- Email: [email protected]
Are Revision Hip Arthroplasty Patients at Higher Risk for Venous Thromboembolic Events Than Primary Hip Arthroplasty Patients?
Accepted Manuscript Are revision hip arthroplasty patients at higher risk for venous thromboembolic events than primary hip arthroplasty patients? P. Maxwell Courtney, MD, Anthony J. Boniello, MD, Brett R. Levine, MD, MS, Neil P. Sheth, MD, Wayne G. Paprosky, MD PII:
S0883-5403(17)30644-7
DOI:
10.1016/j.arth.2017.07.028
Reference:
YARTH 56008
To appear in:
The Journal of Arthroplasty
Received Date: 29 April 2017 Revised Date:
29 June 2017
Accepted Date: 17 July 2017
Please cite this article as: Courtney PM, Boniello AJ, Levine BR, Sheth NP, Paprosky WG, Are revision hip arthroplasty patients at higher risk for venous thromboembolic events than primary hip arthroplasty patients?, The Journal of Arthroplasty (2017), doi: 10.1016/j.arth.2017.07.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Are revision hip arthroplasty patients at higher risk for venous thromboembolic events than primary hip arthroplasty patients?
RI PT
P. Maxwell Courtney MD,1 Anthony J. Boniello MD,2 Brett R. Levine MD, MS,3 Neil P. Sheth MD,4 Wayne G. Paprosky MD3
Study Conducted at Rush University Medical Center, Chicago, IL 1
M AN U
SC
Rothman Institute, Thomas Jefferson University Hospital, Department of Orthopaedic Surgery 2 Drexel University College of Medicine, Department of Orthopaedic Surgery 3 Rush University Medical Center, Department of Orthopaedic Surgery 4 University of Pennsylvania School of Medicine, Department of Orthopaedic Surgery
Source of Funding Each author certifies that he or she, or a member of his or her immediate family, has no funding or commercial associations (eg, consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.
AC C
EP
TE D
Disclosure The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
Corresponding author: P. Maxwell Courtney, MD Rothman Institute, Thomas Jefferson University Hospital Department of Orthopaedic Surgery 925 Chestnut St, 5th Floor Philadelphia, PA 19107 Email: [email protected]
ACCEPTED MANUSCRIPT VTE following Revision THA
1
1
Are revision hip arthroplasty patients at higher risk for venous thromboembolic
2
events than primary hip arthroplasty patients?
3 Abstract:
5
Introduction: The purpose of this study was to determine whether revision total hip
6
arthroplasty (THA) is associated with increased rates of deep venous thrombosis (DVT)
7
and pulmonary embolism (PE) when compared to primary THA.
8
Methods: We queried the American College of Surgeons National Surgical Quality
9
Improvement Program database for all primary and revision THA cases from 2011-2014.
10
Demographic data, medical comorbidities, and venous thromboembolic (VTE) rates
11
within 30 days of surgery were compared between the primary and revision THA groups.
12
Results: Revision THA had a higher rate of DVT than the primary THA (0.6% vs. 0.4%,
13
p=0.016), but there was no difference in rate of PE (0.3% vs. 0.2%, p=0.116). When
14
controlling for confounding variables, revision surgery alone was not a risk factor for
15
DVT (OR 0.833, 95% CI 0.564-1.232) or PE (OR 1.009, 95% CI 0.630-1.616).
16
Independent risk factors for DVT include age > 70 years, malnutrition, infection,
17
operating time > 3 hours, general anesthesia, ASA classification 4 or greater, and kidney
18
disease (all p<0.05). Probability of DVT ranged from 0.2% with zero risk factors to 10%
19
with all risk factors. Independent risk factors for PE included age > 70 years, African
20
American ethnicity, and operating time > 3 hours (all p<0.05) with probabilities of PE
21
postoperatively ranging from 0.2% to 1.1% with all risk factors.
AC C
EP
TE D
M AN U
SC
RI PT
4
ACCEPTED MANUSCRIPT VTE following Revision THA
2
22
Conclusion: Revision surgery alone is not a risk factor for VTE after hip arthroplasty.
23
Surgeons should weigh the risks and benefits of more aggressive anticoagulation in
24
certain high-risk patients.
RI PT
25 26
Keywords: Total hip arthroplasty; complications; revision arthroplasty; pulmonary
27
embolism, deep vein thrombosis
SC
28 29
M AN U
30 31 32 33
37 38 39 40 41 42 43 44
EP
36
AC C
35
TE D
34
ACCEPTED MANUSCRIPT VTE following Revision THA
3
45
Are revision hip arthroplasty patients at higher risk for venous thromboembolic
46
events than primary hip arthroplasty patients?
47 Introduction
RI PT
48
Venous thromboembolic events (VTE), including deep vein thrombosis (DVT)
50
and pulmonary embolism (PE), remain one of the most common complications following
51
total hip arthroplasty (THA).[1] While the clinical significance of lower extremity DVT
52
has been contested, recent studies report the rate of DVT following THA to be around
53
8%,[2] whereas the incidence of in-hospital PE has been reported to be 0.36%.[3] There
54
currently exists a great variability in the exact pharmacologic agents utilized,[1,4–7]
55
however the American Academy of Orthopaedic Surgeons (AAOS) and American
56
College of Chest Physicians (ACCP) recommend that some form of VTE prophylaxis be
57
administered routinely following THA.[8] While rates and risk factors for VTE after
58
primary THA are well studied in the literature,[1,2,5–7,9–11] little has been written about
59
DVT and PE following revision THA. Clinical concerns exist about an increased rate of
60
VTE following revision surgery due to increased surgical exposure, protected weight
61
bearing postoperatively, and poor mobilization postoperatively.
M AN U
TE D
EP
Historically there exist conflicting philosophies between the AAOS and ACCP
AC C
62
SC
49
63
regarding the necessity of and overall safety profile of pharmacoprophylactic
64
regimens.[12] The AAOS has recently raised concerns of postoperative bleeding,
65
hematoma, infection, and potential reoperation with the use of more potent
66
thromboprophylactic medications such as low molecular weight heparin.[13,14] More
67
aggressive anticoagulation in revision patients could lead to poorer outcomes, as revision
ACCEPTED MANUSCRIPT VTE following Revision THA
4
THA is inherently associated with increased rates of bleeding and increased
69
complications.[9] In 2007, the AAOS developed its first clinical practice guideline,
70
focusing on prevention of symptomatic VTE events and limiting perioperative morbidity
71
caused by anticoagulants. As part of this guideline the AAOS recommended risk
72
stratification for VTE events.[13,14] Such stratification may enable physicians to
73
prescribe better-tolerated pharmacoprophylactic regimens with lower adverse affects, and
74
better safety profiles to lower risk individuals.[6] Identifying at-risk patients for VTE
75
events following revision THA remains difficult and has yet to be addressed in the
76
literature.
