- Email: [email protected]
Short-Term Outcomes of Staged Versus Same-Day Surgery for Adult Spinal Deformity Correction
Spine Deformity 7 (2019) 796e803 www.spine-deformity.org
Short-Term Outcomes of Staged Versus Same-Day Surgery for Adult Spinal Deformity Correction Alexander H. Arzeno, MD*, Jayme Koltsov, PhD, Todd F. Alamin, MD, Ivan Cheng, MD, Kirkham B. Wood, MD, Serena S. Hu, MD Department of Orthopaedic Surgery, Stanford University School of Medicine, 450 Broadway St, Pavillion A FL 1 MC6110, Redwood City, CA, 94063, USA Received 3 February 2018; revised 21 October 2018; accepted 17 December 2018
Abstract Study Design: Retrospective cohort study. Objectives: Assess differences between staged (<3 days) and same-day surgery in perioperative factors, radiographic measures, and complications. Summary of Background Data: Surgical adult spinal deformity correction may require combined anterior and posterior approaches. To modulate risk, some surgeons perform surgery that is expected to be longer and/or more complex in two stages. Prior studies comparing staged (>7 days) and same-day surgery demonstrated mixed results and none have examined results with shorter staging intervals. Methods: Retrospective review of adults undergoing combined anterior/posterior approaches for spinal deformity over a 3-year period at a single institution (n592). Univariate differences between staged and same-day surgery were assessed with chi-squared, Fisher exact, and Mann-Whitney U tests. Generalized estimating equations assessed whether differences in perioperative outcomes between groups remained after adjusting for differences in demographic and surgical characteristics. Results: In univariate analyses, staged surgery was associated with a length of stay (LOS) 3 days longer than same-day surgery (9.2 vs. 6.3 days, p ! .001), and greater operative time, blood loss, transfusion requirement, and days in intensive care unit ( p ! .001 for each). Staged surgery had a higher rate of thrombotic events ( p 5 .011) but did not differ in readmission rates or other complications. Radiographically, improvements in Cobb angle (average 13 vs. 17 , p 5 .028), lumbar lordosis (average 14 vs. 23 , p 5 .019), and PI-LL mismatch (average 10 vs. 2 p 5 .018) were greater for staged surgery, likely related to more extensive use of osteotomies in the staged group. After risk adjustment, taking into account the procedural specifics including longer fusion constructs and greater number of osteotomies, LOS no longer differed between staged and same-day surgery; however, the total operative time was 98 minutes longer for staged surgery ( p ! .001). Differences in blood loss between groups was accounted for by differences in operative time and patient and surgical characteristics. Conclusions: Although univariate analysis of our results were in accordance with previously published works, multivariate analysis allowing individual case risk adjustment revealed that LOS was not significantly increased in the staged group as reported in previous studies. There was no difference in infection rates as previously described but an increase in thrombotic events was observed. Level of Evidence: Level III. Ó 2019 Scoliosis Research Society. All rights reserved. Keywords: Adult spine deformity; Fusion; Staged; Complications
Author disclosures: AHA (none), JK (none), TFA (personal fees from Disc Diagnostics, Empirical Spine, Globus Medical, and Spinal Elements, outside the submitted work), IC (personal fees from Spinal Cyte, Spine Wave, Globus Medical, and SpineCraft, outside the submitted work; in addition, he has a patent by NuVasive with royalties paid; including fees from AAOS [board or committee member], Cervical Spine Research Society [board or committee member], Scoliosis Research Society [board or committee member], and Spine: The Spine Journal [editorial or governing board]), KBW (personal fees from TranS1 and Alphatec Spine, outside the submitted work; in addition, he has a patent by Globus Medical with royalties paid; including fees from Spine [editorial or governing board] and
Spine: Journal of Spinal Deformity [editorial or governing board]), SSH (personal fees from NuVasive and DePuy, A Johnson & Johnson Company, outside the submitted work; in addition, she has a patent by NuVasive with royalties paid; including fees from the American Orthopaedic Association [board or committee member] and Scoliosis Research Society [board or committee member]).IRB approval: Research approved by Stanford Institutional Review Board/Privacy Office. *Corresponding author. Department of Orthopaedic Surgery, 300 Pasteur Drive, Room R144, Stanford, CA 94305-534, USA. Tel.: (650) 7217638; fax: (650) 498-6899. E-mail address: [email protected] (A.H. Arzeno).
2212-134X/$ - see front matter Ó 2019 Scoliosis Research Society. All rights reserved. https://doi.org/10.1016/j.jspd.2018.12.008
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
797
Introduction
Materials and Methods
Adult spinal deformity has become increasingly recognized as a cause of disability as our population ages and physical demands continue to increase [1]. Although adult spinal deformity can be managed conservatively, operative intervention is often considered for deformity progression or if symptoms worsen or cannot be improved with conservative management [2]. Surgical management aims to decompress central and foraminal stenosis, correct and stabilize the deformity to improve function and decrease pain, and reduce the subsequent need for analgesics [1,3,4]. Because of the complex nature of spinal deformities, the surgical correction can be lengthy, highly invasive, and may require a combination of surgical approaches [5-9]. Complication rates are between 40% and 80%, with major perioperative complications reported at 20% [5,10-12]. These major complications, including surgical site infection, neurologic deficit, pulmonary embolism (PE), and myocardial infarction (MI), are associated with prolonged length of stay (LOS) and higher rates of intensive care unit admissions [13,14]. For longer surgeries, particularly when different approaches are used, some surgeons consider performing the operation in a staged fashion. Most commonly, the anterior approach is performed on the first day and the posterior approach is performed several days later, dividing the physiologic insult and allowing interval recovery and resuscitation [11,15-17]. Same-day surgery has been associated with shorter LOS relative to staged surgery in previous studies, but estimated blood loss (EBL) and complication rate results are mixed [18-21]. In these studies, all staging was greater than 5 days, and most were over 7 days. Longer periods of staging directly increases hospital length of stay and possibly increases the risks for other complications due to longer periods of immobilization [22]. In addition, nutritional depletion often occurs after surgery, which has been shown to result in higher infection rates for staged procedures [23-25]. Shorter intervals between stages of adult spinal deformity cases may mitigate some of the risks associated with staged surgery and the increased hospitalization length, while still having the benefit of partitioning the physiologic stress of complex and extensive surgery. However, to date, no studies have examined the effects of shorter staging on LOS or perioperative outcomes. Therefore, the purpose of this study is to compare the outcomes of adult spinal deformity correction requiring more than one approach between same-day surgery versus staged surgery with an interval of three or fewer days. Outcomes assessed included 30-day complications and readmissions, LOS, perioperative outcomes, and radiographic deformity correction. We hypothesized that there would be similar outcomes, with no increase in complication rate in the staged group.
This retrospective cohort study was approved by the Institutional Review Board, and no external funding was obtained. All cases with both anterior (including lateral trans-psoas and anterolateral approaches) and posterior procedures from four surgeons at a single academic practice between January 1, 2014, and December 31, 2016, were reviewed. Inclusion criteria were as follows: age greater than 18 years, posterior fusion of at least five levels in the thoracic/lumbar/sacral spine with at least one interbody fusion performed via a separate incision on either the same day or within three days, and a minimum postoperative follow-up of six months. Surgeries for acute spinal trauma, malignancy, infection, or spondyloarthropathy were excluded. A total of 92 patients, 47 same day and 45 staged, met the inclusion and exclusion criteria (Table 1). Primary outcomes collected from the electronic medical record were complications within 30 days (including medical complications, neurologic deficits, infection, proximal junctional kyphosis, pseudoarthrosis, return to operating room), readmissions within 30 days, LOS, total operative time, and EBL. Secondary surgical and admission outcomes collected from the electronic medical record included total anesthetic time, transfusions, use of clotting factors (freshfrozen plasma, platelets), days of intensive care unit admission, and reoperations within 6 months. Additionally, pre- and postoperative radiographs were reviewed for all cases, and the following radiographic parameters were recorded: Cobb angle, center sacral vertical line, pelvic incidence (PI), lumbar lordosis (LL), PI-LL mismatch, pelvic tilt, sacral slope, and sagittal vertical axis. Sagittal modifiers, as described and correlated to health-related quality of life in SRS-Schwab classification, were also organized by grade [17] (Table 2). Postoperative measures were obtained at hospital discharge, 6 weeks postoperatively, and 3, 6, 12, and 24 months postoperatively. Baseline demographic data, including age, sex, body mass index, American Society of Anesthesiologists score, surgical diagnosis, history of prior instrumented fusion, and specific comorbidities (diabetes, history of smoking, history of MI, history of cerebrovascular accident) was collected from the electronic medical record. The approach, number of anterior and posterior instrumented levels, and other
Table 1 Inclusion and exclusion criteria. Inclusion criteria - Age O18 - Both anterior (including lateral and anterolateral approaches) and posterior fusion of at least 5 levels in thoracic/ lumbar/sacral spine - Procedure performed same day or staged ! 5 days apart - Minimum follow- up 6 months
Exclusion criteria - Diagnosis of acute spinal column trauma (fracture or dislocation) - Spinal malignancy, primary or metastatic - Spondyloarthropathies - Spinal osteomyelitis
798
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Table 2 Sagittal modifier grades. Grade 0 1 2
PI-LL Mismatch
!10 10 e20 O20
SVA
PT
!4 cm 4e9.5 cm O9.5 cm
!20 20 e30 O30
Table modified from the SRS-Schwab classification [26]. LL, lumbar lordosis; PI, pelvic incidence; PT, pelvic tilt; SVA, sagittal vertical axis.
