Venous thromboembolism (VTE)–encompassing both deep venous thrombosis (DVT) and pulmonary embolism (PE)–is a leading cause of death in patients with cancer [2, 11, 23]. Patients with cancer are thought to have a four- to sevenfold increased risk for developing a symptomatic VTE when compared with patients without cancer [7, 20]. Spine surgery is also known to be an independent risk factor for symptomatic VTE [8, 15]. Consequently, patients undergoing surgery for spine metastases may be at an even greater risk for developing these complications . In a recent study among patients undergoing surgery for tumors, 22% developed VTE .
Determining which factors are associated with symptomatic VTE and assessing the consequences of these conditions on survival may identify those who could maximally benefit from symptomatic VTE prevention strategies in the perioperative period. The role of chemoprophylaxis in preventing symptomatic VTE in patients undergoing high-risk spine surgery remains controversial [16, 17]. Spine surgeons must weigh the risk of chemoprophylaxis, which includes a higher risk of bleeding, with the benefits of preventing symptomatic VTE. When considering this relationship, one must have an accurate understanding of the incidence and consequences of symptomatic VTE. The incidence of postoperative spinal epidural hematomas and symptomatic VTE–regardless of the use of chemical anticoagulants–appears to be low after spine surgery [16, 18], but it is unclear if the same is true after spine metastases surgery.
In this study, we therefore sought to (1) identify the proportion of patients who develop symptomatic VTE within 90-days of surgical treatment for spine metastases; (2) identify the factors associated with the development of symptomatic VTE among patients receiving surgery for spine metastases; (3) assess the association between the development of postoperative symptomatic VTE and 1-year survival among patients who underwent surgery for spine metastases; and (4) assess if chemoprophylaxis increases the risk of wound complications among patients who underwent surgery for spine metastases.
Patients and Methods
Study Design and Setting
Our institutional review board approved a waiver of consent between January 2002 and January 2014 for this retrospective study at the authors’ hospitals.
We included 637 patients who were 18 years of age or older and had surgery for cervical, thoracic, or lumbar metastases (inclusive of lymphoma and multiple myeloma) . We excluded patients with (1) kyphoplasty and vertebroplasty; (2) revision procedures, defined as any subsequent procedure after the index surgery addressing the metastatic lesion; and (3) a symptomatic and confirmed VTE within 2 weeks before surgery because this would interfere with the main aim of the study to identify the risk of developing postoperative VTE.
Demographics, Description of Study Population
The patients included 371 males (58%) and 266 females (42%) with a median age of 60 years (interquartile range [IQR], 52–68 years; Table 1). The median duration of surgery was 6 hours (IQR, 5–8 hours) and median hospitalization was 8 days (IQR, 6–12 days). Of the 637 spine metastatic operations, 371 (58%) involved the thoracic region; 141 (22%) involved the lumbar area; 86 (14%) the cervical; and 39 (6%) the combined region. IVC filters were placed in 41 patients: 34 (6%) in the nonVTE and seven (10%) in the VTE group. Most common primary tumor types included lung (18%), kidney (13%), and breast cancer (12%; Table 2).
Accounting for All Patients/Study Subjects
We identified 1330 patients using the ICD-9 code for the diagnosis of pathologic vertebrae fracture (733.13). Additionally, we further identified 796 patients using a word-based inquiry of operative reports in our medical database. We manually inspected the medical charts of these 2126 patients for eligibility by two independent research fellows (OQG, PTO), and ultimately included 637 patients . We verified followup until October 4, 2016. At followup after 90 days and 1 year, respectively, 21 of 637 (3%) and 41 of 637 (6%) were lost to followup.
