A retrospective analysis was conducted on 547 patients who received TLIF treatment from August 2018 to December 2021. Obtaining informed consent from each patient, this study was approved by the Ethics Review Committee of our hospital and obtained a unique study registration identification number (the research registration number is 20241111).

The inclusion criteria consisted of patients who: (1) were aged 18 years or older, (2) presence of mechanical back and unilateral radicular leg pain caused by lumbar disc herniation, (3) failure of conservative treatment to alleviate the radicular pain, (4) confirmation of lumbar disc herniation through magnetic resonance imaging (MRI) findings, and (5) follow up to study endpoint. Exclusion criteria were: (1) Other severe spinal diseases such as spinal tumors, spinal tuberculosis, and pyogenic spinal infections, (2) spinal mechanical instability, such as lumbar spondylolisthesis or segmental instability, (3) unwillingness to participate in the study, and (4) Lumbar spine surgery history. Figure 1 shows the research process of this study.

After general anesthesia, the patient is placed in a prone position with the abdomen suspended to reduce intra-abdominal pressure, thereby reducing bleeding from the venous plexus in the spinal canal during decompression operations. A midline incision on the lower back should be made, cutting through the skin and subcutaneous tissue. Then, the paraspinal muscles should be carefully removed from under the periosteum to reveal the vertebral plate and facet joints. Pedicle screws were placed, and then the lower articular process of one side of the upper vertebral body and the upper articular process of the lower vertebral body was bitten off by the vertebral plate clamp. When removing the ligamentum flavum, the dural sac is protected by the nerve peeler to prevent tearing the dura mater. The deep ligamentum flavum was removed and the lateral recess was decompressed. Then the hypertrophic ligamentum flavum and residual joint capsule were removed from the intervertebral foramen and lateral recess.

Expose the intervertebral disc from the intervertebral foramen area. Use a sharp knife to cut through the fibrous ring on the surface of the intervertebral disc. Use nucleus pulposus forceps to remove degenerated nucleus pulposus tissue from the intervertebral disc. Scrape off the upper and lower cartilage endplates, and gradually increase the height of the trial model to select the appropriate size of cage (Cage contacts the upper and lower vertebral plates). Wash the intervertebral space with sterile physiological saline, implant autologous bone, allogeneic bone fragments, and fusion cage and ensure the presence of bone tissue around the fusion cage. Place the appropriate length of the bowl rod and bending, and place the screw tail cap. Cover the surface of the dural sac with gelatin sponge. Place drainage tube and suture the layers.

This study is to comprehensively collect and analyze clinical indicator data of cases, and track and record their follow-up results for up to 2 years after surgery. These include age, gender, osteoporosis, body mass index (BMI), painful limbs, hypertension, diabetes, occupation, lesion segment (multi segment, high-level segment), preoperative pain time, operation time, intraoperative blood loss, postoperative wound drainage, wound infection, deep vein thrombosis (DVT), imaging outcomes were assessed via lumbar x-rays and computed tomography (CT) at 3 months, 1 year, and 2 years postoperatively to evaluate fusion status (solid fusion vs. non-union) and cage position (subsidence defined as ≥2 mm vertical displacement or ≥10° angular deviation from the initial placement), preoperative visual analog scale (VAS)-back, postoperative VAS-back preoperative and postoperative oswestry disability index (ODI) scores and postoperative complications including symptoms recurrence, postoperative hematoma, wound infection and poor wound healing complications.

