- Email: [email protected]
Predictors of Nonneurologic Complications and Increased Length of Stay After Cervical Spine Osteotomy
Original Article
Predictors of Nonneurologic Complications and Increased Length of Stay After Cervical Spine Osteotomy J. Mason DePasse1, Wesley Durand3, Alan H. Daniels2
BACKGROUND: Although previous studies have used National Surgical Quality Improvement Program (NSQIP) data to study complications of thoracolumbar spinal deformity surgery, investigation of cervical spine deformity surgery has been limited. We performed a retrospective analysis of the NSQIP database to identify predictors of complications after cervical spine osteotomy.
-
METHODS: Patients undergoing cervical spine osteotomy were identified in the NSQIP dataset using Current Procedural Terminology codes from years 2007e2016. For each patient, patient and case clinical characteristics, length of stay (LOS), and diagnosis of a nonneurologic complication (including reoperation and readmission) were abstracted. Patient and case clinical predictors of any of the reported complications and increased LOS were identified in multivariate logistic and Poisson regression analyses, respectively.
operative duration (P < 0.0001), and orthopaedic surgeon (vs. neurosurgeon) (P [ 0.0156). CONCLUSIONS: This study is the largest to date of patients undergoing cervical osteotomy and provides useful clinical data for patient selection and counseling and 30day reoperation and readmission rates.
-
-
RESULTS: There were 950 patients identified with mean age 56.1 12.4 years and mean body mass index 29.9 6.8. Mean LOS was 3.5 4.9 days. Overall medical complication rate was 15.8%. The most common complications were transfusion (78; 8.2%), readmission (45; 4.7%), reoperation (32; 3.4%), and reintubation (28; 3.0%). Risk factors for any complication included increased age (P [ 0.0467), American Society of Anesthesiologists classification III (P [ 0.0023) and IV (P [ 0.0013), and increased operative duration (P < 0.0001). Risk factors for increased LOS were decreased functional status (P [ 0.0037), disseminated cancer (P [ 0.0061), American Society of Anesthesiologists classification III and IV (P < 0.0001), increased
-
Key words Cervical spine - Complications - Deformity - Length of stay -
Abbreviations and Acronyms ASD: Adult spinal deformity ASA: American Society of Anesthesiologists IC: Incremental change LOS: Length of stay NIS: National Inpatient Sample
INTRODUCTION
W
ith advances in surgical technique, instrumentation, and understanding of sagittal alignment, operative correction of cervical deformity with osteotomy has become more common.1 Recent studies have reported that surgery improves multiple measures of pain and function2 and likely reaches cost-effectiveness at 4 years.3 However, currently few data are available on the complications of cervical deformity correction and associated risk factors. A study examining >360,000 patients in the National Inpatient Sample database who underwent surgical treatment for any cervical spine pathology found that pulmonary circulation disorders had the greatest impact on complication risk, although age 65 years and other comorbidities, such as renal disease, cancer, and congestive heart failure, were also predictive.4 The authors also noted that performance of an osteotomy was a risk factor for surgical complications, such as wound dehiscence, but it did not have as large an impact as fusion of 9 levels.4 Although these data are useful, it is not specific for cervical deformity. However, the finding that osteotomy increases the risk of surgical complications emphasizes the importance of identifying factors that could potentially be controlled to mitigate that risk.
NSQIP: National Surgical Quality Improvement Program OR: Odds ratio From the 1Department of Orthopaedics, 2Department of Orthopaedics, Division of Spine Surgery, and 3Alpert Medical School of Brown University, Providence, Rhode Island, USA To whom correspondence should be addressed: J. Mason DePasse, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.07.029 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY -: e1-e4, - 2018
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
Table 1. Patient and Case Characteristics, Complication Rate, and Length of Stay Variable
Overall
Complications
Continuous
Mean (SD)
Mean (SD)
Age, years
56.1 (12.4)
60.8 (12.5)
N/A
BMI
29.9 (6.8)
29.1 (7.0)
Operative duration, hours
3.1 (2.3)
5.6 (3.2)
Number (%)
Number (%)
Categorical
Number (%)
Number (%)
Mean (SD)
No
931 (98.0)
147 (15.8)
3.5 (4.8)
Yes
19 (2.0)
3 (15.8)
5.2 (5.2)
I or II
478 (50.4)
40 (8.4)
2.2 (2.5)
III
442 (46.6)
98 (22.2)
4.6 (5.6)
IV
28 (3.0)
11 (39.3)
9.0 (10.8)
Bleeding disorder Length of Stay
Categorical
Table 1. Continued
Mean (SD)
Surgical specialty Neurosurgery
622 (65.5)
97 (15.6)
3.2 (4.1)
Orthopaedics
328 (34.5)
53 (16.2)
4.1 (5.9)
788 (83.0)
122 (15.5)
4.7 (6.4)
Insulin-dependent
67 (7.1)
13 (19.4)
3.4 (4.9)
Noneinsulin-dependent
95 (10.0)
15 (15.8)
3.1 (3.0)
No
709 (74.6)
114 (16.1)
3.6 (5.0)
Yes
241 (25.4)
36 (14.9)
3.3 (4.5)
No
897 (94.4)
138 (15.4)
3.4 (4.8)
Yes
53 (5.6)
12 (22.6)
4.8 (6.1)
Dependent
42 (4.4)
17 (40.5)
8.3 (10.2)
Independent
908 (95.6)
133 (14.7)
3.3 (4.3)
No
897 (94.4)
136 (15.2)
3.5 (4.9)
Yes
53 (5.6)
14 (26.4)
3.9 (3.9)
ASA classification
BMI, body mass index; COPD, chronic obstructive pulmonary disease; ASA, American Society of Anesthesiologists.