procedural specifics (including use of osteotomies and decompression) were also recorded for each case. Univariate differences in categorical variables between staged and same-day surgery were assessed with chisquared and Fisher exact tests. Univariate differences in continuous measures between staged and same-day surgery were assessed with independent samples t tests, or MannWhitney U tests if nonnormally distributed. Generalized estimating equations were used to assess whether stages differed from same-day surgery in LOS, blood loss, and operative time after adjusting for key differences in patient and surgical characteristics. Variables with p ! .25 from the univariate tests for staged versus same-day surgery were considered as candidates in the multivariable models. The initial models included all candidate variables, and subsequently variables were removed stepwise in order of decreasing p value until only trending ( p ! .10) or significant ( p ! .05) terms remained. All analyses were performed in SAS, version 9.4 (SAS Institute, Cary, NC), with a two-sided level of significance of a 5 0.05. Based on previous reports of LOS and complications with staged (O5e7 days) and same-day spinal deformity correction, the 92 patients provided more than 80% power (with a 5 0.05) to detect a difference between groups of 3 days in LOS or 20% in the proportion having 30-day complications. Results Patient demographics did not differ significantly between groups (Table 3). Staged surgery had a much greater proportion of interbody fusion using an anterior approach, relative to the same-day surgery (84.4% vs. 46.8%, p ! .001; Table 3). The staged surgeries also had a greater number of Smith-Peterson osteotomies (87% vs. 51%, p ! .001), three-column osteotomies (including vertebral column resection or pedicle subtraction osteotomy) (16% vs. 2%, p 5 .029), O-arm (64% vs. 13%, p ! .001), neuromonitoring (91% vs. 21%, p ! .001), and decompression (76% vs. 34%, p ! .001) relative to the same-day group. The staged group had a greater number of levels fused relative to the same-day group (9 vs. 8, p ! .001). Perioperative outcomes The LOS for staged surgery (9.2 days, 95% CI 7.7, 10.6), was nearly three days longer than that for same-day
Table 3 Univariate comparisons of baseline parameters between same-day and staged surgery. Same-day Demographics Sex (female) Age, mean (95% CI) BMI, mean (95% CI) Smoking ASA class 1 2 3 Comorbidities Diabetes MI Stroke DVT/PE Prior instrumented surgery Prior spine surgery Surgical details Days between surgeries 1 2 3 Approach Anterior Lateral Anterolateral TLIF Ponte osteotomy Three- column osteotomy O-arm Neuromonitoring Decompression No. of posterior levels fused, mean (95% CI) No. of interbody levels, mean (95% CI) No. of osteotomy levels, mean (95% CI) No. of decompression levels, mean (95% CI) Discharge disposition Home Rehabilitation facility Skilled nursing facility
30 68 25.3 17
(63.8%) (61, 78) (23.1, 30.4) (36.2%)
Staged 34 68 25.5 19
(75.6%) (62, 72) (22.8, 28.7) (42.2%)
p-value .222 .122 .227 .410
1 (2.1%) 24 (51.1%) 22 (46.8%)
d 19 (42.2%) 26 (57.8%)
.414
3 (6.4%) d 4 (8.5%) 4 (8.5%) 16 (34%)
8 (17.8%) 2 (4.5%) 6 (13.6%) d 22 (48.9%)
.116 .231 .514 .118 .148
24 (51.1%)
27 (60%)
.389
d d d
2 (4.4%) 34 (75.6%) 9 (20.0%)
N/A
22 19 6 7 24 1
(46.8%) (40.4%) (12.8%) (14.9%) (51.1%) (2.1%)
38 (84.4%) 16 (35.6%) d d 39 (86.7%) 7 (15.6%)
!.001*
6 10 16 8
(12.8%) (21.3%) (34.0%) (5, 9)
29 41 34 9
!.001* !.001* !.001* !.001*
(64.4%) (91.1%) (75.6%) (9, 9)
.012* !.001* .029*
2 (2, 3)
3 (2, 3)
.506
1 (0, 3)
4 (3, 6)
!.001*
0 (0, 2)
2 (1, 2)
!.001*
.338 19 (40.4%) 3 (6.4%)
17 (38.6%) 7 (15.9%)
25 (53.2%)
20 (45.5%)
ASA, American Society of Anesthesiologists; BMI, body mass index; DVT, deep vein thrombosis; MI, myocardial infarction; PE, pulmonary embolism; TLIF, transforaminal lumbar interbody fusion. Values are n (%) unless otherwise noted. * p ! .05.
surgery (6.3, 95% CI 5.4, 7.2) ( p ! .001; Figure 1). The longer LOS for staged surgery was largely due to the interval between the first and second surgery, which was two
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
799
Fig. 1. Total (A) LOS, (B) operative time, and (C) anesthesia time, and (D) blood loss for same-day and staged surgery (raw, unadjusted values). Bar graphs show means and 95% confidence intervals. Central lines in box plots represent medians, whereas central markers indicate means. *p ! .05 from univariate tests. (LOS, length of stay; Op, operative.)
days for more than 75% of the staged cases, with a range of one to three days. The total operative time and EBL were substantially greater for staged versus same-day surgery ( p ! .001 for both; Figure 1). Secondary perioperative outcomesdpRBC units, pRBC units, platelets, fresh-frozen plasma, and days in the intensive care unitdwere also greater for staged relative to same-day surgery ( p ! .001 for each). After adjustment for patient demographics and surgical characteristics using multivariate regression models, the LOS for staged surgery was 0.8 (95% CI e0.8, 2.4) days longer for staged relative to same-day surgery, and this difference was not statistically significant ( p 5 .303; Supplemental Table 1). Staged surgery was associated with 98.4 (95% CI 55.8, 141.1) minutes longer total operative time ( p ! .001) and a trend toward greater blood loss (1,023.5 mL, 95% CI e154.1, 2201.2, p 5 .088) after risk adjustment (Supplemental Table 2). The trend toward greater blood loss with staged surgery was fully explained by the longer operative time, as adding this into the risk adjustment for the blood loss model reduced the blood-loss difference between staged versus same-day surgery to less than 50 mL ( p 5 .939). Complications The 30-day complication rate for staged surgery was 26.7%, and for same-day surgery was 14.9% (Table 4); however, this difference was not statistically significant ( p 5 .163). When looking at individual complications,
13.6% of the staged surgery patients had a DVT (deep vein thrombosis) or PE (pulmonary embolism), whereas none of the same-day patients experienced this complication ( p 5 .011). Of the six who developed a DVT/PE, five were in cases that had an anterior lumbar interbody fusion as a part of their operation. The rate of DVT/PE in anterior lumbar interbody fusion patients was 7.6% (5/66 patients), whereas the rate in surgeries with no anterior lumbar interbody fusion was 3.8% (1/26 patients); however, our sample was underpowered to attribute the difference to the difference in approaches. In contrast, rates of infection, proximal junctional kyphosis, pseudarthrosis, and revision surgery differed by only 2% to 4% between groups and were not statistically significant. Readmission rates were 10.5% greater for same-day versus staged surgery, but this difference was not statistically significant ( p 5 .112). Radiographic outcomes Preoperatively, there were no differences in radiographic parameters between groups (Figure 2). Postoperatively, sagittal vertical axis was 25.6 mm lower in staged relative to same-day surgery ( p 5 .024), and the postoperative mismatch between pelvic incidence and lumbar lordosis (PI-LL mismatch) was 8.8 degrees less for staged patients ( p 5 .002). Regarding the pre- to postoperative change in radiographic parameters, patients had a larger change in Cobb angle ( p 5 .028), lumbar lordosis ( p 5 .019), pelvic tilt ( p 5 .008), and PI-LL mismatch ( p 5 .018) with staged surgery when compared with same-day surgery. The total
800
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Table 4 Univariate comparisons of adverse events and SRS-Schwab Classification sagittal parameters between same-day and staged surgery. Same-day, n (%)
Staged, n (%)
Adverse events Reoperation 7 (14.9%) 5 Readmission 6 (12.8%) 1 Any complication 7 (14.9%) 12 Dural tear 2 (4.3%) 8 Neurologic d 4 defiecit DVT/PE d 6 Infection 3 (6.4%) 2 PJK 2 (4.3%) 3 Pseudoarthrosis 2 (4.3%) Sagittal imbalance (SRS-Schwab classification) PI minus LL Pre 0 10 (22.2%) 10 1 9 (20.0%) 7 2 26 (57.8%) 26 Post 0 23 (52.3%) 34 1 13 (29.5%) 7 2 8 (18.2%) 1 Change 2 7 (15.9%) 19 1 16 (36.4%) 11 0 21 (47.7%) 11 SVA Pre 0 2 (5.1%) 1 1 1 (2.6%) 2 2 36 (92.3%) 39 Post 0 4 (10.8%) 11 1 1 (2.7%) 2 2 32 (86.5%) 27 Change 2 3 (8.1%) 10 1 1 (2.7%) 1 0 31 (83.8%) 27 1 1 (2.7%) 1 2 1 (2.7%) PT Pre 0 12 (26.7%) 9 1 14 (31.1%) 15 2 19 (42.2%) 19 Post 0 11 (25.0%) 15 1 20 (45.5%) 18 2 13 (29.5%) 9 Change 2 1 (2.3%) 4 1 9 (20.5%) 10 0 28 (63.6%) 24 1 6 (13.6%) 3
p- value
(11.4%) (2.3%) (26.7%) (17.8%) (8.9%)
.619 .112 .163 .048* .053
(13.6%) (4.5%) (6.8%) d
.011* O.999 .670 .495
(23.3%) (16.3%) (60.5%)
.903
(81.0%) (16.7%) (2.4%)
.009*
(46.3%) (26.8%) (26.8%)
.009*
(2.4%) (4.8%) (92.9%)
.865
(27.5%) (5.0%) (67.5%)
.122
