- Email: [email protected]
Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study
G Model JINJ 7477 No. of Pages 5
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Full length article
Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study Sean T. Campbella,* , Abiram Balaa , Sam Y. Jiangb , Michael J. Gardnera , Julius A. Bishopa a b
Department of Orthopaedic Surgery, Stanford University, Stanford, CA, United States School of Medicine, Stanford University, Stanford, CA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 29 October 2017
Introduction: The purpose of this study was to evaluate the effectiveness of Factor Xa inhibitors (XaI) for thromboprophylaxis following hip fracture surgery in a large cohort of patients, and compare XaI against warfarin and enoxaparin. Methods: Patients undergoing hip fracture surgery from 2007 to 2015 were identified in a large claims database. Patients prescribed warfarin, XaI, or enoxaparin within 2 weeks of surgery were identified and grouped into cohorts. Medical comorbidities and complication incidences, including deep venous thrombosis (DVT), pulmonary embolism (PE), and bleeding complications were calculated. Chi-square analysis was performed and adjusted residuals calculated to determine significant differences. Results: DVT rates were significantly different between groups at thirty days only (5.03% warfarin, 2.91% XaI, 3.48% enoxaparin, p = 0.047). PE rates were significantly different at all time points; enoxaparin had the lowest rates. There were no differences in the rates of other complications. Discussion: XaI are an option for thromboprophylaxis in hip fracture patients, although their possible decreased effectiveness against PE compared to enoxaparin should be considered. Conclusions: This study compares the effectiveness of Factor Xa inhibitors to warfarin and enoxaparin for hip fracture patients, using a large national database. In this study, Factor Xa inhibitors had similar effectiveness for DVT prophylaxis compared to these agents. © 2017 Published by Elsevier Ltd.
Keywords: Hip fracture Anticoagulation Deep venous thrombosis Factor Xa inhibitors
Introduction Hip fracture is a common cause of hospital admission and nonelective surgery among geriatric patients, with an annual mean of 958 hip fractures per 100,000 individuals in the United States [1]. Venous thromboembolism (VTE) is a well-known complication that may occur following any orthopaedic injury or procedure, including surgical treatment of hip fractures. Pharmacologic VTE prophylaxis is often used in this setting to mitigate risk, and is recommended by the American Academy of Orthopaedic Surgeons (AAOS) Clinical Practice Guidelines [2]. A variety of anticoagulants have been shown to be effective at reducing the rate of VTE, including warfarin (which interferes with clotting factors II, VII, IX, and X via vitamin K antagonism), enoxaparin (which acts directly on the protein antithrombin), and aspirin (an anti-platelet), among others [3–7].
* Corresponding author at: Orthopaedic Surgery Stanford Hospital and Clinics, 300 Pasteur Drive, Room R144, Stanford, CA 94305-5341, United States. E-mail address: [email protected] (S.T. Campbell).
Despite the evidence in favor of VTE prophylaxis for hip fracture patients, there is limited data to guide surgeons when selecting a pharmacologic agent. In addition, newer anticoagulants such as Factor Xa inhibitors (XaI) are gaining popularity. These have some advantages compared with more traditional prophylaxis agents, including rapid onset, lack of required laboratory monitoring, and in some cases oral administration. They have also been shown to be effective for VTE prophylaxis in total joint arthroplasty [8–12]. However, data on their use in hip fracture patients is limited [6,13,14], and to our knowledge no large databases have been utilized to assess the efficacy of various chemoprophylactic agents, including Factor Xa inhibitors, for VTE prevention among hip fracture patients. The purpose of this study was to evaluate the efficacy of Factor Xa inhibitors for VTE prophylaxis following surgical treatment of hip fracture using a large national database, and to directly compare this against enoxaparin and warfarin. Secondarily, we sought to compare the rates of complications related to each of these three anticoagulants.
https://doi.org/10.1016/j.injury.2017.10.044 0020-1383/© 2017 Published by Elsevier Ltd.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
2
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Patients and methods A retrospective cohort study was designed using a large, combined private payer-Medicare national claims database (PearlDiver, Colorado Springs, CO) containing publicly available, HIPPA compliant patient data. All patient records were queried to identify patients who underwent surgical treatment of a hip fracture between 2007 and 2015, using Current Procedural Terminology (CPT) and International Classification of Diseases, Ninth Revision (ICD-9) codes (Table 1). Patients were age sixty or older at the time of surgery. We then identified those hip fracture patients who had a VTE chemoprophylaxis agent prescribed within two weeks of surgery and divided them into three cohorts: those who received 1) warfarin, 2) Factor Xa inhibitors, or 3) enoxaparin. Any patient who had been prescribed an antiplatelet or anticoagulant within 1 year prior to surgery was excluded. Cohorts were matched by age and gender by proportions. We then compared Charlson Comorbidity Indices (CCI), standard demographic factors and comorbidities as described previously [15], and rates of deep vein thrombosis (DVT), pulmonary embolism (PE), significant bleeding events, postoperative anemia, and post-operative transfusion at 2 weeks, thirty days, 6 weeks, and ninety days between the three cohorts, which were identified using CPT and ICD-9 codes (Table 1). Chi-square tests with significance set to p < 0.05 were used to compare the comorbidities and complication rates between the cohorts at each time point. When Chi-square testing demonstrated a significant difference in the three-way comparison between anticoagulation cohorts, adjusted residuals (AR) were calculated to measure the contributions of individual cells, with values 2 or 2 considered statistically significant [16]. Results There were 4090 hip fracture patients included in the study. 