Impact of patient-reported allergies and psychiatric history on lumbar interbody fusion outcomes: a retrospective cohort study in the United States
Article information
Abstract
Study Design
Retrospective cohort study.
Purpose
To determine whether patient-reported allergies (PRAs) are associated with pain, mobility, and functional outcomes following lumbar interbody fusion (LIF).
Overview of Literature
Prior studies suggest PRAs may correlate with worse postoperative pain, disability, and satisfaction; however, findings are inconsistent, and their impact on LIF outcomes remains unclear.
Methods
We retrospectively analyzed 325 adults undergoing elective LIF for degenerative pathology (November 2019–January 2023). Patients were stratified into ≥1 PRA (n=242) and no known allergies (n=83). Demographics, psychiatric history, preoperative opioid use, and allergy type were recorded. Primary outcomes included changes in Patient-reported outcome measures (PROMs; PROMIS overall, physical, and mental health) and Oswestry Disability Index (ODI) from baseline to 1 year. Secondary outcomes included Visual Analog Scale (VAS) and Activity Measure for Post-Acute Care (AMPAC) mobility. Multivariate regression assessed associations with outcomes. Minimal clinically important difference (MCID) thresholds were ODI ≥10 and PROMIS ≥8. Receiver operating characteristic analysis evaluated PRA count as a predictor of MCID.
Results
Patients with PRAs were more frequently female (57%; p=0.002) and had higher rates of psychiatric comorbidities (p=0.02). The PRA group demonstrated greater improvement in PROMIS overall (p=0.049) and mental health (p=0.026) scores, with no difference in ODI change. PRA count was not associated with VAS, AMPAC, or MCID attainment. Psychiatric history independently predicted lower odds of achieving MCID for PROMIS physical (odds ratio [OR], 0.52; p=0.048) and ODI (OR, 0.27; p=0.002). PRA count showed poor predictive value for MCID (area under the curve<0.600).
Conclusions
PRAs were not associated with worse outcomes after LIF and were linked to greater improvement in PROMIS scores. Psychiatric history, but not allergy count, was a strong predictor of reduced MCID attainment.
Introduction
Lumbar interbody fusion (LIF) is a well-established surgical intervention often considered when conservative treatments fail to alleviate symptoms. The indications for LIF include degenerative disc disease, spondylolisthesis, recurrent disc herniation, spinal instability, and deformities such as scoliosis [1–3]. Success rates for LIF surgery have generally been favorable, with most patients experiencing significant improvements in pain and function. However, the presence of existing conditions, such as patient-reported allergies (PRAs) and psychiatric disorders, can potentially negatively impact these outcomes [4].
Current evidence regarding the correlation between PRAs and postoperative surgical outcomes have been controversial. Some studies have linked multiple PRAs to worse pain and disability scores [4,5], whereas others have shown that patients with more PRAs experienced improvements in postoperative pain and disability burden [6]. Nonetheless, the conflicting evidence surrounding the correlation between PRAs and postoperative outcome scores provides an avenue for further investigation.
Patients with a psychiatric history generally experience worse postsurgical outcomes compared to those without such disorders. Several studies have demonstrated that preexisting mental health conditions were associated with increased risks of postoperative complications, extended hospital stays, and higher readmission rates [6–10]. These findings underscore the importance of preoperative mental health assessment and optimization in improving surgical outcomes among patients with psychiatric disorders.
Therefore, the current study examined whether PRAs were associated with worse postoperative pain, mobility, and functional improvement following LIF. We hypothesize that patients with PRAs would experience less favorable outcomes. Clarifying this relationship could provide insights that could enhance preoperative planning, risk stratification, and postoperative management.