Description of Experiment, Treatment or Surgery
During the period in question, we generally used either 40 mg of enoxaparin or 5000 IUs subcutaneous heparin every 12 hours. Other general thromboembolic prophylaxis approaches and dosages used were: 325 mg of aspirin, 5000 IUs dalteparin daily, and warfarin dependent to maintain an international normalized ratio of 2.0:2.5. There may have been between-surgeon differences in prescribing choices; in particular we found that three surgeons prescribed more heparin relative to low-molecular-weight heparin (LMWH) than others. We therefore performed a more-detailed analysis on the patient populations treated by these surgeons and found them not to be different in terms of age, sex, body mass index (BMI), modified Charlson Comorbidity Index, or American Spinal Injury Association (ASIA) impairment scale (more details on this below, in Statistical Analysis, and in Appendix, Supplemental Digital Content 1, https://links.lww.com/CORR/A174).
Patients on preoperative chemoprophylaxis continued their initial medication postoperatively. All chemoprophylaxis was started 48 hours after surgery and continued day to day but was discontinued if a bleeding complication developed. When chemoprophylaxis was contraindicated, an inferior vena cava (IVC) filter was placed before surgery. Chemical anticoagulants, given postoperatively with a maximum range of 14 days, were considered prophylactic. When regimens overlapped, the most aggressive chemoprophylaxis regimen was considered in our analysis. Mechanical prophylaxis was routinely employed in the form of sequential compression devices and compression stockings at both institutions in all patients during the hospitalization period and therefore not included as a potential treatment variable. LMWH (including enoxaparin and dalteparin, in general dosages of respectively 40 mg and 5000 IUs daily) was the most commonly used chemoprophylaxis in 308 (48%) patients (Table 3). Subcutaneous heparin was injected into 127 patients (20%); aspirin was used for 92 patients (14%) patients; and warfarin was administered in 21 patients (3.3%). No form of chemoprophylaxis was prescribed for 89 patients (14%).
Variables, Outcome Measures, Data Sources, and Bias
Our primary outcome, VTE, was defined as any symptomatic pulmonary embolism (PE) or symptomatic distal or proximal deep venous thromboembolism (DVT) that occurred within 90 days of surgery, presenting with swelling, calf tenderness, tachycardia, pain in the lower extremities, haemoptysis, or tachypnea. Patients were tested at the earliest possible opportunity after the development of these symptoms. The diagnosis was confirmed by one of the following diagnostic procedures: pulmonary arteriography, vascular ultrasound, venography or chest CT. Our secondary outcome was survival after surgery. The date of death was obtained from the Social Security Index and medical charts. Our third outcome was a documented wound complication within 90 days of surgery, defined as a wound complication that might be attributable to chemoprophylaxis and resulted in longer hospitalization. We excluded wound complications such as wound inflammation resulting in the use of antibiotics. Sixty-six patients (10%) had a documented wound complication, consisting of 34 deep infections (5.3%) that were treated by irrigation and débridement; 28 superficial wound complications (4.4%), such as wound dehiscence resulting in surgery; and 12 deep wound complications (1.9%), including seven symptomatic spinal epidural hematomas (1.1%), four seromas (0.6%), and one splenic bleed nine days postoperatively (0.2%). Patients with symptomatic spinal epidural hematomas presented with back pain and/or progressive neurologic dysfunction for which the diagnosis was confirmed with MRI. Decompression surgery resulted in six neurologically intact patients (0.9%) and one patient (0.2%) who maintained a deteriorated neurological outcome. Ten patients (1.6%) had a wound complication followed by a symptomatic VTE.
Preoperative local radiotherapy was administered in 29 patients (40%) of the VTE group. Disease factors included primary tumor type, pathologic fracture, number of bone and visceral metastases, preoperative ASIA impairment scale, and time from primary tumor diagnosis to operation for metastatic disease. Clinical factors included the preoperative comorbidity status defined using the modified Charlson Comorbidity Index  and the use of preoperative local radiotherapy or systemic therapy. Treatment factors included prior embolization, IVC filter placement, vertebral levels included in surgery, surgical approach, surgery duration in hours, estimated blood loss during surgery (liters), total perioperative transfusions, and hospitalization days. Laboratory factors included preoperative hemoglobin levels (g/dL), preoperative white blood cell count (1000/mm3), preoperative platelet count (1000/mm3), creatinine levels (mg/dL) and calcium levels (mg/dL). We obtained preoperative laboratory values by choosing laboratory values nearest to surgery with a maximum range of 7 days. We used the modified Charlson Comorbidity Index to determine the comorbidity status based on an algorithm of the ICD-9 codes classifying 12 comorbidities. We dichotomized the comorbidity status into any additional comorbidity or none (in addition to the metastases). We determined any preoperative neurological deficits (grade A, B, C, or D) or none (score E, including patients with prior but no present deficits) using the preoperative ASIA impairment scale . The Eastern Cooperative Oncology Group (ECOG) performance status was dichotomized into good (0, 1, or 2) or poor scores (3 to 4) [31, 34]. We defined previous systemic therapy as all types of nonradiotherapeutic adjuvants or nonsurgical adjuvants, for example, immunologic, cytotoxic, metabolic or hormonal therapy, administered before surgery. We considered the presence of an IVC filter before surgery or within 90 days postoperatively as prophylactic, except when it was placed after a symptomatic VTE event.