Diabetes status No
Smoking
Dyspnea
Functional status
MATERIALS AND METHODS
COPD
Hypertension No
484 (51.0)
67 (13.8)
3.2 (4.7)
Yes
466 (49.1)
83 (17.8)
3.8 (5.0)
No
941 (99.1)
146 (15.5)
3.4 (4.8)
Yes
9 (1.0)
4 (44.4)
10.2 (8.6)
No
907 (95.5)
137 (15.1)
3.4 (4.5)
Yes
43 (4.5)
13 (30.2)
6.2 (9.4)
Disseminated cancer
Long-term steroid use
Continues
Two studies have investigated complications after surgical treatment for cervical deformity specifically. One evaluated 23 adult patients with cervical deformity treated with 3-column osteotomy and reported a complication rate of 56.5%, with
e2
neurologic deficit, wound infection, distal junctional kyphosis, and cardiorespiratory failure the most common complications.5 A second study of 120 patients, of which 28 underwent 3-column osteotomy, found that cardiopulmonary failure and myocardial infarction were the most common causes of death after surgery.6 Neither of the 2 studies was powered for identifying the causes of complications. To improve patient selection and counseling, we used the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database to identify predictors of complications and increased length of stay (LOS) after cervical spine osteotomy.
www.SCIENCEDIRECT.com
Data Sources This study employed the NSQIP dataset for the years 2007e2016. NSQIP collects data from member hospitals for major surgical procedures, including patient and procedure characteristics, morbidity and mortality outcomes, and reoperation and readmission status for 30 days postoperatively. NSQIP does not include data for pediatric cases, minor cases, trauma cases, and cases with American Society of Anesthesiologists (ASA) score of VI. Given the publicly available and deidentified nature of NSQIP data, this investigation was not considered human subjects research. Patient Selection Patients undergoing cervical spine osteotomy were identified based on the presence of Current Procedural Terminology code of 22210 or 22220. Exclusion criteria were surgeon specialty other than orthopaedic or neurologic surgery and emergent surgery. Dependent Variables The primary variables in the study were any complication and LOS measured in days. Complications abstracted from the database included transfusion, wound disruption, surgical site infection, reintubation, pneumonia, urinary tract infection, sepsis, thromboembolic event, major cardiac event, stroke, reoperation, readmission, and death.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.07.029
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
Table 2. Complications After Cervical Osteotomy Variable
Table 4. Risk Factors for Increased Length of Stay Number (%)
Variable
IC
95% CI
P Value
1.69
1.19e2.41
0.0037
2.47
1.29e4.71
0.0061
III
1.61
1.39e1.86
<0.0001
IV
2.72
1.72e4.31
<0.0001
Functional status (reference ¼ independent)
Any complication
150 (15.8)
Transfusion
78 (8.2)
Readmission
45 (4.7)
Reoperation
32 (3.4)
Postoperative intubation
28 (3.0)
SSI
20 (2.1)
Pneumonia
16 (1.7)
Sepsis
15 (1.6)
UTI
12 (1.3)
Orthopaedics
1.22
1.04e1.44
0.0156
Thromboembolic event
12 (1.3)
Operative duration (per hour)
1.17
1.14e1.20
<0.0001
Wound disruption
5 (0.5)
Stroke/CVA
3 (0.3)
Major cardiac event
2 (0.2)
Death
2 (0.2)
SSI, surgical site infection; UTI, urinary tract infection; CVA, cerebrovascular accident.
