Abstract
Failed back surgery syndrome (FBSS), recently redesignated as persistent spinal pain syndrome type 2, is a common complication after lumbar spine surgery, yet population-level data on its incidence and preoperative determinants remain limited. This study aimed to quantify the incidence of FBSS under multiple definitions and to identify preoperative risk factors for its development. This retrospective cohort study analyzed claims data from the Korean National Health Insurance Service (2009–2023), including 1,029,500 adults who underwent lumbar spine surgery. FBSS was operationalized using two mutually exclusive criteria assigned hierarchically: patients who underwent any subsequent lumbar spine procedure were first classified as the Reoperation group, and the remaining patients who received an International Classification of Diseases-10 M96.1 (post-laminectomy syndrome) diagnosis after the index surgery were classified as the Disease-code group. Cox proportional hazards models adjusted for demographic, clinical, and lifestyle covariates were used to identify preoperative risk factors. The overall FBSS incidence was 26.7% (n = 274,955) based on the combined Disease-code and Reoperation definitions. In multivariable analysis, advancing age, peripheral polyneuropathy, obesity, diabetes mellitus, hypertension, and current smoking were independently associated with increased FBSS risk. Decompression surgery and spinal canal stenosis also conferred elevated risk. Higher income, employment-based insurance, and moderate physical activity (500–1499 MET-minutes/week) were protective. FBSS affects approximately one-quarter of lumbar surgery patients. Multiple modifiable and non-modifiable preoperative factors were identified. These findings support individualized preoperative risk assessment to optimize surgical decision-making and patient counseling.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-026-55626-2.
Keywords: Failed back surgery syndrome, Persistent spinal pain syndrome, Incidence, Risk factors, Cohort studies, Low back pain
Subject terms: Diseases, Health care, Medical research, Risk factors
Introduction
Low back pain is among the foremost contributors to global disability and affects most individuals at some point during their lifetime1. When conservative management proves insufficient, lumbar spine surgery is frequently pursued. However, a notable subset of surgical patients continues to experience persistent or recurrent pain, a condition widely referred to as failed back surgery syndrome (FBSS)2,3. More recently, the International Association for the Study of Pain and the International Classification of Diseases (ICD)−11 taxonomy have proposed ‘persistent spinal pain syndrome type 2’ (PSPS-2) as a more neutral term that avoids the implicit negativity of ‘failure’ and aligns chronic post-surgical spinal pain with the contemporary chronic pain classification4. For consistency with the predominant prior epidemiologic literature cited throughout this manuscript, we retain ‘FBSS’ hereafter while acknowledging PSPS-2 as the contemporary nomenclature. Published estimates suggest that FBSS develops in approximately 10–40% of patients following spinal procedures5,carrying considerable implications for physical function, psychological well-being, and healthcare expenditure6,7.
Despite substantial research into the pathophysiology and treatment of FBSS, critical gaps remain in our understanding of who is most susceptible to developing this condition8. A major challenge lies in the absence of a standardized definition: diagnostic criteria, terminology, and outcome measures differ markedly across studies, making direct comparisons difficult and limiting accurate epidemiological characterization9,10. Furthermore, while several clinical and surgical factors have been proposed as contributors to FBSS, evidence regarding their relative importance—particularly at the population level—is scarce. Existing investigations of risk factors for poor outcomes after lumbar surgery have predominantly focused on specific diagnostic or procedural subgroups, such as revision discectomy after primary microdiscectomy11, reoperation after degenerative lumbar spondylolisthesis surgery12,and new vertebral fracture following percutaneous vertebroplasty or kyphoplasty for osteoporotic vertebral compression fractures13. These disease-specific studies have provided valuable insights into how risk factor profiles vary by clinical indication. However, complementary population-level evidence spanning the full spectrum of lumbar surgical patients remains scarce, particularly from large single-payer healthcare systems with longitudinal follow-up.
Identifying pre-operative predictors of FBSS is of considerable clinical value. If patients at heightened risk can be recognized before surgery, clinicians may be better equipped to counsel patients regarding realistic postoperative expectations, select appropriate surgical approaches, and implement early preventive measures14. Such risk stratification could ultimately reduce the burden of FBSS and improve long-term outcomes after lumbar spine surgery.
South Korea provides a particularly informative setting for studying FBSS. The country has one of the world’s most rapidly aging populations15,and lumbar spine surgery is exceptionally common; over 200,000 spinal procedures were performed in 2023 alone16. The Korean National Health Insurance Service (NHIS) database, which captures healthcare utilization data for the entire population under a single-payer system, offers a rare opportunity for large-scale, longitudinal investigation of FBSS epidemiology17.
Accordingly, the objectives of this study were twofold: first, to establish the claims-based incidence of FBSS using the two complementary, mutually exclusive proxies (diagnostic coding and lumbar reoperation) in a national-payer cohort with longitudinal follow-up; and second, to identify preoperative demographic, clinical, and lifestyle-related factors associated with FBSS development. In doing so, we aim to provide a population-scale, claims-based foundation that complements the disease-specific FBSS literature and supports preoperative risk stratification in routine clinical practice.
Materials and methods
Study population
This retrospective cohort study was based on nationwide claims data from the Korean NHIS, covering the period from 2009 to 2023. South Korea operates a mandatory single-payer insurance system through the NHIS, which has provided universal coverage since 1989. All citizens contribute income-based premiums, and medical costs are shared between the insurer and the patient. Consequently, healthcare providers submit reimbursement claims to the NHIS for all covered services, generating a comprehensive administrative dataset encompassing the entire Korean population.
The initial study pool consisted of 2,346,164 individuals who had at least one lumbar spine surgical claim during the observation window. To assemble the analytic cohort, we applied the following exclusion criteria: (1) erroneous death-date records; (2) incomplete demographic data; (3) primary diagnoses of spinal cord neoplasm ICD-10 codes C41.2, C72, C79.5, D33.4, D43.4, C70.1, D32.1, D48.9, G99.2, C79.4, D42.1); (4) a pre-existing post-laminectomy syndrome diagnosis (M96.1) at enrollment; (5) age younger than 20 years; (6) no available health screening within two years of the index surgery; (7) missing information on smoking, alcohol intake, physical activity, or body mass index; and (8) any spinal surgery or health screening record predating 2009. After applying these criteria, 1,029,500 participants constituted the final cohort (Fig. 1). For each individual, the index lumbar surgery was defined as the first lumbar surgical claim recorded on or after January 1, 2009. The exclusion of any spinal surgery record predating 2009 (criterion 8) served as a look-back washout window of up to 7 years (2002–2008), ensuring that all cohort members underwent their first observed lumbar operation within the analytic period (2009–2023).