We used multivariable logistic regression analysis controlling for confounding variables with a p value < 0.10 from bivariate testing and presumed to be relevant to VTE to assess independent risk factors for symptomatic VTE . Odds ratios for continuous variables are interpreted in terms of each unit increase or decrease on the scale (that is, 1 to 2, 2 to 3, etc; each one-unit/hour increment of longer duration of surgery with an odds ratio of > 1 corresponds to an increased risk of the outcome in question, in this case, symptomatic VTE). In bivariate analysis, two of the 33 variables examined were associated with increased and decreased risk of symptomatic VTE development (see Appendix, Content 2, https://links.lww.com/CORR/A175), respectively: longer duration of surgery (odds ratio [OR], 1.17; 95% confidence interval (CI), 1.07–1.28; p = 0.001) and the absence of metastatic lesion in the cervical region (OR, 0.34; 95% CI, 0.12–0.96; p = 0.042). Two additional variables with p values < 0.10 were included in the multivariable analysis: preoperative white blood cell count (OR, 1.03; 95% CI, 1.00–1.07; p = 0.064) and BMI of > 30 kg/m2 (OR, 1.72; 95% CI, 0.98–3.02; p = 0.058). Additional variables controlled for were age, the modified Charlson Comorbidity Index, the preoperative ASIA impairment scale, visceral metastases, chemoprophylaxis, and estimated blood loss during surgery. Multivariate logistic regression was also performed to examine the association between chemoprophylaxis and wound complication controlling for age, sex and the modified Charlson Comorbidity Index. Bivariate logistic regression was used to assess chemoprophylaxis regimens and surgeons (Appendix, Supplemental Digital Content 1, https://links.lww.com/CORR/A174). The following baseline characteristics were assessed for differences between surgeons, none were significant: age (p = 0.821), sex (p = 0.344), BMI (p = 0.067), the modified Charlson Comorbidity Index (p = 0.077), and ASIA impairment scale (p = 0.253). We used Cox regression analysis after controlling for the confounding factors age, sex, BMI, the modified Charlson Comorbidity Index, visceral metastases, primary tumor type, previous systemic therapy, and operation type to assess a difference in survival between the symptomatic VTE and nonVTE group. Kaplan-Meier plots demonstrated the survival curves for both groups. Multiple chained imputation was used to estimate missing values to retain all values for multivariable analysis. The dataset was recreated multiples times (40 in our cohort) by multiple imputation and the missing values were estimated with plausible values based on the residual variables accounting for uncertainty. Statistical software estimated missing values for: BMI (14% [89 of 637]), duration of surgery (11% [71 of 637]), total estimated blood loss during surgery (11% [71 of 637]), and preoperative white blood cell count (1.7% [11 of 637]). Two-tailed p values of < 0.05 were considered significant. We used Stata 13 (StataCorp LP, College Station, TX, USA) to perform all statistical analyses.
