Complications of Unilateral Biportal Endoscopic Spinal Surgery for Lumbar Spinal Stenosis: A Systematic Review of the Literature and Meta‐analysis of Single‐arm Studies

Abstract

Purpose

This article aims to summarize the incidence of these complications through a meta‐analysis, analyze the causes of complications and provide clinical promotion and recommendations.

Methods

Databases and retrieval platform including PubMed, Web of science, Springer link, Cochrane clinical trials, ProQuest, ScienceDirect, Europe PMC, Wiley online, OVID, Clinical trials, CNKI and WanFang, and supplement the literature through Google Scholar, collect all the unilateral biportal endoscopy (UBE) controlled trials and non‐controlled trials of UBE in the treatment of lumbar spinal stenosis (LSS). The search time limit is from January 1, 2000 to December 25, 2021. After two reviewers independently screened the literature, extracted data and evaluated the risk of bias in the included studies, meta‐analysis was performed using Stata 15.1 software.

Results

Finally, 24 studies were included, including 999 patients. The results of a single‐arm rate meta‐analysis showed that the overall complication rate of UBE treatment of LSS was 6.27% [95% CI (0.0412, 0.0876)], and the incidence of dural tear was 2.49% [95% CI (0.0133, 0.0390)], the incidence of transient paresthesia was 0.14% [95% CI (0.0000, 0.0072)], the incidence of postoperative spinal epidural hematoma was 0.27% [95% CI (0.0000, 0.0096)], the incidence of postop headache, inadequate decompression, root injury and infection was 0.00%.

Conclusion

Current evidence shows that the complication rate of UBE in the treatment of LSS is low, mainly due to dural tears. Limited by the number and quality of included studies, the above conclusions still need to be confirmed by more studies.

The graphical abstract of Single‐arm meta analysis of UBE for LSS.

• Analysis of the complication rate of unilateral biportal endoscopic in the treatment of lumbar spinal stenosis.

• Analyze the causes of complications.

• Advice on reducing complications.

Introduction

Lumbar spinal stenosis (LSS) is a common degenerative disease of the lumbar spine, accounting for 8% to 11% of all degenerative diseases requiring surgical treatment. ¹ The disease can be caused by any type of spinal canal, nerve root canal or intervertebral foramen, the symptoms are lower back or lower extremity pain caused by nerve root compression. Although most patients choose conservative treatment in the early stages, as the disease progresses, due to the serious impact on the quality of life, most patients have to choose surgical treatment to relieve symptoms.

Clinically, the treatment of LSS has undergone the transformation and development from open to minimally invasive, mainly because open surgery may cause greater damage to paravertebral anatomical structures such as muscles and facet joints, which may lead to postoperative complications. The biomechanical stability of the spine is reduced, and long‐term complications such as degeneration of adjacent vertebral bodies occur, and minimally invasive spine surgery emerges as the times require. ² , ³ At present, the most popular minimally invasive surgery is mainly the foraminal technique. However, the intraoperative operation space of this technique is small, the decompression may be incomplete, and the learning curve is steep, which is not friendly to young clinicians. ⁴ Therefore, unilateral biportal endoscopy (UBE), which was proposed by Kambin and Sampson in 1986, has re‐entered people's field of vision in recent years under the improvement of Korean scholars. ⁵ , ⁶

The reason why UBE is currently recognized and popular is mainly because: (i) it has the characteristics of flexible operation in open surgery and a clear magnified field of view in minimally invasive surgery; (ii) it can be used similar to conventional foraminal scope; (iii) because of the multifidus interstitial approach, the damage to the paraspinal muscles is less; (iv) the number of intraoperative fluoroscopy is less, which reduces the radiation level to the patient; and (v) surgery has a wide range of indications and can be used to treat degenerative diseases such as lumbar disc herniation, I ~ II° lumbar spondylolisthesis. With the popularization of UBE technology, some problems have also arisen, mainly in the complications of surgery. After our literature search and screening, we found that these complications are dural tear, transient paresthesia, postoperative spinal epidural hematoma (PSEH), postop headaches, inadequate decompression, root injury and infection. The purpose of this study is as follows: (i) to review the current clinical studies on the use of UBE for the treatment of LSS, and to provide experience for clinicians who perform this procedure; (ii) to summarize the complications of LSS using UBE technology and analyze the incidence of these complications by meta‐analysis; and (iii) to analyze the possible causes of intraoperative complications in UBE.