(25.6%) (2.6%) (69.2%) (2.6%) d
.141
(20.9%) (34.9%) (44.2%)
.812
(35.7%) (42.9%) (21.4%)
.496
(9.8%) (24.4%) (58.5%) (7.3%)
.432
changes were considerably higher than for the same-day group ( p ! .001). Looking at sagittal imbalance modifiers as classified by the SRS-Schwab classification, the staged and same-day patients were similar preoperatively, with no differences in grade of PI-LL mismatch, sagittal vertical axis, or pelvic tilt. Postoperatively, staged surgery resulted in a larger change in PI-LL mismatch grade ( p 5 .009), with 46.3% of staged patients having a decrease of 2 grades, compared with 15.9% of same-day patients. Thus, staged surgery had more patients returned to a PI-LL mismatch grade of 0 (80.0%) compared with same-day surgery (52.3%, p 5 .009). When looking only at those patients who were sagittally imbalanced before surgery, the findings were similar, with staged surgery patients having a larger change in PI-LL mismatch grade (54.3% staged vs. 18.9% same-day decreased two classes, p 5 .002) and having restored sagittal balance (71.4% staged vs. 35.1% same-day PI-LL mismatch grade 0, p 5 .005) postoperatively relative to same-day surgery patients.
Discussion
DVT, deep vein thrombosis; LL, lumbar lordosis; PE, pulmonary embolism; PI, pelvic incidence; PJK, proximal junctional kyphosis; PT, pelvic tilt; SVA, sagittal vertical axis. * p ! .05.
Operative interventions for adult spinal deformity correction are extensive surgical procedures that can carry a relatively high risk for complications. However, for patients with significant disability due to their spinal condition, surgery can result in improved functional outcomes when compared with medical/interventional management [1]. The surgeon and the patient must carefully weigh the benefits of undergoing surgical correction compared with the risks of intervention. It is imperative that we continue to try and maximize benefit while decreasing risk. Staged versus sameday procedures for those that require both anterior and posterior procedures is one such variable. Several studies have previously examined the effects of staging of surgical procedures in spinal deformity correction; however, mixed patient populations, surgical technique, and varying staging protocols have been used and, thus, results have varied. No study has compared staging with this brief an interval and we hypothesized that LOS and potential complications would be similar between groups. A better understanding of the risks and benefits associated with staged procedures will allow more informed decision making and surgical planning for both the surgeon and the patient. In this study, we examined adult patients who underwent spinal deformity correction requiring at least five levels as well as both anterior and posterior fusion through separate approaches. Patients whose procedure had been performed the same day in a sequential manner under the same anesthetic were compared to those who were staged three or fewer days apart. After taking into account the influence of other variables such as patient demographics, comorbidities, and
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
801
Fig. 2. Radiographic results. (A) Cobb angle, (B) pelvic tilt, (C) lumbar lordosis, (D) pelvic incidence-lumbar lordosis mismatch, and (E) sagittal vertical axis. *, pre- to postoperative change in parameter differs between same-day and staged surgery ( p ! .05); þ, postoperative parameter differs between sameday and staged ( p ! 0.05). (LL, lumbar lordosis; PI, pelvic incidence; PT, pelvic tilt; SVA, sagittal vertical axis.)
procedural specifics, we found no difference in EBL attributable to staging alone. Although there was a trend toward greater EBL with staged surgery, after risk adjustment, the difference was explained by the 98 additional minutes of operative time required for staged surgery. Previous studies by Shufflebarger [18] and Tsirikos [19] reported greater blood loss with staged surgeries, as did a more recent study by Maddox using a posterior-only approach. These studies all performed the second stage of their surgery at approximately one week. Tsirikos and Shen [20] interestingly found no increase in EBL with staged when comparing to a longer period of staging. The impact of staging on LOS is also interesting. Surprisingly, we found that LOS was not significantly different (less than one day) between the two groups after adjusting for other factors. Thus, despite the additional two to three days between stages, staging did not result in further prolongation of LOS when patient and operative characteristics were taken into consideration. In those studies where procedures were staged for more than one week without interval discharge, it makes intuitive sense that LOS would thus be longer because of an additional week in hospital as demonstrated in Shufflebarger and Shen. In the study by
Tsirikos, there was no difference in LOS despite a four- to seven-day staging protocol and, similarly, in the study by Hassanzadeh [21] in which patients were discharged between stages, no difference was found. The overall complication rate in our study did not differ between groups, nor did 30-day readmission and revision rates. Reported complication rates have been mixed in previous studies. Shufflebarger and Tsirikos reported higher complication rates in staged versus same-day protocols, whereas Hassanzadeh and Maddox reported no difference. Our study similarly found no difference in the overall complication rate between the two protocols. In particular, prior studies with longer intervals between staging reported higher rates of infection in staged surgery secondary to the two open operations as well as nutritional deficiency after staged surgeries [23-25]. However, in our study, we did not find a significantly increased risk of infection in staged surgery. In our study, we did find a higher rate of thrombotic events in the staged surgery group. However, 84.4% of staged cases were performed with an anterior or anterolateral approach versus 59.6% of same-day cases ( p ! .001), and anterior approaches have previously been associated with higher rates of DVT/PE [22]. In our study, the
802
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
sample size and small number of these events did not allow us to determine whether the increased DVT/PE rate in staged cases was attributable to the greater use of anterior or anterolateral approaches. The staged group also had significantly larger correction in radiographic parameters for both coronal and sagittal balance. Few studies have studied the radiographic results of staging spinal deformity surgery. Shufflebarger et al. found increased correction with a same-day operation, with a significantly higher percentage of coronal correction with same-day relative to staged surgery [18]. In our study, we measured more radiographic alignment variables, including those for sagittal balance, the significance of which is more recently appreciated. Contrary to Shufflebarger’s results, we found higher correction in Cobb angle in the staged group. Additionally, correction in the sagittal plane such as lumbar lordosis correction was greater and final PI-LL mismatch was lower in the staged group. In our study, the improved correction likely reflects the surgical goal of greater deformity correction for those patients, rather than the staging per se. This is evidenced by the greater use of osteotomies and number of fused levels in the staged group. The surgeries that were staged were, on average, more complex procedures, requiring more levels of decompression and fusion, more Smith Petersen osteotomies, and more three-column osteotomies. Hence, cases for which the surgeon decides are clinically indicated for more deformity correction require more complex procedures to achieve that correction, and are therefore more likely to be performed in a staged manner in our center. We did not make ageadjustments to the sagittal balance goals; however, there was no difference in ages between staged and sameday patients. Limitations of our study include the fact that it was a retrospective review. Operations were performed by four different surgeons and the distribution of staged versus same-day cases by surgeon was not uniform and reflected surgeon preferences. The goals of the surgeons in terms of deformity correction and desired postoperative radiographic goals were also likely to have been different, which may explain many of the differences seen in the two different groups such as the amount of planned correction and subsequent use of osteotomies, number of levels being fused, and the distribution of lateral versus anterior procedures as well as the radiographic outcomes achieved. Without a prospective randomized trial, it is impossible to control for the effect of the surgeon’s approach, training, experience, and biases. Additionally, we included different types of anterior interbody fusions and approaches. Although there is likely some degree of confounding, we attempted to account for these variances using a multivariate risk-adjusted analysis. The multivariate model allows us to attribute differences in outcomes based on each specific variable included in the model. When the variables that made the groups different were statistically accounted for (number of levels, osteotomies, etc.) were accounted
