Abstract

Background: Spinal arteriovenous shunts (sAVSs) consist of direct arterial-venous connections without intervening capillaries. Although rare, if left untreated, permanent neurological injury can occur. The present study aims to assess preoperative risks associated with 30-day unplanned readmission, 30-day unplanned reoperation (RTOR), nonhome discharge, and postoperative complications.

Methods: The American College of Surgeons National Surgical Quality Improvement Project database was employed to isolate cases of surgically treated sAVSs from 2012 to 2022. Univariate comparisons and multivariate logistic regression analyses were performed.

Results: Among the cohort, there were 18 (5.0%) 30-day readmissions, 18 (5.0%) RTOR, 145 (40.0%) nonhome discharges, and 52 (14.0%) cases with a postoperative complication. Preoperative hypoalbuminemia was a risk for 30-day readmission (P = 0.03), nonhome discharge (P P = 0.003) in univariate testing, while normoalbuminemia decreased the odds of nonhome discharge in multivariate analysis (OR = 0.2 [0.05–0.89], P = 0.03). Postoperative complications were associated with RTOR in both univariate and multivariate analyses (OR = 5.1 [1.44–17.94], P = 0.01). 30-day readmissions (72.2%, P = 0.008), RTOR (70.6%, P = 0.01), postoperative complications (63.5%, P P = 0.004) resulted in more nonhome discharges in univariate analyses, while thoracic (odds ratio [OR] = 15.2 [1.08–213.74], P = 0.04) and thoracolumbar (OR = 20.9 [1.32–330.05], P = 0.03) sAVS and preoperative steroid use (OR = 11.1 [1.19–103.73], P = 0.04) increased the odds of postoperative complications in multivariate analysis.

Conclusion: Preoperative hypoalbuminemia increases the odds of 30-day readmissions, nonhome discharges, and postoperative complications, stressing the importance of preoperative optimization. 30-day readmission and RTOR are associated with increased rates of postoperative complications, while 30-day readmissions, RTOR, and postoperative complications resulted in more nonhome discharges.

Keywords: Arteriovenous malformation, Fistula, Malnutrition, Spine, Surgery

INTRODUCTION

Spinal arteriovenous shunts (sAVSs) are a group of pathologies first described in 1888 that includes direct arterial-venous fistulous connections and true nidal malformations, both of which occur without intervening capillaries.[ 25 , 45 ] sAVSs are rare with an incidence as low as 0.2/100,000 and arise most frequently in men within the thoracic spine.[ 19 , 45 ] Over the decades, numerous classification systems to describe variants of sAVS have been developed, but the most commonly used is the Kim-Spetzler classification.[ 1 , 9 , 45 ]

The Kim-Spetzler classification of sAVS, described by Kim and Spetzler[ 25 ] in 2006, consisted of 6 sub-types of sAVS based on anatomy and pathophysiology. Largely, each of these subtypes causes neurological injury through fistulous connections leading to venous engorgement, mass effect, venous hypertension, or vascular steal causing ischemia. The ensuing neurological deficits often include pain, numbness, paralysis, and bladder/bowel dysfunction, leading to significant loss of independence and quality of life if the sAVS is left untreated.[ 34 , 39 ] However, due to their rarity, diagnosis may be elusive. The gold standard of diagnosis often consists of magnetic resonance imaging and digital subtraction angiography.[ 8 ] Once identified, the standard treatment is open surgical disconnection, though consideration for adjunctive or isolated endovascular embolization has been increasing.[ 39 ]

The study of sAVS surgical outcomes remains difficult due to their low incidence. Population-based databases provide a unique opportunity to amass a sample size with adequate power for the study of rare pathology. As such, the current study looks to tap into this potential by utilizing the American College of Surgeons’ National Surgical Quality Improvement Project (NSQIP) database. Bhimani et al.[ 8 ] have performed the only other studies utilizing NSQIP to assess surgical outcomes of sAVS. However, this study is limited to assessing rates of postoperative complications as they relate to spinal level, without analyzing the association of preoperative risk factors with 30-day readmission, unplanned reoperation within 30 days of the index surgery (RTOR), postoperative complications, or nonhome discharge. In addition, the data used is limited to the year 2017, and the subsequent analysis is limited to largely univariate testing.

The current study aims to assess the rates of preoperative risks associated with 30-day unplanned readmission, RTOR, nonhome discharge, and postoperative complications. This study is the first to assess preoperative risks associated with 30-day unplanned readmission, RTOR, nonhome discharge, and postoperative complications to characterize surgical outcomes following laminectomy for sAVS ligation utilizing NSQIP data.[ 41 ]

MATERIALS AND METHODS

This retrospective cohort study was reported following STROBE guidelines.