M AN U
SC
RI PT
68
The purpose of this study is to determine 1) Is revision THA associated with
78
higher rates of postoperative venous thromboembolic events than primary THA?
79
Secondary study questions included 2) What are the independent risk factors for VTE
80
after THA to identify patients who may benefit from more aggressive anticoagulation and
81
3) Can a simple risk score be used to predict the probability of DVT and PE after THA?
82 Methods
EP
83
TE D
77
We retrospectively queried the American College of Surgeons-National Surgical
85
Quality Improvement Program (ACS-NSQIP) database for all patients who underwent
86
primary or revision total hip arthroplasty from January 1, 2011 to December 31, 2014.
87
Patients in the database were identified based on the primary procedure consisting of
88
Current Procedural Terminology (CPT) codes for elective primary THA (27130) and
89
revision THA (27134, 27137, and 27138). Emergent cases were excluded. Patients
90
undergoing hip arthroplasty with other procedures listed as the primary surgery were also
AC C
84
ACCEPTED MANUSCRIPT VTE following Revision THA
5
91
excluded from the study. This study was exempt from Institutional Review Board (IRB)
92
approval as all data was de-identified. No external funding was received for this study.
93 The NSQIP database is a validated, national database including cases from over 650
95
hospitals.[15] Demographic variables, medical comorbidities, operative time, deep
96
venous thrombosis, and pulmonary embolism rates were noted from the database. The
97
International Classification of Diseases, 9th edition (ICD-9) code for primary diagnosis
98
was also documented. Deep venous thrombosis was defined as a “new diagnosis of blood
99
clot or thrombus within the venous system (superficial or deep) which may be coupled
100
with inflammation and requires treatment.”[16] Pulmonary embolism was defined as “A.
101
New diagnosis of a new blood clot in a pulmonary artery AND B. The patient has a V-Q
102
scan interpreted as high probability of pulmonary embolism or a positive CT exam, TEE,
103
pulmonary arteriogram, CT angiogram, or any other definitive imaging modality
104
(including direct pathology examination such as autopsy).”[17] Days from surgery to the
105
VTE were also recorded. Complications, reoperations, and readmissions within 30 days
106
of surgery were noted. Complete NSQIP methodology has been reported previously in
107
the literature.13
SC
M AN U
TE D
EP
We defined malnutrition as any patient with a preoperative albumin less than 3.5
AC C
108
RI PT
94
109
g/dL and preoperative kidney disease as any patient with a preoperative creatinine greater
110
than 1.5 mg/L. Patients requiring prior coronary intervention were classified as have a
111
history of cardiac disease. Low hematocrit was defined by a preoperative value of less
112
than 33%. Complications, reoperations, and readmissions within 30 days of surgery were
113
noted. We defined complications from the database within 30 days of surgery as any
ACCEPTED MANUSCRIPT VTE following Revision THA
6
patient having a recorded surgical site infection, pneumonia, respiratory complication
115
requiring reintubation, pulmonary embolism, deep venous thrombosis, renal insufficiency
116
or failure, urinary tract infection, stroke, cardiac arrest, bleeding requiring transfusion,
117
sepsis or septic shock. Specific definitions are found in the NSQIP participant use data
118
file.[16]
RI PT
114
119 Statistical Analysis
SC
120
An a priori power analysis was first conducted to determine the appropriate
122
sample size. Our primary outcome variable was whether revision THA was associated
123
with a higher 30-day DVT and PE rate than primary THA. Based upon prior data
124
suggesting a 1.5% VTE rate among revision hip arthroplasty,[4] in order to detect a 0.5%
125
difference in the incidence of DVT and PE, we would need to enroll a total of 70,528
126
patients assuming a type I error rate of 0.05 to achieve a power of 0.80.
TE D
M AN U
121
Data analysis was first performed comparing patients undergoing revision TJA
128
with those who had a primary THA. Categorical variables were analyzed using chi-
129
square analyses. Continuous variables such as age and BMI were analyzed using
130
unpaired, two-tailed, Student’s t-tests. Statistical significance was set at p < 0.05.
131
Multivariate regression was then performed to identify independent risk factors for DVT
132
and PE. A forward, stepwise, multiple logistic regression analysis was then performed to
133
create a model to identify high-risk risk factors for DVT and PE. We started with the
134
most significant variable and added risk factors in order of ascending p-values. The
135
model was rechecked after each addition and any alpha > 0.10 was removed from the
136
analysis. In order to design this model for practical use, we developed a simplified score
AC C
EP
127
ACCEPTED MANUSCRIPT VTE following Revision THA
7
137
weighted by the coefficients and odds ratios of each independent variable. Statistical
138
analysis was performed using Microsoft Excel (Bellvue, WA) and IBM SPSS version
139
24.0 (Armonk, NY).
141
RI PT
140 Results
Of the 74,405 patients included in the study, there were 7,566 revision THA
143
procedures (10%). The mean age for all patients was 64.9 years (Standard deviation [SD]
144
1.7), while the mean body mass index (BMI) was 29.8 kg/m2 (SD 7.0). There were
145
33,058 males (44%) with the majority of all patients being of Caucasian ethnicity (59,034
146
patients, 79%). There were 324 patients (0.4%) who experienced a DVT within 30 days,
147
while 208 patients (0.3%) had a PE. The mean time from surgery to DVT was 9.1 days
148
(SD 9.1) and to PE was 13.2 days (SD 8.5). Complete descriptive statistics and
149
demographic data of the study population are detailed in Table 1.
TE D
M AN U
SC
142
Revision THA had a higher incidence of DVT than primary THA (0.6% vs. 0.4%,
151
p=0.016), but there was no statistical difference in the rate of PE (0.3% vs. 0.2%,
152
p=0.116). There was no difference in the mean time to DVT (12.9 vs. 14.8 days,
153
p=0.890) or PE (9.2 vs. 8.6 days, p=0.215). Among all revision patients, there was no
154
difference in DVT rate between the both component revisions, acetabular component
155
only, and femoral component only groups (0.7% vs. 0.4% vs. 0.3%, p=0.128). Similarly,
156
there was no difference in the rate of PE between the types of revision groups (0.4% vs.