for, we were able to examine the influence of the staging alone on our outcomes. An additional limitation is that this study was powered for the primary, univariate analyses and was underpowered for the multivariable generalized estimating equation models. Although we attempted to maximize power by reducing the number of covariates through backwards selection, covariates not retained in the final models may be associated with these outcomes, but at a level not able to be detected with the present sample size. Further study will include long-term follow-up of both groups to examine radiographic as well as pain and patientreported outcomes. The higher rate of PE/DVTs for staged surgery patients is a cause for concern and we are considering additional prophylactic modalities for these patients. Conclusion When correcting adult spinal deformity with anterior and posterior approaches, it is reasonable to consider staging the operations if more extensive surgery is indicated and the procedures are anticipated to be of a long total duration. In our study, cases that were more complex and used more osteotomies to enable a greater deformity correction were more likely to be staged. We used a multivariate analysis to allow for individual case risk adjustment for these differences. We show that there was no additional increase in LOS that was attributable to the staging itself. Additionally, there was no difference in infection rates as previously described. An increase in thrombotic events was observed in the staged group, which we believe is likely attributable to the greater use of anterior approach for interbody fusions in the staged cohort. The staged group had significantly greater surgical correction of spinal imbalance when radiographic parameters were compared, likely because of the greater use of osteotomies in those patients. References [1] Kluba T, Dikmenli G, Dietz K, et al. Comparison of surgical and conservative treatment for degenerative lumbar scoliosis. Arch Orthop Trauma Surg 2009;129:1e5. [2] Glassman SD, Hamill CL, Bridwell KH, et al. The impact of perioperative complications on clinical outcome in adult deformity surgery. Spine (Phila Pa 1976) 2007;32:2764e70. [3] Wang G, Hu J, Liu X, Cao Y. Surgical treatments for degenerative lumbar scoliosis: a meta analysis. Eur Spine J 2015;24:1792e9. [4] Glassman SD, Schwab FJ, Bridwell KH, et al. The selection of operative versus nonoperative treatment in patients with adult scoliosis. Spine (Phila Pa 1976) 2007;32:93e7. [5] Cho KJ, Suk SI, Park SR, et al. Complications in posterior fusion and instrumentation for degenerative lumbar scoliosis. Spine (Phila Pa 1976) 2007;32:2232e7. [6] Good CR, Lenke LG, Bridwell KH, et al. Can posterior-only surgery provide similar radiographic and clinical results as combined anterior (thoracotomy/thoracoabdominal)/posterior approaches for adult scoliosis? Spine (Phila Pa 1976) 2010;35:210e8. [7] Charosky S, Guigui P, Blamoutier A, et al; Study Group on Scoliosis. Complications and risk factors of primary adult scoliosis surgery: a
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
[8]
[9]
[10]
[11] [12]
[13]
[14]
[15]
[16]
[17]
multicenter study of 306 patients. Spine (Phila Pa 1976) 2012;37: 693e700. Carreon LY, Puno RM, Dimar JR, et al. Perioperative complications of posterior lumbar decompression and arthrodesis in older adults. J Bone Joint Surg Am 2003;85-A:2089e92. Sansur CA, Smith JS, Coe JD, et al. Scoliosis Research Society morbidity and mortality of adult scoliosis surgery. Spine (Phila Pa 1976) 2011;36:593. Maddox JJ, Pruitt DR, Agel J, Bransford RJ. Unstaged versus staged posterior-only thoracolumbar fusions in deformity: a retrospective comparison of perioperative complications. Spine J 2014;14:1159e65. Rhee JM, Bridwell KH, Lenke LG, et al. Staged posterior surgery for severe adult spinal deformity. Spine (Phila Pa 1976) 2003;28:2116e21. Daubs MD, Lenke LG, Cheh G, et al. Adult spinal deformity surgery: complications and outcomes in patients over age 60. Spine (Phila Pa 1976) 2007;32:2238e44. Howe CR, Agel J, Lee MJ, et al. The morbidity and mortality of fusions from the thoracic spine to the pelvis in the adult population. Spine (Phila Pa 1976) 2011;36:1397e401. Spivak JM, Neuwirth MG, Giordano CP, Bloom N. The perioperative course of combined anterior and posterior spinal fusion. Spine (Phila Pa 1976) 1994;19:520e5. Byrd JA, Scoles PV, Winter RB, et al. Adult idiopathic scoliosis treated by anterior and posterior spinal fusion. J Bone Joint Surg Am 1987;69:843e50. Passias PG, Ma Y, Chiu YL, et al. Comparative safety of simultaneous and staged anterior and posterior spinal surgery. Spine (Phila Pa 1976) 2012;37:247e55. Wright N. Single-surgeon simultaneous versus staged anterior and posterior spinal reconstruction: a comparative study. J Spinal Disord Tech 2005;18:48.
803
[18] Shufflebarger HL, Grimm JO, Bui V, Thomson JD. Anterior and posterior spinal fusion. Staged versus same-day surgery. Spine (Phila Pa 1976) 1991;16:930e3. [19] Tsirikos AI, Chang W, Dabney KW, Miller F. Comparison of onestage versus two-stage anteroposterior spinal fusion in pediatric patients with cerebral palsy and neuromuscular scoliosis. Spine (Phila Pa 1976) 2003;28:1300. [20] Shen J, Qiu G, Wang Y, et al. Comparison of 1-stage versus 2-stage anterior and posterior spinal fusion for severe and rigid idiopathic scoliosisda randomized prospective study. Spine (Phila Pa 1976) 2006;31:2525e8. [21] Hassanzadeh H, Gjolaj JP, El Dafrawy MH, et al. The timing of surgical staging has a significant impact on the complications and functional outcomes of adult spinal deformity surgery. Spine J 2013;13:1717e22. [22] Dearborn JT, Hu SS, Tribus CB, Bradford DS. Thromboembolic complications after major thoracolumbar spine surgery. Spine 1999;24:1471. [23] Hu SS, Fontaine F, Kelly B, Bradford DS. Nutritional depletion in staged spinal reconstructive surgery. the effect of total parenteral nutrition. Spine 1998;23:1401e5. [24] Mandelbaum BR, Tolo VT, McAfee PC, Burest P. Nutritional deficiencies after staged anterior and posterior spinal reconstructive surgery. Clin Orthop Relat Res 1988;234:5e11. [25] Dick J, Boachie-Adjei O, Wilson M. One-stage versus two-stage anterior and posterior spinal reconstruction in adults. Comparison of outcomes including nutritional status, complications rates, hospital costs, and other factors. Spine (Phila Pa 1976) 1992;17:310. [26] Smith JS, Klineberg E, Schwab F, et al. Change in classification grade by the SRS-Schwab adult spinal deformity classification predicts impact on health-related quality of life measures: prospective analysis of operative and nonoperative treatment. Spine 2013;38:1663.
803.e1
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Appendix Supplemental Table 1 Results from the multivariable model for LOS. Variable
LOS (days) Parameter estimate
Staged (yes vs. no) No. of posterior fusions (each additional) No. of interbody fusions (each additional) Level Level (Lumbosacral vs. lumbar) (Thoracolumbar vs. lumbar) (Thoracolumbosacral vs. lumbar) O-arm (yes vs. no) Prior instrumented surgery (yes vs. no) Comorbidity: MI (yes vs. no) Comorbidity: DVT/PE (yes vs. no)
p- value 95% confidence interval
0.8 0.4 1.8
0.8 0.1 1.1
2.4 0.7 2.5
.303 .006* !.001*
3.3 1.9 3.1 2.3 2.8 11.4 3.7
6.9 2.4 6.9 0.8 1.4 7.4 0.7
0.2 6.1 0.7 3.7 4.2 15.4 6.8
.064 .383 .113 .002* !.001* !.001* .018*
DVT, deep vein thrombosis; PE, pulmonary embolism; MI, myocardial infarction; PT, pelvic tilt; SVA, sagittal vertical axis. Staged surgery was not associated with increased LOS after accounting for operative time and other covariates. * p ! .05.
Supplemental Table 2 Results from the multivariable model for operative time. Variable
Staged (yes vs. no) No. of posterior fusions (each additional) No. of interbody fusions (each additional) Prior instrumented surgery (yes vs. no) Anterior or anterolateral surgery (yes vs. no) Diabetes mellitus (yes vs. no)
Operative time (minutes) 95% confidence interval
98.4 26.9 46.5 49.8 47.0 52.6
55.8 19.2 25.0 7.0 4.1 4.8
Staged surgery was associated with longer operative time after adjusting for other covariates. p ! .05.
