Hip osteoarthritis (OA) and lumbar spine pathology (LSP) frequently coexist, sharing common risk factors, such as age, obesity, and prior trauma. Up to 50% of patients undergoing total hip arthroplasty (THA) report concurrent lower back pain [1]. The hip and spine are biomechanically linked; pathology in one region can increase stress and induce compensatory or abnormal motion in the other [2]. These interactions affect the global sagittal balance (GSB), which is typically assessed by the sagittal vertical axis, or the intersection between a plumb line dropped from the center of the C7 vertebra and the posterior superior endplate of S1. Ideally, this line should pass just posterior to the femoral heads [3]. In individuals with preserved flexibility, the hip and spine can compensate for pathologic malalignment, thereby reducing pain and functional limitation. However, as stiffness and degeneration progress, compensatory mechanisms fail, resulting in worsening symptoms. On initial presentation, patients may exhibit spine-predominant, hip-predominant, or mixed symptom patterns, necessitating a comprehensive evaluation of both regions.

Both THA and lumbar fusion (LF) can substantially alter GSB. Consequently, in patients with coexisting disease, determining the optimal sequence of surgical intervention is critical. Evidence addressing the sequence of surgery and its impact on outcomes remains limited, and no clear consensus exists. A large national database study suggested that performing THA first in patients with concurrent spine disease was associated with fewer subsequent surgeries, lower rates of postoperative instability, and reduced opioid use [1]. More recently, a systematic review found insufficient high-quality evidence, noting that although some data support performing THA first, the literature remains inconclusive [4].

The objective of this study was to compare lumbar spine revision rates and postoperative complications among patients who underwent THA before or after LF, as well as among those who underwent LF but not THA. Additionally, we evaluated outcomes of patients who underwent both operations relative to those who underwent LF alone despite known concurrent hip OA.

A retrospective cohort study was conducted using the PearlDiver (PearlDiver Technologies Inc., Colorado Springs, CO, USA) national database, which contains over 170 million multi-payer administrative records from the United States, covering the period from 2010 to the first quarter of 2023. The database has been widely used in orthopedic and spine research [5,6] and is exempt from Institutional Review Board approval.

Patients were identified using International Classification of Diseases, Ninth and 10th Revisions and Current Procedural Terminology codes. All patients who underwent index LF surgery, regardless of approach, were included. To ensure cohort homogeneity, cases involving lumbosacral or pelvic fixation were excluded. Only adult patients aged 18 years or older who remained active in the database for at least 90 days were included in the analysis. Patients with a history of hip fracture fixation were eligible for inclusion. For revision rate analyses at 3 and 5 years, only patients who were active in the database for the corresponding duration were included to minimize bias due to censoring [7]. Patients with a history of spine trauma, infection, or malignancy at or before the time of LF were excluded. Patients were divided into three cohorts: (1) OA_LF: patients with hip OA who underwent LF only; (2) THA_LF: patients who underwent THA first and then LF; and (3) LF_THA: patients who underwent LF first and then THA.

Baseline demographic and clinical variables included age, sex, insurance plan, region, Elixhauser comorbidity index (ECI), and diabetes status. Outcomes assessed included lumbar revision surgery rates at 3 and 5 years, as well as 90-day postoperative complications, such as emergency department (ED) visits, hospital readmission, delirium, sepsis, stroke, superficial and deep wound infection, wound disruption, deep vein thrombosis (DVT), pulmonary embolism (PE), hematoma, and pneumonia.

Statistical analysis was performed using R Studio RStudio (Posit Software; PBC, Boston, MA, USA) within the Bellwether software platform of PearlDiver (PearlDiver Technologies Inc., Colorado Springs, CO, USA). Continuous variables were reported as means±standard deviation and compared using analysis of variance. Categorical variables were presented as frequencies and percentages and compared using the chi-square test. Adjusted odds ratios (aORs) with 95% confidence intervals were calculated using multivariable binary logistic regression, adjusting for age, sex, ECI, insurance plan, region, and diabetes status. Multivariable regression models were run twice: first, including all three cohorts with OA_LF as the reference group, and second, including only LF_THA and THA_LF, with THA_LF as the reference. A p-value <0.05 was considered indicative of statistical significance.