Data collection

All data analyzed were acquired from the American College of Surgeons’ NSQIP database. NSQIP is a database compiled of data submitted by over 700 hospitals in the United States (US) with a focus on the 30-day postoperative time point.[ 6 , 17 ] Data are publicly available to researchers within participating institutions. Dedicated individuals within each participating institution are specifically trained in NSQIP methods and data entry and are responsible for collecting and submitting their institution’s respective patient data. All data are audited to ensure accuracy at regular intervals.[ 6 ] NSQIP is the only multi-institutional, multispecialty, clinically based, prospectively collected quality improvement program for surgical specialties.[ 17 ] It is considered by many to be the most accurate database for surgical outcome research and has been shown to improve surgical outcomes.[ 3 ]

Cases of open laminectomy performed for ligation of sAVS between the years of 2012 and 2022 were identified through current procedural terminology (CPT) codes 63250 (laminectomy, excision/occlusion arteriovenous malformation [AVM], spinal cord; cervical), 63251 (laminectomy, excision/occlusion AVM, spinal cord; thoracic) and 63252 (laminectomy, excision/occlusion AVM, spinal cord; thoracolumbar) similar to previous studies.[ 8 , 41 ] Cases of endovascular embolization were not included to ensure a homogenous surgical strategy within the cohort. Due to the fact that CPT codes for open laminectomy for ligation of sAVS are used exclusively for the treatment of sAVSs, other spinal pathologies treated with laminectomy, such as neurogenic claudication or tumor, were inherently excluded.

Data organization

Cases were grouped into binary primary outcomes of 30-day readmission (+ or −), reoperation within 30 days of the index surgery (RTOR) (+ or −), and nonhome versus home discharge, in addition to a binary secondary outcome of any postoperative complication (+ or −). Body mass index (BMI) was calculated using each patient’s height in inches and weight in pounds (BMI = 703 × [weight in pounds÷height in inches²]) then classified into a categorical variable as follows: underweight ≤18.5, optimal weight = 18.5–24.99, overweight = 2–29.99, obese ≥30. Preoperative sodium (hyponatremia: ≤134 mEq/L, normonatremia: 135–145 mEq/L, hypernatremia: ≥146 mEq/L), preoperative albumin (hypoalbuminemia: ≤3.4 g/dL, normoalbuminemia: 3.5–5.5 g/dL, hyperalbuminemia: ≥5.6 g/dL), preoperative platelet count (thrombocytopenia: ≤149,999 platelets/µL, normal platelet count: 150,000–450,000 platelets/µL, thrombocytosis: ≥450,001 platelets/µL), preoperative partial thromboplastin time (PTT) (hypercoagulable: ≤24 s, normal PTT: 25–35 s, hypercoagulable: ≥36 s), and preoperative normalized international ratio (INR) (hypercoagulable: ≤0.7, normal INR: 0.–1.1, hypocoagulable: ≥1.2) were also classified into categorical variables. A full list of variables included in the analysis is available in Table 1 .

Table 1:

Univariate summary of preoperative patient demographics, characteristics, and laboratory testing.

Statistical analysis

Univariate and multivariate analyses were both performed. For univariate comparison analyses, Mann–Whitney U analysis was utilized for continuous variables given a nonparametric distribution, while a χ² test was utilized for categorical variables. If statistical significance was found, a post hoc pairwise comparison with Bonferroni correction was then utilized to identify where in the variable this significance resides. Covariates with P < 0.2 within the univariate comparison models were then included in a multivariate binary logistic regression analysis to control for confounding variables. Covariates with significant outliers or that demonstrate collinearity as determined by a variance inflation factor >10 were not included in the multivariate logistic regression.[ 7 ] Odds ratios (OR) were calculated as the output for the logistic regression. Missing data were addressed in a listwise fashion consistent with standard methods.[ 24 ] Statistical significance was determined by a P < 0.05 and 95% confidence intervals not including 1. These statistical methods have been previously published in prior studies similarly utilizing NSQIP data.[ 5 , 14 , 21 , 26 , 27 , 33 , 40 ] All analyses were performed using the Statistical Package for the Social Sciences Statistics version 28.0 (IBM, Armonk, NY).

RESULTS

A total of 363 cases of sAVS in which the patient underwent laminectomy for ligation were isolated from the NSQIP database between the years 2012 and 2022 and included in the analysis. A total of 18/363 (5.0%) had an unplanned 30-day readmission, 18/363 (5.0%) underwent unplanned RTOR, 145/363 (40.0%) had a nonhome discharge, and 52/363 (14.3%) suffered a postoperative complication. Cumulative patient demographics, in addition to a summary of patient characteristics and preoperative laboratory testing for univariate analyses, can be found in Table 1 . Table 2 reports the results of univariate analysis for the surgical course. Table 3 reports the frequencies of individual postoperative complications and reasons for RTOR. Summary of multivariate analyses for 30-day readmission, ROTR, nonhome discharge, and postoperative complication can be found in Figures 1 - 4 , respectively. Data are presented as median [interquartile range], and 95% CIs are presented as [lower bound-upper bound] unless otherwise specified. ORs are reported with results of the multivariate binary logistic regression analysis.

