Abstract
Introduction
Aspirin is commonly used for the primary and secondary prevention of cardiovascular disease and stroke. Controversy exists concerning whether and when is the optimal time to stop aspirin before spinal surgery. Previous studies on this topic mainly focused on patients who received thoracolumbar spine surgeries. There are only a few literatures concerning the safety of aspirin use in cervical spine surgery patients.
Methods
This pilot study recruited patients who received cervical laminoplasty from January 2010 to December 2021. The operation time, intraoperative blood loss, and postoperative complications of the patients who had taken aspirin during the perioperative period were compared with age, sex, and comorbidity-matched control patients. Propensity score matching was utilized in the selection of control to minimize bias.
Results
Twenty-one patients who have received cervical laminoplasty while taking aspirin during the perioperative period were included. The control group included 21 age, sex, and comorbidity-matched patients who have not taken aspirin. The operation time and intraoperative and postoperative blood loss were higher in patients taking aspirin but were not statistically significant. There was no statistically significant difference in the wound infection rate. No case of epidural hematoma was found.
Conclusions
In patients undergoing cervical laminoplasty, a continuation of aspirin is safe and does not increase the difficulty of the surgery, wound complication, or hemorrhage.
Keywords: Aspirin, cervical spine surgery, laminoplasty, complications, blood loss
Introduction
Low-dose aspirin is commonly used for the primary and secondary prevention of cardiovascular disease and stroke1). Aspirin exerts its antithrombotic effect through its irreversible inhibition of cyclooxygenase (COX). The effects of aspirin last for the duration of the life of the platelet (≈10 days). Although it may take 10 days for the total platelet population to be renewed, it has been shown that if as little as 20% of platelets have normal COX activity, hemostasis may be normal2). It is a common practice to stop aspirin before spine surgery3), but controversy exists concerning whether and when is the optimal time to stop aspirin before spinal surgery4-6). Several studies and systematic reviews had failed to show an increase in bleeding-related complications with the continual use of aspirin in the perioperative period in spine surgery7-14). Nevertheless, most of these studies focused on patients who received thoracolumbar spine surgeries. It was not until 2022 that studies focusing on the use of aspirin in cervical spine surgeries were published13,14). In this pilot study, we aim to investigate the safety of the perioperative use of aspirin in patients undergoing cervical laminoplasty. Propensity score matching (PSM)15,16) was used in the selection of the control group to avoid confounding biases such as patient's demographic or comorbidities affecting the outcome.
Materials and Methods
1 Study design and patient cohort
This was a retrospective cohort study conducted in a single institution. The study was approved by our institution's review board. Patients who received cervical laminoplasty from January 2010 to December 2021 were identified through electronic patient records. Only patients who had received laminoplasty from C3 to C7 segments were included in the study. Patients were excluded if they had taken aspirin but withheld before operation, had taken any other anticoagulant, or had incomplete patient records. The cohort included 114 patients (87 males, 27 females) with a mean age of 63.7±10.5 years (range: 40-85 years) at the time of operation. The diagnosis of the patients was either cervical myelopathy or central cord syndrome.
2 Surgical techniques
All patients received a double door laminoplasty from C3 to C7 spinal segments using Kurokawa's technique17). The patient's head was stabilized using a skull tong with 3 kg traction on a spinal table. A midline incision was made spanning from C3 to C7, and paraspinal muscles were dissected subperiosteally until the facets were exposed, with the preservation of the facet joints capsule. Longitudinal grooves were made bilaterally at the lamina-facet junction by resecting the outer cortex and a part of the cancellous bone using a 3.5 mm diamond burr. The interspinous ligaments and the tip of the spinous process of C3 to C7 were removed. The lamina of C3 to C7 were split in the midline at the spinous process using a 2.5 mm high-speed diamond burr under operating microscope magnification. Subsequently, the split laminas were opened in the midline, and the constricting fibrous band above the dura mater and the hypertrophied yellow ligament were resected from C3 to C7 levels. Two-millimeter-sized holes were made at the split spinous process using a high-speed drill, and trapezoidal-shaped hydroxyapatite spacer blocks were anchored to the split spinous process using ETHICON™ PDSⓇ (polydioxanone) and ETHICON™ VICRYLⓇ (polyglactin) sutures through the drill holes. Fig. 1 shows the intraoperative photographs. A suction drain was put in the subfascial layer before wound closure. The wound was closed with absorbable sutures. The drain was kept postoperatively until the daily output is less than 50 mL/24 h. All patients were given a rigid neck collar after the operation for 6 weeks. Patients were discharged from the hospital if they could walk safely with a walking aid or transferred to a rehabilitation unit if further training is required.