Methods and Materials

Search Strategy

To perform comprehensive retrieval strategy, we systematically searched relevant studies published in electronic databases including PubMed, Web of science, Springer link, Cochrane clinical trials, ProQuest, ScienceDirect, Europe PMC, Wiley online, OVID, Clinical trials CNKI and WanFang, and supplement the literature through Google Scholar, collect all the unilateral dual channels RCT, non‐RCT and single‐arm trials of endoscopic technique in the treatment of LSS. The search time limit is from January 1, 2000 to December 25, 2021. The search is carried out by combining subject terms and free words. The search strategy is determined after multiple pre‐searches, and the search terms are adjusted according to the specific database. Search terms include: “biportal endoscopic spinal surgery” OR “two portal endoscopic spinal surgery” AND “Lumbar spinal stenosis”.

Inclusion and Exclusion Criteria

The criteria for inclusion article: (i) research type: RCT, non‐RCT or single‐arm clinical research on the treatment of LSS with UBE; (ii) research objects: patients who have clear clinical symptoms and are diagnosed with LSS such by MRI or X‐ray; and (iii) intervention measures: All patients with LSS were treated with UBE surgery.

Exclusion criteria included: (i) non‐English literature; (ii) multiple reports from a single center, studies with the largest sample size without repetition after reading the full text; (iii) for multiple similar studies in the same unit, exclude duplicate cases after reading the full text; and (iv) the patients have serious cardiovascular and cerebrovascular diseases, mental diseases, malignant tumors, etc.

Data Extraction

Two investigators independently screened the literature according to the inclusion and exclusion criteria, extracted data and cross‐checked them. In cases of disagreement, they were discussed and resolved, and the third investigator's opinion was sought if necessary. If the information in the research report is incomplete, the author was contacted to obtain further information; a pre‐designed data extraction table was used to extract the information. The main information extracted from the data includes: (i) general information of the included study, including title, author, year of publication, etc.; (ii) research characteristics, including study area, sample size, age, operation time, follow‐up time, etc.; (iii) concerned outcome indicators and outcome measurement data; and (4) relevant data for bias risk evaluation.

Quality Assessment

In this study, the ethodological index for non‐randomized studies (MINORS) ⁷ was used to evaluate the risk of bias. Since this study only evaluated the single‐arm rate outcome, only the first eight items MINORS were used for evaluation.

Outcome Indictor

In the present meta‐analysis, outcome indicators are included: (i) postop headache: Anesthesia, insufficient body fluid supplement and other reasons cause postoperative headaches in patients; (ii) dural tear: intraoperative or postoperative cerebrospinal fluid leakage; (iii) transient paresthesiay: transient limb numbness after surgery; (iv) PSEH: Postoperative radiographic data showed epidural hematoma; (v) inadequate decompression: postoperative symptoms did not improve significantly; (vi) root injury: injury to nerve roots during surgery, resulting in corresponding sensory and motor dysfunction; and (viii) infection: postoperatively, the incision of the patient did not heal well, there was an inflammatory reaction, and the bacterial test of the drainage tube was positive.

Statistical Analysis

The statistical software Stata15.1 (StataCorp, College Station, TX, USA) was used to analyze the complications. The I² test was used to test the heterogeneity of the included literature. According to the Cochrane manual, I² > 50% was considered heterogeneity, and a random effects model was selected for combined statistical analysis; when I² < 50%, heterogeneity did not exist. A fixed‐effects model was chosen for data processing and analysis. Since the incidence of complications in this meta‐analysis is relatively small (0 < P < 0.2), metaprop is used to perform double arcsine transformation of the data and then statistically merge. ⁸ , ⁹

Sensitivity Analysis

Sensitivity analysis was performed on the included studies using the metain command in tata15.1 software.

Publication Bias

The metafunnel command in the Stata15.1 software was used to draw the funnel diagram, and judge the publication bias subjectively by observing the symmetry of the funnel diagram. If the funnel diagram is symmetrical, there is no publication bias, and the Egger's test is used to objectively judge whether there is publication bias, P > 0.05 Explain funnel chart symmetry.

Results

Selection of Studies

A flow chart for the inclusion of studies is shown in Fig. 1. A total of 577 related documents were obtained in the preliminary inspection, including PubMed: 103, Web of science: 100, Cochrane clinical trials: nine, ProQuest: 29, ScienceDirect: one, Europe PMC: 144, Wiley online: two, Clinical trials: one, CNKI:28, WanFang: 16, Google scholar: 144. After excluding duplicate documents and reading the title and abstract, 462 articles were obtained; 115 articles remained; further reading the full text for re‐screening, the remaining 81 articles were excluded, 24 studies were included eventually.