Fig. 1.
Flowchart depicting patient selection and exclusion criteria.
Because this study relied exclusively on de-identified, retrospective administrative data, the Institutional Review Board of the National Health Insurance Service Ilsan Hospital (NHIMC 2024-CR-031) granted an exemption from ethical review. The Institutional Review Board of the National Health Insurance Service Ilsan Hospital also waived informed consent due to the retrospective nature of this study. All procedures conformed to the Declaration of Helsinki.
Outcome definitions
The primary outcome was incident FBSS, operationalized through two complementary definitions. The Disease-code group comprised patients who received an ICD-10 M96.1 (post-laminectomy syndrome) at any point after the index lumbar surgery. The Reoperation group included those who underwent any additional lumbar spine procedure during the follow-up period. To avoid double-counting across categories, we first identified Reoperation cases based on surgical claim codes, then classified the remaining patients into the Disease-code.
In administrative claims data, FBSS is most commonly operationalized as either diagnostic coding (ICD-10 M96.1, post-laminectomy syndrome) or subsequent lumbar reoperation. We adopted both proxies in parallel for four reasons. First, M96.1 and reoperation are outcome events that clinicians enter in response to persistent post-surgical pain. In contrast, alternative candidates available in claims data—postoperative imaging, opioid or gabapentinoid prescriptions, pain procedure claims, and rehabilitation utilization—are utilization markers with broad differential indications and would inflate sensitivity at the cost of specificity. Second, both M96.1 and reoperation are subject to reimbursement validation, minimizing recall and ascertainment bias. Third, these two proxies are the most consistently applied in the prior FBSS literature, enabling direct comparison with earlier cohorts. Fourth, applying the two definitions hierarchically (reoperation first, M96.1 in the remainder) yields mutually exclusive subgroups and a coherent combined incidence.
Covariates
Demographic characteristics included age, sex, income quintile (derived from insurance premium levels), insurance type (Medical Aid, Self-employed, or Employee), and region of residence (Seoul, Gyeonggi province, metropolitan cities, or other areas).
Lifestyle variables—alcohol consumption, smoking status, weekly physical activity (expressed as metabolic equivalent of task [MET] minutes), and body mass index—were obtained from biennial national health screening performed within the 24 months preceding the index surgery; patients without a qualifying screening in this window were excluded from the analytic cohort. Comorbid conditions assessed included hypertension (ICD-10: I10–I15), diabetes mellitus (ICD-10: E10–E14), chronic kidney disease (ICD-10: N18 or dialysis claim codes O2013, O2016), and peripheral polyneuropathy (ICD-10: G61–G63). Osteoporosis was ascertained through prescription records for bisphosphonates, selective estrogen receptor modulators, or denosumab (Supplementary Table 1).
Index lumbar procedures were classified as fusion, decompression, discectomy, or other based on surgical claim codes (Supplementary Table 2). The “other surgeries” category included vertebroplasty, kyphoplasty, spinal neurolysis, and non-specific spinal column procedures defined within the NHIS reimbursement system. Each procedure category was modeled as a non-mutually exclusive binary variable (yes/no), allowing patients to contribute to multiple procedural categories during a single index operation. Surgical complexity at the index operation was characterized by the number of concurrent lumbar surgical claim codes recorded for the index admission. Patients were classified as undergoing a single procedure (one lumbar surgical claim code) or combined procedures (two or more concurrent lumbar surgical claim codes, indicating, for example, decompression combined with fusion, or multilevel procedures coded separately). Principal diagnoses were categorized into disc herniation, spinal canal stenosis, spondylolisthesis, low back pain, compression fracture, spondylolysis, infection, scoliosis, radiculopathy, lumbar fracture, multiple fracture, and others using the corresponding ICD-10 codes (Supplementary Table 3). Receipt of preoperative electrodiagnostic testing was identified from relevant procedural claim codes (F6123, F6124, E6501–E6503, E6507–E6509, F6111–F6114, F6116, FA111, FA112, FA116, FY681).
Statistical analysis
The analysis proceeded in three phases. First, baseline differences across FBSS subgroups were evaluated with chi-squared tests. Second, time-to-event analyses were performed: Kaplan–Meier curves illustrated cumulative FBSS incidence over time, and Cox proportional hazards models estimated adjusted hazard ratios (aHRs) with 95% confidence intervals (CIs) for each candidate risk factor. Third, subgroup-specific Cox models were fitted separately for the Disease-code and Reoperation definitions to evaluate consistency of risk factor effects across FBSS subtypes. All models were adjusted for the full set of demographic, clinical, and lifestyle covariates described above. Statistical significance was defined as a two-sided p < 0.05. Analyses were conducted in SAS version 9.4.2 (SAS Institute, Cary, NC, USA).
Results
FBSS incidence and baseline profile
Demographic and clinical characteristics of the study population stratified by FBSS status are presented in Table 1, and surgery-related features are summarized in Table 2.
Table 1.
