Optimal surgery sequence in the treatment of degenerative hip-spine syndrome: a propensity score-based inverse probability of treatment weighting analysis

Abstract

Background

The coexistence of spinal degenerative diseases and hip joint degeneration is common among middle-aged and elderly individuals, causing significant suffering and economic burden for patients. Total hip arthroplasty (THA) and lumbar fusion (LF) are primary treatment options for this combined condition, but the impact of the surgical sequence on patient outcomes remains unclear. Hence, this study aims to evaluate the effects of the surgical sequence of THA and LF on symptom recovery, changes in sagittal spine-pelvis parameters, and the incidence of long-term complications in patients.

Methods

A retrospective analysis was conducted on 104 patients diagnosed with hip-spine syndrome (HSS) who underwent THA and LF at the Chinese PLA General Hospital. IPTW was implemented to control potential confounding factors. The impact of surgical sequence on clinical function scores, radiological parameters, and long-term complications was evaluated before and after conducting IPTW. Patients who underwent THA surgery first, followed by LF surgery, were categorized as the THA→LF group; conversely, those who underwent LF surgery first, followed by THA surgery, were categorized as the LF→THA group. Clinical function scores included the Oswestry Disability Index (ODI), Japanese Orthopedic Association (JOA) score, and Harris Hip Score for both groups. Long-term complications assessed in this study included instances of Proximal Junctional Kyphosis (PJK), internal fixation loosening or failure, as well as hip prosthesis dislocation. Radiological parameters included Pelvic Tilt (PT), Pelvic Incidence (PI), Pelvic Incidence minus Lumbar Lordosis (PI-LL), Sacral Slope (SS), and Lumbar Lordosis (LL).

Results

Before conducting IPTW, there were significant differences between the two groups across multiple variables, including age (P = 0.035), fixation stage (P = 0.042), preoperative PT (P = 0.005), preoperative PI-LL (P = 0.004), and preoperative LL (P = 0.040). After conducting IPTW, all baseline data variables had P-values greater than 0.50, indicating that the baseline characteristics between the two groups were comparable. Following IPTW, the study found significant improvements in postoperative ODI, JOA score, and Harris Hip Score for both groups (P < 0.001), indicating that both surgical sequences were effective in enhancing clinical functional activity. However, there were no significant differences between the groups. Additionally, there was no significant difference in the long-term complication rates between the two groups. Regarding radiological parameters, the PT in the THA→LF group was significantly lower than that in the LF→THA group (P = 0.043), while the SS was significantly higher (P = 0.028) at the last follow up after conducting IPTW.

Conclusion

Compared to preoperative assessments, both surgical sequences significantly improve postoperative clinical function scores and radiological parameters for patients. When comparing between the two groups, the surgical sequence exhibits equivalent effects on functional improvement and complication rates. The radiological outcomes indicate that the THA→LF group shows more pronounced effects on PT and SS, suggesting a more favorable impact on pelvic alignment in patients undergoing this sequence. The use of the IPTW method successfully eliminates differences in baseline characteristics, enhancing the reliability of the results. Future research is warranted to further explore these findings and their implications for surgical decision-making in HSS patients.

Introduction

Hip-spine syndrome (HSS) is a complex clinical condition characterized by reciprocal biomechanical and pathological interactions between hip and spinal disorders, leading to debilitating pain, functional impairment, and diminished quality of life [1]. Total hip arthroplasty (THA) and lumbar fusion (LF) represent effective surgical interventions for HSS management. THA effectively alleviates hip-related pain, restores joint mobility, and mitigates compensatory spinal strain, while LF stabilizes spinal segments, corrects deformities, and addresses neurological compression [2]. Despite their individual benefits, the optimal surgical sequence for patients requiring both procedures remains contentious [3].