Table 2:

Univariate summary of surgical course.

Table 3:

Individual postoperative complications and reasons for unplanned re-operation within 30 days of index surgery.

Figure 1:

Forest plot of multivariate analysis results demonstrating odds of unplanned readmission within 30 days of index surgery. *Represents statistical significance. Results reported as odds ratio [95% confidence interval lower bound-upper bound], P-value. The overall multivariate model is statistically significant (P < 0.001).

Figure 2:

Forest plot of multivariate analysis results demonstrating odds of unplanned re-operation within 30-days of index surgery. *Represents statistical significance. Results reported as odds ratio [95% confidence interval lower bound-upper bound], P-value. The overall multivariate model is statistically significant (P < 0.001).

Figure 3:

Forest plot of multivariate analysis results demonstrating odds of nonhome discharge. *Represents statistical significance. Results reported as odds ratio [95% confidence interval lower bound-upper bound], P-value. The overall multivariate model is statistically significant (P < 0.001).

Figure 4:

Forest plot of multivariate analysis results demonstrating odds of any postoperative complication. *Represents statistical significance. Results reported as odds ratio [95% confidence interval lower bound-upper bound], P-value. The overall multivariate model is statistically significant (P < 0.001).

30-day readmissions

Univariate analyses: Among preoperative risk factors, cases with 30-day readmission also had statistically more hypoalbuminemia (P = 0.03), more hyponatremia (P = 0.01), and lower preoperative hematocrit (P = 0.02). Cases with 30-day readmission have significantly more discharges to skilled nursing facilities (SNF) (13.3%) and rehab (10.0%) compared to those who did not (P = 0.03) as well as more postoperative complications (P < 0.001).

Multivariate analysis: The presence of a postoperative complication is significantly associated with higher odds of 30-day readmission (OR = 10.6, 95% CI [1.58–70.94], P = 0.02).

Reoperation within 30 days of index surgery (RTOR)

Univariate analyses: Among preoperative risk factors, cases with RTOR had more preoperative partial disability (P = 0.02) and preoperative hypertension (P = 0.04). Cases with RTOR have significantly more discharges to SNF (13.3%) (P = 0.04), longer length of stay (LOS) (7 [5–15.5] vs. 6 [3–9] days) (P = 0.01), and more postoperative complications (18.8% vs. 2.9%) (P < 0.001) compared to cases without RTOR.

Multivariate analysis: Increased LOS (OR = 1.1, 95% CI [1.001–1.16], P = 0.04) and postoperative complications (OR = 5.1, 95% CI [1.44–17.94], P = 0.01) are significantly associated with higher odds of RTOR.

Nonhome discharge

Univariate: Among demographic factors, significantly more patients aged 70–79 years (59.7%) and 80–89 years (85.7%) (P < 0.001) as well as those with sAVS located in the cervical and thoracolumbar spine (cervical: 33 (50.0%); thoracic: 64 (34.0%); thoracolumbar 44 (48.4%); P = 0.02) had a nonhome discharge. Among preoperative risk factors, hypoalbuminemia (77.1%, P < 0.001), partial disability (85.2%, P < 0.001), diabetes mellitus (59.6%, P = 0.006), hypertension (50.3%, P = 0.002), bleeding disorder (87.5%, P = 0.007), sepsis (66.7%, P = 0.04), lower hematocrit (P = 0.004), hypocoagulable INR (61.5%, P = 0.04), and ASA classes 3 (50.8%) and 4 (72.0%) (P < 0.001) were significant risk factors for nonhome discharge. In addition, being transferred from another acute care hospital or outside emergency department resulted in higher rates of nonhome discharge (acute care hospital: home discharge = 30.8% vs. nonhome discharge = 69.2%; outside emergency department: home discharge = 28.6% vs. nonhome discharge = 71.4%, P < 0.001), while more urgent/emergent cases who had nonhome discharge (52.4%) compared to elective cases (35.1%) (P = 0.002). Those with a nonhome discharge had a significantly longer LOS (nonhome discharge: 8 [6–13] days; home discharge: 4 [2–7] days, P < 0.0005), higher percentage of cases with a 30-day readmission (70.6% vs. 29.5%, P = 0.008), RTOR (70.6% vs. 29.4%, P = 0.01), and more postoperative complications (63.5% vs. 36.5%, P < 0.001) compared to a home discharge. Increased days from admission to surgery were also statistically significantly associated with increased nonhome discharge (nonhome discharge: 0 [0–5] days; home discharge: 0 [0–1] days, P < 0.001); however, this is unlikely to be clinically significant due to similar absolute values.