2326 of these received enoxaparin, 929 received XaI, and 835 received warfarin. The mean CCI for each group was 3 (Table 2). There were some significant differences in demographics between the three cohorts (Table 2). There were no significant differences in the rates of bleeding events, post-operative anemia, or post-operative transfusion between the three groups at any time point studied. There was a significant difference in DVT rates between the three groups at thirty days (5.03% warfarin, 2.91% XaI, and 3.48% enoxaparin, p = 0.047. Higher rates of DVT in the warfarin group contributed significantly to this difference (AR = 2.35 warfarin). AR in the XaI and enoxaparin cells were non-significant, but less than zero, indicating fewer observed DVTs than statistically expected ( 1.41 XaI, 0.72 enoxaparin) (Fig. 1). There were no significant differences in DVT rates at 2 weeks, 6 weeks, or ninety days (Fig. 1). For PE, there were significant differences between the groups at all time points studied following Chi-square analysis (2 weeks:
1.80% warfarin, 1.72% XaI, 0.60% enoxaparin, p = 0.002; thirty days: 2.04% warfarin, 1.83% XaI, 0.95% enoxaparin, p = 0.026; 6 weeks: 2.04% warfarin, 1.83% XaI, 0.99% enoxaparin, p = 0.037; ninety days: 2.63% warfarin, 2.15% XaI, 1.25% enoxaparin, p = 0.017)(Fig. 2). At 2 weeks, adjusted residuals were significant for all cells: the warfarin and XaI groups had significantly more PE events than expected (2.16 warfarin, 2.07 XaI), while the enoxaparin group had significantly fewer ( 3.51 enoxaparin). At the thirty day and six week time points, only lower rates than expected in the enoxaparin groups were found significant after calculation of adjusted residuals (thirty days: 1.86 warfarin, 1.37 XaI, 2.68 enoxaparin; 6 weeks: 1.77 warfarin, 1.29 XaI, 2.54 enoxaparin). At ninety days, there were significantly more events in the warfarin group than expected and significantly fewer in the enoxaparin group (AR = 2.23 warfarin, 1.11 XaI, 2.75 enoxaparin)(Fig. 2). Discussion The most important findings of this study are that in a large claims database review of hip fracture surgery patients: 1) DVT rates were not significantly different between warfarin, Factor Xa inhibitors, and enoxaparin, except at the thirty day time point, where high rates in the warfarin cohort contributed to a statistically significant difference, 2) rates of pulmonary embolism were significantly different between the three groups, with enoxaparin having lower rates than XaI and warfarin at all time points, and 3) there were no significant differences in rates of bleeding, post-operative anemia, and blood transfusion. There are relatively few high quality therapeutic studies assessing the effect of chemoprophylaxis on VTE rates among hip fracture patients, and even fewer data available which compare the efficacy of different agents. One of the studies with the highest methodologic quality on the topic is a multicenter, multinational, randomized controlled trial by Rodgers et al. from 2000, in which 13,356 hip fracture patients were randomized to either an antiplatelet or placebo, and complications were tracked for 35 days [3]. The authors found that the rates of DVT and PE at 35 days were 1.0% (compared to 1.5% in the placebo group) and 0.7% (compared to 1.2%), respectively, and concluded that chemoprophylaxis was effective for VTE risk reduction [3]. In the current study, each of the patient cohorts received some form of chemoprophylaxis, but we found 30 day DVT rates between 2.91–5.03% and PE rates between 0.95–2.04%, which are slightly higher than the Rodgers et al. data [3]. The reasons for this are unknown, and likely multifactorial. There is little information in the Rodgers study regarding the baseline medical comorbidities and general health status of the patients [3], and it is possible our American patient population had poorer overall health status at the time of surgery, which may have increased their risk for VTE. Additionally, up to 44% of patients in the previous study received both the randomized treatment of an antiplatelet in addition to unfractionated heparin or enoxaparin [3]; it is possible this use of dual chemoprophylaxis reduced VTE rates in that study relative to our cohort.
Table 1 List of CPT and IDC-9 codes used to identify patients and complications. Diagnosis/Procedure
Codes
Surgical fixation hip fracture DVT PE Post-operative anemia Bleeding complications Transfusion
ICD-9: P-7935, P-8152 CPT: 27245, 27244, 27248, 27269 ICD-9: 451.11, 451.19, 451.2, 451.81, 451.9, 453.40, 453.41, 453.42, 453.8, 453.9 ICD-9: 415.1, 415.11, 415.13, 415.19 ICD-9: 285.1 ICD-9: 998.1, 998.11, 998.12, 998.13, 719.10, 719.16, 719.16, 39.98 CPT-36430 ICD-9: 99.0, 99.00, 99.02, 99.03, 99.04