*
p- value
Parameter estimate
141.1 34.5 67.9 92.6 90.0 110.0
!.001* !.001* !.001* .023* .032* .072
Short-Term Outcomes of Staged Versus Same-Day Surgery for Adult Spinal Deformity Correction Alexander H. Arzeno, MD*, Jayme Koltsov, PhD, Todd F. Alamin, MD, Ivan Cheng, MD, Kirkham B. Wood, MD, Serena S. Hu, MD Department of Orthopaedic Surgery, Stanford University School of Medicine, 450 Broadway St, Pavillion A FL 1 MC6110, Redwood City, CA, 94063, USA Received 3 February 2018; revised 21 October 2018; accepted 17 December 2018
Abstract Study Design: Retrospective cohort study. Objectives: Assess differences between staged (<3 days) and same-day surgery in perioperative factors, radiographic measures, and complications. Summary of Background Data: Surgical adult spinal deformity correction may require combined anterior and posterior approaches. To modulate risk, some surgeons perform surgery that is expected to be longer and/or more complex in two stages. Prior studies comparing staged (>7 days) and same-day surgery demonstrated mixed results and none have examined results with shorter staging intervals. Methods: Retrospective review of adults undergoing combined anterior/posterior approaches for spinal deformity over a 3-year period at a single institution (n592). Univariate differences between staged and same-day surgery were assessed with chi-squared, Fisher exact, and Mann-Whitney U tests. Generalized estimating equations assessed whether differences in perioperative outcomes between groups remained after adjusting for differences in demographic and surgical characteristics. Results: In univariate analyses, staged surgery was associated with a length of stay (LOS) 3 days longer than same-day surgery (9.2 vs. 6.3 days, p ! .001), and greater operative time, blood loss, transfusion requirement, and days in intensive care unit ( p ! .001 for each). Staged surgery had a higher rate of thrombotic events ( p 5 .011) but did not differ in readmission rates or other complications. Radiographically, improvements in Cobb angle (average 13 vs. 17 , p 5 .028), lumbar lordosis (average 14 vs. 23 , p 5 .019), and PI-LL mismatch (average 10 vs. 2 p 5 .018) were greater for staged surgery, likely related to more extensive use of osteotomies in the staged group. After risk adjustment, taking into account the procedural specifics including longer fusion constructs and greater number of osteotomies, LOS no longer differed between staged and same-day surgery; however, the total operative time was 98 minutes longer for staged surgery ( p ! .001). Differences in blood loss between groups was accounted for by differences in operative time and patient and surgical characteristics. Conclusions: Although univariate analysis of our results were in accordance with previously published works, multivariate analysis allowing individual case risk adjustment revealed that LOS was not significantly increased in the staged group as reported in previous studies. There was no difference in infection rates as previously described but an increase in thrombotic events was observed. Level of Evidence: Level III. Ó 2019 Scoliosis Research Society. All rights reserved. Keywords: Adult spine deformity; Fusion; Staged; Complications
Author disclosures: AHA (none), JK (none), TFA (personal fees from Disc Diagnostics, Empirical Spine, Globus Medical, and Spinal Elements, outside the submitted work), IC (personal fees from Spinal Cyte, Spine Wave, Globus Medical, and SpineCraft, outside the submitted work; in addition, he has a patent by NuVasive with royalties paid; including fees from AAOS [board or committee member], Cervical Spine Research Society [board or committee member], Scoliosis Research Society [board or committee member], and Spine: The Spine Journal [editorial or governing board]), KBW (personal fees from TranS1 and Alphatec Spine, outside the submitted work; in addition, he has a patent by Globus Medical with royalties paid; including fees from Spine [editorial or governing board] and
Spine: Journal of Spinal Deformity [editorial or governing board]), SSH (personal fees from NuVasive and DePuy, A Johnson & Johnson Company, outside the submitted work; in addition, she has a patent by NuVasive with royalties paid; including fees from the American Orthopaedic Association [board or committee member] and Scoliosis Research Society [board or committee member]).IRB approval: Research approved by Stanford Institutional Review Board/Privacy Office. *Corresponding author. Department of Orthopaedic Surgery, 300 Pasteur Drive, Room R144, Stanford, CA 94305-534, USA. Tel.: (650) 7217638; fax: (650) 498-6899. E-mail address: [email protected] (A.H. Arzeno).
2212-134X/$ - see front matter Ó 2019 Scoliosis Research Society. All rights reserved. https://doi.org/10.1016/j.jspd.2018.12.008
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
797
Introduction
Materials and Methods
Adult spinal deformity has become increasingly recognized as a cause of disability as our population ages and physical demands continue to increase [1]. Although adult spinal deformity can be managed conservatively, operative intervention is often considered for deformity progression or if symptoms worsen or cannot be improved with conservative management [2]. Surgical management aims to decompress central and foraminal stenosis, correct and stabilize the deformity to improve function and decrease pain, and reduce the subsequent need for analgesics [1,3,4]. Because of the complex nature of spinal deformities, the surgical correction can be lengthy, highly invasive, and may require a combination of surgical approaches [5-9]. Complication rates are between 40% and 80%, with major perioperative complications reported at 20% [5,10-12]. These major complications, including surgical site infection, neurologic deficit, pulmonary embolism (PE), and myocardial infarction (MI), are associated with prolonged length of stay (LOS) and higher rates of intensive care unit admissions [13,14]. For longer surgeries, particularly when different approaches are used, some surgeons consider performing the operation in a staged fashion. Most commonly, the anterior approach is performed on the first day and the posterior approach is performed several days later, dividing the physiologic insult and allowing interval recovery and resuscitation [11,15-17]. Same-day surgery has been associated with shorter LOS relative to staged surgery in previous studies, but estimated blood loss (EBL) and complication rate results are mixed [18-21]. In these studies, all staging was greater than 5 days, and most were over 7 days. Longer periods of staging directly increases hospital length of stay and possibly increases the risks for other complications due to longer periods of immobilization [22]. In addition, nutritional depletion often occurs after surgery, which has been shown to result in higher infection rates for staged procedures [23-25]. Shorter intervals between stages of adult spinal deformity cases may mitigate some of the risks associated with staged surgery and the increased hospitalization length, while still having the benefit of partitioning the physiologic stress of complex and extensive surgery. However, to date, no studies have examined the effects of shorter staging on LOS or perioperative outcomes. Therefore, the purpose of this study is to compare the outcomes of adult spinal deformity correction requiring more than one approach between same-day surgery versus staged surgery with an interval of three or fewer days. Outcomes assessed included 30-day complications and readmissions, LOS, perioperative outcomes, and radiographic deformity correction. We hypothesized that there would be similar outcomes, with no increase in complication rate in the staged group.
This retrospective cohort study was approved by the Institutional Review Board, and no external funding was obtained. All cases with both anterior (including lateral trans-psoas and anterolateral approaches) and posterior procedures from four surgeons at a single academic practice between January 1, 2014, and December 31, 2016, were reviewed. Inclusion criteria were as follows: age greater than 18 years, posterior fusion of at least five levels in the thoracic/lumbar/sacral spine with at least one interbody fusion performed via a separate incision on either the same day or within three days, and a minimum postoperative follow-up of six months. Surgeries for acute spinal trauma, malignancy, infection, or spondyloarthropathy were excluded. A total of 92 patients, 47 same day and 45 staged, met the inclusion and exclusion criteria (Table 1). Primary outcomes collected from the electronic medical record were complications within 30 days (including medical complications, neurologic deficits, infection, proximal junctional kyphosis, pseudoarthrosis, return to operating room), readmissions within 30 days, LOS, total operative time, and EBL. Secondary surgical and admission outcomes collected from the electronic medical record included total anesthetic time, transfusions, use of clotting factors (freshfrozen plasma, platelets), days of intensive care unit admission, and reoperations within 6 months. Additionally, pre- and postoperative radiographs were reviewed for all cases, and the following radiographic parameters were recorded: Cobb angle, center sacral vertical line, pelvic incidence (PI), lumbar lordosis (LL), PI-LL mismatch, pelvic tilt, sacral slope, and sagittal vertical axis. Sagittal modifiers, as described and correlated to health-related quality of life in SRS-Schwab classification, were also organized by grade [17] (Table 2). Postoperative measures were obtained at hospital discharge, 6 weeks postoperatively, and 3, 6, 12, and 24 months postoperatively. Baseline demographic data, including age, sex, body mass index, American Society of Anesthesiologists score, surgical diagnosis, history of prior instrumented fusion, and specific comorbidities (diabetes, history of smoking, history of MI, history of cerebrovascular accident) was collected from the electronic medical record. The approach, number of anterior and posterior instrumented levels, and other
Table 1 Inclusion and exclusion criteria. Inclusion criteria - Age O18 - Both anterior (including lateral and anterolateral approaches) and posterior fusion of at least 5 levels in thoracic/ lumbar/sacral spine - Procedure performed same day or staged ! 5 days apart - Minimum follow- up 6 months
Exclusion criteria - Diagnosis of acute spinal column trauma (fracture or dislocation) - Spinal malignancy, primary or metastatic - Spondyloarthropathies - Spinal osteomyelitis
798
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Table 2 Sagittal modifier grades. Grade 0 1 2
PI-LL Mismatch
!10 10 e20 O20
SVA
PT
!4 cm 4e9.5 cm O9.5 cm
!20 20 e30 O30
Table modified from the SRS-Schwab classification [26]. LL, lumbar lordosis; PI, pelvic incidence; PT, pelvic tilt; SVA, sagittal vertical axis.