Multivariate: Preoperative normoalbuminemia was associated with significantly lower odds of nonhome discharge compared to preoperative hypoalbuminemia (normoalbuminemia vs. hypoalbuminemia: OR = 0.2, 95% CI [0.05–0.89], P = 0.03). Increased LOS (OR = 1.5, 95% CI [1.12–1.98], P = 0.006) was also associated with increased odds of nonhome discharge. Increased days from admission to surgery were also statistically significantly associated with less odds of nonhome discharge compared to a home discharge (OR = 0.7, 95% CI [0.51–0.98], P = 0.04); however, this is unlikely to be clinically significant due to similar absolute values.

Postoperative complication

Univariate analysis: Among preoperative risk factors, patients with preoperative hypoalbuminemia (41.7% vs. 17.7%, P = 0.003), hyponatremia (42.3% vs. 12.7%, P < 0.001), lower hematocrit (P < 0.001), preoperative steroid use (30.4% vs. 13.7%, P = 0.03), and preoperative sepsis (33.3% vs. 13.7%, P = 0.04) were found to have higher rates of postoperative complications than those who did not. In addition, those presenting after transfer from another hospital experienced more postoperative complications than those who presented from home (30.0% vs. 11.0%, P = 0.003, respectively), while normal preoperative platelet count (14.3% vs. 11.5%, P = 0.04) was associated with slightly higher rates of postoperative complications compared to preoperative thrombocytopenia. Those who suffered a postoperative complication had a LOS 4 days longer than those who did not (P < 0.001). Patients who underwent RTOR (52.9% vs. 12.5%, P < 0.001) and had a 30-day readmission (72.2% vs. 11.6%, P < 0.001) had significantly higher rates of postoperative complications compared to those who did not. There were statistically significantly more postoperative complications in patients who were ventilator dependent pre-operatively (P = 0.02) as well as those with preoperative CHF status (P = 0.02), preoperative hemodialysis (P = 0.02), preoperative transfusion (P = 0.02), and median days from admission to surgery (P = 0.01); however, these results are not clinically significant due to the low absolute value of n.

Multivariate analysis: Within the multivariate model, spinal level gained significance, as thoracic and thoracolumbar sAVSs had higher odds of a postoperative complication compared to cervical sAVSs (thoracic vs. cervical: OR = 15.2, 95% CI [1.08–213.74], P = 0.04; thoracolumbar vs. cervical: OR = 20.9, 95% CI [1.32–330.06], P = 0.03), while hypoalbuminemia lost significance (OR = 0.36, 95% CI [0.05–2.69], P = 0.32). 30-day readmission (OR = 164.02, 95% CI [2.65–10163.63], P = 0.02), preoperative steroid use (OR = 11.1, 95% CI [1.19–103.73], P = 0.03), and preoperative sepsis (OR = 8.7, 05% CI [1.03–73.14], P = 0.04) continued to have higher odds of postoperative complications.

DISCUSSION

The study of sAVSs is challenging due to low incidence.[ 41 ] As a result, the vast majority of studies are limited to case reports, case series, or single-institution retrospective reviews, limiting current knowledge of surgical outcomes.[ 1 , 9 , 11 , 22 , 28 , 32 , 34 , 41 , 44 ] The current study attempts to fill this gap by querying NSQIP data to study rates of, and preoperative risks for, 30-day readmission, RTOR, nonhome discharges, and postoperative complications following open surgical ligation of sAVS. Within this cohort of 363 surgically treated sAVSs, there was a rate of 5% for 30-day readmissions, 5% for RTOR, 39.9% for nonhome discharges, and 14.3% for the presence of a postoperative complication. Novel to this study is the finding that preoperative hypoalbuminemia is associated with higher rates of 30-day readmissions, nonhome discharges, and postoperative complications.

