Prevalence of Risk Factors for Hospital-Acquired Venous Thromboembolism in Neurosurgery and Orthopedic Spine Surgery Patients

INTRODUCTION

Venous thromboembolism (VTE) is the number 1 preventable hospital-acquired cause of morbidity and mortality in the United States.¹ In 2003, over 12 million patients, comprising 31% of U.S. hospital discharges, were at risk of VTE.² It is estimated that there are nearly 300,000 VTE-related deaths annually in the United States, and approximately two-thirds of symptomatic VTE events are hospital acquired.³ A 2015 study found that 72% of VTE patients had emboli that were potentially preventable with prophylaxis.⁴ At the same time, poor adherence to thromboprophylaxis, whether pharmacologic or mechanic, correlates with increased VTE events. Previous studies have reported adherence rates between 38% and 60% with American College of Chest Physicians guidelines.^5–8 Even after a successfully treated episode, subsequent complications affecting the pulmonary, renal, cardiac, and nervous systems may persist for years.⁹ Up to half of lower-extremity deep vein thrombosis (DVT) patients develop postthrombotic syndrome and chronic venous insufficiency, struggling with pain, swelling, skin necrosis, and ulceration.^10,11 Even after a standard course of anticoagulant therapy, one-third of VTE patients experience a recurrence within 10 years of the initial event.¹⁰ While the highest risk occurs within the first year, patients with a previous VTE remain at increased risk for life. Thus, many VTE patients require long-term anticoagulation to prevent additional clots, and this treatment itself decreases quality of life and increases risk for bleeding episodes.¹⁰

Another significant challenge presented by VTEs is a resulting increase in health care spending. It is estimated that the total annual health care cost attributable to VTE ranges from $7594 to $16,644 per patient, up to $10 billion annually. These costs are often driven significantly by increased lengths of stay; reoperations and unplanned patient readmissions are generally even more costly.^9,11,12 There are known risk factors for hospital-acquired VTEs (HA-VTE), but the prevalence of these risk factors is not always reviewed in patients who have sustained DVT or pulmonary embolism (PE). Such factors include genetic (family history, thrombophilic diseases), acquired (age, cancer, obesity, chronic diseases), and transiently acquired (pregnancy, oral contraceptives, trauma, immobilization) variables.¹⁰

Several studies exist evaluating the prevalence of DVT and PE following orthopedic surgery.^13–15 Rates of venographic DVT and proximal DVT 1–2 weeks following major orthopedic surgery in patients receiving no prophylaxis are between 40% and 60% and 10% and 30%, respectively.¹² However, most existing research pertaining to VTEs and orthopedic procedures has explored total hip and knee arthroplasties and hip fractures, and limited data are available on spine surgeries. While there is strong evidence for recommendations on prophylaxis in spine procedures, little information addresses more nuanced details regarding risk factors.¹² Furthermore, analyses associating categories of diagnoses and procedures in spine surgery with VTE development are lacking. Therefore, this study aimed to review HA-VTE rates at our institution and evaluate the prevalence of known risk factors in patients who developed HA-VTE among both neurosurgical and orthopedic spine surgery patients.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Costs for the management of VTE, the most common preventable, hospital-acquired cause of morbidity and mortality in the United States, are substantial and have increased over time.^1,11 VTE is estimated to be the second most common cause of excess length of hospital stay and the third most common cause of excess mortality, and readmissions may be up to 48% more costly than the initial event.^11,12 Given the economic burden of VTEs and their negative impact on patient quality of life, further exploration on the prevalence of patient risk factors for HA-VTE could improve preventive efforts. Twenty-four percent of VTEs are attributable to hospitalizations following surgery, of which orthopedic procedures make up a significant portion.¹⁶ Spine surgery in particular lacks analyses associating categories of diagnoses and procedures with VTE development. This retrospective analysis aimed to review HA-VTE rates at our institution and evaluate the prevalence of known risk factors in patients who developed HA-VTE among both neurosurgical and orthopedic spine surgery patients. A total of 7156 patients undergoing spine surgery at our institution were included in this study.

We report an overall HA-VTE rate of 0.94%. In the current literature, spine fracture patients have been found to experience VTE at 4 times the rate of other spine surgery patients despite being more likely to receive chemoprophylaxis.¹⁷ Although our sample size of VTE patients was small, diagnosis of fracture was found to be an independent risk factor for VTE development. While several studies have found a higher VTE risk in anterior approaches, our study did not find anterior approach to be an independent risk for HA-VTE.^18,19 It is not surprising that patients with malignancy would have a higher HA-VTE risk and that a DRG of noncervical malignancy was a significant independent risk factor for VTE.^19,20 Cervical procedure levels were associated with significantly lower risk of HA-VTE, likely because of the lower metabolic burden of such operations, allowing patients to mobilize and be discharged more quickly than in thoracolumbar cases. Operation by the neurosurgery department was also an independent HA-VTE risk factor, though this is likely attributable to the significantly higher rates of fracture patients and patients with malignancy seen in the neurosurgery patient population as demonstrated by Table 3.