Osteoporosis was diagnosed with the aid of x-ray or computed tomography (CT) findings due to the infrequency of bone density testing in patients lumbar disc herniation (12, 13). In continuous variables, except for age and BMI, all other variables are dichotomized at the median as their high and low values. Obesity was defined as having a BMI of no less than 28. Advanced age is defined as the patient's age not less than 60 years old. Patients were divided into mental workers and manual workers according to their occupation. The patients who undertook a small amount of manual labor and worked mainly indoor were regarded as mental workers. They can also be called non-manual workers. The patients who undertook a large amount of manual labor and worked mainly outdoor were regarded as manual workers. Lesion segment was classified as “multi segment” if the lumbar disc herniation involved two or more adjacent spinal segments (e.g., L3–4 and L4–5), and “high-level segment” was defined as herniation occurring at the L1–2, L2–3 or L3–4 level, distinguishing it from the more common lower segments (L4–5 or L5-S1). DVT records did not include cases of lower extremity intermuscular thrombosis because this type of thrombosis was classified as a peripheral variant of DVT and essentially limited to the venous plexus of soleus and gastrocnemius. Studies have shown that lower extremity venous intermuscular thrombosis has almost no impact on patients (14). Symptom recurrence is manifested as the initial improvement of lower back pain and neuralgia caused by lumbar disc herniation after surgical treatment, followed by the recurrence of lower back pain and lower limb neuralgia associated with lumbar vertebrae (whether due to adjacent segment degeneration or contralateral foraminal stenosis) at 6 months postoperatively (15, 16). According to Centers for Disease Control and Prevention definition (17), Wound infection, including superficial wound infection, was defined as infection involving only the skin or subcutaneous tissue occurring within 30 days postoperatively, while deep wound infection was defined as infection occurring within a year post-surgery, confirmed to be operation-related, and involving deep soft tissues. In this research, no deep wound infection was found, so the wound infection refers to superficial wound infection. According to the process of wound healing (18), poor wound healing was diagnosed based on surgical wound rupture, scar hyperplasia, sinus formation, skin or flap necrosis and surgical wound exudation but there was no bacterial growth after 3 days of cultivation. Postoperative hematoma, also known as epidural hematoma, is the accumulation of blood in the spinal canal after surgery that compresses the cauda equina, nerve roots, or spinal cord that can result in devastating neurologic consequences and it was diagnosed by MRI and ultrasound.

Preoperatively, a comprehensive evaluation of the patient's cardiopulmonary function, coagulation mechanism, and nutritional status should be conducted. Additionally, for patients without contraindications, intramuscular injection of thrombin may be administered to help prevent intraoperative bleeding. Secondly, preoperative guidance is given to patients to perform axial turning to adapt to postoperative position limitations. At the same time, psychological counseling is used to alleviate patients' anxiety and enhance their confidence in treatment. During the operation, general anesthesia was chosen while monitoring the patient's vital signs. After surgery, the patient was given nebulization to assist in sputum discharge. Vital signs were continuously monitored within 24 h after surgery, and lung function exercise was initiated. Low molecular weight heparin was used 24 h after surgery to prevent lower limb thrombosis, and the patient was guided to turn over axially while keeping the wound dressing clean and dry.

Rehabilitation training is mainly aimed at training the core muscles of the lumbar spine. Specifically, within one day after surgery, actively perform ankle pump on the bed and passively perform straight leg lift test. From 1 to 7 days after surgery, actively engage in straight leg raise exercises. Wear an abdominal brace to provide support, and gradually begin to get out of bed and walk slowly, progressively increasing the walking distance over time. One week after surgery, gradually initiate the following exercises: five-point support, three-point support, plank support, and prone knee and hip flexion, among others. After discharge, the patient continued to complete the aforementioned exercises under the supervision of their family members until 3 months post-surgery and attended follow-up examinations as scheduled.

Follow up of patients after discharge is necessary. Generally, patients are scheduled for regular follow-up examinations at one month, three months, six months, one year and two years after surgery. However, it is important to note if patients experience sudden situations such as significant back pain and lower limb neuralgia, they can come for diagnosis at any time. This study would have two endpoints. One is that during a 2-year period, the patient experienced severe back pain and lower limb neuralgia again and after being diagnosed with symptom recurrence through MRI examination, the time and VAS-back were recorded. The other is 2 years after surgery, at this point, all patients except those who have already completed the study would be evaluated.