Materials and Methods
Patient population
This research study was conducted retrospectively using data obtained for clinical purposes. Prior to data collection, we consulted extensively with the Institutional Review Board (IRB) of Geisinger, which determined that our study met the criteria for exemption from ethical approval. An official waiver was granted, and ongoing review of the study protocol was provided by the Geisinger IRB. Between November 2019 and January 2023, a retrospective analysis was conducted on 325 adult patients who underwent elective single- and multi-level LIF for degenerative pathology within a large, rural integrated healthcare system. The exclusion criteria included patients <18 years old; those undergoing surgery for trauma, infection, malignancy, or revision procedures; and those with incomplete preoperative or postoperative measures of interest in data collection.
Data collection
Baseline demographic and clinical data, which included age, gender, race, body mass index (BMI), psychiatric history, smoking status, alcohol use, preoperative opioid use, and all reported allergies, were collected via manual chart review. PRAs were defined as any self-reported allergy documented in the medical records and were categorized into drug allergies (e.g., antibiotics, analgesics), environmental allergies (e.g., food, latex, and animals), and others. Patients with a single documented allergy, regardless of type, were included in the PRA group. Those without any reported allergies of any kind were included into the no known allergies (NKA) group.
Surgical variables included the number of lumbar levels fused, surgical approach (open vs. minimally invasive), and postoperative complications. Patient-reported outcome measures (PROMs) were collected at standardized time intervals (before surgery and 6 weeks, 12 weeks, 6 months, and 1 year after surgery) and included the Oswestry Disability Index (ODI), Patient-Reported Outcomes Measurement Information System (PROMIS) physical, mental, and overall health scores, and Visual Analog Scale (VAS) for pain. Immediate postoperative mobility was assessed using the Activity Measure for Post-acute Care (AMPAC) score.
Statistical analysis
For comparative analysis, patients were stratified into two cohorts (i.e., PRAs or NKA groups) based on their history. Independent t-tests and chi-square tests were used to compare the means and proportions for continuous and categorical data, respectively. Univariate and multivariate linear regression analyses were performed using VAS and AMPAC as the dependent variables. PRAs, psychiatric history, preoperative opioid use, and number of levels fused were used as independent variables. To assess the clinical relevance of the observed changes, the mean delta differences in ODI and PROMIS between the latest follow-up and before surgery were calculated to determine the minimal clinically important difference (MCID), with thresholds for ODI (≥10) and PROMIS (≥8) defined based on previous literature [11]. Univariate and multivariate logistic regression analyses were then performed with ΔPROMIS and ΔODI MCID as the dependent variables and PRAs, psychiatric history, preoperative opioid use, and number of levels fused as the independent variables. The level of statistical significance was set at p<0.05. Given the multiple MCID endpoints, Bonferroni correction was applied across the four primary MCID comparisons (PROMIS overall, physical, mental, and ODI), with the level of statistical significance set at p<0.0125. All statistical analyses were conducted using IBM SPSS ver. 25.0 (IBM Corp., Armonk, NY, USA).
Results
Demographic data
A total of 325 patients (242 with one or more PRA and 83 with NKA) met the inclusion criteria. PRAs included any self-reported allergy documented in the medical records, including drug allergies, environmental allergies, and others.
The PRA group (n=242) was 57% female, with a mean age of 58.3±13.2 years and a BMI of 30.9±6.3 kg/m2. The NKA group (n=83) was 37% female, with a mean age of 56.7±13.1 years and a BMI of 31.1±5.3 kg/m2. Patients with PRAs were more likely to be female (p=0.002) and had a higher mean number of psychiatric disorders (0.9±1.1 vs. 0.7±1.0, p=0.02). Patients with NKA were more likely to report alcohol use (51% vs. 29%, p=0.008). No significant differences in tobacco use, preoperative opioid use, surgical approach, number of fusion levels, complications, or mortality were observed between the groups (Table 1).