Independent Variables The following patient variables were analyzed: age, body mass index, functional status, smoking status, diabetes, chronic obstructive pulmonary disease, hypertension requiring medication, long-term steroid use, disseminated cancer, bleeding disorder, dyspnea, and ASA classification. In addition, the procedural variables operative time and surgical subspecialty were analyzed. Statistical Analysis Descriptive statistics were calculated for continuous variables, and the total number and frequency of all categorical variables were recorded for all patients as well as for patients who sustained any complication. The total number and frequency of each complication were also recorded. Additionally, mean LOS was calculated overall and for patients with each independent categorical variable. An analysis of risk factors for any complication and increased LOS was conducted. Bivariate analyses used c2, Fisher exact, Table 3. Risk Factors for Any Complication Variable
OR
95% CI
P Value
Age (per year)
1.02
1.00e1.04
0.0467
1.05e29.81
0.0433
Disseminated cancer (reference ¼ no) Yes
5.60
ASA (reference ¼ I or II) III
2.13
1.31e3.45
0.0023
IV
5.02
1.88e13.39
0.0013
1.60
1.47e1.74
<0.0001
Operative duration, per hour
OR, odds ratio; CI, confidence interval; ASA, American Society of Anesthesiologists.
WORLD NEUROSURGERY -: e1-e4, - 2018
Dependent Disseminated cancer (reference ¼ no) Yes ASA (reference ¼ I or II)
Surgical specialty (reference ¼ neurosurgery)
IC, incremental change; CI, confidence interval; ASA, American Society of Anesthesiologists.
Wilcoxon-Mann-Whitney, and Kruskal-Wallis tests, as appropriate. Risk factor analysis for total complications was performed with multiple stepwise logistic regression in a multivariate approach. Stepwise variable selection used thresholds of a ¼ 0.20 and a ¼ 0.10 for entry and retention, respectively. The final logistic regression models were evaluated using the HosmerLemeshow goodness-of-fit test. Multivariate analyses for LOS were performed with Poisson regression with robust error variance. All analysis was completed with SAS Version 9.4 (SAS Institute Inc., Cary, North Carolina). Statistical significance was defined as P < 0.05. RESULTS There were 950 patients identified with mean age of 56.1 years (SD 12.4) and mean body mass index of 29.9 (SD 6.8). Of patients, 162 (17.1%) had diabetes, 241 (25.4%) were smokers, and 53 (5.6%) had chronic obstructive pulmonary disease. ASA classification was as follows: ASA I or II in 478 (50.4%) patients, ASA III in 442 (46.6%) patients, and ASA IV in 28 (3.0%) patients. Neurosurgeons performed cervical spine osteotomy in 622 (65.5%) of cases. Mean operative duration was 3.1 hours (SD 2.3), and mean LOS was 3.5 days (SD 4.9) (Table 1). Overall complication rate was 15.8%. The most common complications included transfusion (78; 8.2%), readmission (45; 4.7%), reoperation (32; 3.4%), and reintubation (28; 3.0%) (Table 2). Multivariate analysis demonstrated that risk factors for any complication included increased age (odds ratio [OR] ¼ 1.02 per year, P ¼ 0.0467), disseminated cancer (OR ¼ 5.60, P ¼ 0.0433), ASA classification III (OR ¼ 2.13 vs. ASA I, P ¼ 0.0023) and IV (OR ¼ 5.02, P ¼ 0.0013), and increased operative duration (OR ¼ 1.60 per hour, P < 0.0001) (Table 3). Risk factors for increased LOS were decreased functional status (incremental change [IC] ¼ 1.69, P ¼ 0.0037), disseminated cancer (IC ¼ 2.47, P ¼ 0.0061), ASA classification III (IC ¼ 1.61, P < 0.0001) and IV (IC ¼ 2.72, P < 0.0001), orthopaedic surgeon (vs. neurosurgeon) (IC ¼ 1.22, P ¼ 0.0156), and
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
increased operative duration (IC ¼ 1.17 per hour, P < 0.0001) (Table 4). DISCUSSION With the increasing use of cervical spine osteotomy for correction of adult cervical deformity,1,7 understanding predictors of complications is valuable for patient selection and counseling. We found that increased age, ASA classification of III or IV, and increased operative duration were associated with increased complications, whereas increased age, decreased functional status, cancer, ASA classification III or IV, orthopaedic specialty, and increased operative duration were predictive for increased LOS. In their study of complications associated with surgical treatment for any cervical spine pathology, Passias et al.4 reported that pulmonary circulation disorders were most predictive of morbidity, although age and other comorbidities were also associated with complications. Although increased age was a significant risk factor in our study, only ASA classification greater than II predicted complications; there was no association with any specific disease process. Our findings are more consistent with studies of complications after thoracolumbar adult spinal deformity (ASD) surgery. In a study of 5803 patients, Lee et al.8 found that age >60 years, ASA classification greater than II, and operative time >4 hours portended a higher risk of complications.8 The authors also found that decreased functional status (which we found increased LOS), insulindependent diabetes, and bleeding disorders were associated with complications.8 Similarly, Yoshida et al.9 found increased age, operative time, and estimated blood loss were independent predictors of complications in 304 patients with ASD, whereas De la Garza Ramos et al.10 reported that increased age, operative time >6 hours, and bleeding disorders were predictors of prolonged intubation or reintubation in a study of 1250 patients with ASD. It is possible that a history of bleeding disorder, which was a predictor of complications in studies by both Lee et al.8 and De
REFERENCES 1. Smith JS, Shaffrey CI, Bess S, Shamji MF, Brodke D, Lenke LG, et al. Recent and emerging advances in spinal deformity. Neurosurgery. 2017; 80:S70-S85. 2. Ailon T, Smith JS, Shaffrey CI, et al. Outcomes of operative treatment for adult cervical deformity: a prospective multicenter assessment with 1-year follow-up. Accessed April 12, 2018. [e-pub ahead of print]. Neurosurgery. https://doi.org/10.1093/ neuros/nyx574.