Figure 1.
(A) Longitudinal groove created at the junction of lamina and facet with a high-speed burr. (B) Spinous process split in the midline using a high-speed burr. (C) Split laminas opened. (D) Hydroxyapatite spacer blocks anchored to split spinous process with sutures.
3 Outcome measures
The main outcome measures were the operation time, intraoperative blood loss, preoperation and postoperation hemoglobin levels, postoperation drain output, and blood transfusion requirement defined as the total unit of blood transfused either intraoperatively or postoperatively. Postoperative complications including wound infection, epidural hematoma, or medical complications such as chest or urinary infection, cardiac arrhythmia, and blood electrolyte disturbances were also compared.
4 Study design and statistical analysis
We compared the operation time, intraoperative blood loss, postoperation hemoglobin, and postoperation drain output between those who had taken aspirin in the perioperative period and those who had never taken aspirin. We also compared the incidence of postoperative medical or surgical complications between these patient groups.
PSM was used for the selection of nonaspirin-taking control patients to reduce bias due to an imbalance in the numbers of aspirin-taking patients and nonaspirin-taking patients; 1:1 PSM (aspirin-taking patients: nonaspirin-taking patients) was used to balance the following confounding variables: age, sex, preexisting comorbidities, and smoking status. The clipper value for the PSM for all confounding factors was set as 0.1. Propensity scores of individuals were calculated with logistic regression analysis (IBM SPSS 27, Armonk, NY: IBM Corp). Fig. 2 shows the flowchart in patients' selection.
Figure 2.
Flowchart showing the procedures in patients’ selection.
Statistical analysis was performed using IBM SPSS Statistics software ver. 25.0 (IBM Corp., Armonk, NY, USA). Student's t-test was used for the comparison of continuous data, whereas the Chi-square test was used for the comparison of categorical data. A p-value of ≤0.05 was considered statistically significant. The normality of continuous data was checked using the “Shapiro-Wilk test.” The goodness-of-fit test was used to test the null hypothesis of equiprobability for categorical data.
5 Sample size calculation
The intraoperative blood loss (mL) was considered of clinically meaningful difference. Assuming that the standard deviation was 10% (i.e., population variance is 0.01), the required sample size of each group to achieve an 80% power (β=0.2) at α=0.05 for correctly detecting such difference of μ2−μ1=0.062 change obtained by normal approximation as N=41. Consequently, we matched 21 aspirin-taking patients with 21 nonaspirin-taking patients.
Results
1 Patient's characteristics
Of the 114 patients eligible to be recruited, 21 patients (16.4%) were taking aspirin. The reasons for these patients taking aspirin were as follows: coronary artery disease or history of myocardial infarction (n=11), history of cerebrovascular accident (n=6), and primary prevention (n=4). The doses of aspirin ranged from 80 to 160 mg per day. Patients taking aspirin were significantly older and had a significantly higher prevalence of diabetes or cerebrovascular or cardiovascular disease before PSM. After matching the propensity score, the 21 PSM-matched nonaspirin-taking patients did not have any significant difference in demographics, comorbidities, and smoking status compared with the 21 aspirin-taking patients (Table 1).
Table 1.