Demographic characteristics.
| Variables | Total n |
FBSS groups | p-value | ||
|---|---|---|---|---|---|
| No FBSS n (%) |
Disease-code n (%) |
Re-operation n (%) |
|||
| Total | 1,029,500 | 754,545 (73.3) | 105,958 (10.3) | 168,997 (16.4) | |
| Age, years | |||||
| 20–39 | 60,595 | 50,012 (82.5) | 4,084 (6.7) | 6,499 (10.7) | < 0.001 |
| 40–64 | 446,229 | 333,068 (74.6) | 49,917 (11.2) | 63,244 (14.2) | |
| 65–75 | 336,504 | 233,619 (69.4) | 40,210 (11.9) | 62,675 (18.6) | |
| ≥76 | 186,172 | 137,846 (74.0) | 11,747 (6.3) | 36,579 (19.6) | |
| Male | 452,076 | 341,005 (75.4) | 45,110 (10.0) | 65,961 (14.6) | < 0.001 |
| Income | |||||
| Q1 (lowest) | 202,817 | 148,176 (73.1) | 21,496 (10.6) | 33,145 (16.3) | < 0.001 |
| Q2 | 127,123 | 93,028 (73.2) | 13,721 (10.8) | 20,374 (16.0) | |
| Q3 | 170,989 | 125,846 (73.6) | 17,626 (10.3) | 27,517 (16.1) | |
| Q4 | 226,912 | 166,370 (73.3) | 23,375 (10.3) | 37,167 (16.4) | |
| Q5 (highest) | 301,659 | 221,125 (73.3) | 29,740 (9.9) | 50,794 (16.8) | |
| Type of insurance | |||||
| Medical Aid | 26,273 | 18,118 (69.0) | 3,405 (13.0) | 4,750 (18.1) | < 0.001 |
| NHI Self-Employed | 285,473 | 206,510 (72.3) | 30,546 (10.7) | 48,417 (17.0) | |
| NHI Employee | 717,754 | 529,917 (73.8) | 72,007 (10.0) | 115,830 (16.1) | |
| Region | |||||
| Seoul | 147,357 | 108,640 (73.7) | 15,650 (10.6) | 23,067 (15.7) | < 0.001 |
| Gyeonggi (around Seoul) | 224,781 | 170,247 (75.7) | 19,715 (8.8) | 34,819 (15.5) | |
| Metropolitan | 240,134 | 175,740 (73.2) | 27,004 (11.2) | 37,390 (15.6) | |
| Other regions | 417,228 | 299,918 (71.9) | 43,589 (10.4) | 73,721 (17.7) | |
| Smoking status | |||||
| Never | 688,814 | 499,121 (72.5) | 72,210 (10.5) | 117,483 (17.1) | < 0.001 |
| Former | 216,555 | 169,260 (78.2) | 19,458 (9.0) | 27,837 (12.9) | |
| Current | 124,131 | 86,164 (69.4) | 14,290 (11.5) | 23,677 (19.1) | |
| Alcohol | |||||
| No | 635,795 | 450,721 (70.9) | 69,300 (10.9) | 115,774 (18.2) | < 0.001 |
| Yes | 393,705 | 303,824 (77.2) | 36,658 (9.3) | 53,223 (13.5) | |
| Physical activity (METs) | |||||
| <500 | 553,416 | 397,839 (71.9) | 58,018 (10.5) | 97,559 (17.6) | < 0.001 |
| 500–999 | 199,087 | 142,174 (71.4) | 22,148 (11.1) | 34,765 (17.5) | |
| 1000–1499 | 92,873 | 68,576 (73.8) | 10,085 (10.9) | 14,212 (15.3) | |
| ≥1500 | 184,124 | 145,956 (79.3) | 15,707 (8.5) | 22,461 (12.2) | |
| Body mass index | |||||
| Under weight (< 18.5) | 22,082 | 16,388 (74.2) | 1,404 (6.4) | 4,290 (19.4) | < 0.001 |
| Normal (18.5–24.9) | 557,972 | 410,452 (73.6) | 56,477 (10.1) | 91,043 (16.3) | |
| Over weight (25.0–29.9) | 384,554 | 279,599 (72.7) | 41,494 (10.8) | 63,461 (16.5) | |
| Obesity (≥ 30.0) | 64,892 | 48,106 (74.1) | 6,583 (10.1) | 10,203 (15.7) | |
| Hypertension | 134,479 | 93,475 (69.5) | 15,019 (11.2) | 25,985 (19.3) | < 0.001 |
| Diabetes | 301,396 | 215,522 (71.5) | 32,797 (10.9) | 53,077 (17.6) | < 0.001 |
| Chronic kidney disease | 17,371 | 12,989 (74.8) | 1,657 (9.5) | 2,725 (15.7) | < 0.001 |
| Dialysis | 4,229 | 3,149 (74.5) | 348 (8.2) | 732 (17.3) | < 0.001 |
| Peripheral polyneuropathy | 872,794 | 633,282 (72.6) | 96,319 (11.0) | 143,193 (16.4) | < 0.001 |
| Osteoporosis | 388,360 | 266,902 (68.7) | 33,713 (8.7) | 87,745 (22.6) | < 0.001 |
FBSS, failed back surgery syndrome; MET, metabolic equivalents task; NHI, national health insurance.
Table 2.
Surgery-related characteristics.