Symptomatic VTE was diagnosed in 72 patients (11%), DVT in 40 patients (6.2%), and PE in 38 patients (6.0%) within 90 days after spine surgery for metastases (Table 3). The median age of the 72 patients was 61 years (IQR, 54–68 years), and 40 patients (56%) were men. Six patients (0.9%) had concurrent evidence for a PE and DVT, and eight (1.3%) PEs were fatal (Table 4). Most symptomatic VTEs developed after hospital discharge: the median time between surgery and symptomatic VTE was 21 days (IQR, 7–39 days), and the median postoperative hospitalization of these patients was 8 days (IQR, 6–13 days). The median postoperative day of developing a fatal PE was 24 (IQR, 11–41 days).
After controlling for potentially relevant confounding variables such as age, the modified Charlson Comorbidity Index, visceral metastases, and chemoprophylaxis, we found that longer duration of surgery (OR, 1.15 for each additional hour of surgery; 95% CI, 1.04–1.28; p = 0.009) was independently associated with the development of symptomatic VTE (Table 5). Symptomatic VTE developed in 42 of 374 patients (11%) in the group that used any chemoprophylaxis and in 30 of 263 patients (11%) who received no chemoprophylaxis, demonstrating no association after controlling for age, sex, and the Modified Charlson Comorbidity Index (OR, 0.96; 95% CI, 0.58–1.59; p = 0.863).
Patients with symptomatic VTE compared with those without had lower 1-year survival after controlling for the following potentially confounding variables: age, sex, BMI, the modified Charlson Comorbidity Index, visceral metastases, primary tumor type, previous systemic therapy, and operation type (VTE: 38%; 95% CI, 27–49 versus without VTE: 47%; 95% CI, 42–51; p = 0.044; Fig. 1). The probability of developing a symptomatic VTE rose gradually over the 90-day postoperative period with a notable increase at 30 days after surgery (Fig. 2). Timing of fatal PEs ranged from postoperative day 1 to 78.
After controlling for age, sex, and the modified Charlson Comorbidity Index, we found no association between any of the different chemoprophylaxis regimens and the occurrence of 66 wound complications, consisting of 31 (47%) for LMWH (reference value), 17 (26%) for subcutaneous heparin (OR, 0.94; 95% CI, 0.21–4.23; p = 0.936), nine (14%) for aspirin (OR, 0.97; 95% CI, 0.44–2.1 p = 0.937), seven (11%) for no form of chemoprophylaxis (OR, 0.76; 95% CI, 0.32–1.80; p = 0.535), and two (3%) for warfarin (OR, 1.38; 95% CI, 0.73–2.60; p = 0.316). Likewise, an additional subanalysis between the usage of a chemical anticoagulant versus no chemical anticoagulant showed no difference, but this was underpowered.
Malignant disease and surgery are two major risk factors for symptomatic VTE [4, 5, 26], and the development of symptomatic VTE in patients with cancer is associated with poor survival . Spine surgeons must weigh the risk of chemoprophylaxis, which includes hemorrhagic complications, with the benefits of preventing symptomatic VTE in patients undergoing surgery for spine metastases. Our goal in this study was to investigate the risk of symptomatic VTE, the association between postoperative symptomatic VTE development and 1-year survival and assess the relationship between chemoprophylaxis and proportion of wound complications. A total of 11% of patients developed symptomatic VTE (including 6% who developed symptomatic PE); 1.3% of the patients in this series died of PE. After controlling for potential confounding variables, we found that longer duration of surgery was independently associated with an increased risk of symptomatic VTE and that patients with symptomatic VTE had worse 1-year survival. We did not find an association between the usage of chemical anticoagulants and the development of postoperative wound complications or symptomatic VTE. However, our study was underpowered to show this difference.
This study had several limitations. The most important limitation in this series was the inconsistent use of chemoprophylaxis regimens. In general, we used either 40 mg of enoxaparin or 5000 IUs subcutaneous heparin every 12 hours. Other general thromboembolic prophylactic dosages used were: 325 mg of aspirin, 5000 IUs dalteparin daily, and warfarin dependent to maintain an international normalized ratio of 2.0:2.5. Additional analysis demonstrated that three surgeons prescribed more heparin relative to LMWH than others, despite the lack of identifiable differences in baseline characteristics (Appendix, Supplemental Digital Content 1, https://links.lww.com/CORR/A174); this was most likely based on personal preference of these specific surgeons. While an obvious limitation, LMWH likely was chosen over heparin due to more predictable pharmacokinetics and fewer nonhemorrhagic side-effects . A meta-analysis in medically ill patients showed no difference in major bleeding events between the two, which leads us to believe that the different prescribing pattern of these surgeons likely had only a negligible impact on our study’s results .