Patient’s Demographic Data.
| Group | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| On aspirin (n=21) | Without aspirin (n=93) | P value | On aspirin (n=21) | Without aspirin (n=21) | P value | |
| Sample size | 21 | 93 | 21 | 21 | ||
| Age (years) | 68.24±10.91 (44, 83) | 62.71±10.15 (40, 85) | 0.03 | 68.24±10.91 (44, 83) | 66.76±10.17 (44, 83) | 0.65 |
| Sex (male/female) | 15/6 | 72/21 | 0.58 | 15/6 | 15/6 | 1.00 |
| Medical comorbidities | ||||||
| HTN | 15/21 | 53/93 | 0.33 | 15/21 | 13/21 | 1.00 |
| CAD | 2/21 | 1/93 | 0.09 | 2/21 | 0/21 | 0.49 |
| With cardiac stent | 4/21 | 0/93 | <0.01 | 4/21 | 0/21 | 0.11 |
| With CABG | 2/21 | 0/93 | 0.03 | 2/21 | 0/21 | 0.49 |
| DM | 12/21 | 24/93 | 0.01 | 12/21 | 8/21 | 0.35 |
| Hyperlipidemia | 8/21 | 23/93 | 0.28 | 8/21 | 7/21 | 1.00 |
| CVA | 6/21 | 0/93 | <0.01 | 6/21 | 0/21 | 0.12 |
| ICA stenosis | 1/21 | 0/93 | 0.18 | 1/21 | 0/21 | 1.00 |
| Smoker | 4/21 | 38/93 | 0.31 | 4/21 | 8/21 | 0.57 |
NS=not significant
HTN, hypertension; CAD, coronary artery disease; CABG, coronary artery bypass graft; DM, diabetes mellitus; CVA, cerebrovascular accident; ICA, stenosis internal carotid artery stenosis
* Age (years) by Student’s T-test
* Others (categorical variables) by Chi-square test
2 Comparison of the operation time, intraoperative blood loss, preoperative and postoperative hemoglobin, and postoperative drain output
There was no statistically significant difference between aspirin-taking and nonaspirin-taking patients during the operation time (181.81±47.42 vs. 167.9±36.22 min P=0.29), intraoperative blood loss (403.81±432.38 vs. 354.76±358.78 mL P=0.69), postoperative drain output (523.81±231.46 vs. 489.52±204.71 mL P=0.61), and average total blood transfusion requirement (0.33±0.73 vs. 0.1±0.3 unit P=0.18). Table 2 summarizes the results, and Table 3 shows the test for data normality.
Table 2.
Summary of Surgical Results (Mean±SD).
| Parameter | On aspirin (n=21) | Without aspirin (n=21) | P value |
|---|---|---|---|
| Operation time (mins) | 181.81±47.42 | 167.90±36.22 | 0.29 |
| Preoperative Hb (g/dL) | 13.29±1.39 | 13.69±1.70 | 0.41 |
| Intraoperative blood loss (mL) | 403.81±432.38 | 354.76±358.78 | 0.69 |
| Postoperative Hb (g/dL) | 11.28±1.86 | 12.01±1.80 | 0.20 |
| Postoperative drain output (mL) | 523.81±231.46 | 489.52±204.71 | 0.61 |
| Total transfusion requirement (unit) | 0.33±0.73 | 0.10±0.30 | 0.18 |
SD, standard deviation
Hb, hemoglobin
* All comparisons by Student’s T-test
Table 3.
Test for Data Normality.
| Parameter | Shapiro-Wilk Test | Skewness | Kurtosis | |
|---|---|---|---|---|
| On aspirin (n=21) | W (df=21) | P value | ||
| Operation time (min) | 0.152 | 0.065 | 0.481 | 0.244 |
| Preoperative Hb (g/dL) | 0.150 | 0.062 | 0.108 | 0.301 |
| Intraoperative blood loss (mL) | 0.156 | 0.066 | 0.589 | 0.459 |
| Postoperative Hb (g/dL) | 0.150 | 0.063 | 0.394 | 0.761 |
| Postoperative drain output (mL) | 0.149 | 0.061 | 0.822 | 0.759 |
| Total transfusion requirement (unit) | 0.141 | 0.059 | 0.923 | 0.850 |
| Without aspirin (n=21) | ||||
| Operation time (min) | 0.179 | 0.078 | 0.527 | 0.197 |
| Preoperative Hb (g/dL) | 0.180 | 0.079 | 0.126 | 0.711 |
| Intraoperative blood loss (mL) | 0.182 | 0.080 | 0.540 | 0.089 |
| Postoperative Hb (g/dL) | 0.178 | 0.076 | 0.650 | 0.292 |
| Postoperative drain output (mL) | 0.189 | 0.058 | 0.646 | 0.798 |
| Total transfusion requirement (unit) | 0.193 | 0.090 | 0.975 | 0.962 |
W: Test statistics for Shapiro-Wilk Test
3 Comparison of postoperative medical and surgical complications
There was a significant difference in the incidence of medical complications between aspirin-taking vs. nonaspirin-taking patients (23.8% vs. 0%, P=0.05). Table 4 shows the details of the medical complications.