la Garza Ramos et al.,10 is less predictive in cervical deformity cases owing to decreased incision size and the potential need for more meticulous hemostasis. Otherwise, the risk factors identified in these thoracolumbar ASD studies are consistent with our findings in cervical deformity. It is noteworthy that decreased functional status was predictive of increased LOS but was not predictive of increased complications, especially given that Lee et al.8 found decreased functional status was associated with complications in patients with ASD. Incidence of major cardiac events and thromboembolic events was lower in our study than in their study,8 so it is possible that patients with cervical deformity are less susceptible to vascular events than patients with thoracolumbar ASD and therefore do not experience complications from stasis as frequently. The finding that orthopaedic surgery patients have significantly longer LOS is difficult to explain with our data, but it may be due to procedural differences. Further study will be required to better understand that result. Our study has several significant limitations. It is an uncontrolled retrospective study, and we could not compare patients undergoing osteotomy with patients undergoing cervical fusions without osteotomy. Additionally, the NSQIP database does not record neurologic complications, which suggests our findings underestimate the complication rate. Because we used cervical osteotomy Current Procedural Terminology codes to ensure our sample included only patients with cervical deformity, we also could not identify number of levels fused. NSQIP does not provide data beyond 30 days, so we could not study late complications, such as proximal junctional kyphosis or late failure. CONCLUSIONS Despite the above-mentioned limitations, our study provides predictive data for medical complications and reoperation and readmission rates. These data are valuable for patient selection and counseling for clinicians considering cervical osteotomy for deformity correction.
5. Smith JS, Shaffrey CI, Lafage R, Lafage V, Schwab FJ, Kim HJ, et al. Three-column osteotomy for correction of cervical and cervicothoracic deformities: alignment changes and early complications in a multicenter prospective series of 23 patients. Eur Spine J. 2017;26:2128-2137. 6. Smith JS, Shaffrey CI, Kim HJ, Passias P, Protopsaltis T, Lafage R, et al. Prospective multicenter assessment of all-cause mortality following surgery for adult cervical deformity. Accessed April 10, 2018. [e-pub ahead of print]. Neurosurgery. https://doi.org/10.1093/neuros/nyx605.
3. Poorman GW, Passias PG, Qureshi R, et al. Costutility analysis of cervical deformity surgeries using 1-year outcome. Accessed April 10, 2018. [e-pub ahead of print]. Spine J. https://doi.org/10. 1016/j.spinee.2018.01.016.
7. Kim HJ, Piyaskulkaew C, Riew KD. Comparison of Smith-Petersen osteotomy versus pedicle subtraction osteotomy versus anterior-posterior osteotomy types for the correction of cervical spine deformities. Spine (Phila Pa 1976). 2015;40: 143-146.
4. Passias PG, Diebo BG, Marascalchi BJ, Jalai CM, Horn SR, Zhou PL, et al. A novel index for quantifying the risk of early complications for patients undergoing cervical spine surgeries. J Neurosurg Spine. 2017;27:501-507.
8. Lee NJ, Kothari P, Kim JS, Shin JI, Phan K, Di Capua J, et al. Early complications and outcomes in adult spinal deformity surgery: an NSQIP study based on 5803 patients. Global Spine J. 2017;7: 432-440.
e4
www.SCIENCEDIRECT.com
9. Yoshida G, Hasegawa T, Yamato Y, Kobayashi S, Oe S, Banno T, et al. Predicting perioperative complications in adult spinal deformity surgery using a simple sliding scale. Spine (Phila Pa 1976). 2018;43:562-570. 10. De la Garza Ramos R, Nakhla J, Nasser R, Jada A, Purvis TE, Sciubba DM, et al. Factors associated with prolonged ventilation and reintubation in adult spinal deformity surgery. J Clin Neurosci. 2017; 43:188-191. Conflict of interest statement: A.H. Daniels reports personal fees from Stryker, Orthofix, Spineart, Globus Medical, Depuy Synthes, and Springer. The remaining authors report no conflicts of interest. Received 4 May 2018; accepted 3 July 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.07.029 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.07.029
Predictors of Nonneurologic Complications and Increased Length of Stay After Cervical Spine Osteotomy J. Mason DePasse1, Wesley Durand3, Alan H. Daniels2
BACKGROUND: Although previous studies have used National Surgical Quality Improvement Program (NSQIP) data to study complications of thoracolumbar spinal deformity surgery, investigation of cervical spine deformity surgery has been limited. We performed a retrospective analysis of the NSQIP database to identify predictors of complications after cervical spine osteotomy.