157
0.3% vs. 0.6%, p=0.474). The revision group was older (mean age 65.8 years vs. 64.7
158
years, p<0.001) and had more medical comorbidities than the primary group, including
159
diabetes, kidney disease, cardiac disease, and higher ASA classification (all p<0.001).
AC C
EP
150
ACCEPTED MANUSCRIPT VTE following Revision THA
8
160
Mortality (0.7% vs. 0.2%, p<0.001) and complication rates (37% vs. 17%, p<0.001) were
161
higher in revision THA as well. Data comparing the revision and primary THA groups is
162
shown in Table 2. Multivariate analysis identified independent risk factors for DVT as age over 70
164
years (OR 1.455, 95% CI 1.157-1.829, p=0.001), general anesthesia (OR 1.305, 95% CI
165
1.030-1.654, p=0.028), arthroplasty procedure for infection (OR 4.163, 95% CI 2.180-
166
7.951, p<0.001), operating time over three hours (OR 1.735, 95% CI 1.179-2.553,
167
p=0.005), and American Society of Anesthesiologists (ASA) Class IV or greater (OR
168
1.734, 95% CI 1.056-2.848, p=0.030). Independent risk factors for PE included age over
169
70 years (OR 1.701, 95% CI 1.271-2.277, p<0.001) and operating time over three hours
170
(OR 1.845, 95% CI 1.124-3.026, p=0.015). Revision surgery was not an independent risk
171
factor for DVT (OR 0.833, 95% CI 0.564-1.232, p=0.360) or PE (OR 1.009, 95% 0.630-
172
1.616, p=0.971). Complete regression analysis is listed in Table 3.
TE D
M AN U
SC
RI PT
163
Based upon the coefficients of the forward, stepwise multivariate logistic
174
regression analysis, we created a weighted risk score to calculate the probability of DVT
175
after THA. A statistically significant risk factor’s weighted risk score was based on the
176
odds ratio and assigned a corresponding ordinal value to generate a simple, easy to use
177
score which is more practical or surgeons estimating VTE risk following THA. Risk
178
factors included age over 70 years, general anesthesia, kidney disease, malnutrition,
179
operating time over three hours, ASA IV classification or greater (all one point), and
180
infection (three points). The risk score was significantly associated with probability of
181
DVT after THA (OR 1.527, 95% CI 1.396-1.671, p<0.001). Probability of DVT ranged
182
from 0.2% with zero risk factors to 10% with all risk factors (Figure 1). Multivariate risk
AC C
EP
173
ACCEPTED MANUSCRIPT VTE following Revision THA
9
factors for PE included Age (should this be caps?) > 70 years, African American
184
ethnicity, and operating time > three hours (all one point) and the score was significantly
185
associated with the probability of PE after THA (OR 1.802, 95% CI 1.448-2.241,
186
p<0.001). The probability of PE postoperatively ranges from 0.2% with zero risk factors
187
to 1.1% with all risk factors (Figure 2).
188 Discussion
SC
189
RI PT
183
While rates and risk factors for VTE after primary hip arthroplasty are well
191
studied in the literature,[1,2,5–7,9–11] little has been written about DVT and PE
192
following revision THA. Clinical concerns exist about an increased rate of VTE
193
following revision surgery due to increased surgical exposure, protected weight bearing
194
postoperatively, and poor mobilization postoperatively. However, more aggressive
195
anticoagulation in revision patients could lead to poorer outcomes, as revision THA is
196
inherently associated with increased rates of bleeding and increased complications.[9]
197
We present a large series from a validated, multi-center, national database suggesting that
198
revision surgery alone is not a risk factor for VTE following total hip arthroplasty.
EP
TE D
M AN U
190
While there are limitations to data drawn from large databases, the NSQIP is a
200
validated, accurate national database. NSQIP collects large patient numbers from over
201
650 hospitals nationwide, which helps make findings generalizable to most physicians
202
across the country considering instituting an outpatient TJA protocol. The NSQIP
203
database is stronger than data derived from billing codes in other administrative databases.
204
Data collectors are certified each year, and random audits have found rates of discrepancy
205
less around 2% within each site.[17] One such limitation is in the availability of baseline
AC C
199
ACCEPTED MANUSCRIPT VTE following Revision THA
10
comorbidities. Information regarding history of DVT and PE was unavailable. In the
207
current paper “cardiac disease” was characterized by a history of percutaneous coronary
208
intervention. This excludes a variety of clinically significant cardiac diseases that have
209
been shown to effect outcomes.
RI PT
206
The NSQIP database only includes complications, reoperation and readmission
211
information for 30-days, which may under-report complications when compared to 90-
212
day follow up. Additionally, the database lacks orthopaedic specific complications
213
including dislocation, fracture, and poor functional outcome scores. The database did not
214
track which VTE prophylaxis was administered to each patient. As some data suggests a
215
trend toward less aggressive anticoagulation,[4,7] the variability in chemoprophylaxis
216
could affect VTE rates. Weight-bearing status following revision surgery was also not
217
reported. Patients with restricted weight bearing may be at risk for higher VTE rates
218
because of prolonged immobilization. There is no differentiation regarding diagnostic
219
modalities for DVT in the NSQIP database, therefore only clinically symptomatic DVTs
220
and DVTs identified by ultrasound were included in the same group. In the absence of
221
routine ultrasound screening, asymptomatic DVTs were likely overwhelmingly
222
underreported. While we tracked types of revisions by CPT code, there could still be
223
variability in the surgical procedures ranging from a simple head and liner exchange to a
224
more complex both component revision requiring an extended trochanteric osteotomy.
225
Caution should also be made when interpreting certain differences between the primary
226
and revision groups such as age, which is statistically different but not clinically different.
227
Our study found a higher prevalence of DVT in revision versus primary THA, but
228
no difference in the rate of PE. While a higher prevalence of DVT existed in revision
AC C
EP
TE D
M AN U
SC
210
ACCEPTED MANUSCRIPT VTE following Revision THA
11
THA, the most significant finding of the present study is that revision THA is not an
230
independent risk factor for DVT and PE. These findings agree with previous finding by
231
Wakabayashi et al, which concluded that the prevalence of DVT was higher in the
232
revision THA than for the primary cases, however following a multivariate analysis,
233
revision THA was not an independent risk factor.[11] These findings suggest that
234
confounding variables exist such as increased age, comorbidities, and increased operative
235
time that are associated with revision surgery. While revision THA is not itself an
236
independent risk factor, many of these confounding variables associated with revision
237
THA were found to increase the risk of DVT, and must be considered when selecting,
238
counseling and treating this patient population.