Emerging evidence suggests divergent outcomes based on procedural prioritization. For example, studies have proposed that prioritizing THA may reduce subsequent spinal surgery demands, opioid dependence, and postoperative functional instability [4]. Furthermore, initial LF has been associated with elevated hip dislocation risk and reoperation of THA following subsequent THA [5, 6]. In contrast, Eneqvist et al. [7] reported superior health-related quality-of-life improvements in patients undergoing LF prior to THA, and Parilla et al. [8] found no significant difference in dislocation rates between sequencing groups. Hence, critical limitations persist in current evidence. In detail, the above studies derive from real-world datasets hampered by small sample sizes, variable imbalances between groups, and heterogeneous surgical protocols, undermining result reliability and generalizability [3]. Although randomized controlled trials (RCTs) remain the gold standard for therapeutic evaluation, ethical and logistical constraints may have precluded RCTs investigating the outcome of THA-LF surgery sequencing. Consequently, there is an urgent need for robust statistical methodologies that minimize confounding and emulate randomized conditions in observational data.

Under such circumstance, inverse probability of treatment weighting (IPTW) offers a compelling solution. By assigning propensity score-derived weights to patients, IPTW balances baseline clinical characteristics between THA→LF and LF→THA groups, effectively simulating randomization [9]. This approach reduces selection bias inherent in real-world data while maximizing statistical power through efficient sample utilization [10]. Despite its advantages, IPTW has not yet been applied to investigate surgical sequencing in HSS—a knowledge gaps this study seeks to address.

Therefore, this study employed IPTW methodology to evaluate the impact of THA-LF surgical sequence on symptom resolution, spine-pelvis parameter evolution, and long-term complication profiles in HSS patients.

Patients and methods

Patients

This study retrospectively enrolled a cohort of patients diagnosed with HSS at the PLA General Hospital who underwent both posterior lumbar discectomy with cage implantation, autologous bone graft interbody fusion, and pedicle screw internal fixation surgery and posterolateral approach total hip arthroplasty (THA) using cementless implant system between January 2011 and June 2021. The enrolled patients were comprehensively diagnosed based on symptoms, physical signs and radiological findings. Specifically, patients indicated for lumbar fusion surgery were diagnosed with degenerative spinal conditions including lumbar spondylolisthesis, lumbar spinal stenosis or degenerative scoliosis, while those indicated for THA were diagnosed with hip osteoarthritis. Patients who underwent THA first, followed by lumbar fusion (LF), were categorized as the THA→LF group, while those who underwent LF first, followed by THA, were categorized as the LF→THA group. The study was approved by the Ethics Committee of the First Medical Center of the PLA General Hospital (approval number: S2024-556-01). All procedures adhered to the Declaration of Helsinki and relevant institutional and national guidelines.

Inclusion and exclusion criteria

The inclusion criteria for this study were as follows: (1) an interval of 3 to 9 months between the two surgeries; (2) cases selected through the electronic medical record system with a follow-up period of two years or more after the last surgery; and (3) availability of complete follow-up data on symptom changes and imaging. Patients were excluded if they met any of the following criteria: (1) aged less than 40 years; (2) diagnosed with congenital or idiopathic scoliosis; (3) exhibited poor recovery after traumatic spinal or joint injuries; (4) had secondary spinal-pelvic-joint deformities due to conditions such as spinal tuberculosis or spinal tumors; (5) were adults with congenital hip dysplasia or congenital scoliosis; (6) presented with idiopathic scoliosis or neuromuscular spinal deformities causing severe imbalance; (7) had incomplete or poor-quality imaging data, or imaging that did not include the femoral head or the uppermost fixed vertebra, preventing accurate measurement of spinal-pelvic parameters; or (8) could not be contacted after at least three follow-up attempts or had passed away.

Collection of baseline characteristics

Baseline characteristics were collected from the electronic medical record system, including age, sex, height, weight, body mass index (BMI, kg/m²), joint laterality (unilateral/bilateral), spinal surgical levels, preoperative functional symptoms including preoperative Oswestry Disability Index (ODI) score, Japanese Orthopaedic Association (JOA) score, and Harris Hip Score, and preoperative radiological parameters, such as pelvic tilt (PT), pelvic incidence (PI), pelvic incidence minus lumbar lordosis (PI-LL), sacral slope (SS), and lumbar lordosis (LL). All radiological measurements were performed by trained observers using Surgimap software (Nemaris Inc., version 2.3.2.1, New York). Radiological changes were assessed using lumbar lateral X-rays taken preoperatively, one week postoperatively, and during follow-up.