| Variables | Total | FBSS groups | p-value | ||
|---|---|---|---|---|---|
| No FBSS n (%) |
Disease-code n (%) |
Re-operation n (%) |
|||
| Total | 1,029,500 | 754,545 (73.3) | 105,958 (10.3) | 168,997 (16.4) | |
| Fusion | 230,419 | 156,120 (67.8) | 43,536 (18.9) | 30,763 (13.4) | < 0.001 |
| Decompression | 238,403 | 167,766 (70.4) | 34,048 (14.3) | 36,589 (15.3) | < 0.001 |
| Discectomy | 187,067 | 140,115 (74.9) | 2,102 (1.1) | 44,850 (24.0) | < 0.001 |
| Other surgeries | 640,403 | 468,492 (73.2) | 75,233 (11.7) | 96,678 (15.1) | < 0.001 |
| Principal diagnosis | |||||
| Disc herniation | 419,788 | 316,548 (75.4) | 42,212 (10.1) | 61,028 (14.5) | < 0.001 |
| Canal Stenosis | 242,379 | 166,168 (68.6) | 37,464 (15.5) | 38,747 (16.0) | |
| Spondylolisthesis | 90,438 | 63,404 (70.1) | 16,031 (17.7) | 11,003 (12.2) | |
| Low back pain | 6,490 | 4,909 (75.6) | 315 (4.9) | 1,266 (19.5) | |
| Compression fracture | 1,122 | 758 (67.6) | 15 (1.3) | 349 (31.1) | |
| Spondylolysis | 2,507 | 1,673 (66.7) | 520 (20.7) | 314 (12.5) | |
| Infection | 451 | 307 (68.1) | 55 (12.2) | 89 (19.7) | |
| Scoliosis | 2,023 | 1,358 (67.1) | 348 (17.2) | 317 (15.7) | |
| Radiculopathy | 1,185 | 991 (83.6) | 83 (7.0) | 111 (9.4) | |
| Fracture (other) | 118,795 | 90,429 (76.1) | 1,950 (1.6) | 26,416 (22.2) | |
| Multiple fractures | 877 | 644 (73.4) | 24 (2.7) | 209 (23.8) | |
| Others | 143,445 | 107,356 (74.8) | 6,941 (4.8) | 29,148 (20.3) | |
| Pre-electrodiagnosis | 304,532 | 215,736 (70.8) | 39,158 (12.9) | 49,638 (16.3) | < 0.001 |
| Type of hospital | |||||
| Tertiary hospital | 110,643 | 77,746 (70.3) | 18,677 (16.9) | 14,220 (12.9) | < 0.001 |
| General hospital | 224,235 | 162,302 (72.4) | 22,494 (10.0) | 39,439 (17.6) | |
| Hospital | 671,928 | 497,909 (74.1) | 63,278 (9.4) | 110,741 (16.5) | |
| Other | 22,694 | 16,588 (73.1) | 1,509 (6.6) | 4,597 (20.3) | |
| Hospital’s area | |||||
| Seoul | 283,166 | 205,213 (72.5) | 34,019 (12.0) | 43,934 (15.5) | < 0.001 |
| Gyeonggi (around Seoul) | 206,347 | 156,694 (75.9) | 16,855 (8.2) | 32,798 (15.9) | |
| Metropolitan | 320,540 | 231,718 (72.3) | 36,921 (11.5) | 51,901 (16.2) | |
| Rural | 219,447 | 160,920 (73.3) | 18,163 (8.3) | 40,364 (18.4) | |
| Surgical complexity at index | |||||
| Single procedure | 786,414 | 591,689 (75.2) | 61,644 (7.8) | 133,081 (16.9) | < 0.001 |
| Combined procedures (≥ 2) | 243,086 | 162,856 (67.0) | 44,314 (18.2) | 35,916 (14.8) | |
| Gabapentinoid prescription | 493,371 | 290,198 (58.8) | 34,176 (6.9) | 168,997 (34.3) | < 0.001 |
FBSS, failed back surgery syndrome.
Among the 1,029,500 individuals in the final cohort, 274,955 (26.7%) developed FBSS according to the combined Disease-code and Reoperation criteria. Of these, 105,958 (10.3%) were classified through the Disease-code definition and 168,997 (16.4%) through the Reoperation definition.
Disc herniation was the most frequent principal diagnosis (n = 417,788). A total of 243,086 participants underwent combined procedures (≥ 2 concurrent lumbar surgical claim codes). Notably, all patients in the Reoperation group had also been prescribed gabapentinoids at some point during follow-up.
Preoperative determinants of FBSS
The Kaplan–Meier cumulative incidence curve for FBSS following lumbar spine surgery is depicted in Fig. 2. Cox regression results for the overall FBSS outcome are reported in Table 3, and models for the Disease-code and Reoperation subgroups are provided in Supplementary Tables 4 and 5, respectively.
Fig. 2.
Kaplan–Meier cumulative incidence curve illustrating the time-dependent proportion of patients developing failed back surgery syndrome after lumbar spine surgery.
Table 3.
Cox-proportional hazard model for entire FBSS groups (both the Disease-code and Re-operation groups).
| Variables | aHR | 95% CI | p-value | |
|---|---|---|---|---|
| Lower | Upper | |||
| Age | ||||
| 20–39 | 1.00 | |||
| 40–64 | 1.30 | 1.27 | 1.32 | < 0.001 |
| 65–75 | 1.48 | 1.45 | 1.51 | < 0.001 |
| ≥76 | 1.63 | 1.59 | 1.67 | < 0.001 |
| Female | 0.82 | 0.81 | 0.83 | < 0.001 |
| Income | ||||
| Q1 (lowest) | 1.00 | |||
| Q2 | 1.01 | 1.00 | 1.02 | 0.22 |
| Q3 | 1.01 | 0.99 | 1.02 | 0.48 |
| Q4 | 0.97 | 0.96 | 0.98 | < 0.001 |
| Q5 (highest) | 0.93 | 0.92 | 0.95 | < 0.001 |
| Type of insurance | ||||
| Medical Aid | 1.00 | |||
| NHI Self-Employed | 0.79 | 0.77 | 0.81 | < 0.001 |
| NHI Employee | 0.76 | 0.74 | 0.77 | < 0.001 |
| Region | ||||
| Seoul | 1.00 | |||
| Gyeonggi (around Seoul) | 0.99 | 0.97 | 1.00 | 0.15 |
| Metropolitan | 1.00 | 0.98 | 1.02 | 0.98 |
| Other regions | 1.01 | 1.00 | 1.02 | 0.21 |
| Smoking status | ||||
| Never | 1.00 | |||
| Former | 1.09 | 1.08 | 1.11 | < 0.001 |
| Current | 1.06 | 1.04 | 1.07 | < 0.001 |
| Alcohol | 1.01 | 1.00 | 1.02 | 0.27 |
| Physical activity (METs) | ||||
| <500 | 1.00 | |||
| 500–999 | 0.97 | 0.96 | 0.98 | < 0.001 |
| 1000–1499 | 0.97 | 0.96 | 0.98 | < 0.001 |
| ≥1500 | 1.04 | 1.03 | 1.05 | < 0.001 |
| Body mass index | ||||
| Under weight (< 18.5) | 1.00 | |||
| Normal (18.5–24.9) | 0.98 | 0.95 | 1.00 | 0.08 |
| Over weight (25.0–29.9) | 1.04 | 1.02 | 1.07 | < 0.001 |
| Obesity (≥ 30.0) | 1.13 | 1.10 | 1.16 | < 0.001 |
| Hypertension | 1.05 | 1.04 | 1.06 | < 0.001 |
| Diabetes | 1.07 | 1.07 | 1.08 | < 0.001 |
| Chronic kidney disease | 1.03 | 1.00 | 1.07 | 0.05 |
| Peripheral polyneuropathy | 1.37 | 1.35 | 1.38 | < 0.001 |