239
arthroplasty or higher risk primary arthroplasty patients may have chosen more
240
aggressive chemoprophylaxis, which could explain the comparable VTE rates. We also
241
did not find any difference in DVT or PE rate based on type of revision surgery (both
242
component, acetabular only, femoral only). There is a lack of consensuses in the current
243
literature regarding optimal DVT prophylaxis following THA.[1,2,5–7,9–11] While the
244
optimal prophylaxis for revision surgery is not clear, one study by Deirmengian et al did
245
not find a difference in incidence of VTE between aspirin and warfarin following
246
revision THA.[4] Patients undergoing revision THA without further risk factors for VTE
247
may benefit from less aggressive anticoagulation without being at increased risk for DVT
248
and PE.
SC
M AN U
EP
TE D
Surgeons also treating revision
AC C
249
RI PT
229
The present study found that independent risk factors for DVT were age over 70
250
years, general anesthesia, arthroplasty procedure for infection, operating time over three
251
hours, and ASA IV or greater. Furthermore, independent risk factors for PE included age
ACCEPTED MANUSCRIPT VTE following Revision THA
12
over 70 years, and operating time over three hours. These results are consistent with
253
previous studies. One such national study by Dua et al found that patients with diagnosed
254
DVT following primary TJA were older, more likely to be African American and more
255
likely to have significant comorbidities including congestive heart failure, peripheral
256
artery disease, and end-stage renal disease.[1] Another previous study found an
257
association with anesthesia time with an increased risk of DVT.[5]
RI PT
252
We present an easy to use risk score for VTE events following revision THA. An
259
optimal cutoff has not been established and will depend on each patient and surgeon’s
260
tolerance for risk.
261
anticoagulation have been shown to be effective in primary THA and TKA.13 Our study
262
identifies patients that are at-risk for DVT and PE that may benefit from stronger
263
anticoagulation. The primary limitation of the current VTE risk score presented is that it
264
does not account for differences in VTE prophylaxis protocols in the patients studied.
265
Historically surgeons have treated revisions as high risk cases, and therefore may have
266
prophylactically treated such patients with more aggressive forms of DVT prophylaxis.
267
One notable risk factor missing from the scoring system is history of DVT or PE, as this
268
data was not available in the NSQIP database. Despite this limitation the the present
269
authors conclude that while aggressive anticoagulation is associated with increased
270
adverse events following revision THA, surgeons should consider stratifying their
271
revision THA patients when considering pharmacoprophylaxis against VTE.
272
prospective studies are needed to validate the VTE risk score and subsequent treatment
273
recommendations.
274
SC
258
AC C
EP
TE D
M AN U
Stratification protocols to prevent complications associated with
Further
ACCEPTED MANUSCRIPT VTE following Revision THA 275
13
Conclusion
276
Revision THA alone is not an independent risk factor for DVT and PE when
277
compared to primary hip arthroplasty.
278
operating time > 3 hours, and age > 70 years are at higher risk for VTE events. Surgeons
279
should weigh the risks and benefits of more aggressive anticoagulation in these high-risk
280
patients. Further study is needed to determine the effectiveness of risk stratification
281
protocols for thromboprophylaxis following revision THA.
282
References
[1]
RI PT
SC
M AN U
283 284
Patients undergoing a THA for infection,
Dua A, Desai SS, Lee CJ, Heller JA. National Trends in Deep Vein Thrombosis
285
following Total Knee and Total Hip Replacement in the United States. Ann Vasc
286
Surg 2016. doi:10.1016/j.avsg.2016.05.110. [2]
Wong KL, Daguman R, Lim KH, Shen L, Lingaraj K. Incidence of deep vein
TE D
287 288
thrombosis following total hip arthroplasty: a Doppler ultrasonographic study. J
289
Orthop Surg (Hong Kong) 2011;19:50–3. [3]
Memtsoudis SG, Besculides MC, Gaber L, Liu S, González Della Valle A. Risk
EP
290
factors for pulmonary embolism after hip and knee arthroplasty: A population-
292
based study. Int Orthop 2009;33:1739–45. doi:10.1007/s00264-008-0659-z.
293 294 295
AC C
291
[4]
Can Be Used as Prophylaxis for Prevention of Venous Thromboembolism Following Revision Hip and Knee Arthroplasty. J Arthroplasty 2016;31:11–4.
296 297
Deirmengian GK, Heller S, Smith EB, Maltenfort M, Chen AF, Parvizi J. Aspirin
doi:10.1016/j.arth.2016.03.031. [5]
Jaffer AK, Barsoum WK, Krebs V, Hurbanek JG, Morra N, Brotman DJ. Duration
ACCEPTED MANUSCRIPT VTE following Revision THA
14
298
of anesthesia and venous thromboembolism after hip and knee arthroplasty. Mayo
299
Clin Proc 2005;80:732–8. doi:10.1016/S0025-6196(11)61526-7.
300
[6]
Nam D, Nunley RM, Johnson SR, Keeney JA, Clohisy JC, Barrack RL. Thromboembolism Prophylaxis in Hip Arthroplasty: Routine and High Risk
302
Patients. J Arthroplasty 2015;30:2299–303. doi:10.1016/j.arth.2015.06.045.
303
[7]
RI PT
301
Forster R, Stewart M. Anticoagulants (extended duration) for prevention of venous thromboembolism following total hip or knee replacement or hip fracture repair.
305
Cochrane Database Syst Rev 2016. doi:10.1002/14651858.CD004179.pub2. [8]
Eikelboom JW, Karthikeyan G, Fagel N, Hirsh J. American Association of
M AN U
306
SC
304
Orthopedic Surgeons and American College of Chest Physicians guidelines for
308
venous thromboembolism prevention in hip and knee arthroplasty differ: what are
309
the implications for clinicians and patients? Chest 2009;135:513–20.
310
doi:10.1378/chest.08-2655.
311
[9]
TE D
307
Kester BS, Merkow RP, Ju MH, Peabody TD, Bentrem DJ, Ko CY, et al. Effect of post-discharge venous thromboembolism on hospital quality comparisons
313
following hip and knee arthroplasty. J Bone Joint Surg Am 2014;96:1476–84.
314
doi:10.2106/JBJS.M.01248.