Measurement methods for radiological parameters

PI was defined as the angle between the line connecting the femoral head center to the midpoint of the sacral endplate and the perpendicular to the sacral endplate. PT was defined as the angle between the line connecting the femoral head center to the midpoint of the sacral endplate and the vertical line. SS was defined as the angle between the sacral endplate and the horizontal line [11]. LL was defined as the angle between the superior endplate of L1 and the sacral endplate [12]. All angles were measured using the Cobb method to ensure accuracy and consistency.

Clinical functional prognosis

Spinal symptoms were evaluated using the ODI and JOA, while hip symptoms and functional activity were assessed using the Harris Hip Score [13,14,15]. The ODI is a self-reported questionnaire that evaluates the impact of low back pain on daily activities, such as pain intensity, personal care, lifting, walking, sitting, standing, sleeping, social life, and traveling [13]. It consists of 10 sections, each scored from 0 to 5, with higher scores indicating greater disability. The JOA score is a clinician-administered assessment tool used to evaluate the severity of symptoms and functional impairment in patients with spinal conditions [16], particularly lumbar spine disorders. It includes domains such as subjective symptoms (e.g., low back pain, leg pain, and gait), clinical signs (e.g., sensory and motor function), and restrictions in daily activities (e.g., turning over, standing, and walking). Scores range from 0 to 29, with lower scores indicating more severe impairment. Changes in symptoms before and after surgery were recorded and analyzed through the electronic medical record system and telephone surveys.

Long-term complications

Long-term postoperative complications, including proximal junctional kyphosis (PJK), loosening or failure of internal fixation, and hip prosthesis dislocation, were assessed during the most recent telephone follow-up in June 2023. PJK was defined as a sagittal Cobb angle of ≧ 10 degree between the lower endplate of the upper instrumented vertebra (UIV) and the upper endplate of the second vertebra above it (UIV + 2), with an increase of at least 10 degree compared to preoperative measurements [17]. Lumbar internal fixation loosening was diagnosed based on lumbar X-rays, identified by a radiolucent zone > 1 mm around the pedicle screws or the “double halo” sign [18]. Hip prosthesis dislocation, a common complication after primary THA, was detected using standard pelvic X-rays.

Implement of inverse probability of treatment weighting (IPTW)

IPTW was employed to balance baseline characteristic differences between the THA→LF and LF→THA groups [10]. A logistic regression model was used to estimate propensity scores (PS), with covariates including sex, age, height, weight, BMI, laterality, number of spinal levels, and multiple clinical scores (preoperative ODI, JOA, and Harris scores) as well as preoperative radiological parameters. Based on the estimated PS, inverse probability weights were calculated for each patient, with weights of 1/PS for the treatment group and 1/(1-PS) for the control group. Weights were rounded to integers for simplification. The weighted dataset was constructed using the “svydesign” function in the R survey package, and a weighted baseline characteristics table was generated using the “svyCreateTableOne” function. The standardized mean difference (SMD) was used to assess balance between groups after weighting.

Statistical analysis

Quantitative data following a normal distribution were presented as mean ± standard deviation (SD), while categorical data were expressed as counts and percentages. For normally distributed data with homogeneity of variance, between-group comparisons were performed using t-tests; otherwise, the Mann-Whitney U test was used. For repeated measures, repeated-measures ANOVA was applied if normality and sphericity assumptions were met, with Bonferroni or LSD correction for between-group comparisons; otherwise, non-parametric methods such as the Friedman test or generalized estimating equations were used. Radiological changes were assessed using lumbar lateral X-rays taken preoperatively, one week postoperatively, and during follow-up. Curves depicting radiological changes were plotted, and curve fitting was performed using LOESS. All statistical analyses were conducted using R software (version 4.1.2). A p-value < 0.05 was considered statistically significant.