CPT = Current Procedural Terminology, ICD-9 = International Classification of Diseases, Ninth Revision DVT = deep venous thrombosis, PE = pulmonary embolism.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
Table 2 Patient Characteristics. Enoxaparin 2326
Coumadin 835
Factor Xa 929
P value
170 (7.3) 413 (17.8) 498 (21.4) 491 (21.1) 348 (15) 190 (8.2) 216 (9.3)
61 (7.3) 149 (17.8) 179 (21.4) 176 (21.1) 125 (15) 67 (8) 78 (9.3)
68 (7.3) 165 (17.8) 199 (21.4) 196 (21.1) 139 (15) 76 (8.2) 86 (9.3)
1.000 0.998 1.000 1.000 1.000 0.990 0.998
Female Male
1634 (70.2) 692 (29.8)
586 (70.2) 249 (29.8)
653 (70.3) 276 (29.7)
0.999 0.999
Congestive Heart Failure Valvular Disease Pulmonary Circulation Disorders Peripheral Vascular Disease HTN (uncomplicated) HTN (complicated) HTN (uncomplicated and complicated) Paralysis Other Neurological Disorders Chronic Pulmonary Disease Diabetes without chronic complications Diabetes with chronic complications Hypothyroidism Renal Failure Liver disease Chronic Peptic Ulcer Disease HIV/AIDS Lymphoma Metastatic Cancer Solid Tumor without Metastasis Rheumatoid Arthritis/Collagen Vascular Diseases Coagulation Deficiency Obesity Weight Loss Fluid and Electrolyte Disorders Blood Loss Anemia Deficiency Anemias Alcohol Abuse Drug Abuse Psychoses Depression
286 (12.3) 374 (16.1) 115 (4.9) 570 (24.5) 1461 (62.8) 429 (18.4) 1471 (63.2) 45 (1.9) 540 (23.2) 696 (29.9) 585 (25.2) 288 (12.4) 522 (22.4) 364 (15.6) 94 (4) 13 (0.6) 10 (0.4) 28 (1.2) 98 (4.2) 338 (14.5) 280 (12) 122 (5.2) 200 (8.6) 342 (14.7) 556 (23.9) 85 (3.7) 705 (30.3) 97 (4.2) 103 (4.4) 361 (15.5) 477 (20.5)
95 (11.4) 125 (15) 38 (4.6) 154 (18.4) 496 (59.4) 104 (12.5) 497 (59.5) 26 (3.1) 155 (18.6) 232 (27.8) 191 (22.9) 86 (10.3) 172 (20.6) 103 (12.3) 24 (2.9) 10 (<1.2) 0 (0) 13 (1.6) 39 (4.7) 127 (15.2) 67 (8) 34 (4.1) 53 (6.3) 114 (13.7) 199 (23.8) 23 (2.8) 223 (26.7) 22 (2.6) 24 (2.9) 75 (9) 146 (17.5)
122 (13.1) 159 (17.1) 45 (4.8) 248 (26.7) 608 (65.4) 174 (18.7) 605 (65.1) 24 (2.6) 216 (23.3) 276 (29.7) 234 (25.2) 119 (12.8) 215 (23.1) 151 (16.3) 31 (3.3) 12 (1.3) 0 (0) 24 (2.6) 42 (4.5) 132 (14.2) 110 (11.8) 57 (6.1) 106 (11.4) 128 (13.8) 256 (27.6) 33 (3.6) 285 (30.7) 30 (3.2) 55 (5.9) 126 (13.6) 191 (20.6)
0.534 0.473 0.902 0.000 0.032 0.000 0.046 0.127 0.015 0.498 0.393 0.206 0.408 0.039 0.259 0.059 0.022 0.018 0.832 0.830 0.005 0.149 0.001 0.670 0.074 0.464 0.110 0.095 0.008 0.000 0.146
n Demographics Age (%) 60 65 70 75 80 85 90
to 64 to 69 to 74 to 79 to 84 to 89 and over
Gender (%)
Co-morbidities Conditions (%)
Comparison of patient demographics and baseline comorbidities across cohorts. Bold text indicates a statistically significant difference. is used to indicated that there are less than 10 patients in a particular cohort; the exact number is hidden in the PearlDiver software for patient privacy reasons. The p-values associated with these entries are indicated with and are uncertain.
A few previous studies have examined Factor Xa inhibitors for VTE prophylaxis in hip fracture patients. In a 2001 paper, Eriksson et al. studied 1250 patients with hip fractures that had been randomized to either postoperative Factor Xa inhibitor or enoxaparin [13]. Each patient received bilateral venography post-operatively to assess for DVT, which ensured that even asymptomatic VTE were detected. The authors found that at postoperative day 11, the rates of DVT were 7.9% in the XaI group and 18.8% in the enoxaparin group, which were significantly different in their statistical analysis [13]. The group also found there were significantly more “minor bleeding” complications in the XaI cohort [13]. More recently, Sasaki et al. randomized a smaller group of 84 hip fracture patients to XaI or enoxaparin following surgical treatment of hip fractures [6]. These authors found a similar trend compared to Eriksson et al [13]: lower rates of VTE in the XaI group compared to the enoxaparin group, with a trade-off of more complications related to bleeding6. In the current study, we were able to include a greater number of patients than either of the previously mentioned studies by drawing on a large national
claims database. However, our work was retrospective in nature compared to the prospective Eriksson and Sasaki studies, and this did introduce some significant differences into the baseline characteristics for our cohorts that are absent in the aforementioned randomized controlled trials. One important methodologic difference is that in this study, outcomes were based on CPT and ICD-9 codes, so VTE events were only detected if they were clinically significant. This may partially account for the differences in VTE rates between our study and Eriksson’s, in which routine venography was performed to detect even asymptomatic DVTs [13]. Additionally, at 2 weeks from surgery (which most corresponds to the time point studied by Eriksson) we found no significant difference between DVT rates for our three cohorts (although we did find a difference at thirty days), and while there was a significant difference in PE rates at all time points, enoxaparin had lower rates than either XaI or warfarin. Finally, in this study there were no significant differences in bleeding complications.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
4
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 1. Rates of deep venous thrombosis (DVT) by treatment cohort (warfarin, Factor Xa inhibitor, and enoxaparin) at 4 time points studied (2 weeks, 30 days, 6 weeks, and 90 days). *indicates significance with p < 0.05 in Chi-square analysis. # indicates the cohort had a significant contribution to the difference observed, with an adjusted residual 2 or 2.