Baseline demographics, perioperative characteristics, and outcome measures
Patient-reported outcome scores
Patients with PRAs demonstrated significantly greater improvements in PROMIS overall health scores (Δ1 year–preoperative; 11.5±12.8 vs. 6.7±10.2, p=0.049), higher PROMIS mental health scores (Δ1 year–preoperative; 13.5±18.7 vs. 6.6±14.6, p=0.026), and higher preoperative ODI scores (46.7±14.5 vs. 41.6±15.2, p=0.038). No significant differences in PROMIS overall health scores were observed before surgery (p=0.122) and 6 weeks (p=0.411), 12 weeks (p=0.583), 6 months (p=0.970), and 1 year (p=0.425) after surgery. No significant differences in PROMIS mental health scores were observed before surgery (p=0.247) and 6 weeks (p=0.814), 12 weeks (p=0.460), 6 months (p=0.971), and 1 year (p=0.320) after surgery. No significant differences in PROMIS physical scores were observed before surgery (p=0.125) and 6 weeks (p=0.114), 12 weeks (p=0.547), 6 months (p=0.819), 1 year (p=0.741) after surgery, as well as for Δ1 year–preoperative (p=0.118). For ODI, there were no significant differences in ODI scores were observed 6 weeks (p=0.916), 12 weeks (p=0.567), 6 months (p=0.518), 1 year (p=0.749) after surgery, as well as for Δ1 year–preoperative (p=0.792) (Table 1).
Minimum clinically important difference analyses
Multivariate logistic regression analysis showed that psychiatric history was associated with lower odds of achieving MCID in PROMIS physical function (odds ratio [OR], 0.517; p=0.048). Patients with a psychiatric history had significantly lower odds of achieving MCID in ODI on univariate (OR, 0.283; p=0.002) and multivariate analyses (OR, 0.273; p=0.002) (Table 2). PRAs, preoperative opioid use, and number of fusion levels were not significantly associated with MCID achievement or changes in PROMIS physical, overall, and mental health scores or the ODI. After Bonferroni correction for four MCID comparisons (α=0.0125), psychiatric history remained independently associated with lower odds of achieving ODI MCID (p=0.002), whereas the association between PRAs and PROMIS physical MCID did not meet the adjusted significance threshold (p=0.048).
Univariate and multivariate logistic regression of clinical factors and ΔPROMIS and ΔODI MCID
Immediate postoperative AMPAC and VAS scores
Univariate and multivariate analyses showed that no variables, including number of allergies, psychiatric history, preoperative opioid use, or number of levels fused, were significantly associated with immediate postoperative VAS scores (Table 3). An increase in the number of levels fused was significantly associated with a decrease in immediate postoperative AMPAC scores on univariate (B=−1.096, p<0.0001) and multivariate regression analyses (B=−1.061, p<0.0001). No other variables were identified as significant predictors of AMPAC scores.
Univariate and multivariate regression of clinical factors and postoperative VAS and AMPAC scores
Discussion
Our study challenges prevailing assumptions regarding the prognostic implications of PRAs in LIF outcomes. Prior spine literature has often linked PRAs with increased postoperative pain, disability, or dissatisfaction [10,12]. In our cohort, however, PRAs were not associated with worse pain (VAS), function (PROMIS/ODI), mobility (AMPAC), or complication rates. In fact, patients with PRAs demonstrated significantly greater improvements in PROMIS mental and overall health scores 1 year after surgery, with comparable physical recovery rates to patients with NKA. These findings suggest that PRAs may reflect psychosocial factors rather than exert direct biological effects on postoperative recovery. As such, the clinical significance of PRAs in preoperative risk stratification may need to be revisited.
The observation that patients with PRAs demonstrated greater improvement in PROMIS mental health scores despite a higher burden of psychiatric comorbidity warrants cautious interpretation. Importantly, our dataset did not capture details regarding perioperative care utilization, including visitation frequency, provider contact, psychological counseling, or formal mental health interventions. As such, we cannot determine whether the patients with PRA received varying perioperative management or follow-up protocols, which could directly explain these findings. Consequently, the observed association should be interpreted as hypothesis-generating rather than indicative of a causal treatment effect.