Results

Baseline characteristics before and after conducting IPTW

In the clinical baseline data of patients before conducting IPTW, significant differences were observed between the THA→LF group and the LF→THA group across multiple variables, including age (P = 0.035), spinal surgical levels (P = 0.042), preoperative PT (P = 0.005), preoperative PI-LL (P = 0.004), and preoperative LL (P = 0.040) (Supplementary Table 1). All these variables showed statistically significant differences, indicating an imbalance in baseline data, which could influence surgical selection and postoperative outcomes in our study.

Notably, after conducting IPTW, the clinical baseline data of patients showed significant improvement in balance between the THA→LF group (85 cases) and the LF→THA group (98 cases). The P-values for all variables were greater than 0.50 (Table 1), indicating no statistically significant differences in baseline data between the two groups. More importantly, changes in the SMD further supported the improved balance between the groups: except for fixation stage (SMD: 0.412) and preoperative PT (SMD: 0.315), the SMDs for all other variables were below 0.20. Specifically, the SMD for age decreased from 0.422 before conducting IPTW to 0.079 after conducting IPTW; fixation stage decreased from 0.830 to 0.412; Preoperative PT decreased from 0.549 to 0.315; Preoperative PI-LL decreased from 0.584 to 0.163; and Preoperative LL decreased from 0.415 to 0.046. These results demonstrate that the differences between the two groups were significantly reduced after inverse probability weighting, and the baseline characteristics were well-balanced. This suggests that the IPTW method effectively eliminated confounding factors between the two groups, providing a more reliable foundation for subsequent analysis.

Table 1 Clinical baseline characteristics after IPTW

Postoperative clinical activity function scores and long-term complications before and after conducting IPTW

Both before and after conducting IPTW, the postoperative ODI scores were significantly lower than the preoperative scores (P < 0.001), the postoperative JOA scores were significantly higher than the preoperative JOA scores (P < 0.001), and the postoperative Harris scores were significantly higher than the preoperative Harris scores (P < 0.001). These results suggest that both surgical sequences significantly improved patients’ clinical activity function. Before conducting IPTW, there were no significant differences in postoperative ODI scores (P = 0.299) or JOA scores (P = 0.987) between the two groups (Supplementary Table 2). However, a significant difference was observed in postoperative Harris scores between the groups (THA→LF group: 92.80 ± 4.93 vs. LF→THA group: 94.72 ± 4.59, P = 0.043) (Fig. 1), indicating that the LF→THA group showed better improvement in hip joint function postoperatively.

Fig. 1

Radar plot for the preoperative and postoperative changes of ODI, JOA, and Harris scores before IPTW

However, after conducting IPTW, no significant differences were observed in postoperative ODI scores (P = 0.075), JOA scores (P = 0.961), or Harris scores (P = 0.563) between the two groups (Table 2). These results indicate that, after balancing baseline data through IPTW, the previously observed superior improvement in hip joint function in the LF→THA group disappeared. The ODI score in the THA→LF group (12.94 ± 3.36) was slightly higher than that in the LF→THA group (11.07 ± 4.93), but the difference only approached significance (P = 0.075) (Fig. 2).

Table 2 The postoperative clinical activity function scores and long-term complications of patients after IPTW

Fig. 2

Radar plot for the preoperative and postoperative changes of ODI, JOA, and Harris scores after IPTW

Long-term complications before and after conducting IPTW

Before conducting IPTW, the incidence of PJK was 3.8% overall, with 6.5% in the THA→LF group and 1.7% in the LF→THA group, showing no significant difference (P = 0.453) (Supplementary Table 2). Similarly, 4.8% of patients experienced internal fixation loosening, with 6.5% in THA→LF and 3.4% in LF→THA (P = 0.790). The rate of hip prosthesis dislocation was low at 2.9% overall, with 4.3% in the THA→LF group and 1.7% in the LF→THA group (0.838). Overall, no significant differences were found between the two surgical groups regarding PJK, internal fixation loosening, or hip prosthesis dislocation. Furthermore, after conducting IPTW, the incidence of long-term surgical complications remained low in both groups (Table 2), with no significant differences between them, indicating the sequences of surgery had no impact on the long-term complication outcome.