316 317 318 319 320
[10]
Nam D, Nunley RM, Johnson SR, Keeney JA, Clohisy JC, Barrack RL. The
AC C
315
EP
312
Effectiveness of a Risk Stratification Protocol for Thromboembolism Prophylaxis
After Hip and Knee Arthroplasty. J Arthroplasty 2016;31:1299–306. doi:10.1016/j.arth.2015.12.007.
[11]
Wakabayashi H, Hasegawa M, Niimi R, Sudo A. Clinical analysis of preoperative deep vein thrombosis risk factors in patients undergoing total hip arthroplasty.
ACCEPTED MANUSCRIPT VTE following Revision THA 321 322
15
Thromb Res 2015;136:855–8. doi:10.1016/j.thromres.2015.06.021. [12]
Burnett RSJ, Clohisy JC, Wright RW, McDonald DJ, Shively RA, Givens SA, et al. Failure of the American College of Chest Physicians-1A Protocol for Lovenox
324
in Clinical Outcomes for Thromboembolic Prophylaxis. J Arthroplasty
325
2007;22:317–24. doi:10.1016/j.arth.2007.01.007. [13]
327 328
Mont MA, Jacobs JJ. Preventing Venous Thromboembolic Disease in Patients Undergoing Elective Hip and Knee Arthroplasty Guideline. vol. 19. 2011.
[14]
SC
326
RI PT
323
Barrack RL, Burnett RSJ. Deep Vein Thrombosis Prophylaxis: Protecting the Patient or the Surgeon? Semin Arthroplasty 2008;19:109–11.
330
doi:10.1053/j.sart.2007.12.013.
331
[15]
332
M AN U
329
American College of Surgeons. ACS National Surgical Quality Improvement Program (ACS NSQIP) n.d.
[16]
American College of Surgeons. NSQIP participant use data file 2014.
334
[17]
American College of Surgeons. Data collection, analysis, and reporting n.d.
EP AC C
335
TE D
333
ACCEPTED MANUSCRIPT
Figure 1. A DVT risk score predicting probability of having a DVT within 30 days of total hip arthroplasty (p<0.001)
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. A PE risk score predicting probability of having a PE within 30 days of total hip arthroplasty (p<0.001)
ACCEPTED MANUSCRIPT
EP
AC C
M AN U
SC
59034 (79) 5002 (7) 1060 (1.4) 7566 (10) 4865 (6.5) 1663 (2.2) 1038 (1.4) 8691 (12) 1840 (3) 2490 (3) 4618 (3) 407 (0.5) 80 (0.1) 2130 (2.9) 3440 (5)
RI PT
Number 64.9 (11.7) 29.8 (7.0) 33058 (44)
TE D
Demographic Data (N=74,405) Mean age (years) (SD) Mean BMI (kg/m^2) (SD) Male Gender (%) Ethnicity (%) White African American Asian Revision Arthroplasty (%) Both components Acetabulum only Femur only Diabetes Mellitus (%) Smoking History (%) Malnourished (%) Kidney Disease (%) Cardiac Disease (%) History of Stroke (%) Kidney Disease (%) Low Preoperative Hematocrit (%) ASA Classification (%) I II III IV V Mean Operative Time (minutes) (SD) Operative Time > 3 hours (%) Return to Operating Room (%) Readmission within 30 days (%) Mortality (%) Any Complication (%) Deep Venous Thrombosis (%) Pulmonary Embolism (%) Mean time to PE (days) (SD) Mean time to DVT (SD)
3039 (4) 39752 (53) 29751 (40) 1773 (2) 5 (<0.1) 99.8 (49.9) 4195 (6) 268 (0.4) 452 (0.6) 220 (0.3) 14200 (19) 324 (0.4) 208 (0.3) 13.2 (8.5) 9.1 (9.1)
Table 1. Descriptive statistics and demographic data on the entire patient population
ACCEPTED MANUSCRIPT
Revision THA N=7,566 65.8 (12.5) 29.1 (8.0) 3383 (45)
P value
53007 (79) 4502 (7) 978 (1) 7706 (12) 37261 (56) 1574 (2) 1762 (3) 1813 (3) 344 (0.5) 64 (0.1) 1813 (3) 2467 (4)
6027 (80) 500 (7) 82 (1) 985 (13) 5464 (72) 266 (4) 728 (10) 317 (4) 63 (0.8) 16 (0.2) 317 (4) 973 (13)
0.028
2887 (4) 36585 (55) 25909 (39) 1373 (2) 4 (<0.1) 94.4 (42) 2233 (3) 189 (0.3) 356 (0.5) 166 (0.2) 11411 (17) 278 (0.4) 180 (0.2) 9.2 (9.1) 12.9 (8.5)
AC C
EP
TE D
<0.001 <0.001 0.601
RI PT
SC
Mean age (years) (SD) Mean BMI (kg/m^2) (SD) Male Gender (%) Ethnicity (%) White African American Asian Diabetes Mellitus (%) General Anesthesia (%) Smoking History (%) Malnourished (%) Kidney Disease (%) Cardiac Disease (%) History of Stroke (%) Kidney Disease (%) Low Preoperative Hematocrit (%) ASA Classification (%) I II III IV V Mean Operative Time (minutes) (SD) Operative Time > 3 hours (%) Return to Operating Room (%) Readmission within 30 days (%) Mortality (%) Any Complication (%) Deep Venous Thrombosis (%) Pulmonary Embolism (%) Mean time to PE (days) (SD) Mean time to DVT (days) (SD)
Primary THA N=66,839 64.7 (11.6) 29.8 (6.9) 29675 (44)
M AN U
Demographic Data
152 (2) 3167 (420 3842 (51) 400 (5) 1 (<0.1) 147.0 (79) 1962 (26) 79 (1) 96 (1) 54 (0.7) 2789 (37) 46 (0.6) 28 (0.3) 8.6 (9.3) 14.8 (7.9)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.004 <0.001 <0.001 <0.001
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.016 0.116 0.215 0.890
Table 2. Comparison of patient characteristics between the primary THA and revision THA groups.