Fig. 1.

Flow diagram for the meta‐analysis method. * database or platform and articles number: PubMed(n = 103); Web of science: (n = 100); Cochrane clinical trials: (n = 9); ProQuest:(n = 29); ScienceDirect: (n = 1); Europe PMC: (n = 144); Wiley online:(n = 2); Clinical trials: (n = 1); CNKI: (n = 28); WanFang: (n = 16); Google scholar: (n = 144)

Characteristics of the Studies Included in the Meta‐analysis

The basic characteristics of the included studies are shown in Table 1. A total of 24 studies were included, including 10 single‐arm studies and14 controlled trials which include two RCT studies, two Prospective studies and 10 retrospective studies. Ten of them were Korean studies, one Thai study, one Hungarian study, eight Chinese studies, one Ukrainian study, one Japanese study and one Saudi Arabian study. A total of 999 patients with lumbar spinal stenosis were included. All patients have similar basic diseases, treatment methods and other basic characteristics.

TABLE 1.

Basic characteristics of included studies

Author	Country/Region	Year	Design	Sample size (M/F)	Age (years)	Operation time (min)	Pressure (mmHg)	Blood loss (mL) /ΔHb (g/L)	Complications
Eum et al. 12	Korea	2016	Retrospective	58 (18/40)	63.4 ± 7.4	68.9 ± 16.1	20–30		A (3); B (2); C (2); D (1)
Torudom et al. 33	Thailand	2016	Retrospective	30 (11/19)	56 ± 6.2	98.3 ± 14.3			C (2); E (1)
Kim et al. 27	Korea	2018	Retrospective	55 (26/29)	70.7 ± 14.3	53 ± 11.5	30		B (2); D (1)
Heo et al. 34	Korea	2018	Prospective	46 (18/28)	65.8 ± 8.9	61.1 ± 5.2			B (1); D (1)
Park et al. 10	Korea	2019	Retrospective	60 (31/29)	67.6 (41–91)	83.8 ± 37.9	40		B (3); D (1); E (2)
Choi et al. 14	Korea	2019	Retrospective	35 (14/21)	65.4 ± 11.8		25–30	(−2.5 ± 0.7) g/L
Heo et al. 35	Korea	2019	Retrospective	37 (15/22)	66.7 ± 9.4	62.4 ± 5.7			B (1); D (1)
Kang et al. 36	Korea	2019	RCT	32 (18/14)	65.1 ± 8.6	36 ± 11	30		D (1)
Zhang et al. 37	China	2019	Retrospective	5 (3/2)	68.5	110.6 ± 18.9	30	15.2 ± 9.7 mL
Pao et al. 11	China (Taiwan)	2020	Retrospective	81 (38/43)	70.2 ± 10.8	89 ± 56.9			B (4); C (1); D (1); E (1)
Czigleczki et al. 13	Hungary	2020	Retrospective	21 (9/12)	66.5 (30–83)		20–30		A (3); B (2); E (1)
Min et al. 38	Korea	2020	Retrospective	54 (27/27)	65.74 ± 10.52	53.68 ± 6.75			B (2); D (1)
Park et al. 39	Korea	2020	RCT	32 (13/19)	66.2 (41–80)	67.2 ± 19.8	30–40		B (2); D (1)
Kim et al. 40	Korea	2020	Retrospective	30 (13/17)	64.23 ± 5.26	58.10 ± 6.04		53.63 ± 10.08 mL	B (1)
Fishchenko et al. 41	Ukraine	2020	Retrospective	67 (37/30)	45.80 ± 14.18	71.3 ± 21.9		34.8 ± 16.2 mL	B (4)
Ito et al. 42	Japan	2021	Retrospective	42 (28/14)	66.3 ± 12.3	57 ± 10.3	30		B (2)
Aygun et al. 43	Turkey	2021	Prospective	77 (44/33)	64.64 ± 10.09	57.74		49.47 mL
Chen et al. 44	China	2021	Retrospective	13 (7/6)	63.5 (48–81)	75.1 (55–100)		48.8 (25–80) mL	B (1)
Tuo et al. 45	China	2021	Retrospective	22 (12/10)	59.1 ± 11.7			48.2 ± 7.2 mL	B (3); D (1)
Wang et al. 46	China	2021	Retrospective	Central: 28 (15/13)	Central: 67.79 ± 6.29	Central: 120.75 ± 9.79		Central: −21.54 ± 9.79 g/L	C (5)
				Lateral: 36 (17/19)	Lateral: 62.75 ± 8.02	Lateral: 106.4 ± 12.99		Lateral: −21.00 ± 7.67 g/L
Zhao et al. 47	China	2021	Retrospective	34 (16/18)	65.71 ± 10.55	65.30 ± 9.58		<50 mL
Kang et al. 48	China	2021	Retrospective	50 (22/28)	64.97 ± 9.83	62.59 ± 7.10		47.15 ± 9.20 mL	B (2); F (3); G (2)
Wang et al. 49	China	2021	Retrospective	23 (13/10)	61.52 ± 4.09	66.52 ± 3.51		46.96 ± 4.19 mL	B (1); C (1)
Gu et al. 50	China	2021	Retrospective	31 (11/20)	62.8 ± 8.1	55.1 ± 11.7			B (1)