-
METHODS: Patients undergoing cervical spine osteotomy were identified in the NSQIP dataset using Current Procedural Terminology codes from years 2007e2016. For each patient, patient and case clinical characteristics, length of stay (LOS), and diagnosis of a nonneurologic complication (including reoperation and readmission) were abstracted. Patient and case clinical predictors of any of the reported complications and increased LOS were identified in multivariate logistic and Poisson regression analyses, respectively.
operative duration (P < 0.0001), and orthopaedic surgeon (vs. neurosurgeon) (P [ 0.0156). CONCLUSIONS: This study is the largest to date of patients undergoing cervical osteotomy and provides useful clinical data for patient selection and counseling and 30day reoperation and readmission rates.
-
-
RESULTS: There were 950 patients identified with mean age 56.1 12.4 years and mean body mass index 29.9 6.8. Mean LOS was 3.5 4.9 days. Overall medical complication rate was 15.8%. The most common complications were transfusion (78; 8.2%), readmission (45; 4.7%), reoperation (32; 3.4%), and reintubation (28; 3.0%). Risk factors for any complication included increased age (P [ 0.0467), American Society of Anesthesiologists classification III (P [ 0.0023) and IV (P [ 0.0013), and increased operative duration (P < 0.0001). Risk factors for increased LOS were decreased functional status (P [ 0.0037), disseminated cancer (P [ 0.0061), American Society of Anesthesiologists classification III and IV (P < 0.0001), increased
-
Key words Cervical spine - Complications - Deformity - Length of stay -
Abbreviations and Acronyms ASD: Adult spinal deformity ASA: American Society of Anesthesiologists IC: Incremental change LOS: Length of stay NIS: National Inpatient Sample
INTRODUCTION
W
ith advances in surgical technique, instrumentation, and understanding of sagittal alignment, operative correction of cervical deformity with osteotomy has become more common.1 Recent studies have reported that surgery improves multiple measures of pain and function2 and likely reaches cost-effectiveness at 4 years.3 However, currently few data are available on the complications of cervical deformity correction and associated risk factors. A study examining >360,000 patients in the National Inpatient Sample database who underwent surgical treatment for any cervical spine pathology found that pulmonary circulation disorders had the greatest impact on complication risk, although age 65 years and other comorbidities, such as renal disease, cancer, and congestive heart failure, were also predictive.4 The authors also noted that performance of an osteotomy was a risk factor for surgical complications, such as wound dehiscence, but it did not have as large an impact as fusion of 9 levels.4 Although these data are useful, it is not specific for cervical deformity. However, the finding that osteotomy increases the risk of surgical complications emphasizes the importance of identifying factors that could potentially be controlled to mitigate that risk.
NSQIP: National Surgical Quality Improvement Program OR: Odds ratio From the 1Department of Orthopaedics, 2Department of Orthopaedics, Division of Spine Surgery, and 3Alpert Medical School of Brown University, Providence, Rhode Island, USA To whom correspondence should be addressed: J. Mason DePasse, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.07.029 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY -: e1-e4, - 2018
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
Table 1. Patient and Case Characteristics, Complication Rate, and Length of Stay Variable
Overall
Complications
Continuous
Mean (SD)
Mean (SD)
Age, years
56.1 (12.4)
60.8 (12.5)
N/A
BMI
29.9 (6.8)
29.1 (7.0)
Operative duration, hours
3.1 (2.3)
5.6 (3.2)
Number (%)
Number (%)
Categorical
Number (%)
Number (%)
Mean (SD)
No
931 (98.0)
147 (15.8)
3.5 (4.8)
Yes
19 (2.0)
3 (15.8)
5.2 (5.2)
I or II
478 (50.4)
40 (8.4)
2.2 (2.5)
III
442 (46.6)
98 (22.2)
4.6 (5.6)
IV
28 (3.0)
11 (39.3)
9.0 (10.8)
Bleeding disorder Length of Stay
Categorical
Table 1. Continued
Mean (SD)
Surgical specialty Neurosurgery
622 (65.5)
97 (15.6)
3.2 (4.1)
Orthopaedics
328 (34.5)
53 (16.2)
4.1 (5.9)
788 (83.0)
122 (15.5)
4.7 (6.4)
Insulin-dependent
67 (7.1)
13 (19.4)
3.4 (4.9)
Noneinsulin-dependent
95 (10.0)
15 (15.8)
3.1 (3.0)
No
709 (74.6)
114 (16.1)
3.6 (5.0)
Yes
241 (25.4)
36 (14.9)
3.3 (4.5)
No
897 (94.4)
138 (15.4)
3.4 (4.8)
Yes
53 (5.6)
12 (22.6)
4.8 (6.1)
Dependent
42 (4.4)
17 (40.5)
8.3 (10.2)
Independent
908 (95.6)
133 (14.7)
3.3 (4.3)
No
897 (94.4)
136 (15.2)
3.5 (4.9)
Yes
53 (5.6)
14 (26.4)
3.9 (3.9)
ASA classification
BMI, body mass index; COPD, chronic obstructive pulmonary disease; ASA, American Society of Anesthesiologists.