Second, screening for and detection of VTE may have been inconsistent over the years, where the threshold for screening may have been lower in more recent years and detection techniques more effective. However, year of surgery was not associated with VTE occurrence and no differences were found in baseline characteristics of the patient population over the years. Also, the proportion of patients with VTE may have been underestimated given the lack of a universal screening protocol and the fact that we were only able to include those with symptomatic events. We anticipate a relatively low number of missed events because these patients are part of a complicated group that remains under enhanced postoperative surveillance. Third, the survival analysis between the symptomatic VTE and nonVTE groups likely consists of uncontrolled for differences. However, we controlled for the most important confounding survival variables, such as age, primary tumor type, and the modified Charlson Comorbidity Index. Fourth, the exact duration and compliance of anticoagulation could not always be confirmed. However, both centers have implemented protocols that call for patients to continue anticoagulation regimens 4 weeks after surgery and employ sequential compression devices and compression stockings during hospitalization. Fifth, history of VTE was excluded in the analysis because of the unreliability of this specific personal history data in a tertiary center. Sixth, lymphoma and multiple myeloma metastasized to the spine were included, which are known for their better prognosis and this could have potentially led to selection bias . Nonetheless, these patients represent a sizeable portion of patients, 90 of 637 (14%), who develop spine metastasis and therefore warrant consideration in a study such as this. Seventh, this cohort represents a heterogeneous population with numerous comorbidities and potential confounders. We attempted to control for these factors by using multivariable regression testing and the modified Charlson Comorbidity Index, ECOG performance status, and ASIA impairment scale as objectification of case complexity, yet we recognize the prospect of residual confounding. Lastly, this remains a retrospective work with all the inherent limitations associated with such a study design, including reliance on chart abstraction and search algorithms to identify eligible patients.
Compared with previous work, this study has a relatively large sample of patients with spinal metastases collected over the last 15 years and extensive followup considering the cohort’s clinical characteristics [3, 44]. In this series of 637 patients, 11% developed symptomatic VTE (72 of 637) with 6.0% (38 of 637) developing PE, and 1.3% (eight of 637) who died of this complication. Another, similar study reported a symptomatic VTE in 1.6% of patients undergoing surgery for symptomatic spinal metastases, but this study was designed to determine survival and not specifically evaluate the risk of symptomatic VTE . Compared with similar VTE studies for spine and musculoskeletal tumor surgery, the findings are relatively high; particularly, the number of fatal PEs is unprecedented [5, 9, 17, 19, 27, 28, 30, 32, 36, 39, 42]. Multiple factors might explain the high risk of symptomatic VTE observed here, including older age, neoplasm, severe venous stasis, prolonged immobilization and paralysis postoperatively, as well as longer operation times [1, 15, 17].
Longer duration of spine surgery was independently associated with an increased risk of postoperative symptomatic VTE in our series, a surgical factor postulated in previous surgical research . Clinically, procedures with prolonged surgery, where each hour corresponds with an increased odds ratio of 1.15 for symptomatic VTE development, may warrant greater consideration for symptomatic VTE prevention such as chemoprophylaxis. Some tumor histologies, especially lymphoma and multiple myeloma, are proven to be associated with hypercoagulability and increased symptomatic VTE risk , but none of them were identified as factors associated with VTE. However, we were not sufficiently powered to address this association.