There remains limited data on preoperative risk assessments and important postoperative quality measures, such as 30-day readmissions and RTOR, following surgery for sAVSs.[ 11 , 22 , 39 ] Varshneya et al.[ 41 ] utilized the MarketScan database to identify 976 patients with sAVS who underwent treatment in the form of radiosurgery, open surgery, endovascular embolization, or both open and adjunctive endovascular embolization between the years 2007 and 2015. Among open surgically treated patients, any postoperative complication occurred at a rate of 31.15%, nearly double that of the current study. Although a LOS of 6 days following open surgery is similar to the LOS reported in the current study, rates of 30-day readmission were approximately 4 times higher in their cohort compared to ours (20.8% vs. 5.0%, respectively). This disparity could be attributed to their longer follow-up interval of 90 days compared to 30 days in the present study. There are also significant differences in data collection methods. MarketScan extracts information through administrative claims primarily from employer-sponsored health plans across all facets of medicine. NSQIP data come from academic centers, are limited to surgical procedures, and are vetted by dedicated nurse reviewers who use standardized definitions and inclusion criteria.[ 3 ] For surgical research, the use of NSQIP data is considered superior. In fact, NSQIP is widely considered the most accurate database for surgical research and has been validated for its ability to improve surgical outcomes.[ 3 ]

NSQIP data are often employed to improve surgical outcomes of common procedures, though the power of its robust sample size can be harnessed for rare surgical pathology as well. Bhimani et al.[ 8 ] queried the NSQIP database to identify patients with sAVS who underwent open surgical ligation between 2008 and 2017. This study is limited to providing rates of postoperative complications and comparing these rates between spinal levels. Preoperative risks for 30-day readmission and RTOR were not included within their analysis, nor did they assess nonhome discharges or control for confounding variables with multivariate analysis. Nevertheless, Bhimani et al.[ 8 ] report a slightly elevated rate of 6.6% for 30-day readmissions as well as a similar ~6-day LOS and 4.6% RTOR rate compared to the current study. 30-day readmission rates have continued to decline as the study cohort becomes more recent than Varshneya et al.[ 41 ] in 2015, Bhimani et al.[ 8 ] In 2017, and finally in 2022, in the current study. Although there is an ongoing need to identify the optimal surgical strategy,[ 43 ] these results continue to represent the positive impacts that research into the knowledge of sAVS’s anatomy, pathophysiology, and treatment options has had on surgical outcomes.

Here, we identified preoperative hypoalbuminemia and lower preoperative hematocrit as risk factors for 30-day readmission, nonhome discharge, and postoperative complications following laminectomy for sAVS. Hypoalbuminemia is a serum laboratory value that represents malnutrition. Preoperative malnutrition is a well-known modifiable risk factor for postoperative complications following oncological spine surgery and spinal fusions.[ 10 , 31 , 35 ] Specifically, hypoalbuminemia has been shown to increase mortality among oncological spine surgery by 7 fold and the risk of SSI from 1.6% to 21.6% following lumbar fusions, while a recent meta-analysis revealed that hypoalbuminemia increases the odds of any postoperative complication, mortality, 30-day readmission, and need for revision following fusions for deformity correction.[ 15 , 29 , 42 ] Unique to the current study’s cohort, however, sAVS ligations utilize a more tissue sparing approach of either laminectomy alone or laminoplasty. Despite these less invasive and physiologically demanding approaches compared to fusion surgeries, preoperative malnutrition still played a critical role in postoperative outcomes despite the majority of patients undergoing elective surgery, not urgent or emergent within the NSQIP database. Malnutrition may be directly or indirectly related to inferior clinical outcomes. For example, hypoalbuminemia may be directly related to a poor outcome in the event of poor wound healing, resulting in RTOR for revision. It may also be indirectly related to poor clinical outcomes as a surrogate marker for poor baseline health and limited physiological reserve leading to inherently higher risk of any complication or the need for sub-acute rehab post-operatively instead of a home discharge. As a result of mounting evidence regarding the importance of preoperative nutritional optimization, nutritional evaluation and intervention have become commonplace in enhanced recovery after surgery protocols.[ 35 ] If progressive neurological decline or a surgical emergency precludes preoperative optimization, however, postoperative nutritional supplementation can still provide value. In particular, optimizing caloric intake with a goal of 25 kcal/kg/day and 1–2 g/kg/day of protein daily can enhance the healing process and lower the risk of complication.[ 18 ] These data support efforts to optimize each patient’s nutrition before surgery for sAVS, though the natural history of sAVSs may not allow sufficient time, in which postoperative nutritional interventions should still be implemented.