patients who had been on aspirin pre-operatively, and this may have eliminated a large number of patients who were instructed to continue their home regimen for prophylaxis following surgery. Finally, we chose to exclude patients who were treated for hip fracture with total hip arthroplasty (THA) from the study. This was decided for two reasons: 1) multiple studies already exist regarding the use of XaI in THA, and 2) THA is perhaps a more “extensive” procedure than others commonly performed for hip fracture, and including these patients could result in a significant number of outliers in terms of outcome measures like anemia, bleeding, and transfusion. Despite these limitations, this study is one of the largest to date to assess the efficacy of Factor Xa inhibitors for VTE prophylaxis in hip fracture patients and directly compare these agents to others commonly used for chemoprophylaxis. In conclusion, in this large national database study of hip fracture patients, warfarin, Factor Xa inhibitors, and enoxaparin were similarly effective for DVT prophylaxis following hip fracture surgery at two weeks, six weeks, and ninety days. At thirty days only, warfarin was statistically inferior. There were significant differences in rates of PE between the groups, with enoxaparin statistically superior to XaI and warfarin at all time points studied. The rates of complications related to bleeding were no different between the groups. Factor Xa inhibitors are a viable option for use as thromboprophylaxis in hip fracture patients, although their increased cost and possible decreased efficacy compared to enoxaparin for PE prevention needs to be weighed against the potential for increased compliance and patient preference. References
Fig. 2. Rates of pulmonary embolism (PE) by treatment cohort (warfarin, Factor Xa inhibitor, and enoxaparin) at 4 time points studied (2 weeks, 30 days, 6 weeks, and 90 days). * indicates significance with p < 0.05 in Chi-square analysis. # indicates the cohort had a significant contribution to the difference observed, with an adjusted residual 2 or 2.
This study has multiple limitations, including those inherent to large database studies based on administrative claims: First, the accuracy of the data collected is dependent on the accuracy of the initial coding; subsequent analysis of this data assumes that the coding is accurate and complete. This also means that it is unknown how VTE events were diagnosed in this study (ultrasound versus venography); the only information available if whether a relevant code was entered. Additionally, unlike a traditional retrospective chart review, individualized patient data is unavailable in the database, and specific variables or outcomes cannot be tied to a particular patient. This means that certain analyses, such as regression or propensity score matching, cannot be performed. Beyond the inherent limitations of database studies, this particular study is limited by the lack of an antiplatelet (aspirin) cohort. A cohort of patients that had received aspirin for VTE prophylaxis would have been ideal to include, since this agent is commonly used for this purpose and was previously studied with high quality methodology [3]. Unfortunately, the number of patients in such a cohort obtained via the claims database used would have been too small to allow for a meaningful comparison. This is likely partially related to the fact that medications tracked by the database must be filled via prescription, and many patients that used aspirin for VTE prophylaxis may have been instructed to obtain their aspirin over-the-counter. Additionally we excluded
[1] Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip fractures in the United States. JAMA 2009;302(14):1573–9. [2] No author listed. Management of Hip Fractures in the Elderly: American Academy of Orthopaedic Surgeons; 2017 [Available from: http://www. orthoguidelines.org/guideline-detail?id=1279 . [3] Rodgers A, MacMahon S, Collins R, Prentice C. Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: pulmonary Embolism Prevention (PEP) trial. Lancet 2000;355(9212):1295. [4] Chotanaphuti T, Jareonarpornwatana A, Laoruengthana A. The mortality rate after thromboembolism prophylaxis in the hip fracture surgery. J Med Assoc Thail 2009;92(Suppl 6):S115–9. [5] Sasaki S, Miyakoshi N, Matsuura H, Saitoh H, Kudoh D, Shimada Y. Prospective randomized controlled trial on the effect of fondaparinux sodium for prevention of venous thromboembolism after hip fracture surgery. J Orthop Sci 2009;14(5):491–6. [6] Sasaki S, Miyakoshi N, Matsuura H, Saito H, Nakanishi T, Kudo Y, et al. Prospective study on the efficacies of fondaparinux and enoxaparin in preventing venous thromboembolism after hip fracture surgery. J Orthop Sci 2011;16(1):64–70. [7] J.Ørgensen PS, Knudsen JB, Broeng L, Josephsen L, Bjerregaard P, Hagen K, et al. The thromboprophylactic effect of a low-Molecular-Weight heparin (Fragmin) in hip fracture surgery: a placebo-controlled study. Clin Orthop Relat Res 1992;278:95–100. [8] Raskob G, Cohen AT, Eriksson BI, Puskas D, Shi M, Bocanegra T, et al. Oral direct factor Xa inhibition with edoxaban for thromboprophylaxis after elective total hip replacement. Thromb Haemost 2010;104(3):642–9. [9] Turpie AG, Gallus AS, Hoek JA. A synthetic pentasaccharide for the prevention of deep-vein thrombosis after total hip replacement. New Engl J Med 2001;344 (9):619–25. [10] Eriksson BI, Borris L, Dahl O, Haas S, Huisman M, Kakkar A, et al. Oral, direct factor Xa inhibition with BAY 59–7939 for the prevention of venous thromboembolism after total hip replacement. J Thromb Haemost 2006;4 (1):121–8. [11] Agnelli G, Haas S, Ginsberg J, Krueger K, Dmitrienko A, Brandt J. A phase II study of the oral factor Xa inhibitor LY517717 for the prevention of venous thromboembolism after hip or knee replacement. J Thromb Haemost 2007;5 (4):746–53. [12] Eriksson BI, Turpie A, Lassen M, Prins M, Agnelli G, Kälebo P, et al. A dose escalation study of YM150, an oral direct factor Xa inhibitor, in the prevention of venous thromboembolism in elective primary hip replacement surgery. J Thromb Haemost 2007;5(8):1660–5. [13] Eriksson BI, Bauer KA, Lassen MR, Turpie AG. Fondaparinux compared with enoxaparin for the prevention of venous thromboembolism after hip-fracture surgery. New Engl J Med 2001;345(18):1298–304.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx [14] Eriksson BI, Lassen MR. Duration of prophylaxis against venous thromboembolism with fondaparinux after hip fracture surgery: a multicenter, randomized, placebo-controlled, double-blind study. Arch Intern Med 2003;163(11):1337–42.