procedural specifics (including use of osteotomies and decompression) were also recorded for each case. Univariate differences in categorical variables between staged and same-day surgery were assessed with chisquared and Fisher exact tests. Univariate differences in continuous measures between staged and same-day surgery were assessed with independent samples t tests, or MannWhitney U tests if nonnormally distributed. Generalized estimating equations were used to assess whether stages differed from same-day surgery in LOS, blood loss, and operative time after adjusting for key differences in patient and surgical characteristics. Variables with p ! .25 from the univariate tests for staged versus same-day surgery were considered as candidates in the multivariable models. The initial models included all candidate variables, and subsequently variables were removed stepwise in order of decreasing p value until only trending ( p ! .10) or significant ( p ! .05) terms remained. All analyses were performed in SAS, version 9.4 (SAS Institute, Cary, NC), with a two-sided level of significance of a 5 0.05. Based on previous reports of LOS and complications with staged (O5e7 days) and same-day spinal deformity correction, the 92 patients provided more than 80% power (with a 5 0.05) to detect a difference between groups of 3 days in LOS or 20% in the proportion having 30-day complications. Results Patient demographics did not differ significantly between groups (Table 3). Staged surgery had a much greater proportion of interbody fusion using an anterior approach, relative to the same-day surgery (84.4% vs. 46.8%, p ! .001; Table 3). The staged surgeries also had a greater number of Smith-Peterson osteotomies (87% vs. 51%, p ! .001), three-column osteotomies (including vertebral column resection or pedicle subtraction osteotomy) (16% vs. 2%, p 5 .029), O-arm (64% vs. 13%, p ! .001), neuromonitoring (91% vs. 21%, p ! .001), and decompression (76% vs. 34%, p ! .001) relative to the same-day group. The staged group had a greater number of levels fused relative to the same-day group (9 vs. 8, p ! .001). Perioperative outcomes The LOS for staged surgery (9.2 days, 95% CI 7.7, 10.6), was nearly three days longer than that for same-day
Table 3 Univariate comparisons of baseline parameters between same-day and staged surgery. Same-day Demographics Sex (female) Age, mean (95% CI) BMI, mean (95% CI) Smoking ASA class 1 2 3 Comorbidities Diabetes MI Stroke DVT/PE Prior instrumented surgery Prior spine surgery Surgical details Days between surgeries 1 2 3 Approach Anterior Lateral Anterolateral TLIF Ponte osteotomy Three- column osteotomy O-arm Neuromonitoring Decompression No. of posterior levels fused, mean (95% CI) No. of interbody levels, mean (95% CI) No. of osteotomy levels, mean (95% CI) No. of decompression levels, mean (95% CI) Discharge disposition Home Rehabilitation facility Skilled nursing facility
30 68 25.3 17
(63.8%) (61, 78) (23.1, 30.4) (36.2%)
Staged 34 68 25.5 19
(75.6%) (62, 72) (22.8, 28.7) (42.2%)
p-value .222 .122 .227 .410
1 (2.1%) 24 (51.1%) 22 (46.8%)
d 19 (42.2%) 26 (57.8%)
.414
3 (6.4%) d 4 (8.5%) 4 (8.5%) 16 (34%)
8 (17.8%) 2 (4.5%) 6 (13.6%) d 22 (48.9%)
.116 .231 .514 .118 .148
24 (51.1%)
27 (60%)
.389
d d d
2 (4.4%) 34 (75.6%) 9 (20.0%)
N/A
22 19 6 7 24 1
(46.8%) (40.4%) (12.8%) (14.9%) (51.1%) (2.1%)
38 (84.4%) 16 (35.6%) d d 39 (86.7%) 7 (15.6%)
!.001*
6 10 16 8
(12.8%) (21.3%) (34.0%) (5, 9)
29 41 34 9
!.001* !.001* !.001* !.001*
(64.4%) (91.1%) (75.6%) (9, 9)
.012* !.001* .029*
2 (2, 3)
3 (2, 3)
.506
1 (0, 3)
4 (3, 6)
!.001*
0 (0, 2)
2 (1, 2)
!.001*
.338 19 (40.4%) 3 (6.4%)
17 (38.6%) 7 (15.9%)
25 (53.2%)
20 (45.5%)
ASA, American Society of Anesthesiologists; BMI, body mass index; DVT, deep vein thrombosis; MI, myocardial infarction; PE, pulmonary embolism; TLIF, transforaminal lumbar interbody fusion. Values are n (%) unless otherwise noted. * p ! .05.
surgery (6.3, 95% CI 5.4, 7.2) ( p ! .001; Figure 1). The longer LOS for staged surgery was largely due to the interval between the first and second surgery, which was two
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
799
Fig. 1. Total (A) LOS, (B) operative time, and (C) anesthesia time, and (D) blood loss for same-day and staged surgery (raw, unadjusted values). Bar graphs show means and 95% confidence intervals. Central lines in box plots represent medians, whereas central markers indicate means. *p ! .05 from univariate tests. (LOS, length of stay; Op, operative.)
days for more than 75% of the staged cases, with a range of one to three days. The total operative time and EBL were substantially greater for staged versus same-day surgery ( p ! .001 for both; Figure 1). Secondary perioperative outcomesdpRBC units, pRBC units, platelets, fresh-frozen plasma, and days in the intensive care unitdwere also greater for staged relative to same-day surgery ( p ! .001 for each). After adjustment for patient demographics and surgical characteristics using multivariate regression models, the LOS for staged surgery was 0.8 (95% CI e0.8, 2.4) days longer for staged relative to same-day surgery, and this difference was not statistically significant ( p 5 .303; Supplemental Table 1). Staged surgery was associated with 98.4 (95% CI 55.8, 141.1) minutes longer total operative time ( p ! .001) and a trend toward greater blood loss (1,023.5 mL, 95% CI e154.1, 2201.2, p 5 .088) after risk adjustment (Supplemental Table 2). The trend toward greater blood loss with staged surgery was fully explained by the longer operative time, as adding this into the risk adjustment for the blood loss model reduced the blood-loss difference between staged versus same-day surgery to less than 50 mL ( p 5 .939). Complications The 30-day complication rate for staged surgery was 26.7%, and for same-day surgery was 14.9% (Table 4); however, this difference was not statistically significant ( p 5 .163). When looking at individual complications,
13.6% of the staged surgery patients had a DVT (deep vein thrombosis) or PE (pulmonary embolism), whereas none of the same-day patients experienced this complication ( p 5 .011). Of the six who developed a DVT/PE, five were in cases that had an anterior lumbar interbody fusion as a part of their operation. The rate of DVT/PE in anterior lumbar interbody fusion patients was 7.6% (5/66 patients), whereas the rate in surgeries with no anterior lumbar interbody fusion was 3.8% (1/26 patients); however, our sample was underpowered to attribute the difference to the difference in approaches. In contrast, rates of infection, proximal junctional kyphosis, pseudarthrosis, and revision surgery differed by only 2% to 4% between groups and were not statistically significant. Readmission rates were 10.5% greater for same-day versus staged surgery, but this difference was not statistically significant ( p 5 .112). Radiographic outcomes Preoperatively, there were no differences in radiographic parameters between groups (Figure 2). Postoperatively, sagittal vertical axis was 25.6 mm lower in staged relative to same-day surgery ( p 5 .024), and the postoperative mismatch between pelvic incidence and lumbar lordosis (PI-LL mismatch) was 8.8 degrees less for staged patients ( p 5 .002). Regarding the pre- to postoperative change in radiographic parameters, patients had a larger change in Cobb angle ( p 5 .028), lumbar lordosis ( p 5 .019), pelvic tilt ( p 5 .008), and PI-LL mismatch ( p 5 .018) with staged surgery when compared with same-day surgery. The total
800
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Table 4 Univariate comparisons of adverse events and SRS-Schwab Classification sagittal parameters between same-day and staged surgery. Same-day, n (%)
Staged, n (%)
Adverse events Reoperation 7 (14.9%) 5 Readmission 6 (12.8%) 1 Any complication 7 (14.9%) 12 Dural tear 2 (4.3%) 8 Neurologic d 4 defiecit DVT/PE d 6 Infection 3 (6.4%) 2 PJK 2 (4.3%) 3 Pseudoarthrosis 2 (4.3%) Sagittal imbalance (SRS-Schwab classification) PI minus LL Pre 0 10 (22.2%) 10 1 9 (20.0%) 7 2 26 (57.8%) 26 Post 0 23 (52.3%) 34 1 13 (29.5%) 7 2 8 (18.2%) 1 Change 2 7 (15.9%) 19 1 16 (36.4%) 11 0 21 (47.7%) 11 SVA Pre 0 2 (5.1%) 1 1 1 (2.6%) 2 2 36 (92.3%) 39 Post 0 4 (10.8%) 11 1 1 (2.7%) 2 2 32 (86.5%) 27 Change 2 3 (8.1%) 10 1 1 (2.7%) 1 0 31 (83.8%) 27 1 1 (2.7%) 1 2 1 (2.7%) PT Pre 0 12 (26.7%) 9 1 14 (31.1%) 15 2 19 (42.2%) 19 Post 0 11 (25.0%) 15 1 20 (45.5%) 18 2 13 (29.5%) 9 Change 2 1 (2.3%) 4 1 9 (20.5%) 10 0 28 (63.6%) 24 1 6 (13.6%) 3
p- value
(11.4%) (2.3%) (26.7%) (17.8%) (8.9%)
.619 .112 .163 .048* .053
(13.6%) (4.5%) (6.8%) d
.011* O.999 .670 .495
(23.3%) (16.3%) (60.5%)
.903
(81.0%) (16.7%) (2.4%)
.009*
(46.3%) (26.8%) (26.8%)
.009*
(2.4%) (4.8%) (92.9%)
.865
(27.5%) (5.0%) (67.5%)
.122
(25.6%) (2.6%) (69.2%) (2.6%) d
.141