| Procedure type a | ||||
| Fusion | 1.11 | 1.08 | 1.14 | < 0.001 |
| Decompression | 1.11 | 1.08 | 1.13 | < 0.001 |
| Discectomy | 0.98 | 0.96 | 1.01 | 0.17 |
| Other surgeries | 1.06 | 1.04 | 1.09 | < 0.001 |
| Principal diagnosis | ||||
| Disc herniation | 1.00 | |||
| Canal Stenosis | 1.10 | 1.08 | 1.11 | < 0.0001 |
| Spondylolisthesis | 0.98 | 0.96 | 0.99 | 0.01 |
| Low back pain | 0.91 | 0.87 | 0.96 | 0.00 |
| Compression fracture | 1.02 | 0.92 | 1.14 | 0.66 |
| Spondylolysis | 1.07 | 1.00 | 1.14 | 0.06 |
| Infection | 1.22 | 1.03 | 1.43 | 0.02 |
| Scoliosis | 1.13 | 1.04 | 1.22 | 0.00 |
| Radiculopathy (other) | 0.97 | 0.84 | 1.11 | 0.65 |
| Fracture (lumbar spine) | 0.88 | 0.87 | 0.90 | < 0.001 |
| Multiple fracture | 0.87 | 0.76 | 0.99 | 0.03 |
| Others | 0.93 | 0.91 | 0.94 | < 0.001 |
| Pre-electrodiagnosis | 1.14 | 1.13 | 1.15 | < 0.001 |
| Dialysis | 0.99 | 0.93 | 1.06 | 0.81 |
| Osteoporosis | 0.99 | 0.98 | 1.00 | 0.05 |
| Type of hospital | ||||
| Tertiary hospital | 1.00 | |||
| General hospital | 0.97 | 0.96 | 0.99 | < 0.001 |
| Hospital | 0.95 | 0.93 | 0.96 | < 0.001 |
| Other | 0.97 | 0.94 | 0.99 | 0.02 |
| Hospital’s area | ||||
| Seoul | 1.00 | |||
| Gyeonggi (around Seoul) | 0.99 | 0.98 | 1.01 | 0.41 |
| Metropolitan | 1.03 | 1.01 | 1.04 | 0.00 |
| Rural | 1.02 | 1.01 | 1.04 | 0.00 |
| Surgical complexity at index | ||||
| Single procedure | 1.00 | |||
| Combined procedures (≥ 2) | 1.01 | 0.99 | 1.04 | 0.32 |
| Gabapentinoid prescription | 3.08 | 3.05 | 3.11 | < 0.001 |
aEach procedure was modeled as a separate binary indicator. The reference group for each estimate is patients not receiving that specific procedure.
aHR, adjusted hazard ratio; CI, confidence interval; MET, metabolic equivalents task; NHI, national health insurance.
Advancing age was strongly associated with FBSS risk, with the oldest group (≥ 76 years) demonstrating the highest risk (aHR, 1.63; 95% CI, 1.59–1.67; p < 0.001). Smoking (Current: aHR, 1.06; 95% CI, 1.04–1.07; p < 0.001) and obesity (BMI ≥ 30: aHR, 1.13; 95% CI, 1.10–1.16; p < 0.001) were also significant risk factors. Among comorbidities, hypertension (aHR, 1.05; 95% CI, 1.04–1.06; p < 0.001), diabetes mellitus (aHR, 1.07; 95% CI, 1.07–1.08; p < 0.001), and peripheral polyneuropathy (aHR, 1.37; 95% CI, 1.35–1.38; p < 0.001) independently elevated the likelihood of FBSS. Patients who had undergone preoperative electrodiagnostic testing also faced increased risk (aHR, 1.14; 95% CI, 1.13–1.15; p < 0.001). These patterns were broadly consistent across both the Disease-code and Reoperation subgroup analyses.
Regarding surgical variables, decompression procedures (aHR, 1.11; 95% CI, 1.08–1.13; p < 0.001) and a principal diagnosis of spinal canal stenosis (aHR, 1.10; 95% CI, 1.08–1.11; p < 0.001) were each associated with greater FBSS risk. These associations remained consistent in subgroup analyses.
Several factors were protective against FBSS. Higher income (Q5 vs. Q1: aHR, 0.93; 95% CI, 0.92–0.95; p < 0.001), employee-based coverage (vs. Medical Aid: aHR, 0.76; 95% CI, 0.74–0.77; p < 0.001), and moderate physical activity (1000–1499 METs: aHR, 0.97; 95% CI, 0.96–0.98; p < 0.001) were each associated with reduced FBSS incidence. These protective trends were similarly observed in the subgroup models. The physical activity–FBSS relationship was nonlinear: while moderate activity was protective, the highest category (≥ 1500 METs: aHR, 1.04; 95% CI, 1.03–1.05; p < 0.001) was associated with a modest increase in adjusted risk.
Female sex was associated with a lower overall FBSS risk, although the direction of this association differed in the Disease-code subgroup, where women showed a higher incidence. Regional differences, alcohol consumption, chronic kidney disease, and surgical complexity at the index operation did not reach statistical significance in the primary analysis.
Discussion
In this nationwide population-based study of over one million lumbar spine surgery recipients, we found that FBSS affected approximately one in four patients when defined by diagnostic codes and reoperation. We further identified a range of preoperative factors—spanning demographics, comorbidities, health behaviors, and surgical characteristics—that were independently associated with FBSS development. These findings provide a comprehensive epidemiological foundation for understanding which patients are most vulnerable to unfavorable outcomes after lumbar surgery.
The observed 26.7% incidence aligns with the widely cited range of 10–40% reported in the literature5,though differences in study populations and case definitions preclude direct comparison. Importantly, the incidence varied depending on how FBSS was defined. This wide range emphasizes the sensitivity of FBSS prevalence estimates to definitional choices and highlights the ongoing need for standardized diagnostic criteria14,18. Under the Korean NHIS reimbursement framework, diagnostic code and reoperation after the index surgery can be longitudinally confirmed, suggesting that multi-criteria approaches like ours offer a more balanced assessment. Although Oh et al.19 reported high rates of gabapentinoid and opioid prescriptions in patients with FBSS, supporting the association between persistent neuropathic symptoms and FBSS, though their study did not examine preoperative risk factors.