5
[15] Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care 1998;36(1):8–27. [16] Sharpe D. Your chi-square test is statistically significant: now what? Practical assessment. Res Eval 2015;20.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Full length article
Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study Sean T. Campbella,* , Abiram Balaa , Sam Y. Jiangb , Michael J. Gardnera , Julius A. Bishopa a b
Department of Orthopaedic Surgery, Stanford University, Stanford, CA, United States School of Medicine, Stanford University, Stanford, CA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 29 October 2017
Introduction: The purpose of this study was to evaluate the effectiveness of Factor Xa inhibitors (XaI) for thromboprophylaxis following hip fracture surgery in a large cohort of patients, and compare XaI against warfarin and enoxaparin. Methods: Patients undergoing hip fracture surgery from 2007 to 2015 were identified in a large claims database. Patients prescribed warfarin, XaI, or enoxaparin within 2 weeks of surgery were identified and grouped into cohorts. Medical comorbidities and complication incidences, including deep venous thrombosis (DVT), pulmonary embolism (PE), and bleeding complications were calculated. Chi-square analysis was performed and adjusted residuals calculated to determine significant differences. Results: DVT rates were significantly different between groups at thirty days only (5.03% warfarin, 2.91% XaI, 3.48% enoxaparin, p = 0.047). PE rates were significantly different at all time points; enoxaparin had the lowest rates. There were no differences in the rates of other complications. Discussion: XaI are an option for thromboprophylaxis in hip fracture patients, although their possible decreased effectiveness against PE compared to enoxaparin should be considered. Conclusions: This study compares the effectiveness of Factor Xa inhibitors to warfarin and enoxaparin for hip fracture patients, using a large national database. In this study, Factor Xa inhibitors had similar effectiveness for DVT prophylaxis compared to these agents. © 2017 Published by Elsevier Ltd.
Keywords: Hip fracture Anticoagulation Deep venous thrombosis Factor Xa inhibitors
Introduction Hip fracture is a common cause of hospital admission and nonelective surgery among geriatric patients, with an annual mean of 958 hip fractures per 100,000 individuals in the United States [1]. Venous thromboembolism (VTE) is a well-known complication that may occur following any orthopaedic injury or procedure, including surgical treatment of hip fractures. Pharmacologic VTE prophylaxis is often used in this setting to mitigate risk, and is recommended by the American Academy of Orthopaedic Surgeons (AAOS) Clinical Practice Guidelines [2]. A variety of anticoagulants have been shown to be effective at reducing the rate of VTE, including warfarin (which interferes with clotting factors II, VII, IX, and X via vitamin K antagonism), enoxaparin (which acts directly on the protein antithrombin), and aspirin (an anti-platelet), among others [3–7].
* Corresponding author at: Orthopaedic Surgery Stanford Hospital and Clinics, 300 Pasteur Drive, Room R144, Stanford, CA 94305-5341, United States. E-mail address: [email protected] (S.T. Campbell).
Despite the evidence in favor of VTE prophylaxis for hip fracture patients, there is limited data to guide surgeons when selecting a pharmacologic agent. In addition, newer anticoagulants such as Factor Xa inhibitors (XaI) are gaining popularity. These have some advantages compared with more traditional prophylaxis agents, including rapid onset, lack of required laboratory monitoring, and in some cases oral administration. They have also been shown to be effective for VTE prophylaxis in total joint arthroplasty [8–12]. However, data on their use in hip fracture patients is limited [6,13,14], and to our knowledge no large databases have been utilized to assess the efficacy of various chemoprophylactic agents, including Factor Xa inhibitors, for VTE prevention among hip fracture patients. The purpose of this study was to evaluate the efficacy of Factor Xa inhibitors for VTE prophylaxis following surgical treatment of hip fracture using a large national database, and to directly compare this against enoxaparin and warfarin. Secondarily, we sought to compare the rates of complications related to each of these three anticoagulants.
https://doi.org/10.1016/j.injury.2017.10.044 0020-1383/© 2017 Published by Elsevier Ltd.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
2
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Patients and methods A retrospective cohort study was designed using a large, combined private payer-Medicare national claims database (PearlDiver, Colorado Springs, CO) containing publicly available, HIPPA compliant patient data. All patient records were queried to identify patients who underwent surgical treatment of a hip fracture between 2007 and 2015, using Current Procedural Terminology (CPT) and International Classification of Diseases, Ninth Revision (ICD-9) codes (Table 1). Patients were age sixty or older at the time of surgery. We then identified those hip fracture patients who had a VTE chemoprophylaxis agent prescribed within two weeks of surgery and divided them into three cohorts: those who received 1) warfarin, 2) Factor Xa inhibitors, or 3) enoxaparin. Any patient who had been prescribed an antiplatelet or anticoagulant within 1 year prior to surgery was excluded. Cohorts were matched by age and gender by proportions. We then compared Charlson Comorbidity Indices (CCI), standard demographic factors and comorbidities as described previously [15], and rates of deep vein thrombosis (DVT), pulmonary embolism (PE), significant bleeding events, postoperative anemia, and post-operative transfusion at 2 weeks, thirty days, 6 weeks, and ninety days between the three cohorts, which were identified using CPT and ICD-9 codes (Table 1). Chi-square tests with significance set to p < 0.05 were used to compare the comorbidities and complication rates between the cohorts at each time point. When Chi-square testing demonstrated a significant difference in the three-way comparison between anticoagulation cohorts, adjusted residuals (AR) were calculated to measure the contributions of individual cells, with values 2 or 2 considered statistically significant [16]. Results There were 4090 hip fracture patients included in the study. 