Comparing such factors among all patients, between the orthopedic surgery and neurosurgery divisions and also between Hospital A and Hospital C allowed us to further characterize our findings. There were significant differences in most procedure types, diagnoses, and DRGs. Hospital A and the orthopedic surgery division generally followed similar trends, while Hospital C and the neurosurgery division followed similar trends. This is likely because the majority of the spine cases performed by the orthopedic surgery division occurred at Hospital A, while those performed by neurosurgery occurred at Hospital C. Overall, comparison between facilities and also between departments generally showed significant differences in the procedures and diagnoses represented by their respective patient populations. These results are consistent with the overall trend that VTEs developed in a higher percentage of neurosurgery patients and in a higher percentage of patients at Hospital C.

Among patients who developed VTEs, orthopedic patients had a higher EBL and percentage of patients undergoing anterior-posterior surgery. Higher intraoperative blood loss is associated with higher rates of VTE in previous studies.¹⁷ However, EBL was not found to be an independent risk factor in our analysis, and the higher EBL noted in the orthopedic department more likely reflects a higher volume and percentage of deformity procedures performed. We found a higher rate of anterior-posterior surgery in orthopedic surgery patients. Neurosurgery patients had higher rates of active cancer and more patients over the age of 60. While having a lower average patient age, it is possible that the neurosurgery patients who were over 60 years had more complicated conditions, such as malignancy and trauma, that made them more predisposed to developing VTE compared to the orthopedics patients. There was also no significant difference in first day to mobilization between orthopedic surgery and neurosurgery patients, so patients of both departments were generally mobilized at the same time postoperatively. In addition, there was no significant difference in invasiveness index scores between departments. While the overall average index score for procedures conducted by neurosurgery would likely be higher than that of orthopedics, many of the orthopedic VTE patients underwent extensive adult deformity surgeries involving arthrodesis and instrumentation of numerous levels as opposed to more common procedures that were less invasive, such as laminectomy and discectomy.

We found no significant difference between orthopedic and neurosurgery patients with regard to risk factors for VTE such as BMI >40, prior VTE, congestive heart failure with ejection fraction <40%, thrombophilic disorder, pulmonary hypertension, restrictive lung disease, smoking history, durotomy, or preoperative neurologic deficit. More surgical levels are associated with increased VTE risk, and cardiac risk factors predict increased readmission rates.¹⁷ The neurosurgery and orthopedic surgery patients who developed VTE had an average number of risk factors of 1.74 and 1.85, respectively, which reflects the overall low rate of operation on patients with a high risk for VTE.

There were a few notable practice differences between the neurosurgery and orthopedic surgery departments. Orthopedic patients had lower rates of inferior vena cava filter placement, while postoperative anticoagulation was initiated sooner in neurosurgery patients. These differences most likely highlight differing philosophies between departments about approaching postprocedure anticoagulation as well as timing. While earlier use of VTE prophylaxis has been proposed to better prevent thromboembolic events without a significantly increased risk of bleeding complications, there is no clear consensus on the most effective timing of initiation.²¹ The neurosurgical patients were started on anticoagulation on postoperative day 1, which demonstrates a routine use for starting shortly after surgery. Orthopedic surgery patients were started on postoperative day 3, which suggests a “wait and see” approach to drain removal and patient mobility. Yet, while neurosurgical patients were started on anticoagulation earlier, there was a higher rate of VTE in their group overall. It may be prudent to further explore prophylaxis regimens in future studies, as these results suggest that expanding on standard prophylaxis for patients at increased risk of VTE, such as trauma and cancer patients, may potentially be beneficial.

Our study had several limitations, mainly its retrospective design with a risk of selection bias and the relatively small sample size for patients who developed HA-VTEs, which may lead to an underestimation of significance for some of the parameters studied. It is also important to recognize that the true incidence of VTE may be higher, with some readmissions for VTE going to outside institutions and also the possibility of asymptomatic VTE. Existing literature suggests that VTEs may be underrecognized following spine surgery and may be as high as 15.5% in patients without prophylaxis.²² This study demonstrates that, at our institution, the true VTE rate is unlikely to be that high. As this analysis focused on a single institution, it also reflects the clinical and operative decision making of our own surgeons with regard to patient selection, surgical procedures, and management. The difference between neurosurgery and orthopedic surgery may additionally be due to practice patterns related to diagnosing VTE. The difference noted may be due simply to a higher likelihood to order ultrasounds on neurosurgery postoperative patients. Our study adds to existing literature pertaining to postoperative VTE in spine surgery by exploring the prevalence of risk factors in greater detail and examining differences between spine disciplines as well as categories of diagnoses and procedures. Future studies may expand on this work, and a multicenter study may further elucidate the reasons why certain procedures and diagnoses are more associated with HA-VTEs as well as differences between spine disciplines.

CONCLUSIONS

The overall rate of HA-VTE is 0.94% for all spine surgery patients at our institution. The HA-VTE rate was 0.61% in orthopedic patients and 1.87% in neurosurgery patients. In the patients who sustained VTE, the neurosurgical patients had higher rates of active cancer and age >60, and orthopedic patients had higher EBL and rates of anterior-posterior surgery. This highlights the different patient populations between the 2 departments and the need for individualized thromboprophylaxis regimens.