The development of postoperative VTE has an association with a decreased 1-year survival rate. This poor survival in patients with symptomatic VTE can be explained by the highly complex patient population with multiple comorbidities and other disease-related factors, in which patients with more advanced cancer develop VTE more easily. VTE may function more as an infirmity marker than as the main cause for poor survival. Another explanation may be the high incidence of fatal PEs (1.3%), suggesting that VTE prevention, such as adequate chemoprophylaxis, could improve short-term survival. However, spine surgeons may be hesitant to use chemical anticoagulants after spine surgery out of concern for severe hemorrhagic complications, such as spinal epidural hematoma [18, 37]. A recent study reported that aspirin is safe regarding wound complications and effective in preventing symptomatic VTE after total joint arthroplasty, which makes it a viable chemoprophylaxis agent in the spine metastases population . Therefore, given the incidence of fatal PEs (1.3%) and symptomatic spinal epidural hematomas (1.1%; including only one (0.2%) patient who developed a deteriorated neurological outcome), further study is desirable to assess more adequate anticoagulation in this population.
With respect to the timing of symptomatic VTE events, risk increased after 30 days postoperatively and kept rising (Fig. 2). In addition, half the symptomatic VTEs occurred after 3 weeks (21 days; IQR, 7–39 days), which is considerably longer compared with postoperative hospitalization for VTE patients (8 days; IQR 6–13 days). We also observed that symptomatic VTE timing exceeded hospitalization in fatal PEs (respectively, 24; IQR, 11–41 versus 14; IQR, 9–20). Similar studies have reported comparable results about this late onset of symptomatic VTE [9, 19, 28, 36, 39, 42]. Both institutions followed protocols that recommend postoperative anticoagulant regimens of about 4 weeks, but considering the late onset of symptomatic VTE, a longer duration of anticoagulant use may be indicated [19, 39, 41]. The national orthopaedic guidelines are unclear about addressing this problem of chemoprophylaxis duration stating that the “patients and physicians discuss the duration of prophylaxis” . Spine surgeons have demonstrated practice variability in high-risk spine surgery patients regarding not only the duration of chemoprophylaxis, but also the use of chemoprophylactic agents . In addition, it is also possible that there is a gap between actual received outpatient chemoprophylaxis and guideline recommendations. Although this study was not designed to specifically address this compliance variable, previous studies report poor compliance in outpatient anticoagulant prophylaxis after major orthopaedic surgery and prophylaxis prescription at discharge [6, 13, 43]. A clear trend is developing toward shorter hospitalizations after major orthopaedic surgery , necessitating more emphasis on compliance of outpatient anticoagulant prophylaxis. Novel oral anticoagulants may fulfil a prominent role, since most patients prefer oral agents . Further study, preferably a randomized control trial with consistent postoperative VTE screening, could help surgeons better balance the risks and benefits as they choose from among the available postoperative prophylactic regimens.
In conclusion, this study demonstrates a high risk of symptomatic 90-day VTE among patients undergoing spine surgery for metastases; 11% of the patients developed symptomatic VTE, 6% developed a symptomatic pulmonary embolism, and 1.3% patients died of that complication. While those with symptomatic VTE were less likely to survive 1-year than those who did not, we recognize that this may reflect overall infirmity as much as anything else, since many of these patients did not die from complications related to VTE. Further studies such as randomized controlled trials with consistent postoperative VTE screening comparing different chemoprophylaxis regimens are required to identify better symptomatic VTE prevention.
1. Al-Dujaili TM, Majer CN, Madhoun TE, Kassis SZ, Saleh AA. Deep venous thrombosis in spine surgery patients: Incidence and hematoma formation. Int Surg.
2. Ambrus JL, Ambrus CM, Mink IB, Pickren JW. Causes of death in cancer patients. J Med.
3. Amelot A, Balabaud L, Choi D, Fox Z, Crockard HA, Albert T, Arts CM, Buchowski JM, Bunger C, Chung CK, Coppes MH, Depreitere B, Fehlings MG, Harrop J, Kawahara N, Kim ES, Lee CS, Leung Y, Liu ZJ, Martin-Benlloch JA, Massicotte EM, Meyer B, Oner FC, Peul W, Quraishi N, Tokuhashi Y, Tomita K, Ulbricht C, Verlaan JJ, Wang M, Mazel C. Surgery for metastatic spine tumors in the elderly. Advanced age is not a contraindication to surgery! Spine J.