Diabetes status No
Smoking
Dyspnea
Functional status
MATERIALS AND METHODS
COPD
Hypertension No
484 (51.0)
67 (13.8)
3.2 (4.7)
Yes
466 (49.1)
83 (17.8)
3.8 (5.0)
No
941 (99.1)
146 (15.5)
3.4 (4.8)
Yes
9 (1.0)
4 (44.4)
10.2 (8.6)
No
907 (95.5)
137 (15.1)
3.4 (4.5)
Yes
43 (4.5)
13 (30.2)
6.2 (9.4)
Disseminated cancer
Long-term steroid use
Continues
Two studies have investigated complications after surgical treatment for cervical deformity specifically. One evaluated 23 adult patients with cervical deformity treated with 3-column osteotomy and reported a complication rate of 56.5%, with
e2
neurologic deficit, wound infection, distal junctional kyphosis, and cardiorespiratory failure the most common complications.5 A second study of 120 patients, of which 28 underwent 3-column osteotomy, found that cardiopulmonary failure and myocardial infarction were the most common causes of death after surgery.6 Neither of the 2 studies was powered for identifying the causes of complications. To improve patient selection and counseling, we used the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database to identify predictors of complications and increased length of stay (LOS) after cervical spine osteotomy.
www.SCIENCEDIRECT.com
Data Sources This study employed the NSQIP dataset for the years 2007e2016. NSQIP collects data from member hospitals for major surgical procedures, including patient and procedure characteristics, morbidity and mortality outcomes, and reoperation and readmission status for 30 days postoperatively. NSQIP does not include data for pediatric cases, minor cases, trauma cases, and cases with American Society of Anesthesiologists (ASA) score of VI. Given the publicly available and deidentified nature of NSQIP data, this investigation was not considered human subjects research. Patient Selection Patients undergoing cervical spine osteotomy were identified based on the presence of Current Procedural Terminology code of 22210 or 22220. Exclusion criteria were surgeon specialty other than orthopaedic or neurologic surgery and emergent surgery. Dependent Variables The primary variables in the study were any complication and LOS measured in days. Complications abstracted from the database included transfusion, wound disruption, surgical site infection, reintubation, pneumonia, urinary tract infection, sepsis, thromboembolic event, major cardiac event, stroke, reoperation, readmission, and death.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.07.029
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
Table 2. Complications After Cervical Osteotomy Variable
Table 4. Risk Factors for Increased Length of Stay Number (%)
Variable
IC
95% CI
P Value
1.69
1.19e2.41
0.0037
2.47
1.29e4.71
0.0061
III
1.61
1.39e1.86
<0.0001
IV
2.72
1.72e4.31
<0.0001
Functional status (reference ¼ independent)
Any complication
150 (15.8)
Transfusion
78 (8.2)
Readmission
45 (4.7)
Reoperation
32 (3.4)
Postoperative intubation
28 (3.0)
SSI
20 (2.1)
Pneumonia
16 (1.7)
Sepsis
15 (1.6)
UTI
12 (1.3)
Orthopaedics
1.22
1.04e1.44
0.0156
Thromboembolic event
12 (1.3)
Operative duration (per hour)
1.17
1.14e1.20
<0.0001
Wound disruption
5 (0.5)
Stroke/CVA
3 (0.3)
Major cardiac event
2 (0.2)
Death
2 (0.2)
SSI, surgical site infection; UTI, urinary tract infection; CVA, cerebrovascular accident.