2326 of these received enoxaparin, 929 received XaI, and 835 received warfarin. The mean CCI for each group was 3 (Table 2). There were some significant differences in demographics between the three cohorts (Table 2). There were no significant differences in the rates of bleeding events, post-operative anemia, or post-operative transfusion between the three groups at any time point studied. There was a significant difference in DVT rates between the three groups at thirty days (5.03% warfarin, 2.91% XaI, and 3.48% enoxaparin, p = 0.047. Higher rates of DVT in the warfarin group contributed significantly to this difference (AR = 2.35 warfarin). AR in the XaI and enoxaparin cells were non-significant, but less than zero, indicating fewer observed DVTs than statistically expected ( 1.41 XaI, 0.72 enoxaparin) (Fig. 1). There were no significant differences in DVT rates at 2 weeks, 6 weeks, or ninety days (Fig. 1). For PE, there were significant differences between the groups at all time points studied following Chi-square analysis (2 weeks:
1.80% warfarin, 1.72% XaI, 0.60% enoxaparin, p = 0.002; thirty days: 2.04% warfarin, 1.83% XaI, 0.95% enoxaparin, p = 0.026; 6 weeks: 2.04% warfarin, 1.83% XaI, 0.99% enoxaparin, p = 0.037; ninety days: 2.63% warfarin, 2.15% XaI, 1.25% enoxaparin, p = 0.017)(Fig. 2). At 2 weeks, adjusted residuals were significant for all cells: the warfarin and XaI groups had significantly more PE events than expected (2.16 warfarin, 2.07 XaI), while the enoxaparin group had significantly fewer ( 3.51 enoxaparin). At the thirty day and six week time points, only lower rates than expected in the enoxaparin groups were found significant after calculation of adjusted residuals (thirty days: 1.86 warfarin, 1.37 XaI, 2.68 enoxaparin; 6 weeks: 1.77 warfarin, 1.29 XaI, 2.54 enoxaparin). At ninety days, there were significantly more events in the warfarin group than expected and significantly fewer in the enoxaparin group (AR = 2.23 warfarin, 1.11 XaI, 2.75 enoxaparin)(Fig. 2). Discussion The most important findings of this study are that in a large claims database review of hip fracture surgery patients: 1) DVT rates were not significantly different between warfarin, Factor Xa inhibitors, and enoxaparin, except at the thirty day time point, where high rates in the warfarin cohort contributed to a statistically significant difference, 2) rates of pulmonary embolism were significantly different between the three groups, with enoxaparin having lower rates than XaI and warfarin at all time points, and 3) there were no significant differences in rates of bleeding, post-operative anemia, and blood transfusion. There are relatively few high quality therapeutic studies assessing the effect of chemoprophylaxis on VTE rates among hip fracture patients, and even fewer data available which compare the efficacy of different agents. One of the studies with the highest methodologic quality on the topic is a multicenter, multinational, randomized controlled trial by Rodgers et al. from 2000, in which 13,356 hip fracture patients were randomized to either an antiplatelet or placebo, and complications were tracked for 35 days [3]. The authors found that the rates of DVT and PE at 35 days were 1.0% (compared to 1.5% in the placebo group) and 0.7% (compared to 1.2%), respectively, and concluded that chemoprophylaxis was effective for VTE risk reduction [3]. In the current study, each of the patient cohorts received some form of chemoprophylaxis, but we found 30 day DVT rates between 2.91–5.03% and PE rates between 0.95–2.04%, which are slightly higher than the Rodgers et al. data [3]. The reasons for this are unknown, and likely multifactorial. There is little information in the Rodgers study regarding the baseline medical comorbidities and general health status of the patients [3], and it is possible our American patient population had poorer overall health status at the time of surgery, which may have increased their risk for VTE. Additionally, up to 44% of patients in the previous study received both the randomized treatment of an antiplatelet in addition to unfractionated heparin or enoxaparin [3]; it is possible this use of dual chemoprophylaxis reduced VTE rates in that study relative to our cohort.
Table 1 List of CPT and IDC-9 codes used to identify patients and complications. Diagnosis/Procedure
Codes
Surgical fixation hip fracture DVT PE Post-operative anemia Bleeding complications Transfusion
ICD-9: P-7935, P-8152 CPT: 27245, 27244, 27248, 27269 ICD-9: 451.11, 451.19, 451.2, 451.81, 451.9, 453.40, 453.41, 453.42, 453.8, 453.9 ICD-9: 415.1, 415.11, 415.13, 415.19 ICD-9: 285.1 ICD-9: 998.1, 998.11, 998.12, 998.13, 719.10, 719.16, 719.16, 39.98 CPT-36430 ICD-9: 99.0, 99.00, 99.02, 99.03, 99.04
CPT = Current Procedural Terminology, ICD-9 = International Classification of Diseases, Ninth Revision DVT = deep venous thrombosis, PE = pulmonary embolism.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
3
Table 2 Patient Characteristics. Enoxaparin 2326
Coumadin 835
Factor Xa 929
P value
170 (7.3) 413 (17.8) 498 (21.4) 491 (21.1) 348 (15) 190 (8.2) 216 (9.3)
61 (7.3) 149 (17.8) 179 (21.4) 176 (21.1) 125 (15) 67 (8) 78 (9.3)
68 (7.3) 165 (17.8) 199 (21.4) 196 (21.1) 139 (15) 76 (8.2) 86 (9.3)
1.000 0.998 1.000 1.000 1.000 0.990 0.998
Female Male
1634 (70.2) 692 (29.8)
586 (70.2) 249 (29.8)
653 (70.3) 276 (29.7)
0.999 0.999
Congestive Heart Failure Valvular Disease Pulmonary Circulation Disorders Peripheral Vascular Disease HTN (uncomplicated) HTN (complicated) HTN (uncomplicated and complicated) Paralysis Other Neurological Disorders Chronic Pulmonary Disease Diabetes without chronic complications Diabetes with chronic complications Hypothyroidism Renal Failure Liver disease Chronic Peptic Ulcer Disease HIV/AIDS Lymphoma Metastatic Cancer Solid Tumor without Metastasis Rheumatoid Arthritis/Collagen Vascular Diseases Coagulation Deficiency Obesity Weight Loss Fluid and Electrolyte Disorders Blood Loss Anemia Deficiency Anemias Alcohol Abuse Drug Abuse Psychoses Depression