4. Baron JA, Gridley G, Weiderpass E, Nyrén O, Linet M. Venous thromboembolism and cancer. Lancet (London, England). 1998;351:1077–80.
5. Benevenia J, Bibbo C, Patel D V, Grossman MG, Bahramipour PF, Pappas PJ. Inferior vena cava filters prevent pulmonary emboli in patients with metastatic pathologic fractures of the lower extremity. Clin Orthop Relat Res.
6. Bergqvist D, Arcelus JI, Felicissimo P. Evaluation of the duration of thromboembolic prophylaxis after high-risk orthopaedic surgery: The ETHOS observational study. Thromb Haemost.
7. Blom JW. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA. 2005;293:715.
8. Clagett GP, Anderson FA Jr., Geerts W, Heit JA, Knudson M, Lieberman JR, Merli GJ, Wheeler HB. Prevention of venous thromboembolism. Chest. 1998;114:531S–560S.
9. Damron TA, Wardak Z, Glodny B, Grant W. Risk of venous thromboembolism in bone and soft-tissue sarcoma patients undergoing surgical intervention: A report from prior to the initiation of SCIP measures. J Surg Oncol.
10. DeStefano V, Za T, Rossi E. Venous thromboembolism in multiple myeloma. Semin Thromb Hemost.
11. Donati MB. Cancer and thrombosis. Haemostasis. 1994;24:128–31.
12. Friedman RJ. Novel oral anticoagulants for VTE prevention in orthopedic surgery: overview of phase 3 Trials. Orthopedics. 2011;34:795–804.
13. Gao Y, Long A, Xie Z, Meng Y, Tan J, Lv H, Zhang L, Zhang L, Tang P. The compliance of thromboprophylaxis affects the risk of venous thromboembolism in patients undergoing hip fracture surgery. Springerplus. 2016;5:1362.
14. Garcia DA, Baglin TP, Weitz JI, Samama MM. Parenteral anticoagulants. Chest. 2012;141:e24S–e43S.
15. Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, Colwell CW. Prevention of venous thromboembolism: American College of Chest Physicians evidence-based clinical practice guidelines (8th edition). Chest. 2008;133:381S–453S.
16. Glotzbecker MP, Bono CM, Harris MB, Brick G, Heary RF, Wood KB. Surgeon practices regarding postoperative thromboembolic prophylaxis after high-risk spinal surgery. Spine (Phila. Pa. 1976).
17. Glotzbecker MP, Bono CM, Wood KB, Harris MB. Thromboembolic disease in spinal surgery: a systematic review. Spine (Phila. Pa. 1976).
18. Glotzbecker MP, Bono CM, Wood KB, Harris MB. Postoperative spinal epidural hematoma: a systematic review. Spine (Phila Pa 1976). 2010;35:E413-20.
19. Groot OQ, Ogink PT, Janssen SJ, Paulino Pereira NR, Lozano-Calderon S, Raskin K, Hornicek F, Schwab JH. High risk of venous thromboembolism after surgery for long bone metastases. Clin Orthop Relat Res.
20. Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ. Risk factors for deep vein thrombosis and pulmonary embolism: a population-based case-control study. Arch Intern Med.
21. Kanaan AO, Silva MA, Donovan JL, Roy T, Al-Homsi AS. Meta-analysis of venous thromboembolism prophylaxis in medically Ill patients. Clin Ther.
22. Katagiri H, Okada R, Takagi T, Takahashi M, Murata H, Harada H, Nishimura T, Asakura H, Ogawa H. New prognostic factors and scoring system for patients with skeletal metastasis. Cancer Med.
23. Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Thromboembolism is a leading cause of death in cancer patients receiving outpatient chemotherapy. J Thromb. Haemost. 2007;5:632–634.
24. Kim JYS, Khavanin N, Rambachan A, McCarthy RJ, Mlodinow AS, De Oliveria GS, Stock MC, Gust MJ, Mahvi DM. Surgical duration and risk of venous thromboembolism. JAMA Surg.