Among the preoperative risk factors identified, several have well-established biological plausibility. Advancing age is recognized as a robust predictor of poor surgical outcomes in the spine literature20,21,likely reflecting cumulative degenerative changes, diminished tissue healing capacity, and a higher comorbidity burden. The association between peripheral polyneuropathy and FBSS risk was particularly strong (aHR 1.37), which is consistent with the notion that pre-existing neuropathic pain pathways may predispose patients to chronic postsurgical pain22,23. Similarly, preoperative electrodiagnostic testing—typically performed when significant neurological compromise is suspected—may serve as a surrogate marker for the severity of neural injury24, thereby identifying patients who are less likely to achieve complete symptom resolution; however, this interpretation should be cautious, as it would be an association that should be interpreted as a marker of the underlying clinical phenotype prompting the test rather than as a direct effect of the test itself.
Modifiable behavioral factors, including smoking and obesity, were independently associated with elevated FBSS risk. Notably, the elevated risk of obesity in the adjusted Cox model contrasts with the crude FBSS proportion, in which the obesity category appears similar to, or lower than, the other body mass index strata. This discrepancy is explained by confounding: obese patients in our cohort were, on average, younger and had a different comorbidity profile than non-obese patients, and the independent contribution of obesity emerges only after multivariable adjustment. Smoking impairs microvascular perfusion and disc nutrition, while obesity places excessive mechanical load on the spine and has been linked to heightened systemic inflammation; Both factors may impede postoperative recovery and tissue repair, promoting chronicity of symptoms25,26. These findings are consistent with existing evidence from smaller clinical cohorts and add population-level support for preoperative lifestyle modification as a potentially meaningful strategy.
The association between metabolic comorbidities—hypertension and diabetes—and FBSS risk likely operates through shared pathways, including microvascular dysfunction, impaired wound healing, and subclinical neuropathy27. Diabetes in particular is known to exacerbate peripheral nerve damage and delay tissue repair28, both of which may promote persistent pain after spinal surgery. These observations reinforce the importance of optimizing medical management of chronic diseases before proceeding with elective lumbar operations.
From a surgical perspective, decompression procedures were associated with a higher FBSS risk compared with other surgical types, and spinal canal stenosis as the principal indication also conferred elevated risk. A previous study conducted by Hajilo et al.29also highlighted fusion surgery and prior operations as major FBSS contributors. In our cohort, increased risk associated with decompression and stenosis likely reflects the advanced degenerative pathology encountered in these patients and the limited capacity of surgery to fully reverse long-standing neural compression30.
The protective effects of higher socioeconomic status and employment-based insurance are noteworthy and may be mediated by better access to timely medical care, more favorable working conditions, and greater health literacy31,32. Earlier disease recognition and access to optimal perioperative management may enable these patients to achieve better surgical outcomes and lower rates of chronic postsurgical pain.
Several considerations regarding the scope of our findings merit emphasis. Our analysis represents a population-level characterization of FBSS risk factors across a clinically heterogeneous lumbar surgical cohort; it should be interpreted as identifying system-level risk indicators rather than universal predictors applicable to every diagnostic subgroup. Prior studies have shown that risk factor profiles for revision surgery differ across clinical entities, including disc herniation, degenerative spondylolisthesis, and osteoporotic compression fracture11–13. Although our Cox model included principal diagnosis as a covariate, formal diagnosis-stratified analyses with full covariate-by-diagnosis interactions would provide additional granularity. Such analyses are prioritized in our planned follow-up NHIS investigation.
In addition, several aspects of this study merit emphasis. The use of a nationally representative database encompassing over one million surgical patients provides robust statistical power and broad generalizability within the Korean population. By applying multiple operational definitions of FBSS, we were able to characterize incidence from complementary angles and demonstrate how definitional choices influence epidemiological estimates. The availability of linked health screening data further allowed adjustment for behavioral risk factors that are typically unavailable in claims-based studies.
However, outcome misclassification is an intrinsic concern of claims-based FBSS definitions, and warrants explicit discussion. ICD-10 M96.1 (post-laminectomy syndrome) has not been formally validated as a standalone FBSS code, and subsequent lumbar surgical claims may reflect indications other than persistent post-surgical pain—including planned staged surgery, adjacent segment disease, recurrent disc herniation, infection, trauma, implant removal, or deformity progression. Several features of our design partially attenuate these concerns: spinal cord neoplasm was excluded at cohort entry; infection, compression fracture, lumbar fracture, and multiple fracture were modeled as separate categories within the principal-diagnosis covariate, so that reoperations occurring predominantly within these diagnostic groups contribute their effect to the corresponding covariate rather than to overall FBSS risk; and patients with pre-existing M96.1 at the index surgery were excluded to preserve incident outcome ascertainment. Nevertheless, we cannot fully exclude residual misclassification, and a series of sensitivity definitions—M96.1 alone, reoperation with a minimum latency of 90 or 180 days, reoperation combined with pain-related medication or procedure claims, and reoperation with infection/fracture/tumor indications removed—would be required to directly quantify its magnitude in future studies.