In all specialties, the risks of surgical intervention must be weighed against the risk of the pathology’s natural history to the patient. In the case of sAVSs, the risks of nonoperative management are drastic. 90% of all patients with untreated sAVS will be unable to ambulate independently, and 50% will have significant neurological impairment within 3 years of symptom onset.[ 2 , 4 , 8 , 13 , 16 , 39 ] The neurological prognosis is particularly poor in the event of hemorrhage, which carries an estimated risk of 4% per year.[ 16 ] Due to these devastating consequences of sAVS going untreated, open surgical ligation has become the gold standard.[ 2 , 11 , 22 , 39 ] Intra-operative adjuncts, such as neuro-monitoring, as well as surgical techniques to limit operative time and blood loss are important factors that can enhance surgical outcomes. As endovascular technology continues to advance, the use of adjunctive or stand-alone embolization has increased.[ 11 , 20 , 34 , 41 ] Despite encouraging results, there are also concerns over the rate of incomplete obliteration and long-term durability when used as an isolated management strategy.[ 38 , 39 , 41 , 43 ] It is important to note, however, that multiple sub-type of sAVSs exist, and surgical nuances to treat the underlying pathology differ from patient to patient. Nonetheless, the current study further supports the uniform use of surgical intervention in the case of sAVS, as rates of 30-day readmission, RTOR, and postoperative complications are low. The risk of a postoperative complication in this cohort of 363 patients is certainly lower than the 50% risk of significant neurological injury with nonoperative management.[ 2 , 8 ]

These data represent the first study in which preoperative risks were associated with 30-day unplanned readmission, RTOR, nonhome discharge, and postoperative complications following surgery for sAVS. In addition, it is the largest sample size extracted from the NSQIP database and represents the most recent cohort to date. These data provide valuable insight into the growing knowledge of sAVS and their surgical treatment, while expanding upon previously published data. In addition, it provides surgeons with data to provide more comprehensive and analytical preoperative counseling to patients with sAVS.

There are several limitations to consider when interpreting these results. First is the retrospective nature of our cohort. Although a limitation, utilizing a prospectively collected population-based database retrospectively allows for an adequate sample size in otherwise rare pathology. For example, with the exception of Varshneya et al.[ 41 ] and Bhimani et al.,[ 8 ] the largest available studies are limited to case reports and institutional reviews consisting of <110 patients compared to the current study’s 363.[ 12 , 19 , 22 , 23 , 30 , 36 - 38 , 44 ] While bolstering significant power, NSQIP data lack granularity at times. For example, details regarding if there were intra-operative neuromonitoring changes, the exact surgical techniques used, the detailed neurological status of the patient, and the reason for readmission are missing, which limits the ability to understand further where improvement in postoperative care can be made and control for confounding factors. In addition, a significant number of cases were missing the reason for RTOR. Postoperative neurological status is also an important variable that is lacking within these data, which would provide helpful inside into the cause of non-home discharges. It is important to note that ORs tend to overestimate risk, and this must be considered when interpreting the results of this study. Finally, sAVSs are heterogenous pathology; however, the Kim-Spetzler classification of each lesion is not available in NSQIP data, limiting our ability to control for this heterogeneity.

Disclosures

The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.

CONCLUSION

Preoperative hypoalbuminemia increases the odds of 30-day readmissions, nonhome discharges, and postoperative complications following surgery for sAVSs. In addition, 5.0% had a 30-day readmission, 5.0% had RTOR, 39.9% had a nonhome discharge, and 14.3% suffered a postoperative complication. The most common postoperative complications were UTI (n = 16), transfusion (n = 12), and venous thrombosis requiring treatment (n = 11). 30-day readmission and RTOR are associated with increased rates of postoperative complications, while 30-day readmissions, RTOR, and postoperative complications resulted in more nonhome discharges. These data stress the importance of optimizing preoperative nutritional status to maximize surgical outcomes.

Ethical approval:

Institutional Review Board approval is not required as it is retrospective study.

Declaration of patient consent:

Patient’s consent is not required as there are no patients in this study.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript, and no images were manipulated using AI.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.

References

1. Abecassis ZA, Emerson S, Zhang F, Ghodke B, Sekhar LN. Embolization and resection of epidural spinal arteriovenous malformations: Case report. Oper Neurosurg (Hagerstown). 2023. 24: e50-4

2. Alkhaibary A, Alharbi A, Alnefaie N, Alammar H, Arishy AM, Alghanim N. Spinal dural arteriovenous fistula: A comprehensive review of the history, classification systems, management, and prognosis. Chin Neurosurg J. 2024. 10: 2

3. Alluri RK, Leland H, Heckmann N. Surgical research using national databases. Ann Transl Med. 2016. 4: 393

4. Aminoff MJ, Logue V. The prognosis of patients with spinal vascular malformations. Brain. 1974. 97: 211-8

5. Arnone GD, Esfahani DR, Papastefan S, Rao N, Kumar P, Slavin KV. Diabetes and morbid obesity are associated with higher reoperation rates following microvascular decompression surgery: An ACS-NSQIP analysis. Surg Neurol Int. 2017. 8: 268

6. Asthana S, Bajaj P, Staub J, Workman C, Khazanchi R, Reyes S. Comparison of RVU reimbursement in anterior or posterior approach for single-and multilevel cervical spinal fusion. Clin Spine Surg. 2025. 38: E141-4