One possible explanation is that PRA status may serve as a marker for heightened symptom awareness or psychosocial vulnerability rather than exert a direct biological influence on postoperative recovery. Prior studies have linked multiple reported allergies with increased health vigilance, somatization, and psychological distress, particularly anxiety and depression [1,4,9]. In this context, patients with higher baseline psychological burden may perceive structured surgical care and standardized postoperative follow-up as particularly beneficial, resulting in greater relative improvement in patient-reported mental health over time. However, this interpretation remains speculative given that the present study failed to assess differences in care engagement.
Despite greater improvement in PROMIS mental health scores, patients with PRAs did not demonstrate superior absolute mental health scores at individual postoperative time points, nor did they exhibit differences in physical function, disability, pain, or mobility outcomes. This pattern suggests that the observed mental health improvements may reflect regression toward the mean or recovery from worse baseline psychosocial status rather than a sustained or treatment-driven advantage [11,13,14]. Prospective studies incorporating validated psychometric instruments, care utilization metrics, and standardized assessments of psychosocial support are needed to better describe the mechanisms underlying these findings.
Our findings do not support a dose–response relationship between the number of reported allergies and poorer outcomes. Unlike prior studies that linked increased PRA counts to decreased functional gains or higher dissatisfaction rates [5], our data showed no such trend. In fact, receiving operating characteristic curve analysis (area under the curve<0.600) demonstrated that PRA count had poor predictive value for achieving MCID in PROMIS or ODI scores, further challenging the notion that PRAs directly impair functional recovery (Fig. 1). This discrepancy may reflect differences in surgical complexity, institutional preoperative counseling strategies, or patient selection criteria. Although Lyons et al. [4] reported worsening ODI scores with increasing number of allergies, their cohort lacked standardized mental health metrics, potentially conflating allergy burden with unmeasured psychiatric morbidity [10]. Our data suggest that PRAs, when decoupled from psychiatric illness, did not independently hinder physical recovery.
Receiver operating characteristic (ROC) curves for the number of patient-reported allergies (PRAs) predicting minimal clinically important difference (MCID) achievement in PROMIS overall (A), PROMIS physical (B), PROMIS mental (C), and Oswestry Disability Index (ODI) scores (D).
In contrast to the limited prognostic utility of PRAs, preexisting psychiatric comorbidities emerged as a significant barrier to achieving optimal postoperative outcomes and remained a robust predictor of reduced ODI MCID attainment after correction for multiple comparisons. Although patients with PRAs had a higher prevalence of psychiatric comorbidities, inclusion of psychiatric history as a covariate during multivariate and MCID analyses demonstrated that psychiatric disorders, rather than PRA status, was the primary driver of reduced functional recovery. In our cohort, patients with documented psychiatric diagnoses demonstrated reduced odds of achieving clinically meaningful improvements in physical function, as evidenced by their lower rates of achieving the MCID for PROMIS physical (OR, 0.517) and ODI (OR, 0.273). These findings are consistent with existing studies in both arthroplasty and spine surgery, which showed that depression and anxiety were associated with a 20%–30% reduction in the likelihood of reaching functional recovery benchmarks [2,15–17]. Prior studies suggest that depression and anxiety may be particularly associated with worse postoperative outcomes; however, diagnosis-specific effects could not be evaluated in the present study and warrant further investigations [2,15–17]. Nonetheless, several mechanisms may explain this relationship, including neurobiological dysregulation involving inflammatory mediators (e.g., interleukin-6 and tumor necrosis factor-alpha), which can impair tissue healing and central pain modulation [18,19].
Behavioral and psychosocial factors, such as diminished adherence to rehabilitation protocols and reduced physical activity, may negatively affect postoperative recovery. Additionally, cognitive distortions and negative emotional states can alter pain perception and increase self-reported disability, thereby complicating clinical assessment. These results reinforce the importance of incorporating validated preoperative mental health screening tools, such as the Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder seven-item scale, into routine surgical risk stratification. As recommended by Enhanced Recovery After Surgery pathways, integrating behavioral health interventions, such as cognitive behavioral therapy, into perioperative care has shown early promise, with pilot studies in spine surgery reporting a 15%–25% improvement in patient-reported outcomes [1,20–22].