This study also has limitations that should be considered. First, the retrospective observational design precludes causal inference; the identified associations should be interpreted as predictive rather than etiological. Second, claims-based data lack granularity regarding surgical technique, implant specifications, intraoperative events, and detailed spine-specific disease severity variables—all of which may substantially influence postoperative outcomes and FBSS risk28,30,32,33. In particular, the database could not capture stenosis severity, spondylolisthesis grade, spinal instability, sagittal alignment, number of decompressed or fused levels, fusion approach, biologic augmentation, implant type, dural tears, blood loss, or perioperative complications. In addition, several surgical categories—particularly the “other surgeries” group—comprised heterogeneous procedures with potentially distinct clinical indications and postoperative risk profiles. Therefore, procedure-specific interpretations should be made cautiously. Further, comorbidity ascertainment relied on single-claim ICD-10 coding without additional validation through repeated claims, medication co-occurrence, or laboratory thresholds, which may introduce non-differential misclassification relative to more conservative pharmacoepidemiologic definitions. Collectively, these limitations may have resulted in substantial residual and unmeasured confounding, and therefore, the identified associations should not be interpreted as causal estimates. Third, patient-reported outcome measures such as pain severity scores, functional status, and psychological profiles were not available, and these factors likely play a meaningful role in FBSS development. Fourth, inclusion was restricted to individuals with available national health screening data, because several key lifestyle covariates were derived from the screening database. This may have introduced selection bias toward patients who were more health-conscious or more engaged with healthcare services, potentially affecting the cohort’s representativeness. Moreover, the actual screening-to-surgery interval varies across participants within this window, and short-term lifestyle changes prior to surgery could not be captured. Fifth, several aspects of our analysis are intrinsically tied to the Korean healthcare and demographic context and have limited direct generalizability to other settings. Korean-specific administrative and healthcare-system structures define insurance type, region of residence, and hospital area categories, and the corresponding hazard ratios should not be transposed directly to other healthcare-financing systems or geographic contexts. Population characteristics, including ethnic homogeneity and country-specific patterns of spinal surgery utilization, further constrain the external applicability of absolute incidence estimates. In contrast, the observed associations of biological and behavioral risk factors, including advancing age, peripheral polyneuropathy, diabetes, hypertension, smoking, and obesity, with FBSS risk are consistent with prior international literature and are likely to have broader generalizability. However, external validation in independent cohorts remains warranted. Sixth, we did not perform various sensitivity analyses with alternative washout periods, latency windows for reoperation, and diagnosis-stratified models. Seventh, formal proportional hazards diagnostics and competing-risk analyses were not performed. Because death may represent a clinically relevant competing event in this elderly surgical population, particularly for reoperation outcomes, the Cox model estimates should be interpreted with appropriate caution.
Conclusion
This large-scale national cohort study established that FBSS develops in a substantial proportion of lumbar spine surgery patients. Multiple preoperative factors were independently associated with FBSS risk, including older age, tobacco use, obesity, hypertension, diabetes, peripheral polyneuropathy, decompression surgery, and spinal canal stenosis as the surgical indication. Conversely, higher socioeconomic status and moderate physical activity appeared protective. These findings support the incorporation of systematic preoperative risk stratification into clinical decision-making for lumbar spine surgery. Future prospective investigations incorporating detailed clinical and patient-reported outcome data are warranted to refine these risk estimates and to evaluate whether targeted preoperative interventions can meaningfully reduce FBSS incidence.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This study utilized claims data provided by the Korean National Health Insurance Service (NHIMC 2024-CR-031).
Author contributions
D.P.: conceptualized the study, designed the methodology, interpreted the data, and wrote the original draft; W.-R.L.: curated the data, performed the statistical analyses, and accessed the database; M.B.: contributed to data curation and validation; S.-J.P.: provided clinical expertise on surgical variables and reviewed the manuscript; J.H.K.: supervised the study, contributed to data interpretation, and critically revised the manuscript. H.S.K.: supervised the study, contributed to study conceptualization, and critically revised the manuscript. All authors reviewed and approved the final version of the manuscript.
Funding
This work was supported by the 2024 Annual Clinical Research Fund of the National Health Insurance Service Ilsan Hospital (NHIMC 2024-CR-031).
Data availability
The data used in this study are not publicly accessible due to the privacy and ethical restrictions governing the Korean NHIS data sharing system. Access is limited to authorized researchers working within the NHIS internal network (W.-R.L.).
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval and informed consent
As this study utilized exclusively de-identified administrative data, the Institutional Review Board of the National Health Insurance Service Ilsan Hospital (NHIMC 2024-CR-031) waived the requirement for ethical approval. The Institutional Review Board of the National Health Insurance Service Ilsan Hospital also waived informed consent due to the retrospective nature of this study. The study was conducted in accordance with the Declaration of Helsinki.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Jong Hun Kim, Email: rehappydoc@gmail.com.
Hyoung Seop Kim, Email: jh7521@naver.com.
References
- 1.Collaborators, G. B. D. O. M. D. Global, regional, and national Burden of other musculoskeletal disorders, 1990–2020, and projections to 2050: A systematic analysis of the Global Burden of Disease Study 2021. Lancet Rheumatol.5, e670–e682. 10.1016/S2665-9913(23)00232-1 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hartvigsen, J. et al. What low back pain is and why we need to pay attention. Lancet391, 2356–2367. 10.1016/S0140-6736(18)30480-X (2018). [DOI] [PubMed] [Google Scholar]
- 3.Daniell, J. R. & Osti, O. L. Failed back surgery syndrome: A review article. Asian Spine J.12, 372–379. 10.4184/asj.2018.12.2.372 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Christelis, N. et al. Persistent spinal pain syndrome: A proposal for failed back surgery syndrome and ICD-11. Pain Med.22, 807–818. 10.1093/pm/pnab015 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Stanton, E. W., Chang, K. E., Formanek, B., Buser, Z. & Wang, J. The incidence of failed back surgery syndrome varies between clinical setting and procedure type. J. Clin. Neurosci.103, 56–61. 10.1016/j.jocn.2022.06.027 (2022). [DOI] [PubMed] [Google Scholar]
- 6.Pandelani, F. F., Nyalunga, S. L. N., Mogotsi, M. M. & Mkhatshwa, V. B. Chronic pain: Its impact on the quality of life and gender. Front. Pain Res.4, 1253460. 10.3389/fpain.2023.1253460 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sylwander, C., Haglund, E., Larsson, I. & Andersson, M. L. E. Health-related quality of life, lifestyle habits and chronic pain in individuals with knee pain - A 2-year follow-up study. Scand. J. Prim. Health Care.43, 422–433. 10.1080/02813432.2025.2452916 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sebaaly, A., Lahoud, M. J., Rizkallah, M., Kreichati, G. & Kharrat, K. Etiology, evaluation, and treatment of failed back surgery syndrome. Asian Spine J.12, 574–585. 10.4184/asj.2018.12.3.574 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Alizadeh, R. & Sharifzadeh, S. R. Pathogenesis, etiology and treatment of failed back surgery syndrome. Neurochirurgie68, 426–431. 10.1016/j.neuchi.2021.09.005 (2022). [DOI] [PubMed] [Google Scholar]
- 10.Park, D. et al. Impact of postoperative chronic back pain syndrome on the incidence of psychiatric and brain disorders, and all-cause mortality. J. Pain Res.19, 582506. 10.2147/JPR.S582506 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ton, A. et al. Predictors and postoperative complication risks for revision discectomies following primary lumbar microdiscectomy. Glob. Spine J.15, 3273–3279. 10.1177/21925682251323274 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen, Y. et al. A systematic review and meta-analysis of risk factors for reoperation after degenerative lumbar spondylolisthesis surgery. BMC Surg.23, 192. 10.1186/s12893-023-02082-8 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lian, Y. S., Chen, L. P. & Chen, J. Y. Risk factors for new vertebral fracture after percutaneous vertebroplasty or kyphoplasty for osteoporotic vertebral compression fractures (OVCFs): an updated systematic review and meta-analysis. Eur. Spine J.34, 5116–5125. 10.1007/s00586-025-09082-7 (2025). [DOI] [PubMed] [Google Scholar]
- 14.Yeo, J. Failed back surgery syndrome-terminology, etiology, prevention, evaluation, and management: a narrative review. J. Yeungnam Med. Sci.41, 166–178. 10.12701/jyms.2024.00339 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Baek, J. Y., Lee, E., Jung, H. W. & Jang, I. Y. Geriatrics fact sheet in Korea 2021. Ann. Geriatr. Med. Res.25, 65–71. 10.4235/agmr.21.0063 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Korean National Health Insurance Service. Major Surgical Statistics Annual Report. 2024. (2023). Available from: https://www.nhis.or.kr/nhis/together/wbhaec06800m01.do. Accessed February 11, 2025.