7. Bayman EO, Dexter F. Multicollinearity in logistic regression models. Anesth Analg. 2021. 133: 362-5

8. Bhimani AD, Rosinski CL, Patel S, Chaudhry NS, Denyer S, Behbahani M. Adult spinal arteriovenous malformations: natural history and a multicenter study of short-term surgical outcomes. World Neurosurg. 2019. 132: e290-6

9. Brinjikji W, Colombo E, Cloft HJ, Lanzino G. Clinical and imaging characteristics of spinal dural arteriovenous fistulas and spinal epidural arteriovenous fistulas. Neurosurgery. 2021. 88: 666-73

10. Camino-Willhuber G, Tani S, Schonnagel L, Caffard T, Haffer H, Chiapparelli E. Association of frailty and preoperative hypoalbuminemia with the risk of complications, readmission, and mortality after spine surgery. World Neurosurg. 2023. 174: e152-8

11. El-Hajj VG, Daller C, Fletcher-Sandersjöö A, Gharios M, Bydon M, Söderman M. The negative impact of treatment delays on the long-term neurological outcomes of spinal dural arteriovenous fistulas: A longitudinal cohort study. Neurosurg Focus. 2024. 56: E14

12. Elkordy A, Endo T, Sato K, Sonoda Y, Takahashi A, Tominaga T. Exclusively epidural spinal metameric arteriovenous shunts: Case report and literature review. Spine J. 2015. 15: e15-22

13. Flores BC, Klinger DR, White JA, Batjer HH. Spinal vascular malformations: Treatment strategies and outcome. Neurosurg Rev. 2017. 40: 15-28

14. Gelfand Y, Benton JA, Longo M, de la Garza Ramos R, Berezin N, Nakhla JP. Comparison of 30-day outcomes in patients with cervical spine metastasis undergoing corpectomy versus posterior cervical laminectomy and fusion: A 2006-2016 ACS-NSQIP database study. World Neurosurg. 2021. 147: e78-84

15. Gelfand Y, De la Garza Ramos R, Nakhla JP, Echt M, Yanamadala V, Yassari R. Predictive value of hypoalbuminemia and severe hypoalbuminemia in oncologic spine surgery. Clin Neurol Neurosurg. 2021. 210: 107009

16. Gross BA, Du R. Spinal glomus (type II) arteriovenous malformations: A pooled analysis of hemorrhage risk and results of intervention. Neurosurgery. 2013. 72: 25-32

17. Hall BL, Hamilton BH, Richards K, Bilimoria KY, Cohen ME, Ko CY. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: an evaluation of all participating hospitals. Ann Surg. 2009. 250: 363-76

18. Hardy EJ, Deane CS, Lund JN, Phillips BE. Loss of muscle mass in the immediate post-operative period is associated with inadequate dietary protein and energy intake. Eur J Clin Nutr. 2023. 77: 503-5

19. Hiramatsu M, Ishibashi R, Suzuki E, Miyazaki Y, Murai S, Takai H. Incidence and clinical characteristics of spinal arteriovenous shunts: Hospital-based surveillance in Okayama, Japan. J Neurosurg Spine. 2022. 36: 670-7

20. Ioannidis I, Kalogeras A, Tasiou A, Vlychou M, Fountas KN. Coil embolization of a ruptured anterior spinal artery aneurysm associated with spinal cord arteriovenous malformation. Neurointervention. 2024. 19: 190-3

21. Jean RA, DeLuzio MR, Kraev AI, Wang G, Boffa DJ, Detterbeck FC. Analyzing risk factors for morbidity and mortality after lung resection for lung cancer using the NSQIP database. J Am Coll Surg. 2016. 222: 992-1000.e1

22. Jing L, Su W, Guo Y, Sun Z, Wang J, Wang G. Microsurgical treatment and outcomes of spinal arteriovenous lesions: Learned from consecutive series of 105 lesions. J Clin Neurosci. 2017. 46: 141-7

23. Kalani MA, Choudhri O, Gibbs IC, Soltys SG, Adler JR, Thompson PA. Stereotactic radiosurgery for intramedullary spinal arteriovenous malformations. J Clin Neurosci. 2016. 29: 162-7

24. Kang H. The prevention and handling of the missing data. Korean J Anesthesiol. 2013. 64: 402-6

25. Kim LJ, Spetzler RF. Classification and surgical management of spinal arteriovenous lesions: Arteriovenous fistulae and arteriovenous malformations. Neurosurgery. 2006. 59: S195-201