ACCEPTED MANUSCRIPT
RI PT
95% Confidence Interval 0.665-1.051 0.731-1.675 0.564-1.232 1.157-1.829 1.030-1.654 0.864-1.502 0.749-1.440 0.759-2.581 0.225-3.805 0.995-2.660 0.905-2.315 2.180-7.951 1.179-2.553 0.889-2.080 1.056-2.848
SC
Odds Ratio 0.836 1.107 0.833 1.455 1.305 1.139 1.039 1.400 0.925 1.626 1.447 4.163 1.735 1.360 1.734
M AN U
Risk Factor Male Gender African American Ethnicity Revision Arthroplasty Age > 70 years General Anesthesia BMI > 35 kg/m2 Diabetes Mellitus Smoking History Cardiac Disease Kidney Disease Malnutrition Infection Operating Time > 3 hours Low Preoperative Hematocrit ASA IV or greater
P value 0.125 0.632 0.360 0.001 0.028 0.355 0.819 0.281 0.913 0.053 0.123 <0.001 0.005 0.156 0.030
AC C
EP
TE D
Table 3. Multivariate logistic regression analysis identifying independent risk factors for deep venous thrombosis within 30 days of hip arthroplasty.
ACCEPTED MANUSCRIPT
95% Confidence Interval 0.702-1.253 0.988-2.534 0.630-1.616 1.271-2.277 0.887-1.606 0.686-1.435 0.806-1.817 0.592-3.022 0.450-2.137 0.785-2.767 0.379-4.337 1.124-3.026 0.817-2.484 0.312-1.934
RI PT
Odds Ratio 0.938 1.582 1.009 1.701 1.193 0.993 1.210 1.337 0.981 1.474 1.281 1.845 1.424 0.777
SC
Risk Factor Male Gender African American Ethnicity Revision Arthroplasty Age > 70 years General Anesthesia BMI > 35 kg/m2 Diabetes Mellitus Smoking History Kidney Disease Malnutrition Infection Operating Time > 3 hours Low Preoperative Hematocrit ASA IV or greater
AC C
EP
TE D
M AN U
Table 4. Multivariate logistic regression analysis identifying independent risk factors for pulmonary embolism within 30 days of hip arthroplasty.
P value 0.665 0.056 0.971 <0.001 0.244 0.968 0.357 0.485 0.961 0.227 0.690 0.015 0.213 0.587
ACCEPTED MANUSCRIPT
Odds Ratio 1.452 1.314 1.651 1.571 3.818 1.668 1.782 1.527
95% Confidence Interval 1.162-1.816 1.038-1.664 1.021-2.672 1.003-2.461 2.132-6.837 1.155-2.407 1.089-2.915 1.396-1.671
P value 0.001 0.023 0.041 0.048 <0.001 0.006 0.021 <0.001
PE Risk Score 1 1 1 3
Odds Ratio 1.744 1.657 2.040 1.802
95% Confidence Interval 1.314-2.315 1.038-2.646 1.294-3.216 1.448-2.241
P value <0.001 0.034 0.002 <0.001
RI PT
Risk Factor for PE Age > 70 years African American Ethnicity Operating Time > 3 hours TOTAL
DVT Risk Score 1 1 1 1 3 1 1 9
SC
Risk Factor for DVT Age > 70 years General Anesthesia Kidney Disease Malnutrition Infection Operating Time > 3 hours ASA IV or greater TOTAL
AC C
EP
TE D
M AN U
Table 5. Forward, stepwise multivariate logistic regression analysis identifying simple risk score for VTE events after total hip arthroplasty.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0883-5403(17)30644-7
DOI:
10.1016/j.arth.2017.07.028
Reference:
YARTH 56008
To appear in:
The Journal of Arthroplasty
Received Date: 29 April 2017 Revised Date:
29 June 2017
Accepted Date: 17 July 2017
Please cite this article as: Courtney PM, Boniello AJ, Levine BR, Sheth NP, Paprosky WG, Are revision hip arthroplasty patients at higher risk for venous thromboembolic events than primary hip arthroplasty patients?, The Journal of Arthroplasty (2017), doi: 10.1016/j.arth.2017.07.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Are revision hip arthroplasty patients at higher risk for venous thromboembolic events than primary hip arthroplasty patients?
RI PT
P. Maxwell Courtney MD,1 Anthony J. Boniello MD,2 Brett R. Levine MD, MS,3 Neil P. Sheth MD,4 Wayne G. Paprosky MD3
Study Conducted at Rush University Medical Center, Chicago, IL 1
M AN U
SC
Rothman Institute, Thomas Jefferson University Hospital, Department of Orthopaedic Surgery 2 Drexel University College of Medicine, Department of Orthopaedic Surgery 3 Rush University Medical Center, Department of Orthopaedic Surgery 4 University of Pennsylvania School of Medicine, Department of Orthopaedic Surgery
Source of Funding Each author certifies that he or she, or a member of his or her immediate family, has no funding or commercial associations (eg, consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.
AC C
EP
TE D
Disclosure The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
Corresponding author: P. Maxwell Courtney, MD Rothman Institute, Thomas Jefferson University Hospital Department of Orthopaedic Surgery 925 Chestnut St, 5th Floor Philadelphia, PA 19107 Email: [email protected]
ACCEPTED MANUSCRIPT VTE following Revision THA
1
1
Are revision hip arthroplasty patients at higher risk for venous thromboembolic
2
events than primary hip arthroplasty patients?
3 Abstract:
5
Introduction: The purpose of this study was to determine whether revision total hip
6
arthroplasty (THA) is associated with increased rates of deep venous thrombosis (DVT)
7
and pulmonary embolism (PE) when compared to primary THA.
8
Methods: We queried the American College of Surgeons National Surgical Quality
9
Improvement Program database for all primary and revision THA cases from 2011-2014.
10
Demographic data, medical comorbidities, and venous thromboembolic (VTE) rates
11
within 30 days of surgery were compared between the primary and revision THA groups.
12
Results: Revision THA had a higher rate of DVT than the primary THA (0.6% vs. 0.4%,
13
p=0.016), but there was no difference in rate of PE (0.3% vs. 0.2%, p=0.116). When
14
controlling for confounding variables, revision surgery alone was not a risk factor for
15
DVT (OR 0.833, 95% CI 0.564-1.232) or PE (OR 1.009, 95% CI 0.630-1.616).
16
Independent risk factors for DVT include age > 70 years, malnutrition, infection,
17
operating time > 3 hours, general anesthesia, ASA classification 4 or greater, and kidney
18
disease (all p<0.05). Probability of DVT ranged from 0.2% with zero risk factors to 10%
19
with all risk factors. Independent risk factors for PE included age > 70 years, African
20
American ethnicity, and operating time > 3 hours (all p<0.05) with probabilities of PE
21
postoperatively ranging from 0.2% to 1.1% with all risk factors.