(20.9%) (34.9%) (44.2%)
.812
(35.7%) (42.9%) (21.4%)
.496
(9.8%) (24.4%) (58.5%) (7.3%)
.432
changes were considerably higher than for the same-day group ( p ! .001). Looking at sagittal imbalance modifiers as classified by the SRS-Schwab classification, the staged and same-day patients were similar preoperatively, with no differences in grade of PI-LL mismatch, sagittal vertical axis, or pelvic tilt. Postoperatively, staged surgery resulted in a larger change in PI-LL mismatch grade ( p 5 .009), with 46.3% of staged patients having a decrease of 2 grades, compared with 15.9% of same-day patients. Thus, staged surgery had more patients returned to a PI-LL mismatch grade of 0 (80.0%) compared with same-day surgery (52.3%, p 5 .009). When looking only at those patients who were sagittally imbalanced before surgery, the findings were similar, with staged surgery patients having a larger change in PI-LL mismatch grade (54.3% staged vs. 18.9% same-day decreased two classes, p 5 .002) and having restored sagittal balance (71.4% staged vs. 35.1% same-day PI-LL mismatch grade 0, p 5 .005) postoperatively relative to same-day surgery patients.
Discussion
DVT, deep vein thrombosis; LL, lumbar lordosis; PE, pulmonary embolism; PI, pelvic incidence; PJK, proximal junctional kyphosis; PT, pelvic tilt; SVA, sagittal vertical axis. * p ! .05.
Operative interventions for adult spinal deformity correction are extensive surgical procedures that can carry a relatively high risk for complications. However, for patients with significant disability due to their spinal condition, surgery can result in improved functional outcomes when compared with medical/interventional management [1]. The surgeon and the patient must carefully weigh the benefits of undergoing surgical correction compared with the risks of intervention. It is imperative that we continue to try and maximize benefit while decreasing risk. Staged versus sameday procedures for those that require both anterior and posterior procedures is one such variable. Several studies have previously examined the effects of staging of surgical procedures in spinal deformity correction; however, mixed patient populations, surgical technique, and varying staging protocols have been used and, thus, results have varied. No study has compared staging with this brief an interval and we hypothesized that LOS and potential complications would be similar between groups. A better understanding of the risks and benefits associated with staged procedures will allow more informed decision making and surgical planning for both the surgeon and the patient. In this study, we examined adult patients who underwent spinal deformity correction requiring at least five levels as well as both anterior and posterior fusion through separate approaches. Patients whose procedure had been performed the same day in a sequential manner under the same anesthetic were compared to those who were staged three or fewer days apart. After taking into account the influence of other variables such as patient demographics, comorbidities, and
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
801
Fig. 2. Radiographic results. (A) Cobb angle, (B) pelvic tilt, (C) lumbar lordosis, (D) pelvic incidence-lumbar lordosis mismatch, and (E) sagittal vertical axis. *, pre- to postoperative change in parameter differs between same-day and staged surgery ( p ! .05); þ, postoperative parameter differs between sameday and staged ( p ! 0.05). (LL, lumbar lordosis; PI, pelvic incidence; PT, pelvic tilt; SVA, sagittal vertical axis.)
procedural specifics, we found no difference in EBL attributable to staging alone. Although there was a trend toward greater EBL with staged surgery, after risk adjustment, the difference was explained by the 98 additional minutes of operative time required for staged surgery. Previous studies by Shufflebarger [18] and Tsirikos [19] reported greater blood loss with staged surgeries, as did a more recent study by Maddox using a posterior-only approach. These studies all performed the second stage of their surgery at approximately one week. Tsirikos and Shen [20] interestingly found no increase in EBL with staged when comparing to a longer period of staging. The impact of staging on LOS is also interesting. Surprisingly, we found that LOS was not significantly different (less than one day) between the two groups after adjusting for other factors. Thus, despite the additional two to three days between stages, staging did not result in further prolongation of LOS when patient and operative characteristics were taken into consideration. In those studies where procedures were staged for more than one week without interval discharge, it makes intuitive sense that LOS would thus be longer because of an additional week in hospital as demonstrated in Shufflebarger and Shen. In the study by
Tsirikos, there was no difference in LOS despite a four- to seven-day staging protocol and, similarly, in the study by Hassanzadeh [21] in which patients were discharged between stages, no difference was found. The overall complication rate in our study did not differ between groups, nor did 30-day readmission and revision rates. Reported complication rates have been mixed in previous studies. Shufflebarger and Tsirikos reported higher complication rates in staged versus same-day protocols, whereas Hassanzadeh and Maddox reported no difference. Our study similarly found no difference in the overall complication rate between the two protocols. In particular, prior studies with longer intervals between staging reported higher rates of infection in staged surgery secondary to the two open operations as well as nutritional deficiency after staged surgeries [23-25]. However, in our study, we did not find a significantly increased risk of infection in staged surgery. In our study, we did find a higher rate of thrombotic events in the staged surgery group. However, 84.4% of staged cases were performed with an anterior or anterolateral approach versus 59.6% of same-day cases ( p ! .001), and anterior approaches have previously been associated with higher rates of DVT/PE [22]. In our study, the
802
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
sample size and small number of these events did not allow us to determine whether the increased DVT/PE rate in staged cases was attributable to the greater use of anterior or anterolateral approaches. The staged group also had significantly larger correction in radiographic parameters for both coronal and sagittal balance. Few studies have studied the radiographic results of staging spinal deformity surgery. Shufflebarger et al. found increased correction with a same-day operation, with a significantly higher percentage of coronal correction with same-day relative to staged surgery [18]. In our study, we measured more radiographic alignment variables, including those for sagittal balance, the significance of which is more recently appreciated. Contrary to Shufflebarger’s results, we found higher correction in Cobb angle in the staged group. Additionally, correction in the sagittal plane such as lumbar lordosis correction was greater and final PI-LL mismatch was lower in the staged group. In our study, the improved correction likely reflects the surgical goal of greater deformity correction for those patients, rather than the staging per se. This is evidenced by the greater use of osteotomies and number of fused levels in the staged group. The surgeries that were staged were, on average, more complex procedures, requiring more levels of decompression and fusion, more Smith Petersen osteotomies, and more three-column osteotomies. Hence, cases for which the surgeon decides are clinically indicated for more deformity correction require more complex procedures to achieve that correction, and are therefore more likely to be performed in a staged manner in our center. We did not make ageadjustments to the sagittal balance goals; however, there was no difference in ages between staged and sameday patients. Limitations of our study include the fact that it was a retrospective review. Operations were performed by four different surgeons and the distribution of staged versus same-day cases by surgeon was not uniform and reflected surgeon preferences. The goals of the surgeons in terms of deformity correction and desired postoperative radiographic goals were also likely to have been different, which may explain many of the differences seen in the two different groups such as the amount of planned correction and subsequent use of osteotomies, number of levels being fused, and the distribution of lateral versus anterior procedures as well as the radiographic outcomes achieved. Without a prospective randomized trial, it is impossible to control for the effect of the surgeon’s approach, training, experience, and biases. Additionally, we included different types of anterior interbody fusions and approaches. Although there is likely some degree of confounding, we attempted to account for these variances using a multivariate risk-adjusted analysis. The multivariate model allows us to attribute differences in outcomes based on each specific variable included in the model. When the variables that made the groups different were statistically accounted for (number of levels, osteotomies, etc.) were accounted