Radiological changes before and after conducting IPTW

Before implementing IPTW, during the first week postoperatively, the PI-LL in the THA→LF group was significantly higher than that in the LF→THA group (P = 0.048) (Supplementary Table 3). However, at the last follow-up, this difference lost statistical significance (P = 0.085). Additionally, at the last follow-up, the SS in the THA→LF group was significantly higher than that in the LF→THA group (P = 0.008), while the differences between the two groups for the PT and PI-LL, though not reaching statistical significance, still showed a trend toward difference (PT: P = 0.097; PI-LL: P = 0.085).

After controlling for all baseline and preoperative radiological characteristics using IPTW, there were no statistically significant differences between groups for any radiological parameters during the first week postoperatively (PT: P = 0.450; PI: P = 0.621; PI-LL: P = 0.638; SS: P = 0.157; LL: P = 0.944) (Table 3). At the last follow-up, however, the PT in the THA→LF group was significantly lower than that in the LF→THA group (6.51 ± 6.06° vs. 10.15 ± 6.65°, P = 0.043), and the SS was significantly higher in the THA→LF group compared to the LF→THA group (38.63 ± 4.88° vs. 33.61 ± 9.89°, P = 0.028). The differences in PI, PI-LL, and LL remained non-significant.

Table 3 The postoperative radiological characteristics of patients after IPTW

Analysis of changes in radiological characteristics before and after surgery with IPTW

This study further presented the dynamic trends of various radiological parameters over time. In detail, before conducting IPTW, the PT showed a decrease in both groups postoperatively, with a more pronounced decline in the THA→LF group (Fig. 3A). After applying IPTW, the preoperative PT differences between the two groups diminished, but the trend of postoperative PT decrease continued to exist in both groups, remaining more pronounced in the THA→LF group. At the last follow-up, the difference between the two groups widened and was statistically significant (P = 0.043) (Fig. 3B).

Fig. 3

Dynamic changes of PT before and after surgery. (A) Before IPTW; (B) After IPTW

Before conducting IPTW, the PI was consistently higher in the THA→LF group compared to the LF→THA group; however, these differences between the two groups did not reach statistical significance (Supplementary Fig. 1A). After conducting IPTW, the difference between the two groups in PI further reduced but remained statistically insignificant (Supplementary Fig. 1B). Before conducting IPTW, the PI-LL was significantly higher in the THA→LF group than in the LF→THA group at both preoperative and one-week postoperative measurements. At the last follow-up, the PI-LL remained significantly higher in the THA→LF group (Supplementary Fig. 2A), approaching statistical significance. After controlling for baseline data using IPTW, the differences between the two groups in PI-LL were no longer statistically significant (Supplementary Fig. 2B). It is worth noting that, regardless of using IPTW, the trends of PI-LL in both groups exhibited a decreasing pattern over time.

The trends of SS over time showed an increase in both groups, with a more marked rise in the THA→LF group. Notably, whether before (Fig. 4A) or after (Fig. 4B) IPTW, at the last follow-up, SS was significantly higher in the THA→LF group than in the LF→THA group. Both groups demonstrated significantly higher LL at one week postoperative compared to preoperative measurements, but this increase was no longer evident at the last follow-up. Before conducting IPTW, the preoperative LL was significantly lower in the THA→LF group compared to the LF→THA group (Fig. 5A). Although LL remained lower in the THA→LF group at both one week postoperative and the last follow-up, the differences did not achieve statistical significance. After conducing IPTW, the differences between the two groups in LL were no longer significant, and the values were closely aligned (Fig. 5B).

Fig. 4

Dynamic changes of SS before and after surgery. (A) Before IPTW; (B) After IPTW

Fig. 5

Dynamic changes of LL before and after surgery. (A) Before IPTW; (B) After IPTW

Discussions

Principal findings

The principal findings of this study indicate that both surgical sequences—THA→LF and LF→THA—significantly improve postoperative clinical function scores, including the ODI, JOA score, and Harris Hip Score, when compared to preoperative assessments. Importantly, no significant differences were observed between the two groups in terms of functional improvements and rates of long-term complications after IPTW, suggesting that both surgical approaches are effective for managing patients with HSS. In addition to functional outcomes, radiological parameters revealed notable differences. The THA→LF group demonstrated a significantly lower PT and a higher SS at the last follow-up after IPTW, highlighting a more favorable impact on pelvic alignment compared to the LF→THA group. These findings suggest that the sequence of surgeries may play a role in optimizing sagittal alignment and enhancing overall postoperative recovery. The use of IPTW effectively balanced baseline characteristics between the groups, thus reinforcing the credibility of the findings. This methodological approach helps mitigate biases related to treatment selection and confounding variables, providing a clearer picture of the impact of surgical sequencing on outcomes.