Neither PRA status nor allergy count significantly impacted postoperative pain (VAS), mobility (AMPAC), or surgical complications. This lack of an association between PRAs and psychiatric history and immediate postoperative metrics (i.e., AMPAC and VAS) implies that intraoperative technical success is achievable regardless of these factors. Moreover, meaningful recovery following LIF may be more closely linked to improvements in mental health outcomes. Additionally, improved engagement with healthcare providers and better postoperative monitoring may have mitigated baseline psychological distress in the PRA subgroup, thereby improving their mental health. This interpretation aligns with the growing body of evidence advocating for integrated mental health support in surgical settings to optimize recovery outcomes, particularly for vulnerable patient populations [4]. However, the long-term divergence between mental and physical outcomes highlights the need for tailored rehabilitation. For example, patients with PRAs may benefit from integrating psychological interventions into physical therapy to sustain early improvements in mental health, whereas patients with psychiatric disorders may require augmented analgesia or activity pacing.
Some limitations inherent to the retrospective design of this study should be acknowledged. First, confounding by unmeasured variables, such as socioeconomic status, health literacy, or surgeon-specific factors, cannot be overlooked. Second, although psychiatric diagnoses were included in the analysis, detailed information regarding disorder severity, treatment adherence, and psychotropic medication use was not available. Additionally, psychiatric diagnoses were analyzed as a composite variable, which precluded the assessment of whether specific conditions (e.g., depression, anxiety, and bipolar disorder) were differentially associated with postoperative outcomes following LIF. Furthermore, allergy data were not verified through allergy testing but rather relied on self-report within the electronic medical records, which may be influenced by subjective interpretation or documentation variability. Third, the generalizability of our findings may be limited considering that the study was conducted in a predominantly rural setting. Differences in patient diversity, access to care, and presentation of psychiatric comorbidities between urban and rural populations may lead to differing results. Furthermore, our data reflect only short- to mid-term outcomes, with follow-up limited to 1-year PROMs and immediate postoperative VAS and AMPAC scores. Future studies incorporating validated psychometric tools, such as the Pain Catastrophizing Scale or PHQ-9, along with biomarker profiling (e.g., cortisol and cytokine levels), could help clarify the mechanisms underlying the association between psychiatric comorbidities, PRAs, and postoperative outcomes.
Conclusions
PRAs were not associated with worse outcomes following elective LIF. On the contrary, patients with PRAs demonstrated greater improvements in PROMIS overall and mental health scores 1 year after surgery. PRA status did not influence postoperative pain (VAS), mobility (AMPAC), complication rates, or MCID achievement in PROMIS or ODI scores. These findings suggest that PRAs should be viewed as potential markers for underlying psychosocial needs, reinforcing the importance of holistic perioperative support and evidence-based behavioral interventions for patients with psychiatric comorbidities.
Key Points
Patient-reported allergies (PRAs) did not predict worse pain, mobility, or disability after lumbar interbody fusion.
The PRA group demonstrated greater improvement in Patient-reported Outcomes Measurement Information System (PROMIS) overall and mental health scores at 1 year than did the no known allergies group.
Psychiatric history independently predicted decreased odds of achieving the minimal clinically important difference (MCID) for PROMIS physical and Oswestry Disability Index.
Allergy count showed poor discriminative ability for MCID (area under the curve<0.600).
Preoperative screening and mental health optimization may
Data Availability
The data supporting this study are available upon reasonable request from the corresponding author.
Notes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Author Contributions
Conceptualization: JS, EF, AJ, TC. Study design: MP. Data acquisition: JS, EF, AJ. Statistical analysis: MP. Supervision of data analysis: TC. Data interpretation: MP, AT. Methodology: MP, AT. Project administration: TC. Writing–original draft: JS, EF, AJ. Writing–review & editing: AT, MP, TC. Supervision: AT. Final approval of the manuscript: all authors.