AC C
EP
TE D
M AN U
SC
RI PT
4
ACCEPTED MANUSCRIPT VTE following Revision THA
2
22
Conclusion: Revision surgery alone is not a risk factor for VTE after hip arthroplasty.
23
Surgeons should weigh the risks and benefits of more aggressive anticoagulation in
24
certain high-risk patients.
RI PT
25 26
Keywords: Total hip arthroplasty; complications; revision arthroplasty; pulmonary
27
embolism, deep vein thrombosis
SC
28 29
M AN U
30 31 32 33
37 38 39 40 41 42 43 44
EP
36
AC C
35
TE D
34
ACCEPTED MANUSCRIPT VTE following Revision THA
3
45
Are revision hip arthroplasty patients at higher risk for venous thromboembolic
46
events than primary hip arthroplasty patients?
47 Introduction
RI PT
48
Venous thromboembolic events (VTE), including deep vein thrombosis (DVT)
50
and pulmonary embolism (PE), remain one of the most common complications following
51
total hip arthroplasty (THA).[1] While the clinical significance of lower extremity DVT
52
has been contested, recent studies report the rate of DVT following THA to be around
53
8%,[2] whereas the incidence of in-hospital PE has been reported to be 0.36%.[3] There
54
currently exists a great variability in the exact pharmacologic agents utilized,[1,4–7]
55
however the American Academy of Orthopaedic Surgeons (AAOS) and American
56
College of Chest Physicians (ACCP) recommend that some form of VTE prophylaxis be
57
administered routinely following THA.[8] While rates and risk factors for VTE after
58
primary THA are well studied in the literature,[1,2,5–7,9–11] little has been written about
59
DVT and PE following revision THA. Clinical concerns exist about an increased rate of
60
VTE following revision surgery due to increased surgical exposure, protected weight
61
bearing postoperatively, and poor mobilization postoperatively.
M AN U
TE D
EP
Historically there exist conflicting philosophies between the AAOS and ACCP
AC C
62
SC
49
63
regarding the necessity of and overall safety profile of pharmacoprophylactic
64
regimens.[12] The AAOS has recently raised concerns of postoperative bleeding,
65
hematoma, infection, and potential reoperation with the use of more potent
66
thromboprophylactic medications such as low molecular weight heparin.[13,14] More
67
aggressive anticoagulation in revision patients could lead to poorer outcomes, as revision
ACCEPTED MANUSCRIPT VTE following Revision THA
4
THA is inherently associated with increased rates of bleeding and increased
69
complications.[9] In 2007, the AAOS developed its first clinical practice guideline,
70
focusing on prevention of symptomatic VTE events and limiting perioperative morbidity
71
caused by anticoagulants. As part of this guideline the AAOS recommended risk
72
stratification for VTE events.[13,14] Such stratification may enable physicians to
73
prescribe better-tolerated pharmacoprophylactic regimens with lower adverse affects, and
74
better safety profiles to lower risk individuals.[6] Identifying at-risk patients for VTE
75
events following revision THA remains difficult and has yet to be addressed in the
76
literature.
M AN U
SC
RI PT
68
The purpose of this study is to determine 1) Is revision THA associated with
78
higher rates of postoperative venous thromboembolic events than primary THA?
79
Secondary study questions included 2) What are the independent risk factors for VTE
80
after THA to identify patients who may benefit from more aggressive anticoagulation and
81
3) Can a simple risk score be used to predict the probability of DVT and PE after THA?
82 Methods
EP
83
TE D
77
We retrospectively queried the American College of Surgeons-National Surgical
85
Quality Improvement Program (ACS-NSQIP) database for all patients who underwent
86
primary or revision total hip arthroplasty from January 1, 2011 to December 31, 2014.
87
Patients in the database were identified based on the primary procedure consisting of
88
Current Procedural Terminology (CPT) codes for elective primary THA (27130) and
89
revision THA (27134, 27137, and 27138). Emergent cases were excluded. Patients
90
undergoing hip arthroplasty with other procedures listed as the primary surgery were also
AC C
84
ACCEPTED MANUSCRIPT VTE following Revision THA
5
91
excluded from the study. This study was exempt from Institutional Review Board (IRB)
92
approval as all data was de-identified. No external funding was received for this study.
93 The NSQIP database is a validated, national database including cases from over 650
95
hospitals.[15] Demographic variables, medical comorbidities, operative time, deep
96
venous thrombosis, and pulmonary embolism rates were noted from the database. The
97
International Classification of Diseases, 9th edition (ICD-9) code for primary diagnosis
98
was also documented. Deep venous thrombosis was defined as a “new diagnosis of blood
99
clot or thrombus within the venous system (superficial or deep) which may be coupled
100
with inflammation and requires treatment.”[16] Pulmonary embolism was defined as “A.
101
New diagnosis of a new blood clot in a pulmonary artery AND B. The patient has a V-Q
102
scan interpreted as high probability of pulmonary embolism or a positive CT exam, TEE,
103
pulmonary arteriogram, CT angiogram, or any other definitive imaging modality
104
(including direct pathology examination such as autopsy).”[17] Days from surgery to the
105
VTE were also recorded. Complications, reoperations, and readmissions within 30 days
106
of surgery were noted. Complete NSQIP methodology has been reported previously in
107
the literature.13
SC
M AN U
TE D
EP
We defined malnutrition as any patient with a preoperative albumin less than 3.5
AC C
108
RI PT
94
109
g/dL and preoperative kidney disease as any patient with a preoperative creatinine greater
110
than 1.5 mg/L. Patients requiring prior coronary intervention were classified as have a
111
history of cardiac disease. Low hematocrit was defined by a preoperative value of less
112
than 33%. Complications, reoperations, and readmissions within 30 days of surgery were
113
noted. We defined complications from the database within 30 days of surgery as any
ACCEPTED MANUSCRIPT VTE following Revision THA
6
patient having a recorded surgical site infection, pneumonia, respiratory complication
115
requiring reintubation, pulmonary embolism, deep venous thrombosis, renal insufficiency
116
or failure, urinary tract infection, stroke, cardiac arrest, bleeding requiring transfusion,
117
sepsis or septic shock. Specific definitions are found in the NSQIP participant use data
118
file.[16]
RI PT
114
119 Statistical Analysis
SC
120
An a priori power analysis was first conducted to determine the appropriate
122
sample size. Our primary outcome variable was whether revision THA was associated
123
with a higher 30-day DVT and PE rate than primary THA. Based upon prior data
124
suggesting a 1.5% VTE rate among revision hip arthroplasty,[4] in order to detect a 0.5%
125
difference in the incidence of DVT and PE, we would need to enroll a total of 70,528
126
patients assuming a type I error rate of 0.05 to achieve a power of 0.80.