Impact of THA and LF surgical sequence on spine-pelvis parameters

The sequence of hip and spinal surgeries exerts significant effects on clinical outcomes. This study systematically explored and compared dynamic changes in spine-pelvis parameters under two surgical sequences while evaluating symptom improvement and long-term complication rates following sequential LF or THA. Prior studies have demonstrated that abnormal spine-pelvis parameters contribute to sagittal imbalance, compromising postural stability and increasing energy expenditure [19]. Lumbar degeneration and hip pathologies often reduce LL while triggering compensatory increases in PT [20]. Preoperatively, both spinal and joint surgeons prioritize optimizing spine-pelvis parameters to ensure postoperative sagittal alignment. Celestre et al. [21] proposed that pelvic rotation, namely, changes in PT, compensates for LL loss, thereby maintaining sagittal balance. Based on IPTW analysis, this study revealed that postoperative changes in sagittal spine-pelvis parameters depend on the sequence of THA and LF implementation. In the THA→LF group, PT decreased significantly postoperatively and remained stable during the two-year follow-up, whereas no such trend was observed in the LF→THA group. Immediate postoperative SS increased in the THA→LF cohort but remained unchanged in the LF→THA group after surgery. Both groups exhibited significant postoperative LL restoration compared to preoperative values; however, limited long-term LL improvement was observed, likely attributable to partial or complete lumbar fusion fixation. This investigation underscores the critical role of surgical sequencing in modulating spine-pelvis alignment. The THA→LF sequence demonstrated superior efficacy in optimizing PT and SS, suggesting its potential advantage in achieving sagittal balance. These findings advocate for tailored surgical planning in patients with concurrent hip-spine pathology to enhance functional recovery and biomechanical stability.

Impact of THA and LF surgical sequence on symptom relief and complications

Both LF and THA demonstrate efficacy in alleviating low back and leg pain symptoms. For example, studies indicate comparable postoperative improvements in health-related quality of life between LF and THA [22, 23]. Okuzu et al. [24] reported that 62.5% of THA patients experienced significant low back pain relief postoperatively. This improvement may be attributed to functional interdependencies within the lumbar-pelvic-hip complex, where both surgical interventions modulate biomechanical interactions. Sequential implementation of these procedures may amplify compensatory effects, influencing symptom recovery trajectories and complication risks [25], the optimal surgical sequence, however, remains controversial. Steven et al. [4] proposed that prioritizing THA reduces subsequent spinal surgery requirements, opioid dependence, and postoperative functional instability. Conversely, Eneqvist et al. [6] observed superior symptom recovery in patients undergoing hip surgery prior to spinal procedures, despite similar preoperative symptom severity. Nevertheless, conflicting evidence suggests higher subsequent THA demand in patients undergoing spinal surgery first [4].