286 (12.3) 374 (16.1) 115 (4.9) 570 (24.5) 1461 (62.8) 429 (18.4) 1471 (63.2) 45 (1.9) 540 (23.2) 696 (29.9) 585 (25.2) 288 (12.4) 522 (22.4) 364 (15.6) 94 (4) 13 (0.6) 10 (0.4) 28 (1.2) 98 (4.2) 338 (14.5) 280 (12) 122 (5.2) 200 (8.6) 342 (14.7) 556 (23.9) 85 (3.7) 705 (30.3) 97 (4.2) 103 (4.4) 361 (15.5) 477 (20.5)
95 (11.4) 125 (15) 38 (4.6) 154 (18.4) 496 (59.4) 104 (12.5) 497 (59.5) 26 (3.1) 155 (18.6) 232 (27.8) 191 (22.9) 86 (10.3) 172 (20.6) 103 (12.3) 24 (2.9) 10 (<1.2) 0 (0) 13 (1.6) 39 (4.7) 127 (15.2) 67 (8) 34 (4.1) 53 (6.3) 114 (13.7) 199 (23.8) 23 (2.8) 223 (26.7) 22 (2.6) 24 (2.9) 75 (9) 146 (17.5)
122 (13.1) 159 (17.1) 45 (4.8) 248 (26.7) 608 (65.4) 174 (18.7) 605 (65.1) 24 (2.6) 216 (23.3) 276 (29.7) 234 (25.2) 119 (12.8) 215 (23.1) 151 (16.3) 31 (3.3) 12 (1.3) 0 (0) 24 (2.6) 42 (4.5) 132 (14.2) 110 (11.8) 57 (6.1) 106 (11.4) 128 (13.8) 256 (27.6) 33 (3.6) 285 (30.7) 30 (3.2) 55 (5.9) 126 (13.6) 191 (20.6)
0.534 0.473 0.902 0.000 0.032 0.000 0.046 0.127 0.015 0.498 0.393 0.206 0.408 0.039 0.259 0.059 0.022 0.018 0.832 0.830 0.005 0.149 0.001 0.670 0.074 0.464 0.110 0.095 0.008 0.000 0.146
n Demographics Age (%) 60 65 70 75 80 85 90
to 64 to 69 to 74 to 79 to 84 to 89 and over
Gender (%)
Co-morbidities Conditions (%)
Comparison of patient demographics and baseline comorbidities across cohorts. Bold text indicates a statistically significant difference. is used to indicated that there are less than 10 patients in a particular cohort; the exact number is hidden in the PearlDiver software for patient privacy reasons. The p-values associated with these entries are indicated with and are uncertain.
A few previous studies have examined Factor Xa inhibitors for VTE prophylaxis in hip fracture patients. In a 2001 paper, Eriksson et al. studied 1250 patients with hip fractures that had been randomized to either postoperative Factor Xa inhibitor or enoxaparin [13]. Each patient received bilateral venography post-operatively to assess for DVT, which ensured that even asymptomatic VTE were detected. The authors found that at postoperative day 11, the rates of DVT were 7.9% in the XaI group and 18.8% in the enoxaparin group, which were significantly different in their statistical analysis [13]. The group also found there were significantly more “minor bleeding” complications in the XaI cohort [13]. More recently, Sasaki et al. randomized a smaller group of 84 hip fracture patients to XaI or enoxaparin following surgical treatment of hip fractures [6]. These authors found a similar trend compared to Eriksson et al [13]: lower rates of VTE in the XaI group compared to the enoxaparin group, with a trade-off of more complications related to bleeding6. In the current study, we were able to include a greater number of patients than either of the previously mentioned studies by drawing on a large national
claims database. However, our work was retrospective in nature compared to the prospective Eriksson and Sasaki studies, and this did introduce some significant differences into the baseline characteristics for our cohorts that are absent in the aforementioned randomized controlled trials. One important methodologic difference is that in this study, outcomes were based on CPT and ICD-9 codes, so VTE events were only detected if they were clinically significant. This may partially account for the differences in VTE rates between our study and Eriksson’s, in which routine venography was performed to detect even asymptomatic DVTs [13]. Additionally, at 2 weeks from surgery (which most corresponds to the time point studied by Eriksson) we found no significant difference between DVT rates for our three cohorts (although we did find a difference at thirty days), and while there was a significant difference in PE rates at all time points, enoxaparin had lower rates than either XaI or warfarin. Finally, in this study there were no significant differences in bleeding complications.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
4
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 1. Rates of deep venous thrombosis (DVT) by treatment cohort (warfarin, Factor Xa inhibitor, and enoxaparin) at 4 time points studied (2 weeks, 30 days, 6 weeks, and 90 days). *indicates significance with p < 0.05 in Chi-square analysis. # indicates the cohort had a significant contribution to the difference observed, with an adjusted residual 2 or 2.
patients who had been on aspirin pre-operatively, and this may have eliminated a large number of patients who were instructed to continue their home regimen for prophylaxis following surgery. Finally, we chose to exclude patients who were treated for hip fracture with total hip arthroplasty (THA) from the study. This was decided for two reasons: 1) multiple studies already exist regarding the use of XaI in THA, and 2) THA is perhaps a more “extensive” procedure than others commonly performed for hip fracture, and including these patients could result in a significant number of outliers in terms of outcome measures like anemia, bleeding, and transfusion. Despite these limitations, this study is one of the largest to date to assess the efficacy of Factor Xa inhibitors for VTE prophylaxis in hip fracture patients and directly compare these agents to others commonly used for chemoprophylaxis. In conclusion, in this large national database study of hip fracture patients, warfarin, Factor Xa inhibitors, and enoxaparin were similarly effective for DVT prophylaxis following hip fracture surgery at two weeks, six weeks, and ninety days. At thirty days only, warfarin was statistically inferior. There were significant differences in rates of PE between the groups, with enoxaparin statistically superior to XaI and warfarin at all time points studied. The rates of complications related to bleeding were no different between the groups. Factor Xa inhibitors are a viable option for use as thromboprophylaxis in hip fracture patients, although their increased cost and possible decreased efficacy compared to enoxaparin for PE prevention needs to be weighed against the potential for increased compliance and patient preference. References
Fig. 2. Rates of pulmonary embolism (PE) by treatment cohort (warfarin, Factor Xa inhibitor, and enoxaparin) at 4 time points studied (2 weeks, 30 days, 6 weeks, and 90 days). * indicates significance with p < 0.05 in Chi-square analysis. # indicates the cohort had a significant contribution to the difference observed, with an adjusted residual 2 or 2.