25. Kirshblum SC, Burns SP, Biering-Sorensen F, Donovan W, Graves DE, Jha A, Johansen M, Jones L, Krassioukov A, Mulcahey MJ, Schmidt-Read M, Waring W. International standards for neurological classification of spinal cord injury (Revised 2011). J Spinal Cord Med.
26. Lee AYY. Cancer and venous thromboembolism: Prevention, treatment and survival. J Thromb Thrombolysis. 2008;25:33–36.
27. Lin PP, Graham D, Hann LE, Boland PJ, Healey JH. Deep venous thrombosis after orthopedic surgery in adult cancer patients. J Surg Oncol.
28. Mitchell SY. Venous thromboembolism in patients with primary bone or soft-tissue sarcomas. J Bone Joint Surg.
29. Mont MA, Jacobs JJ. Preventing venous thromboembolic disease in patients undergoing elective hip and knee arthroplasty guideline. J Am Acad Orthop Surg.
30. Morii T, Mochizuki K, Tajima T, Aoyagi T, Satomi K. Venous thromboembolism in the management of patients with musculoskeletal tumor. J Orthop Sci.
31. Nathan SS, Healey JH, Mellano D, Hoang B, Lewis I, Morris CD, Athanasian EA, Boland PJ. Survival in patients operated on for pathologic fracture: Implications for end-of-life orthopedic care. J Clin Oncol.
32. Nathan SS, Simmons KA, Lin PP, Hann LE, Morris CD, Athanasian EA, Boland PJ, Healey JH. Proximal deep vein thrombosis after hip replacement for oncologic indications. J Bone Joint Surg Am.
33. OECD. OECD Health Data 2009 – comparing health statistics across OECD countries - OECD. Available at: http://www.oecd.org/health/oecdhealthdata2009comparinghealthstatisticsacrossoecdcountries.htm
. Accessed January 26, 2018.
34. Oken M, Creech R, Tormey D, Horton J, Davis T, McFadden E, Carbone P. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol.
35. Parvizi J, Huang R, Restrepo C, Chen AF, Austin MS, Hozack WJ, Lonner JH. Low-dose aspirin is effective chemoprophylaxis against clinically important venous thromboembolism following total joint arthroplasty: A preliminary analysis. J Bone Joint Surg Am.
36. Patel AR, Crist MK, Nemitz J, Mayerson JL. Aspirin and compression devices versus low-molecular-weight heparin and PCD for VTE prophylaxis in orthopedic oncology patients. J Surg Oncol.
37. Paulino Pereira NR, Ogink PT, Groot OQ, Ferrone ML, Hornicek FJ, van Dijk CN, Bramer JAM, Schwab JH. Complications and reoperations after surgery for 647 patients with spine metastatic disease. Spine J.
38. Quan H, Li B, Couris CM, Fushimi K, Graham P, Hider P, Januel JM, Sundararajan V. Updating and validating the charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol.
39. Shallop B, Starks A, Greenbaum S, Geller DS, Lee A, Ready J, Merli G, Maltenfort M, Abraham JA. Thromboembolism after intramedullary nailing for metastatic bone lesions. J Bone Joint Surg Am.
40. Sørensen HT, Mellemkjær L, Olsen JH, Baron JA. Prognosis of cancers associated with venous thromboembolism. N Engl J Med.
41. Sweetland S, Green J, Liu B, Berrington de Gonzalez A, Canonico M, Reeves G, Beral V. Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study. BMJ. 2009;339:b4583–b4583.
42. Tuy B, Bhate C, Beebe K, Patterson F, Benevenia J. IVC filters may prevent fatal pulmonary embolism in musculoskeletal tumor surgery. Clin Orthop Relat Res.
43. Wilke T, Müller S. Nonadherence in outpatient thromboprophylaxis after major orthopedic surgery: a systematic review. Expert Rev Pharmacoecon Outcomes Res.
44. Yoshioka K, Murakami H, Demura S, Kato S, Hayashi H, Inoue K, Ota T, Shinmura K, Yokogawa N, Fang X, Tsuchiya H. Comparative study of the prevalence of venous thromboembolism after elective spinal surgery. Orthopedics. 2013;36:e223–e228.