Independent Variables The following patient variables were analyzed: age, body mass index, functional status, smoking status, diabetes, chronic obstructive pulmonary disease, hypertension requiring medication, long-term steroid use, disseminated cancer, bleeding disorder, dyspnea, and ASA classification. In addition, the procedural variables operative time and surgical subspecialty were analyzed. Statistical Analysis Descriptive statistics were calculated for continuous variables, and the total number and frequency of all categorical variables were recorded for all patients as well as for patients who sustained any complication. The total number and frequency of each complication were also recorded. Additionally, mean LOS was calculated overall and for patients with each independent categorical variable. An analysis of risk factors for any complication and increased LOS was conducted. Bivariate analyses used c2, Fisher exact, Table 3. Risk Factors for Any Complication Variable
OR
95% CI
P Value
Age (per year)
1.02
1.00e1.04
0.0467
1.05e29.81
0.0433
Disseminated cancer (reference ¼ no) Yes
5.60
ASA (reference ¼ I or II) III
2.13
1.31e3.45
0.0023
IV
5.02
1.88e13.39
0.0013
1.60
1.47e1.74
<0.0001
Operative duration, per hour
OR, odds ratio; CI, confidence interval; ASA, American Society of Anesthesiologists.
WORLD NEUROSURGERY -: e1-e4, - 2018
Dependent Disseminated cancer (reference ¼ no) Yes ASA (reference ¼ I or II)
Surgical specialty (reference ¼ neurosurgery)
IC, incremental change; CI, confidence interval; ASA, American Society of Anesthesiologists.
Wilcoxon-Mann-Whitney, and Kruskal-Wallis tests, as appropriate. Risk factor analysis for total complications was performed with multiple stepwise logistic regression in a multivariate approach. Stepwise variable selection used thresholds of a ¼ 0.20 and a ¼ 0.10 for entry and retention, respectively. The final logistic regression models were evaluated using the HosmerLemeshow goodness-of-fit test. Multivariate analyses for LOS were performed with Poisson regression with robust error variance. All analysis was completed with SAS Version 9.4 (SAS Institute Inc., Cary, North Carolina). Statistical significance was defined as P < 0.05. RESULTS There were 950 patients identified with mean age of 56.1 years (SD 12.4) and mean body mass index of 29.9 (SD 6.8). Of patients, 162 (17.1%) had diabetes, 241 (25.4%) were smokers, and 53 (5.6%) had chronic obstructive pulmonary disease. ASA classification was as follows: ASA I or II in 478 (50.4%) patients, ASA III in 442 (46.6%) patients, and ASA IV in 28 (3.0%) patients. Neurosurgeons performed cervical spine osteotomy in 622 (65.5%) of cases. Mean operative duration was 3.1 hours (SD 2.3), and mean LOS was 3.5 days (SD 4.9) (Table 1). Overall complication rate was 15.8%. The most common complications included transfusion (78; 8.2%), readmission (45; 4.7%), reoperation (32; 3.4%), and reintubation (28; 3.0%) (Table 2). Multivariate analysis demonstrated that risk factors for any complication included increased age (odds ratio [OR] ¼ 1.02 per year, P ¼ 0.0467), disseminated cancer (OR ¼ 5.60, P ¼ 0.0433), ASA classification III (OR ¼ 2.13 vs. ASA I, P ¼ 0.0023) and IV (OR ¼ 5.02, P ¼ 0.0013), and increased operative duration (OR ¼ 1.60 per hour, P < 0.0001) (Table 3). Risk factors for increased LOS were decreased functional status (incremental change [IC] ¼ 1.69, P ¼ 0.0037), disseminated cancer (IC ¼ 2.47, P ¼ 0.0061), ASA classification III (IC ¼ 1.61, P < 0.0001) and IV (IC ¼ 2.72, P < 0.0001), orthopaedic surgeon (vs. neurosurgeon) (IC ¼ 1.22, P ¼ 0.0156), and
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE J. MASON DEPASSE ET AL.
PREDICTORS OF COMPLICATIONS AFTER CERVICAL SPINE OSTEOTOMY
increased operative duration (IC ¼ 1.17 per hour, P < 0.0001) (Table 4). DISCUSSION With the increasing use of cervical spine osteotomy for correction of adult cervical deformity,1,7 understanding predictors of complications is valuable for patient selection and counseling. We found that increased age, ASA classification of III or IV, and increased operative duration were associated with increased complications, whereas increased age, decreased functional status, cancer, ASA classification III or IV, orthopaedic specialty, and increased operative duration were predictive for increased LOS. In their study of complications associated with surgical treatment for any cervical spine pathology, Passias et al.4 reported that pulmonary circulation disorders were most predictive of morbidity, although age and other comorbidities were also associated with complications. Although increased age was a significant risk factor in our study, only ASA classification greater than II predicted complications; there was no association with any specific disease process. Our findings are more consistent with studies of complications after thoracolumbar adult spinal deformity (ASD) surgery. In a study of 5803 patients, Lee et al.8 found that age >60 years, ASA classification greater than II, and operative time >4 hours portended a higher risk of complications.8 The authors also found that decreased functional status (which we found increased LOS), insulindependent diabetes, and bleeding disorders were associated with complications.8 Similarly, Yoshida et al.9 found increased age, operative time, and estimated blood loss were independent predictors of complications in 304 patients with ASD, whereas De la Garza Ramos et al.10 reported that increased age, operative time >6 hours, and bleeding disorders were predictors of prolonged intubation or reintubation in a study of 1250 patients with ASD. It is possible that a history of bleeding disorder, which was a predictor of complications in studies by both Lee et al.8 and De
REFERENCES 1. Smith JS, Shaffrey CI, Bess S, Shamji MF, Brodke D, Lenke LG, et al. Recent and emerging advances in spinal deformity. Neurosurgery. 2017; 80:S70-S85. 2. Ailon T, Smith JS, Shaffrey CI, et al. Outcomes of operative treatment for adult cervical deformity: a prospective multicenter assessment with 1-year follow-up. Accessed April 12, 2018. [e-pub ahead of print]. Neurosurgery. https://doi.org/10.1093/ neuros/nyx574.