This study has multiple limitations, including those inherent to large database studies based on administrative claims: First, the accuracy of the data collected is dependent on the accuracy of the initial coding; subsequent analysis of this data assumes that the coding is accurate and complete. This also means that it is unknown how VTE events were diagnosed in this study (ultrasound versus venography); the only information available if whether a relevant code was entered. Additionally, unlike a traditional retrospective chart review, individualized patient data is unavailable in the database, and specific variables or outcomes cannot be tied to a particular patient. This means that certain analyses, such as regression or propensity score matching, cannot be performed. Beyond the inherent limitations of database studies, this particular study is limited by the lack of an antiplatelet (aspirin) cohort. A cohort of patients that had received aspirin for VTE prophylaxis would have been ideal to include, since this agent is commonly used for this purpose and was previously studied with high quality methodology [3]. Unfortunately, the number of patients in such a cohort obtained via the claims database used would have been too small to allow for a meaningful comparison. This is likely partially related to the fact that medications tracked by the database must be filled via prescription, and many patients that used aspirin for VTE prophylaxis may have been instructed to obtain their aspirin over-the-counter. Additionally we excluded
[1] Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip fractures in the United States. JAMA 2009;302(14):1573–9. [2] No author listed. Management of Hip Fractures in the Elderly: American Academy of Orthopaedic Surgeons; 2017 [Available from: http://www. orthoguidelines.org/guideline-detail?id=1279 . [3] Rodgers A, MacMahon S, Collins R, Prentice C. Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: pulmonary Embolism Prevention (PEP) trial. Lancet 2000;355(9212):1295. [4] Chotanaphuti T, Jareonarpornwatana A, Laoruengthana A. The mortality rate after thromboembolism prophylaxis in the hip fracture surgery. J Med Assoc Thail 2009;92(Suppl 6):S115–9. [5] Sasaki S, Miyakoshi N, Matsuura H, Saitoh H, Kudoh D, Shimada Y. Prospective randomized controlled trial on the effect of fondaparinux sodium for prevention of venous thromboembolism after hip fracture surgery. J Orthop Sci 2009;14(5):491–6. [6] Sasaki S, Miyakoshi N, Matsuura H, Saito H, Nakanishi T, Kudo Y, et al. Prospective study on the efficacies of fondaparinux and enoxaparin in preventing venous thromboembolism after hip fracture surgery. J Orthop Sci 2011;16(1):64–70. [7] J.Ørgensen PS, Knudsen JB, Broeng L, Josephsen L, Bjerregaard P, Hagen K, et al. The thromboprophylactic effect of a low-Molecular-Weight heparin (Fragmin) in hip fracture surgery: a placebo-controlled study. Clin Orthop Relat Res 1992;278:95–100. [8] Raskob G, Cohen AT, Eriksson BI, Puskas D, Shi M, Bocanegra T, et al. Oral direct factor Xa inhibition with edoxaban for thromboprophylaxis after elective total hip replacement. Thromb Haemost 2010;104(3):642–9. [9] Turpie AG, Gallus AS, Hoek JA. A synthetic pentasaccharide for the prevention of deep-vein thrombosis after total hip replacement. New Engl J Med 2001;344 (9):619–25. [10] Eriksson BI, Borris L, Dahl O, Haas S, Huisman M, Kakkar A, et al. Oral, direct factor Xa inhibition with BAY 59–7939 for the prevention of venous thromboembolism after total hip replacement. J Thromb Haemost 2006;4 (1):121–8. [11] Agnelli G, Haas S, Ginsberg J, Krueger K, Dmitrienko A, Brandt J. A phase II study of the oral factor Xa inhibitor LY517717 for the prevention of venous thromboembolism after hip or knee replacement. J Thromb Haemost 2007;5 (4):746–53. [12] Eriksson BI, Turpie A, Lassen M, Prins M, Agnelli G, Kälebo P, et al. A dose escalation study of YM150, an oral direct factor Xa inhibitor, in the prevention of venous thromboembolism in elective primary hip replacement surgery. J Thromb Haemost 2007;5(8):1660–5. [13] Eriksson BI, Bauer KA, Lassen MR, Turpie AG. Fondaparinux compared with enoxaparin for the prevention of venous thromboembolism after hip-fracture surgery. New Engl J Med 2001;345(18):1298–304.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044
G Model JINJ 7477 No. of Pages 5
S.T. Campbell et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx [14] Eriksson BI, Lassen MR. Duration of prophylaxis against venous thromboembolism with fondaparinux after hip fracture surgery: a multicenter, randomized, placebo-controlled, double-blind study. Arch Intern Med 2003;163(11):1337–42.
5
[15] Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care 1998;36(1):8–27. [16] Sharpe D. Your chi-square test is statistically significant: now what? Practical assessment. Res Eval 2015;20.
Please cite this article in press as: S.T. Campbell, et al., Are factor Xa inhibitors effective thromboprophylaxis following hip fracture surgery?: A large national database study, Injury (2017), https://doi.org/10.1016/j.injury.2017.10.044