Abbreviations: A, postop headache; B, dural tear; C, transient paresthesia; D, PSEH; E, inadequate decompression; F, root injury; G, infection; RCT, Randomized controlled study.

Methodological Study Quality Assessment

See Table 2 for the evaluation results of the risk of bias of the included studies.

TABLE 2.

Results of risk assessment of bias in included studies

Study	A clearly state aim	Inclusion of consecutive patients	Prospective collection of data	Endpoints appropriate to the aim of the study	Unbiased assessment of the study endpoint	Follow‐up period appropriate to the aim of the study	Loss to follow up less than 5%	Prospective calculation of the study size
Eum et al. 12	2	2	2	2	1	2	0	0
Torudom et al. 33	2	2	2	2	1	2	2	0
Kim et al. 27	2	2	2	2	1	2	0	0
Heo et al. 34	2	2	2	2	1	2	0	0
Park et al. 10	2	2	2	2	2	2	2	2
Choi et al. 14	2	2	2	2	1	2	0	0
Heo et al. 35	2	2	2	2	1	2	1	0
Kang et al. 36	2	2	2	2	1	2	0	0
Zhang et al. 37	2	2	2	2	1	2	2	0
Pao et al. 11	2	2	2	2	1	2	2	0
Czigleczki et al. 13	2	2	2	2	1	2	2	0
Min et al. 38	2	2	2	2	1	2	2	0
Park et al. 39	2	2	2	2	2	2	0	2
Kim et al. 40	2	2	2	2	1	2	2	0
Fishchenko et al. 41	2	2	2	2	1	2	2	0
Ito et al. 42	2	2	2	2	1	2	2	0
Aygun et al. 43	2	2	2	2	1	2	2	0
Chen et al. 44	2	2	2	2	1	2	2	0
Tuo et al. 45	2	2	2	2	1	2	2	0
Wang et al. 46	2	2	2	2	1	2	2	0
Zhao et al. 47	2	2	2	2	1	2	2	0
Kang et al. 48	2	2	2	2	1	2	2	0
Wang et al. 49	2	2	2	2	1	2	2	0
Gu et al. 50	2	2	2	2	1	2	2	0

Note: The items are scored 0 (not reported), 1 (reported but inadequate) or 2 (reported and adequate).

Meta‐Analysis Results

Primary Outcomes

Overall Complication Rate

Except for four studies that did not report surgical complications, ¹⁵ , ¹⁸ , ²⁶ , ³⁰ all other studies reported the occurrence of surgical complications, and the heterogeneity of different studies was statistically different (I² = 44.55%, P = 0.01), but because the single‐arm meta itself has large heterogeneity, and sensitivity analysis and publication bias showed that no literature had a large interference with the results, so subgroup analysis was not performed, and a random‐effects model was used for meta‐analysis. The overall complication rate of UBE treatment of LSS was 6.27% [95% CI (0.0412, 0.0876)] (Fig. 2A).

Fig. 2.

(A) Forest plot of overall complication rate. (B) Forest plot of dural tear rate. (C) Forest plot of transient paresthesia rate. (D) Forest plot of PSEH rate. (E) Forest plot of postop headache rate. (F) Forest plot of inadequate decompression rate. (G) Forest plot of root injury rate. (H) Forest plot of infection rate

Dural Tear Rate

A total of 17 studies reported the occurrence of dural tear during surgery, ¹⁰ , ¹² , ¹³ , ¹⁴ , ¹⁶ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁷ , ²⁸ , ³¹ , ³² , ³³ and there was no heterogeneity among the studies (I² = 13.23%, p = 0.28). A random‐effects model was used for meta‐analysis. The incidence of dural tear was 2.49% [95% CI (0.0133, 0.0390)] (Fig. 2B).