26. King S, Calisi O, Caldwell C, Berger D, Rich AM, Dan Y. Frequency and predictors of preoperative cardiac testing overuse in low-risk patients before laparoscopic bariatric surgery. Am J Cardiol. 2023. 186: 181-5

27. Lakomkin N, Sathiyakumar V, Wick B, Shen MS, Jahangir AA, Mir H. Incidence and predictive risk factors of postoperative sepsis in orthopedic trauma patients. J Orthop Traumatol. 2017. 18: 151-8

28. Lee JM, Park JH, Park JC, Ahn JS, Park W. Treatment of conus medullaris arteriovenous malformation: The role of microsurgical treatment. J Neurosurg Spine. 2024. 41: 115-21

29. Li X, Li H, Huang S, Pan Y. Association between hypoalbuminemia and complications after degenerative and deformity-correcting spinal surgeries: A systematic review and meta-analysis. Front Surg. 2022. 9: 1030539

30. Lv X, Li Y, Yang X, Jiang C, Wu Z. Endovascular embolization for symptomatic perimedullary AVF and intramedullary AVM: A series and a literature review. Neuroradiology. 2012. 54: 349-59

31. Mohamed A, Sheehan C, Kaur P, Schwab F, Butler A. Preoperative serum albumin and TLC as predictors of postoperative complications in spine surgery. Clin Spine Surg. 2024. 37: 467-71

32. O’Reilly ST, Hendriks EJ, Brunet MC, Itsekson Z, Shahrani RA, Agid R. Recognition of the variant type of spinal dural arteriovenous fistula: A rare but important consideration. J Neurosurg Spine. 2022. 37: 624-8

33. Piper KF, Tomlinson SB, Santangelo G, Van Galen J, DeAndrea-Lazarus I, Towner J. Risk factors for wound complications following spine surgery. Surg Neurol Int. 2017. 8: 269

34. Pizzuto S, Cortese J, Sgreccia A, Di Maria F, Consoli A, Rodesch G. Endovascular approach to posterior spinal cord AV shunts via the anterior spinal artery. J Neuroradiol. 2024. 51: 101207

35. Qureshi R, Rasool M, Puvanesarajah V, Hassanzadeh H. Perioperative nutritional optimization in spine surgery. Clin Spine Surg. 2018. 31: 103-7

36. Rangel-Castilla L, Russin JJ, Zaidi HA, Martinez-Del-Campo E, Park MS, Albuquerque FC. Contemporary management of spinal AVFs and AVMs: Lessons learned from 110 cases. Neurosurg Focus. 2014. 37: E14

37. Singh B, Behari S, Jaiswal AK, Sahu RN, Mehrotra A, Mohan BM. Spinal arteriovenous malformations: Is surgery indicated?. Asian J Neurosurg. 2016. 11: 134-42

38. Singh K, Zaben M, Manivannan S, Van Beijnum J, Galea J, Zilani G. Endovascular and surgical obliteration rates of spinal dural arteriovenous fistulae: A single UK Centre experience. Br J Neurosurg. 2023. 37: 1613-8

39. Takai K, Endo T, Yasuhara T, Seki T, Watanabe K, Tanaka Y. Neurosurgical versus endovascular treatment of spinal dural arteriovenous fistulas: A multicenter study of 195 patients. J Neurosurg Spine. 2021. 34: 514-21

40. Turrentine FE, Sohn MW, Jones RS. Congestive heart failure and noncardiac operations: Risk of serious morbidity, readmission, reoperation, and mortality. J Am Coll Surg. 2016. 222: 1220-9

41. Varshneya K, Pendharkar AV, Azad TD, Ratliff JK, Veeravagu A. A Descriptive analysis of spinal cord arteriovenous malformations: Clinical features, outcomes, and trends in management. World Neurosurg. 2019. 131: e579-85

42. Xie J, Liu H, Deng S, Niu T, Wang J, Wang H. Association between immediate postoperative hypoalbuminemia and surgical site infection after posterior lumbar fusion surgery. Eur Spine J. 2023. 32: 2012-9

43. Yuan CW, Wang YJ, Zhang SJ, Shen SL, Duan HZ. Clinical outcomes following microsurgery and endovascular embolization in the management of spinal dural arteriovenous fistula: A meta-analysis study. Beijing Da Xue Xue Bao Yi Xue Ban. 2022. 54: 304-14

44. Zhang HJ, Silva N, Solli E, Ayala AC, Tomycz L, Christie C. Treatment options and long-term outcomes in pediatric spinal cord vascular malformations: A case report and review of the literature. Childs Nerv Syst. 2020. 36: 3147-52

45. Zyck S, Davidson CL, Sampath R, editors. Arteriovenous malformations of the central nervous system. StatPearls. Treasure Island, FL: StatPearls Publishing; 2025. p.