Table 4.
Association between Aspirin Taking and Any Postoperative Medical Complications.
| Medical complication | P value | ||
|---|---|---|---|
| Yes | No | ||
| On aspirin (n=21) | 5 (23.8%) | 16 (76.2%) | 0.05* |
| Without aspirin (n=21) | 0 | 21 (100.0%) | |
| Total | 5 | 37 | |
* Significant difference found
* Chi-square test
Details of medical complications: Urinary tract infection (UTI) n=1; UTI+hyponatremia n=1; bradycardia n=1, hematuria n=1, chest infection n=2
Goodness-of-fit test on testing the null hypothesis of equiprobability (equidistribution): The distribution was not significantly different from an equiprobability model (X2=0.52, P=0.779).
There were three cases of postoperative wound infection, which required surgical debridement. The incidence was higher in the aspirin-taking vs. nonaspirin-taking group (9.5% vs. 4.8%), but the difference was not statistically significant (Table 5). There was no case of symptomatic postoperative epidural hematoma in the entire cohort of 114 patients.
Table 5.
Association between Aspirin Taking and Any Postoperative Wound Infection Complications.
| Wound infection complications | P value | ||
|---|---|---|---|
| Yes | No | ||
| On aspirin (n=21) | 2 (9.5%) | 19 (90.5%) | 1.00* |
| Without aspirin (n=21) | 1 (4.8%) | 20 (95.2%) | |
| Total | 3 | 39 | |
* Chi-square test
Wound infection complications: three wound infection cases required surgical debridement
Goodness-of-fit test on testing the null hypothesis of equiprobability (equidistribution): The distribution was not significantly different from an equiprobability model (X2=0.87, P=0.324).
Discussion
With the aging population, we encounter more and more patients taking aspirin presenting to the spine clinic. A recent review has suggested that for most patients using aspirin for primary cardiovascular disease prevention, preoperative aspirin cessation is safe18). There is controversy concerning the perioperative use of aspirin for those patients who are taking aspirin for secondary prevention of cardiovascular disease. The cessation of aspirin may cause a platelet rebound phenomenon and a prothrombotic state leading to a major adverse cardiovascular event, known as the “Aspirin Withdrawal Syndrome”19). In the 2014 PeriOperative Ischemia Evaluation-2 (POISE-2) trial, which evaluated the outcome of those continuing or stopping low-dose aspirin in “at risk” patients undergoing noncardiac surgeries, the authors concluded that perioperative aspirin continuation dose not confer a benefit in terms of major adverse cardiovascular events but may contribute to excessive blood loss20). However, the 2014 “ACC/AHA guideline on perioperative cardiovascular evaluation and management of patients undergoing noncardiac surgery” suggested that continuing aspirin is reasonable for patients already receiving it for secondary prevention of cardiovascular disease when major noncardiac surgery is planned and the perceived risk of a major adverse cardiovascular event is greater than the risk of bleeding21).
There is controversy concerning the optimal time to stop aspirin before spine surgery, with some studies suggesting stopping aspirin for just 3-7 days before operation might not be sufficient4). Thus, in our control group, we excluded the six patients who withheld aspirin for at least 5 days before the operation as the effect of aspirin might still persist.
Several studies have investigated the effect of perioperative use of aspirin in thoracolumbar spine surgeries6-9,11,12). Two systematic reviews11,12) found that there was no observed increase in bleeding or operation time with continual aspirin use. Nevertheless, this finding cannot be directly applied to cervical spine surgery patients as most of the included subjects received thoracolumbar spine surgeries.