This study systematically evaluated and compared symptom profiles, functional outcomes, and long-term complications between the THA→LF and LF→THA groups at preoperative and two-year postoperative follow-ups. All patients exhibited significant improvements in JOA scores, ODI scores, and Harris Hip Scores at two years postoperatively. Before and after conducting IPTW, no significant differences were observed between the two groups in JOA or ODI scores. Although minor Harris Score discrepancies were noted initially, these differences lost statistical significance following IPTW adjustment. Regarding long-term complications, including PJK, lumbar hardware loosening/failure, and hip prosthesis dislocation, no significant intergroup differences were detected before or after IPTW. Notably, the THA→LF group demonstrated increased LL and reduced PT, biomechanical changes that may correlate with decreased acetabular anteversion and elevated dislocation risk, aligning with prior reports [26, 27]. Previously, a meta-analysis demonstrated that patients with HSS who underwent LF before THA had a significantly higher risk of THA dislocation compared to those who underwent THA first [5]. In our study, pre-IPTW analysis revealed a hip dislocation rate of 6.4% in the THA→LF group versus 1.8% in the LF→THA group. Post-IPTW adjustment, rates narrowed (3.5% vs. 2.0%), and thus this trend warrants validation in larger cohorts. While symptom evolution, lumbar fixation constructs, and prosthetic modifications may associate with sagittal imbalance, causal relationships remain unexplored in this study. In our study, to minimize variability from surgical technique heterogeneity, patients were stratified based on lumbar fixation segments and hip prosthesis laterality during propensity score matching. Spine-pelvis parameters were dynamically assessed at three timepoints: preoperatively, one week postoperatively, and ≥ 2 years postoperatively. This longitudinal approach enhances the accuracy of outcome comparisons between surgical sequences and provides robust insights into long-term biomechanical adaptations.

Clinical implications and future directions

This study systematically evaluated the impact of surgical sequence (THA vs. LF) on outcomes in patients with HSS using IPTW. By applying IPTW, this study effectively addressed baseline variable imbalances inherent in real-world data, establishing a robust analytical framework for investigating surgical sequencing. The findings demonstrate comparable efficacy between THA→LF and LF→THA in functional recovery and complication mitigation. However, the THA→LF sequence exhibited superior potential in optimizing pelvic alignment, as evidenced by reduced PT and increased SS. These insights advance clinical practice by highlighting the importance of surgical sequence selection in enhancing long-term biomechanical stability and patient prognosis. In the future, multicenter, prospective studies with larger cohorts are warranted to confirm these findings and explore long-term outcomes in specific subgroups, such as elderly patients or those with severe spinal deformities or advanced hip pathologies. In addition, integrated biomechanical and dynamic imaging analyses are needed to elucidate how surgical sequencing modulates functional interdependencies within the spine-pelvis-hip complex, particularly the temporal dynamics of PT, SS, and LL and their correlations with symptom resolution. Furthermore, comprehensive assessments of global spinal alignment and whole-body balance should be incorporated to evaluate how surgical interventions influence overall postural stability and compensatory mechanisms across the entire kinematic chain. Lastly, leveraging artificial intelligence and machine learning to integrate multimodal data could enable personalized decision-making models, refining surgical planning and improving patient satisfaction.

Limitations

This study has several limitations that should be acknowledged. First, it is a retrospective analysis, which inherently limits the ability to establish causality and may introduce biases related to data collection and patient selection. Second, while the IPTW method was employed to control for confounding factors, unmeasured variables may still influence the outcomes and remain unaccounted for. Third, the relatively small sample size in each surgical group may limit the generalizability of the findings and reduce statistical power to detect differences in long-term complications. Additionally, the follow-up duration, while sufficient for assessing short-term outcomes, may not adequately capture late-onset complications that could arise in this patient population. Finally, variations in surgical techniques and the experience of different surgeons performing the procedures might introduce variability in outcomes that were not controlled for in this study. These limitations suggest that further prospective studies with larger sample sizes and longer follow-up periods are necessary to validate our findings and draw more definitive conclusions. Overall, this study contributes valuable insights into the management of HSS, indicating that while both surgical sequences yield substantial benefits in terms of clinical function and complications, the choice of sequence could have implications for radiological outcomes and pelvic alignment. Further research is needed to corroborate these findings and explore the above-mentioned aspects.

Conclusions

This study highlights that both surgical sequences, THA→LF and LF→THA, significantly improve postoperative clinical function scores and radiological parameters when compared to preoperative assessments. However, there are no significant differences observed in functional improvement or complication rates between the two groups. Notably, radiological outcomes demonstrates that the THA→LF group exhibited more favorable effects on PT and SS, suggesting better pelvic alignment in this sequence. The application of IPTW effectively control for baseline disparities, thereby enhancing the trustworthiness of the results. These findings underscore the importance of considering surgical sequence in treatment planning for patients with HSS, warranting further research to refine surgical decision-making in this patient population.