Transient Paresthesia Rate

A total of five studies reported transient paresthesia during surgery, ¹⁰ , ¹¹ , ¹⁹ , ²⁹ , ³² and there was no heterogeneity among studies (I² = 0.00%, P = 0.57). A random effects model was used for meta‐analysis. The incidence of transient paresthesia was 0.14% [95% CI (0.0000, 0.0072)] (Fig. 2C).

PSEH Rate

A total of ten studies reported the occurrence of PSEH during surgery, ¹⁰ , ¹² , ¹³ , ¹⁴ , ¹⁶ , ¹⁷ , ¹⁹ , ²¹ , ²² , ²⁸ and there was no heterogeneity between studies (I² = 0.00%, P = 0.98). A random effects model was used for meta‐analysis. The incidence of PSEH was 0.27% [95% CI (0.0000, 0.0096)] (Fig. 2D).

Secondary Outcomes

Postop Headache Rate

A total of two studies reported postop headache during surgery, ¹⁰ , ²⁰ and there was no heterogeneity between studies (I² = 0.00%, P = 0.83), and a random‐effects model was used for meta‐analysis. The incidence of postop headache was 0.00% [95% CI (0.0000, 0.0027)] (Fig. 2E).

Inadequate Decompression Rate

A total of four studies reported inadequate decompression during surgery, ¹¹ , ¹⁴ , ¹⁹ , ²⁰ and there was no heterogeneity among studies (I² = 0.00%, P = 0.99). A random effects model was used for meta‐analysis. The incidence of inadequate decompression was 0.00% [95% CI (0.0000, 0.0036)] (Fig. 2F).

Root Injury Rate

One study reported root injury during surgery, ³¹ and there was no heterogeneity among studies (I² = 0.00%, P = 0.99). A random‐effects model was used for meta‐analysis. The incidence of root injury was 0.00% [95% CI (0.0000, 0.0022)] (Fig. 2G).

Infection Rate

One study reported infection during surgery, ³¹ and there was no heterogeneity among studies (I² = 0.00%, P = 1.00). A random‐effects model was used for meta‐analysis. The incidence of infection was 0.00% [95% CI (0.0000, 0.0013)] (Fig. 2H).

Sensitivity Analysis

Sensitivity analysis showed that none of the studies greatly interfered with the results, the sensitivity analysis results showed in Fig. 3.

Fig. 3.

Sensitivity analysis of overall complication

Publication Bias

In this study, the funnel chart was used to analyze the publication bias of total complication, and the Egger's test ¹⁵ was used to statistically test the degree of asymmetry of the funnel chart. The results showed that there was no publication bias in the total complications (P = 0.454 > 0.05). The Egger's test results were 0.76 [95% CI (−0.38, 0.82)], and the funnel diagram is shown in Fig. 4.

Fig. 4.

Funnel plot of publication bias for overall complication

Discussion

This study was a single‐arm meta‐analysis of surgical complication rates for LSS treated with UBE technology. A total of 24 studies including 999 patients were included. As a minimally invasive surgical method that is gradually being promoted in clinical practice, UBE combines the advantages of traditional minimally invasive techniques and open surgery. However, due to the novelty of this technique, there may be some difficulties that hinder the promotion of the technique. Therefore, this study summarizes the surgical complications of UBE in the treatment of LSS through meta‐analysis, hoping to provide help for clinical promotion. The results showed that the total complication rate of UBE in the treatment of LSS was about was 6.27%. Complications include postop headache, transient paresthesia, dural tear, PSEH, root injury, infection, and inadequate decompression. These articles and data provide valuable experience for the follow‐up clinical UBE operation, and also analyze the causes and precautions that lead to these complications as follows.