In 2022, two studies on aspirin use in cervical spine surgeries were published in Spine Surgery and Related Research13,14). In the Japanese multicenter study by Uehara et al.13), there was no increase in blood loss in elderly patients taking antiplatelets or anticoagulants who underwent cervical spine injury surgeries. However, this study included patients taking antiplatelets or anticoagulants other than aspirin, and a variety of cervical spine procedures with different approaches were included, making the results less specific. The retrospective cohort study by Inoue et al.14) included 73 aspirin-taking and 322 nonaspirin-taking patients who received cervical laminoplasty in a single institution. There was no difference in surgical time, intraoperative blood loss, postoperative drain output, or perioperative complications. However, significant differences in the baseline demographics and medical comorbidities were present between the aspirin- and nonaspirin-taking groups. The aspirin-taking group included more male patients, who were being older and had more cardiovascular or cerebrovascular risk factors and complications, and this could have caused confounding bias in the results.
Our results are similar to the findings of the study by Inoue et al.14) in which taking aspirin in the perioperative period did not cause a significant increase in the operation time, intraoperative and postoperative bleeding, and blood transfusion requirement in patients who receive cervical laminoplasty. There was no significant difference in the incidence of postoperative wound infection between aspirin- and nonaspirin-taking patients. The only significant difference in our study was the higher incidence of postoperative medical complications among aspirin-taking patients, which may be accounted for by the presence of less common medical comorbidities among the aspirin-taking patients not matched by PSM.
To our knowledge, this is the second study that assessed the safety of the perioperative use of aspirin in cervical laminoplasty. The demographics and underlying comorbidities of the controls were carefully matched using PSM analysis thus reducing selection bias. Our results can certainly supplement the findings of the prior similar study by Inoue et al.14) as our patients received the same type of cervical laminoplasty and the outcome measures were similar.
There are several limitations in this study. First, the sample size of this study is small and might be underpowered. However, even if statistical significance is reached for the increase in operation time and intraoperative and postoperative blood loss, the small difference is not clinically significant. Our study serves as a pilot study and a multicenter study with a larger sample size can overcome this problem. Surgeons should be particularly interested in the two bleeding-related complications: wound infection and spinal epidural hematoma (SEH). In our results, the incidence of postoperative wound infection was higher in aspirin-taking patients but not statistically significant, a multicenter study with a larger patient volume would be able to produce more evidence on this question. There was no case of SEH in our patients. However, SEH was reported to be a rare event in the order of 0.2%22), and therefore, a much larger case number would be required to assess the effect of aspirin use on the incidence of postoperative SHE. Second, the retrospective nature of this study makes it difficult to exclude selection bias. Third, despite the use of PSM in the selection of control, only common medical comorbidities such as hypertension, diabetes, and cardiovascular or cerebrovascular disease were matched. The presence of less common diseases such as renal failure and autoimmune disease in either group may affect the outcome.
Conclusions
Our results suggest that in patients undergoing cervical laminoplasty, a continuation of aspirin is safe and does not increase the difficulty of surgery, wound complication, or hemorrhage. A multicenter study with a larger sample size should be done in the near future to give a conclusion.
Conflicts of Interest: The authors declare that there are no relevant conflicts of interest.
Sources of Funding: None.
Author Contributions: (1) Concept or design: Yu Chung Wong
(2) Acquisition of data: Tsun Kit Lau and Yu Chung Wong
(3) Analysis or interpretation of data: Wai Wang Chau
(4) Drafting of the manuscript: Yu Chung Wong
(5) Critical revision for important intellectual content: Kin On Kwok and Sheung Wai Law
All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.
Ethical Approval: The study was approved by the ethics review board of the Joint New Territories East Cluster/Chinese University of Hong Kong Ethics Committee (reference number: 2021.302).
Informed Consent: This study is a retrospective cohort study. Our institution did not require informed consent for retrospective studies.