- 17.Cheol Seong, S. et al. Data resource profile: The national health information database of the National Health Insurance Service in South Korea. Int. J. Epidemiol.46, 799–800. 10.1093/ije/dyw253 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Miekisiak, G. Failed back surgery syndrome: No longer a surgeon’s defeat-a narrative review. Medicina (Kaunas)10.3390/medicina59071255 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Oh, J. & Yeo, J. Trends in the prescription of opioids and gabapentinoids in patients with failed back surgery syndrome in Korea: A population-based study. Korean J. Pain37, 73–83. 10.3344/kjp.23246 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Parenteau, C. S., Lau, E. C., Campbell, I. C. & Courtney, A. Prevalence of spine degeneration diagnosis by type, age, gender, and obesity using Medicare data. Sci. Rep.11, 5389. 10.1038/s41598-021-84724-6 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cole, K. L. et al. Frailty predicts worse outcomes for spine surgery patients with interhospital transfer status: Analysis of 295,875 patients from the National Surgical Quality Improvement Program (NSQIP) 2015–2019. Clin. Neurol. Neurosurg.224, 107519. 10.1016/j.clineuro.2022.107519 (2023). [DOI] [PubMed] [Google Scholar]
- 22.Baron, R., Binder, A. & Wasner, G. Neuropathic pain: Diagnosis, pathophysiological mechanisms, and treatment. Lancet Neurol.9, 807–819. 10.1016/S1474-4422(10)70143-5 (2010). [DOI] [PubMed] [Google Scholar]
- 23.Shamji, M. F. & Shcharinsky, A. Use of neuropathic pain questionnaires in predicting persistent postoperative neuropathic pain following lumbar discectomy for radiculopathy. J. Neurosurg. Spine. 24, 256–262. 10.3171/2015.4.SPINE141310 (2016). [DOI] [PubMed] [Google Scholar]
- 24.Lee, J. H. & Lee, S. H. Clinical usefulness of electrodiagnostic study to predict surgical outcomes in lumbosacral disc herniation or spinal stenosis. Eur. Spine J.24, 2276–2280. 10.1007/s00586-014-3475-8 (2015). [DOI] [PubMed] [Google Scholar]
- 25.Siccoli, A., Staartjes, V. E., Klukowska, A. M., Muizelaar, J. P. & Schroder, M. L. Overweight and smoking promote recurrent lumbar disk herniation after discectomy. Eur. Spine J.31, 604–613. 10.1007/s00586-022-07116-y (2022). [DOI] [PubMed] [Google Scholar]
- 26.Lener, S., Wipplinger, C., Hartmann, S., Thome, C. & Tschugg, A. The impact of obesity and smoking on young individuals suffering from lumbar disc herniation: A retrospective analysis of 97 cases. Neurosurg. Rev.43, 1297–1303. 10.1007/s10143-019-01151-y (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhang, X. et al. Association of diabetes mellitus with postoperative complications and mortality after non-cardiac surgery: A meta-analysis and systematic review. Front Endocrinol (Lausanne)13, 841256. 10.3389/fendo.2022.841256 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Steinmetz, M. P. et al. Adverse impact of diabetes on spine fusion and patient-reported outcomes: A systematic review and meta-analysis. Spine (Phila Pa 1976)50, E347–E360. 10.1097/BRS.0000000000005396 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hajilo, P., Imani, B., Zandi, S., Mehrafshan, A. & Khazaei, S. Risk factors analysis and risk prediction model for failed back surgery syndrome: A prospective cohort study. Heliyon11, e40607. 10.1016/j.heliyon.2024.e40607 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Dimitriou, D. et al. A simple preoperative score predicting failure following decompression surgery for degenerative lumbar spinal stenosis. Spine (Phila Pa 1976)48, 610–616. 10.1097/BRS.0000000000004584 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Yap, Z. L., Summers, S. J., Grant, A. R., Moseley, G. L. & Karran, E. L. The role of the social determinants of health in outcomes of surgery for low back pain: a systematic review and narrative synthesis. Spine J.22, 793–809. 10.1016/j.spinee.2021.11.013 (2022). [DOI] [PubMed] [Google Scholar]
- 32.Touponse, G., Li, G., Rangwalla, T., Beach, I. & Zygourakis, C. Socioeconomic effects on lumbar fusion outcomes. Neurosurgery92, 905–914. 10.1227/neu.0000000000002322 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hejazi-Garcia, C. et al. The social deprivation index and lumbar spine surgery outcomes. J. Neurosurg. Spine42, 579–588. 10.3171/2024.11.SPINE241048 (2025). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data used in this study are not publicly accessible due to the privacy and ethical restrictions governing the Korean NHIS data sharing system. Access is limited to authorized researchers working within the NHIS internal network (W.-R.L.).