Dural Tear

Among them, the incidence of spinal dural tear is about 2.49%. There are four main reasons for spinal dural tear caused by UBE spine endoscopy: (i) The visual field under endoscopy is a two‐dimensional plane and is easily blurred., Beginners are easy to make mistakes; ³⁵ (ii) UBE does not require retraction of the anatomical structure to expose the dura mater, which is quite different from other techniques; ³⁶ (iii) patients with complex conditions require too long operation time and increase the risk of spinal membrane tears ²⁰ , ³⁷ ; (4) There is a spinal ligament that connects the dorsal side of the dura mater with the lamina and the ligamentum flavum on the surface of the central area of ​​the dura mater. This ligament is a mesh structure with thickness and shape ranging from thin strips to thick sheets It is covered and hidden by the epidural fat. During the operation, the injected saline will squeeze both sides of the dura mater, causing the area to fold. At this time, during the ligamentum flavum resection, the central area may be damaged. Spinal dura tear; ³⁸ , ⁴⁰ and (v) when using high‐speed drills, the peripheral fibrous bands and vascular bundles of the dura may be stretched around the drill neck, causing larger tears. ³⁹ , ⁴⁰ , ⁴¹

PSEH

PSEHs incidence is about 0.27%. The use of an infusion pump during surgery may be an unavoidable risk factor. When the outflow of saline solution is blocked, the pump continues to infuse saline to increase the pressure in the surgical field, cover up bleeding points, and cause intraoperative hemostasis. Lack of saline solution may cause PSEH. ⁴² Although symptomatic PSEH is relatively rare (the incidence rate is 0.02% to 4.6%), ⁴³ it can lead to serious consequences such as cauda equina syndrome and even lower limb paralysis, which affects the life of patients. Therefore, early detection and early detection are important. Handling is extremely important.

Transient Paresthesia

The incidence of transient paresthesia is about 0.14%. The main reason for transient paresthesia after surgery is that palsy and pain are both conducted by sensory nerves. The pain is transmitted by small unmyelinated fibers, and the conductive palsy is thick and thick. ¹⁰ , ⁴⁴ , ⁴⁵ The structure of myelin fibers, unmyelinated fibers is relatively simple, and the postoperative recovery is faster, while myelin fibers need to undergo a longer and more complex repair process; in addition, pain is a more acute and uncomfortable sensation than palsy, so surgery the former is often covered up. After the postoperative pain is weakened or recovered, the palsy is exposed. ⁴⁶ , ⁴⁷ , ⁴⁸ Most patients will relieve the palsy. However, because the palsy is positively related to the illness time and the degree of stenosis, the recovery time of different patients is different. ⁴⁹

Comparing with Previously Published Clinical Meta‐Analysis

There are three clinical studies on meta‐analysis. The article by Lin et al. ⁵⁰ showed that the complication rate of UBE in the treatment of LSS was 6.7%, which is higher than our study (6.27%), and the disadvantage of this article is that it only averages the results of the included articles, and is for a variety of lumbar spine diseases, not specifically for LSS, and does not analyze specific complications; Pairuchvej et al. and Pranata et al. ⁵¹ , ⁵² both concluded that 6.2%, But did not give analysis results for specific complications, and because the articles included in the above three studies are controlled studies, the number of articles is small, and the results are less convincing.

Strengths and Limitations

So far, this is the first summary of the postoperative complications of UBE treatment of LSS, and put forward the causes for the complications. Unlike previously published clinical meta‐analysis, our latest article retrieves 16 eligible clinical studies, including controlled and single‐arm studies, and it provides more high‐level literature through evidence‐based medical analysis to obtain more reliable data from raw data. The results are designed to arouse the attention of clinicians and provide advice on surgical procedures and postoperative prevention.

This meta‐analysis has certain limitations: (i) the single‐group rate meta‐analysis has its own methodological limitations, because it is a comparison of rates, not a simple mean, so although weight indicators are added, there is no design for controlled studies, the accuracy of the results is more likely to be reduced; ⁹ (ii) there are a large number of studies in South Korea included in this study, and the extrapolation of the results has certain limitations; and (iii) some studies only conduct descriptive analysis and cannot extract accurate clinical data. In short, there are certain postoperative complications in the treatment of LSS with UBE. Limited by the number and quality of the studies included, the above conclusions still need to be confirmed by more studies.

Conclusions

Our systematic review of the meta‐analysis concluded that the UBE technique has a lower postoperative complication rate in general, and is safer than other minimally invasive techniques that have been reported. It is worth noting that, the incidence of spinal dural tears is higher than other complications (2.49%) and requires more detailed surgical operations. At the same time, because epidural PSEH may cause serious complications, timely follow‐up is required after an imaging review after surgery.

Author Contributions

Wang Bin, He Peng, and Wu Zhengfang designed the systematic review. Wang Bin and He Peng drafted the protocol, and Liu Xiaowei and Xu Bin revised the manuscript. Wang Bin and He Peng will independently screen the potential studies, extract data, assess the risk of bias, and finish data synthesis. Wang Bin and He Peng will arbitrate any disagreements during the review. All authors approved the publication of the manuscript.