In this large national database study of 70,265 patients, patients with hip OA who underwent both LF and THA had higher odds of lumbar revision surgery at both 3- and 5-year follow-ups. Among patients who underwent both procedures, those who underwent LF before THA were at a greater risk of revision lumbar surgery compared with those who underwent THA first. In addition, the LF_THA cohort demonstrated higher odds of ED visits, readmission, superficial wound infection, PE, and hematoma within 90 days, compared with the THA_LF and OA_LF cohorts.

The hip and spine function as a biomechanical unit, and pathology in one can alter alignment and loading patterns in the other. This interdependence, often referred to as the spine-hip or hip-spine syndrome, explains the frequent coexistence of these disorders [3,8]. Surgical correction at either site leads to abrupt biomechanical changes, which may compromise outcomes if the compensatory capacity of the other segment is impaired. For example, prior studies suggest that a stiff or fused spine limits pelvic motion, increasing the risk of prosthetic impingement and dislocation after THA [9]. However, there is little evidence of primary spine surgery failure due to preexisting hip OA. We hypothesize that hip OA primarily affects spinal mechanics through chronic rather than acute biomechanical adaptations. Altered pelvic tilt, posture, and gait patterns associated with hip OA may increase mechanical stress on adjacent segments, accelerating degeneration over time but not directly compromising the integrity of fused spinal segments. Therefore, hip OA may contribute to adjacent segment disease rather than acute fusion failure. In contrast, THA alters spinopelvic alignment more abruptly, potentially affecting spinal balance and predisposing to mechanical complications. Although THA has been associated with changes in spinal alignment, prior studies have shown that back pain tends to improve after THA [10–12]. Collectively, these findings support performing THA before LF when both pathologies are present, as worsening of hip OA after spinal fusion appears more likely than worsening of back pain after THA.

Our results align with previous studies that favor performing THA before LF. A retrospective cohort study reported significantly lower rates of subsequent surgery in patients who underwent THA first [13]. Similarly, a national database analysis found THA-first sequencing in patients requiring both surgeries was associated with reduced opioid use, hip dislocation, and need for further surgery on the spine or hip [1]. While that study did not identify significant differences in the rates of infection or thromboembolic complications, our findings expand on the literature by demonstrating distinct postoperative outcome profiles, including higher risks of wound and thromboembolic complications in LF_THA patients. Interestingly, patients in our LF_THA cohort were generally younger and had fewer comorbidities, which may reflect lower tolerance for spinal symptoms compared with hip OA. However, future research should focus on gathering more empirical data on this aspect. A notable observation in our analysis was the reversal of odds of lumbar revision when the interval between LF and THA was less than 2 years. This suggests that temporal proximity of the procedures may influence biomechanical adaptation and revision risk. Future studies incorporating radiographic and spinopelvic alignment data are needed to validate this relationship.

The present study has some limitations. Administrative databases such as PearlDiver lack detailed clinical information, including the severity of hip or lumbar pathology, the number of fused levels, surgical indications, and spinopelvic parameters. It is possible that patients who underwent LF but not THA had milder hip OA than patients who underwent both surgeries. Moreover, variations in surgical technique, perioperative protocols, and surgeon experience could not be controlled for. Furthermore, patient-specific unmeasured confounders included the absence of presurgical and postsurgical functional status and surgical complexity. Lastly, the database does not report the causes of revision, such as pseudarthrosis, adjacent segment disease, or infection. Despite these limitations, the large nationally representative cohort strengthens the generalizability of the findings. The study demonstrated markedly higher odds of lumbar revision (aOR >2) among patients undergoing both THA and LF, with the highest risk observed when THA preceded LF at 5-year follow-up.

Based on these findings, patients requiring both LF and THA should be counseled about their elevated risk of postoperative complications and LF reoperation compared with patients undergoing LF alone. When feasible, performing THA before LF may reduce complications and revision risk. Future research should focus on biomechanical modeling to help minimize the impact on the spine when operating on the hip and vice versa. Future randomized controlled trials are warranted to provide higher-level evidence, along with studies adequately powered to detect differences in the severity of pathology rather than the binary diagnostic results (diagnosis/no diagnosis).

Patients who underwent THA before LF had lower odds of lumbar revision surgery compared with those who underwent LF first. These findings support the notion that in patients requiring both procedures for concurrent hip OA and LSP, performing THA prior to LF may lead to more favorable outcomes.