Incidence and Risk Factors Associated With Venous Thromboembolism After Orthopaedic Below-Knee Surgery

Reinout R.O. Heijboer, MD; Bart Lubberts, MD; Daniel Guss, MD, MBA; A. Holly Johnson, MD; Christopher W. DiGiovanni, MD


J Am Acad Orthop Surg. 2019;27(10):e482-e490. 

In This Article


In contrast to the hip and knee arthroplasty literature, data to guide surgeons about the prevention of thromboembolic events in orthopaedic patients undergoing below-knee surgery are scarce.[14,15] Providers struggle to navigate a landscape in which VTE is increasingly seen as a "never event," and yet little is known about whether such patients require prophylaxis, which prophylactic agent to use, in what dosage it should be given, the ideal time and duration to administer prophylaxis, and whether the benefits of use outweigh the risks. An improved understanding of the incidence and risk factors of symptomatic VTE among different types of pharmacologic thromboprophylactic strategies may help orthopaedic surgeons guide their patients through a shared decision-making process postoperatively. In our cohort, the overall incidence rate of a symptomatic thromboembolic event was 2.5%. Factors associated with developing VTE regardless of thromboprophylaxis use were male sex, age greater than 65 years, lower leg surgery, combination surgery, history of VTE, family history of cardiovascular disease, and a Charlson Comorbidity Index greater than 2. The risk for developing VTE in these populations was not found to be lowered by thromboprophylaxis.

The strengths of the current study are the large sample size and analytic approaches used to investigate the incidence of VTE and determine associated risk factors adjusted for the variability in type of pharmacologic thromboprophylaxis. However, several limitations exist. First, billing and diagnostic codes were used to define the patient cohort and identify comorbidities. Coding inaccuracy could alter this analysis. However, the development of DVT or PE as well as the thromboprophylactic agent used in each case was attained by the direct chart review—and we think that this review increased the accuracy of our data. Second, the dose and length of thromboprophylaxis is likely to have varied within the antiplatelet and anticoagulant subgroups. Third, we were not able to capture data on type of surgery (eg, reconstructive, traumatic) and postoperative immobilization, which have been described as potential risk factors.[5,16,17] An increased number of trauma patients in the current database could potentially have caused the higher rate of VTE compared with those in other previously analyzed databases. Fourth, we were not able to capture data on the location of the DVT. These data could have been valuable to see whether a relationship with PE or mortality rate exists. Lastly, this study relied on symptomatic VTE rather than screening methods, which may arguably be more clinically relevant but certainly may underreport incidence of VTE events. In addition, patients may have sought care for a VTE event outside the three hospitals, further underestimating the incidence of VTE.

Previous large retrospective studies with similar patient demographics have reported a lower incidence of symptomatic VTE compared with our findings.[18–21] Shibuya et al[19] studied 75,664 foot and ankle trauma patients and found that the incidence of DVT and PE was 0.28% and 0.21%, respectively. Jameson et al[21] reported on VTE after foot and ankle surgery from the English National Health Service database including 88,241 patients. The incidence of DVT and PE was 0.12% and 0.17% after ankle fracture, 0.01% and 0.02% after first metatarsal osteotomy, and 0.03% and 0.11% after hindfoot fusion. The incidence of PE after 1,633 total ankle replacements was 0.06% with no DVTs being reported. Correspondingly, SooHoo et al[20] found a DVT incidence of 0.05% and a PE incidence of 0.34% in 57,173 patients who were surgically treated for an ankle fracture. A recent meta-analysis performed by Calder et al[18] which included assessment of 43,381 foot and/or ankle surgical patients reported a symptomatic VTE incidence rate of 0.6% in foot and ankle patients without prophylaxis and 1.0% in patients with prophylaxis. A possible explanation may be that these studies only included patients with isolated foot and ankle procedures, excluding patients with lower leg surgery. In our study, patients with lower leg surgery presented with higher VTE rates (4.3%) compared with patients who had isolated foot/midfoot or hindfoot/ankle surgery (1.3% and 1.2%, respectively). It is also proposed that the increased incidence is a reflection of healthcare providers having a greater suspicion for the complication.[22] In recent years, healthcare organizations and institutes such as the National Institute for Health and Care Excellence quality standards have emphasized that providers should make patients aware of VTE by providing verbal and written information on admission and discharge. As a result, patients are more frequently educated to recognize symptoms such as swelling or soreness of the leg, skin redness, or a cough with or without blood.[23,24] Thus, newer studies of VTE incidence may identify a higher incidence of VTE.

Notably, we found that 5.9% of patients diagnosed with VTE died within 90 days and that the associated relative risk of mortality after developing a VTE event was 3.9 (P < 0.001; 95% CI, 3.5 to 6.7). Previous studies found a 30- to 90-day mortality rate between 1.1% and 6.3% following VTE after orthopaedic foot and ankle surgery and between 7.0% and 13% after orthopaedic below-knee surgery.[25–28] However, these studies did not mention the relative risk of mortality. In patients undergoing fracture or replacement surgery of the hip or knee, the mortality rate has been found to be 6.7% to 25%.[27,29–33] The variable mortality rate in patients in whom VTE develops between different lower extremity orthopaedic procedures may be because of the location of the thromboembolism, although additional study is needed to better define this finding. Proximal lower extremity DVT has a higher chance of causing a PE and related increased changes of mortality.[34,35] Numerous retrospective studies have failed to demonstrate reduction in the risk of VTE with the use of thromboprophylaxis compared with no prophylaxis among the foot and ankle surgical population.[18,19,21,36–39] Interestingly, we found the highest VTE rate in patients who received anticoagulant agents, and suspect this finding, like with previous studies, to most likely be the result of selection bias wherein anticoagulants were preferentially used in those patients deemed most at risk. In fact, a large retrospective study by Gaskill et al[36] concerning the incidence and associated risk factors of VTE after arthroscopically assisted anterior cruciate ligament reconstruction identified a similar finding. Patients who were postoperatively treated with anticoagulants showed a notable increase in the overall VTE (OR, 98.32; 95% CI, 61.63 to 156.86; P < 0.001). Contrarily, when a randomized trial by Lassen et al[40] was performed, they found a reduction in the risk of DVT with the use of reviparin compared with placebo in patients who were immobilized after a leg fracture or rupture of the Achilles tendon. Thus, given the predominantly level IV methodology identified in our current body of literature, drawing any conclusion about the use of pharmacologic thromboprophylaxis to prevent VTE after foot and ankle surgery is difficult.[38,39] Furthermore, as our results underscore, patients with comorbid potential risk factors for VTE are more likely to receive prophylaxis compared with patients with fewer comorbidities. Large-scale, properly powered, prospective, randomized controlled or pragmatic studies are necessary to help patients and providers alike better understand who needs prophylaxis, which prevention strategies are most effective, and the risk-benefit profile inherent to each strategy.