for, we were able to examine the influence of the staging alone on our outcomes. An additional limitation is that this study was powered for the primary, univariate analyses and was underpowered for the multivariable generalized estimating equation models. Although we attempted to maximize power by reducing the number of covariates through backwards selection, covariates not retained in the final models may be associated with these outcomes, but at a level not able to be detected with the present sample size. Further study will include long-term follow-up of both groups to examine radiographic as well as pain and patientreported outcomes. The higher rate of PE/DVTs for staged surgery patients is a cause for concern and we are considering additional prophylactic modalities for these patients. Conclusion When correcting adult spinal deformity with anterior and posterior approaches, it is reasonable to consider staging the operations if more extensive surgery is indicated and the procedures are anticipated to be of a long total duration. In our study, cases that were more complex and used more osteotomies to enable a greater deformity correction were more likely to be staged. We used a multivariate analysis to allow for individual case risk adjustment for these differences. We show that there was no additional increase in LOS that was attributable to the staging itself. Additionally, there was no difference in infection rates as previously described. An increase in thrombotic events was observed in the staged group, which we believe is likely attributable to the greater use of anterior approach for interbody fusions in the staged cohort. The staged group had significantly greater surgical correction of spinal imbalance when radiographic parameters were compared, likely because of the greater use of osteotomies in those patients. References [1] Kluba T, Dikmenli G, Dietz K, et al. Comparison of surgical and conservative treatment for degenerative lumbar scoliosis. Arch Orthop Trauma Surg 2009;129:1e5. [2] Glassman SD, Hamill CL, Bridwell KH, et al. The impact of perioperative complications on clinical outcome in adult deformity surgery. Spine (Phila Pa 1976) 2007;32:2764e70. [3] Wang G, Hu J, Liu X, Cao Y. Surgical treatments for degenerative lumbar scoliosis: a meta analysis. Eur Spine J 2015;24:1792e9. [4] Glassman SD, Schwab FJ, Bridwell KH, et al. The selection of operative versus nonoperative treatment in patients with adult scoliosis. Spine (Phila Pa 1976) 2007;32:93e7. [5] Cho KJ, Suk SI, Park SR, et al. Complications in posterior fusion and instrumentation for degenerative lumbar scoliosis. Spine (Phila Pa 1976) 2007;32:2232e7. [6] Good CR, Lenke LG, Bridwell KH, et al. Can posterior-only surgery provide similar radiographic and clinical results as combined anterior (thoracotomy/thoracoabdominal)/posterior approaches for adult scoliosis? Spine (Phila Pa 1976) 2010;35:210e8. [7] Charosky S, Guigui P, Blamoutier A, et al; Study Group on Scoliosis. Complications and risk factors of primary adult scoliosis surgery: a
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
[8]
[9]
[10]
[11] [12]
[13]
[14]
[15]
[16]
[17]
multicenter study of 306 patients. Spine (Phila Pa 1976) 2012;37: 693e700. Carreon LY, Puno RM, Dimar JR, et al. Perioperative complications of posterior lumbar decompression and arthrodesis in older adults. J Bone Joint Surg Am 2003;85-A:2089e92. Sansur CA, Smith JS, Coe JD, et al. Scoliosis Research Society morbidity and mortality of adult scoliosis surgery. Spine (Phila Pa 1976) 2011;36:593. Maddox JJ, Pruitt DR, Agel J, Bransford RJ. Unstaged versus staged posterior-only thoracolumbar fusions in deformity: a retrospective comparison of perioperative complications. Spine J 2014;14:1159e65. Rhee JM, Bridwell KH, Lenke LG, et al. Staged posterior surgery for severe adult spinal deformity. Spine (Phila Pa 1976) 2003;28:2116e21. Daubs MD, Lenke LG, Cheh G, et al. Adult spinal deformity surgery: complications and outcomes in patients over age 60. Spine (Phila Pa 1976) 2007;32:2238e44. Howe CR, Agel J, Lee MJ, et al. The morbidity and mortality of fusions from the thoracic spine to the pelvis in the adult population. Spine (Phila Pa 1976) 2011;36:1397e401. Spivak JM, Neuwirth MG, Giordano CP, Bloom N. The perioperative course of combined anterior and posterior spinal fusion. Spine (Phila Pa 1976) 1994;19:520e5. Byrd JA, Scoles PV, Winter RB, et al. Adult idiopathic scoliosis treated by anterior and posterior spinal fusion. J Bone Joint Surg Am 1987;69:843e50. Passias PG, Ma Y, Chiu YL, et al. Comparative safety of simultaneous and staged anterior and posterior spinal surgery. Spine (Phila Pa 1976) 2012;37:247e55. Wright N. Single-surgeon simultaneous versus staged anterior and posterior spinal reconstruction: a comparative study. J Spinal Disord Tech 2005;18:48.
803
[18] Shufflebarger HL, Grimm JO, Bui V, Thomson JD. Anterior and posterior spinal fusion. Staged versus same-day surgery. Spine (Phila Pa 1976) 1991;16:930e3. [19] Tsirikos AI, Chang W, Dabney KW, Miller F. Comparison of onestage versus two-stage anteroposterior spinal fusion in pediatric patients with cerebral palsy and neuromuscular scoliosis. Spine (Phila Pa 1976) 2003;28:1300. [20] Shen J, Qiu G, Wang Y, et al. Comparison of 1-stage versus 2-stage anterior and posterior spinal fusion for severe and rigid idiopathic scoliosisda randomized prospective study. Spine (Phila Pa 1976) 2006;31:2525e8. [21] Hassanzadeh H, Gjolaj JP, El Dafrawy MH, et al. The timing of surgical staging has a significant impact on the complications and functional outcomes of adult spinal deformity surgery. Spine J 2013;13:1717e22. [22] Dearborn JT, Hu SS, Tribus CB, Bradford DS. Thromboembolic complications after major thoracolumbar spine surgery. Spine 1999;24:1471. [23] Hu SS, Fontaine F, Kelly B, Bradford DS. Nutritional depletion in staged spinal reconstructive surgery. the effect of total parenteral nutrition. Spine 1998;23:1401e5. [24] Mandelbaum BR, Tolo VT, McAfee PC, Burest P. Nutritional deficiencies after staged anterior and posterior spinal reconstructive surgery. Clin Orthop Relat Res 1988;234:5e11. [25] Dick J, Boachie-Adjei O, Wilson M. One-stage versus two-stage anterior and posterior spinal reconstruction in adults. Comparison of outcomes including nutritional status, complications rates, hospital costs, and other factors. Spine (Phila Pa 1976) 1992;17:310. [26] Smith JS, Klineberg E, Schwab F, et al. Change in classification grade by the SRS-Schwab adult spinal deformity classification predicts impact on health-related quality of life measures: prospective analysis of operative and nonoperative treatment. Spine 2013;38:1663.
803.e1
A.H. Arzeno et al. / Spine Deformity 7 (2019) 796e803
Appendix Supplemental Table 1 Results from the multivariable model for LOS. Variable
LOS (days) Parameter estimate
Staged (yes vs. no) No. of posterior fusions (each additional) No. of interbody fusions (each additional) Level Level (Lumbosacral vs. lumbar) (Thoracolumbar vs. lumbar) (Thoracolumbosacral vs. lumbar) O-arm (yes vs. no) Prior instrumented surgery (yes vs. no) Comorbidity: MI (yes vs. no) Comorbidity: DVT/PE (yes vs. no)
p- value 95% confidence interval
0.8 0.4 1.8
0.8 0.1 1.1
2.4 0.7 2.5
.303 .006* !.001*
3.3 1.9 3.1 2.3 2.8 11.4 3.7
6.9 2.4 6.9 0.8 1.4 7.4 0.7
0.2 6.1 0.7 3.7 4.2 15.4 6.8
.064 .383 .113 .002* !.001* !.001* .018*
DVT, deep vein thrombosis; PE, pulmonary embolism; MI, myocardial infarction; PT, pelvic tilt; SVA, sagittal vertical axis. Staged surgery was not associated with increased LOS after accounting for operative time and other covariates. * p ! .05.
Supplemental Table 2 Results from the multivariable model for operative time. Variable
Staged (yes vs. no) No. of posterior fusions (each additional) No. of interbody fusions (each additional) Prior instrumented surgery (yes vs. no) Anterior or anterolateral surgery (yes vs. no) Diabetes mellitus (yes vs. no)
Operative time (minutes) 95% confidence interval
98.4 26.9 46.5 49.8 47.0 52.6
55.8 19.2 25.0 7.0 4.1 4.8
Staged surgery was associated with longer operative time after adjusting for other covariates. p ! .05.
*
p- value
Parameter estimate
141.1 34.5 67.9 92.6 90.0 110.0
!.001* !.001* !.001* .023* .032* .072