To conduct a comparative analysis on patients' dissatisfaction as the dependent variable, we utilized a chi-square test for univariate analysis. Preoperative and postoperative ODI scores and VAS scores using t-test. All continuous variables are presented as mean ± standard deviation (SD). Variables that yielded a P value of less than 0.05 in the univariate analysis were included as input in the multivariate logistic regression model. For each variable, we computed the odds ratio (OR) with its 95% confidence interval (CI). In the chi-square test and multivariate logistic regression model, except for age and BMI, all the other continuous variables were dichotomized at the median. P value less than 0.05 was considered significant. All statistical analyses were done using SPSS software version 27.0 (SPSS, Inc., Chicago, IL, USA).

A total of 547 patients who met the inclusion criteria from August 2018 to December 2021 were included in this study. Among them, there were 261 male patients (47.7%) and 286 female patients (52.3%). The average age and BMI were 55.2 years old and 26.89 kg/m², respectively. Among all patients, 51.7% (283/547) suffer from left lower limb pain and 48.3% (264/547) suffer from right lower limb pain. The average preoperative pain time is 5.7 months. Please refer to Table 1 for detailed basic information of patients.

We compared the preoperative and postoperative ODI scores and VAS scores of the collected data. The results revealed that both ODI and VAS scores demonstrated significant improvement after surgery compared to preoperative levels. This indicates that TLIF can effectively enhance patients' quality of life. The specific data is shown in Table 2.

The most common complication, symptom recurrence, was analyzed using time series distribution plots to identify temporal patterns of postoperative symptom recurrence. Additionally, Kaplan-Meier survival analysis was employed to further assess the recurrence rates over time. By examining Figures 2,3 together, it is evident that the time of symptom recurrence in patients follows a normal distribution, with the peak incidence occurring at approximately 17.6 months post-surgery.

Among all 547 patients, a total of 89 (16.3%) experienced complications and the incidence of postoperative complications in patients is shown in Table 3. In this study, the most common complication was symptom recurrence, which was seen in 13.3% (73/547) of patients. There were 4 patients who developed postoperative hematomas, all of which occurred within one week after surgery. After the occurrence of postoperative hematoma, patients underwent a second surgical debridement, and all recovered well. Four patients developed wound infections, and all four patients had their waist wounds contaminated with urine. Although dressing was changed promptly upon detection, infection ultimately occurred, and Escherichia coli was cultured in all wounds after secretion culture. The infected wound did not improve after dressing change, and after undergoing debridement surgery again, the wound ultimately achieved Grade A healing. This study reported a total of 8 patients who experienced poor wound healing. All 8 patients underwent regular dressing changes, wound disinfection, and enhanced suturing to achieve final healing. The wound ultimately achieved Grade A healing after suture removal. The results of the imaging evaluation showed that intervertebral fusion had reached a solid state in all cases, and there was displacement or angulation of the fusion cages over time.

The univariate analysis of postoperative complications found that the recurrence of symptoms was related to BMI, preoperative pain time, High-level segment, intraoperative bleeding volume and postoperative VAS-back. Postoperative hematoma was related to hypertension and wound drainage volume. Poor wound healing was related to BMI and Wound drainage volume. However, this study failed to find the related factors of wound infection. Table 4 shows the detailed relations between the adjusted univariate analysis parameters and complications.

Perform multicollinearity tests on significant variables in univariate analysis using tolerance (TOL) and variance inflation factor (VIF). TOL > 0.10 or VIF < 10.0 indicates that there is no significant multicollinearity between variables. The results show that the TOL values range from 0.944–0.994 and the VIF values range from 1.006–1.060 among the variables. The variables can be included as predictive factors in the logistic regression model, and the specific results are shown in Table 5.

After binary logistic analysis of the above single factors, we found that BMI and preoperative pain time were independent risk factors for symptom recurrence, and BMI were independent risk factors for poor wound healing. Seen Table 6 for details.