la Garza Ramos et al.,10 is less predictive in cervical deformity cases owing to decreased incision size and the potential need for more meticulous hemostasis. Otherwise, the risk factors identified in these thoracolumbar ASD studies are consistent with our findings in cervical deformity. It is noteworthy that decreased functional status was predictive of increased LOS but was not predictive of increased complications, especially given that Lee et al.8 found decreased functional status was associated with complications in patients with ASD. Incidence of major cardiac events and thromboembolic events was lower in our study than in their study,8 so it is possible that patients with cervical deformity are less susceptible to vascular events than patients with thoracolumbar ASD and therefore do not experience complications from stasis as frequently. The finding that orthopaedic surgery patients have significantly longer LOS is difficult to explain with our data, but it may be due to procedural differences. Further study will be required to better understand that result. Our study has several significant limitations. It is an uncontrolled retrospective study, and we could not compare patients undergoing osteotomy with patients undergoing cervical fusions without osteotomy. Additionally, the NSQIP database does not record neurologic complications, which suggests our findings underestimate the complication rate. Because we used cervical osteotomy Current Procedural Terminology codes to ensure our sample included only patients with cervical deformity, we also could not identify number of levels fused. NSQIP does not provide data beyond 30 days, so we could not study late complications, such as proximal junctional kyphosis or late failure. CONCLUSIONS Despite the above-mentioned limitations, our study provides predictive data for medical complications and reoperation and readmission rates. These data are valuable for patient selection and counseling for clinicians considering cervical osteotomy for deformity correction.
5. Smith JS, Shaffrey CI, Lafage R, Lafage V, Schwab FJ, Kim HJ, et al. Three-column osteotomy for correction of cervical and cervicothoracic deformities: alignment changes and early complications in a multicenter prospective series of 23 patients. Eur Spine J. 2017;26:2128-2137. 6. Smith JS, Shaffrey CI, Kim HJ, Passias P, Protopsaltis T, Lafage R, et al. Prospective multicenter assessment of all-cause mortality following surgery for adult cervical deformity. Accessed April 10, 2018. [e-pub ahead of print]. Neurosurgery. https://doi.org/10.1093/neuros/nyx605.
3. Poorman GW, Passias PG, Qureshi R, et al. Costutility analysis of cervical deformity surgeries using 1-year outcome. Accessed April 10, 2018. [e-pub ahead of print]. Spine J. https://doi.org/10. 1016/j.spinee.2018.01.016.
7. Kim HJ, Piyaskulkaew C, Riew KD. Comparison of Smith-Petersen osteotomy versus pedicle subtraction osteotomy versus anterior-posterior osteotomy types for the correction of cervical spine deformities. Spine (Phila Pa 1976). 2015;40: 143-146.
4. Passias PG, Diebo BG, Marascalchi BJ, Jalai CM, Horn SR, Zhou PL, et al. A novel index for quantifying the risk of early complications for patients undergoing cervical spine surgeries. J Neurosurg Spine. 2017;27:501-507.
8. Lee NJ, Kothari P, Kim JS, Shin JI, Phan K, Di Capua J, et al. Early complications and outcomes in adult spinal deformity surgery: an NSQIP study based on 5803 patients. Global Spine J. 2017;7: 432-440.
e4
www.SCIENCEDIRECT.com
9. Yoshida G, Hasegawa T, Yamato Y, Kobayashi S, Oe S, Banno T, et al. Predicting perioperative complications in adult spinal deformity surgery using a simple sliding scale. Spine (Phila Pa 1976). 2018;43:562-570. 10. De la Garza Ramos R, Nakhla J, Nasser R, Jada A, Purvis TE, Sciubba DM, et al. Factors associated with prolonged ventilation and reintubation in adult spinal deformity surgery. J Clin Neurosci. 2017; 43:188-191. Conflict of interest statement: A.H. Daniels reports personal fees from Stryker, Orthofix, Spineart, Globus Medical, Depuy Synthes, and Springer. The remaining authors report no conflicts of interest. Received 4 May 2018; accepted 3 July 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.07.029 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.07.029