and faster recovery for lumbar decompression [6], but its growing use has highlighted UBE-specific complications [7]. Among these, PSEH remains a potentially devastating but underreported event [6,8]. A PSEH is defined as a contained intraspinal and/or paraspinal hemorrhagic collection found anywhere along the course of the operative tract [9]. The key clinical challenge is identifying which hematomas require urgent surgical evacuation versus those that can be safely observed [10].
In our study, postoperative PSEH was identified in 126 of 238 patients (52.9%), a rate that is higher than previously reported figures, including 24.7% after UBE surgery [3] and 14.6% after microscopic lumbar surgery [11]. This comparatively higher incidence after UBE is likely multifactorial. Proposed explanations include irrigation-related masking of ongoing venous oozing during endoscopy [5], the inherent difficulty of achieving consistent hemostasis from cancellous bone surfaces through a minimally invasive corridor, and challenges in obtaining a tight, watertight fascial closure following multiportal access, all of which may facilitate postoperative epidural blood accumulation [5].
Surgical drains have been reported to reduce the risk of postoperative epidural hematoma [12,13]. Our study is one of the first to systematically link drain output patterns and post operative MRI grading of hematoma related to UBE surgery. We observed a paradoxical inverse relationship between drain output and hematoma grade: higher-grade collections on MRI were associated with lower POD-1, POD-2, and total drain outputs. This supports the concept that suboptimal drainage either due to drain blockage, loss of negative pressure, malposition, or space tamponade, may allow progressive epidural accumulation despite drain placement [14]. Monitoring early drain dynamics could therefore serve as a practical surrogate marker for impending PSEH.
Despite the high MRI-detected incidence of postoperative epidural hematoma (52.9%), only 2 patients developed symptomatic PSEH. The clinical impact of an epidural collection depends more on the degree to which it compromises postoperative canal expansion. The critical ratio defined as the ratio of postoperative-to-preoperative cross-sectional area within the lumbar canal—has been proposed as a direct radiological surrogate of thecal sac compression [15]. In our cohort, the critical ratio decreased stepwise with increasing MRI hematoma grade, indicating progressively poorer radiological decompression in the presence of larger postoperative collections.
Biportal endoscopy has been proposed as a potential independent risk factor for PSEH [5]. Consistent with our findings, Bekki et al. [16] reported that upper lumbar decompression—particularly in severely stenotic, tight canals—may be especially susceptible to postoperative epidural hematoma formation. In our cohort, male sex was associated with MRI-detected PSEH occurrence, aligning with the male predominance reported by Domenicucci et al. [17]. Although operative duration did not differ significantly across grades, grade 3 showed a numerically longer mean operative time, possibly reflecting more extensive decompression, greater soft-tissue manipulation, and higher bleeding/hemostatic demand. Preoperative PT also tended to be higher with increasing hematoma severity, suggesting that baseline coagulation abnormalities may predispose to larger postoperative epidural collections.
In our cohort, age, hypertension, anticoagulant use, drain duration, and procedure type were not significantly associated with PSEH. This contrasts with broader spine-surgery literature linking postoperative SEH to older age, obesity, comorbidities, coagulation abnormalities, and more extensive procedures (e.g., multilevel or revision surgery) [18]. This discrepancy may reflect our relatively homogeneous population, dominated by single-level UBE performed with standardized hemostasis and drain protocols.
At our institute, several intraoperative strategies are routinely employed to minimize the risk of PSEH. These include meticulous hemostasis using radiofrequency ablation for coagulating bleeders at the end of the procedure, strict control of systolic blood pressure to around 90 mmHg, and the adjunctive use of topical agents such as epinephrine, bone wax for osseous bleeders, and hemostatic sealants. In addition, a Jackson-Pratt drain (Barovac evacuator closed wound drainage system, tube size 1.6×3.2) is placed at the conclusion of surgery (Figure 3).
This study has several notable strengths. It uniquely includes postoperative MRI for all 238 UBE cases. It also represents one of the largest single-center cohorts with routine MRI after UBE surgery. To our knowledge, no other prior studies have highlighted postoperative drain output as a practical indicator for early hematoma detection. Finally, the introduction of the “critical ratio” in context of UBE adds a novel and impactful parameter that enhances early recognition of hematoma risk. This study is limited by its retrospective design, single-center setting, and single-observer MRI grading, which may introduce selection and measurement bias. Antithrombotic agents were not classified by type (antiplatelet vs. anticoagulant, single vs. dual therapy), precluding medication specific risk stratification. Additionally, hematoma evolution beyond the first POD was not systematically assessed in asymptomatic patients. Future prospective, multicenter studies with standardized imaging intervals and inclusion of clinical outcome scores are needed to validate these results.
CONCLUSION
MRI-detected postoperative epidural hematoma is common after UBE, whereas clinically significant symptomatic hematoma requiring evacuation is uncommon. Hematoma occurrence was associated with sex and surgical level, and hematoma severity varied by surgical level. Lower early drain output—particularly reduced POD-1 output—was associated with higher MRI compression grades and should prompt closer clinical observation and consideration of early MRI, but it should not be used as a stand-alone threshold. The MRI critical ratio provides an objective metric to interpret the clinical relevance of postoperative collections.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