References
- 1.Patrono C, Baigent C. Role of aspirin in primary prevention of cardiovascular disease. Nat Rev Cardiol. 2019;16(11):675-86. [DOI] [PubMed] [Google Scholar]
- 2.Awtry EH, Loscalzo J. Aspirin. Circulation. 2000;101(10):1206-18. [DOI] [PubMed] [Google Scholar]
- 3.Korinth MC, Gilsbach JM, Weinzierl MR. Low-dose aspirin before spinal surgery: results of a survey among neurosurgeons in Germany. Eur Spine J. 2007;16(3):365-72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Epstein NE. When and if to stop low-dose aspirin before spine surgery? Surg Neurol Int. 2018;9:154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Porto GBF, Jeffrey Wessell DO, Alvarado A, et al. Anticoagulation and spine surgery. Global Spine J. 2020;10(1 Suppl):53S-64S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kang SB, Cho KJ, Moon KH, et al. Does low-dose aspirin increase blood loss after spinal fusion surgery? Spine J. 2011;11(4):303-7. [DOI] [PubMed] [Google Scholar]
- 7.Lee JS, Son DW, Sung SK, et al. Effects of discontinuance of preoperative anti-platelet medication in multi-level thoracolumbar spine surgery. Turk Neurosurg. 2018;28(1):99-104. [DOI] [PubMed] [Google Scholar]
- 8.Soleman J, Baumgarten P, Perrig WN, et al. Non-instrumented extradural lumbar spine surgery under low-dose acetylsalicylic acid: a comparative risk analysis study. Eur Spine J. 2016;25(3):732-9. [DOI] [PubMed] [Google Scholar]
- 9.Kulkarni AG, Patel J, Khandge A, et al. The practice of continuation of anti-platelet therapy during the perioperative period in lumbar minimally invasive spine surgery (MISS): how different is the morbidity in this scenario? Spine. 2020;45(10):673-8. [DOI] [PubMed] [Google Scholar]
- 10.Cuellar JM, Petrizzo A, Vaswani R, et al. Does aspirin administration increase perioperative morbidity in patients with cardiac stents undergoing spinal surgery? Spine. 2015;40(9):629-35. [DOI] [PubMed] [Google Scholar]
- 11.Cheng A, Poon MTC, Demetriades AK. Aspirin therapy discontinuation and intraoperative blood loss in spinal surgery: a systematic review. Neurosurg Rev. 2018;41(4):1029-36. [DOI] [PubMed] [Google Scholar]
- 12.Zhang C, Wang G, Liu X, et al. Safety of continuing aspirin therapy during spinal surgery: a systematic review and meta-analysis. Medicine. 2017;96(46):e8603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Uehara M, Ikegami S, Takizawa T, et al. Is blood loss greater in elderly patients under antiplatelet or anticoagulant medication for cervical spine injury surgery? A Japanese multicenter survey. Spine Surg Relat Res. 2022;6(4):366-72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Inoue T, Mizutamari M, Hatake K. Safety of continuous low-dose aspirin therapy for cervical laminoplasty. Spine Surg Relat Res. 2021;6(3):240-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika 1983;70(1):41-55. [Google Scholar]
- 16.Kane LT, Fang T, Galetta MS, et al. Propensity score matching: a statistical method. Clin Spine Surg. 2020;33(3):120-2. [DOI] [PubMed] [Google Scholar]
- 17.Hirabayashi S. Recent surgical methods of double-door laminoplasty of the cervical spine (Kurokawa's method). Spine Surg Relat Res. 2018;2(2):154-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Gerstein NS, Albrechtsen CL, Mercado N, et al. A comprehensive update on aspirin management during noncardiac surgery. Anesth Analg. 2020;131(4):1111-23. [DOI] [PubMed] [Google Scholar]
- 19.Gerstein NS, Schulman PM, Gerstein WH, et al. Should more patients continue aspirin therapy perioperatively?: clinical impact of aspirin withdrawal syndrome. Ann Surg. 2012;255(5):811-9. [DOI] [PubMed] [Google Scholar]
- 20.Devereaux PJ, Mrkobrada M, Sessler DI, et al. Aspirin in patients undergoing noncardiac surgery. N Engl J Med. 2014;370(16):1494-503. [DOI] [PubMed] [Google Scholar]
- 21.Fleisher LA, Fleischmann KE, Auerbach AD, et al. 2014 ACC/AHA guideline on perioperative cardiovascular evaluation and management of patients undergoing noncardiac surgery: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. J Am Coll Cardiol. 2014;64(22):e77-137. [DOI] [PubMed] [Google Scholar]
- 22.Glotzbecker MP, Bono CM, Wood KB, et al. Postoperative spinal epidural hematoma: a systematic review. Spine. 2010;35(10):E413-20. [DOI] [PubMed] [Google Scholar]