TE D
M AN U
121
Data analysis was first performed comparing patients undergoing revision TJA
128
with those who had a primary THA. Categorical variables were analyzed using chi-
129
square analyses. Continuous variables such as age and BMI were analyzed using
130
unpaired, two-tailed, Student’s t-tests. Statistical significance was set at p < 0.05.
131
Multivariate regression was then performed to identify independent risk factors for DVT
132
and PE. A forward, stepwise, multiple logistic regression analysis was then performed to
133
create a model to identify high-risk risk factors for DVT and PE. We started with the
134
most significant variable and added risk factors in order of ascending p-values. The
135
model was rechecked after each addition and any alpha > 0.10 was removed from the
136
analysis. In order to design this model for practical use, we developed a simplified score
AC C
EP
127
ACCEPTED MANUSCRIPT VTE following Revision THA
7
137
weighted by the coefficients and odds ratios of each independent variable. Statistical
138
analysis was performed using Microsoft Excel (Bellvue, WA) and IBM SPSS version
139
24.0 (Armonk, NY).
141
RI PT
140 Results
Of the 74,405 patients included in the study, there were 7,566 revision THA
143
procedures (10%). The mean age for all patients was 64.9 years (Standard deviation [SD]
144
1.7), while the mean body mass index (BMI) was 29.8 kg/m2 (SD 7.0). There were
145
33,058 males (44%) with the majority of all patients being of Caucasian ethnicity (59,034
146
patients, 79%). There were 324 patients (0.4%) who experienced a DVT within 30 days,
147
while 208 patients (0.3%) had a PE. The mean time from surgery to DVT was 9.1 days
148
(SD 9.1) and to PE was 13.2 days (SD 8.5). Complete descriptive statistics and
149
demographic data of the study population are detailed in Table 1.
TE D
M AN U
SC
142
Revision THA had a higher incidence of DVT than primary THA (0.6% vs. 0.4%,
151
p=0.016), but there was no statistical difference in the rate of PE (0.3% vs. 0.2%,
152
p=0.116). There was no difference in the mean time to DVT (12.9 vs. 14.8 days,
153
p=0.890) or PE (9.2 vs. 8.6 days, p=0.215). Among all revision patients, there was no
154
difference in DVT rate between the both component revisions, acetabular component
155
only, and femoral component only groups (0.7% vs. 0.4% vs. 0.3%, p=0.128). Similarly,
156
there was no difference in the rate of PE between the types of revision groups (0.4% vs.
157
0.3% vs. 0.6%, p=0.474). The revision group was older (mean age 65.8 years vs. 64.7
158
years, p<0.001) and had more medical comorbidities than the primary group, including
159
diabetes, kidney disease, cardiac disease, and higher ASA classification (all p<0.001).
AC C
EP
150
ACCEPTED MANUSCRIPT VTE following Revision THA
8
160
Mortality (0.7% vs. 0.2%, p<0.001) and complication rates (37% vs. 17%, p<0.001) were
161
higher in revision THA as well. Data comparing the revision and primary THA groups is
162
shown in Table 2. Multivariate analysis identified independent risk factors for DVT as age over 70
164
years (OR 1.455, 95% CI 1.157-1.829, p=0.001), general anesthesia (OR 1.305, 95% CI
165
1.030-1.654, p=0.028), arthroplasty procedure for infection (OR 4.163, 95% CI 2.180-
166
7.951, p<0.001), operating time over three hours (OR 1.735, 95% CI 1.179-2.553,
167
p=0.005), and American Society of Anesthesiologists (ASA) Class IV or greater (OR
168
1.734, 95% CI 1.056-2.848, p=0.030). Independent risk factors for PE included age over
169
70 years (OR 1.701, 95% CI 1.271-2.277, p<0.001) and operating time over three hours
170
(OR 1.845, 95% CI 1.124-3.026, p=0.015). Revision surgery was not an independent risk
171
factor for DVT (OR 0.833, 95% CI 0.564-1.232, p=0.360) or PE (OR 1.009, 95% 0.630-
172
1.616, p=0.971). Complete regression analysis is listed in Table 3.
TE D
M AN U
SC
RI PT
163
Based upon the coefficients of the forward, stepwise multivariate logistic
174
regression analysis, we created a weighted risk score to calculate the probability of DVT
175
after THA. A statistically significant risk factor’s weighted risk score was based on the
176
odds ratio and assigned a corresponding ordinal value to generate a simple, easy to use
177
score which is more practical or surgeons estimating VTE risk following THA. Risk
178
factors included age over 70 years, general anesthesia, kidney disease, malnutrition,
179
operating time over three hours, ASA IV classification or greater (all one point), and
180
infection (three points). The risk score was significantly associated with probability of
181
DVT after THA (OR 1.527, 95% CI 1.396-1.671, p<0.001). Probability of DVT ranged
182
from 0.2% with zero risk factors to 10% with all risk factors (Figure 1). Multivariate risk
AC C
EP
173
ACCEPTED MANUSCRIPT VTE following Revision THA
9
factors for PE included Age (should this be caps?) > 70 years, African American
184
ethnicity, and operating time > three hours (all one point) and the score was significantly
185
associated with the probability of PE after THA (OR 1.802, 95% CI 1.448-2.241,
186
p<0.001). The probability of PE postoperatively ranges from 0.2% with zero risk factors
187
to 1.1% with all risk factors (Figure 2).
188 Discussion
SC
189
RI PT
183
While rates and risk factors for VTE after primary hip arthroplasty are well
191
studied in the literature,[1,2,5–7,9–11] little has been written about DVT and PE
192
following revision THA. Clinical concerns exist about an increased rate of VTE
193
following revision surgery due to increased surgical exposure, protected weight bearing
194
postoperatively, and poor mobilization postoperatively. However, more aggressive
195
anticoagulation in revision patients could lead to poorer outcomes, as revision THA is
196
inherently associated with increased rates of bleeding and increased complications.[9]
197
We present a large series from a validated, multi-center, national database suggesting that
198
revision surgery alone is not a risk factor for VTE following total hip arthroplasty.
EP
TE D
M AN U
190
While there are limitations to data drawn from large databases, the NSQIP is a
200
validated, accurate national database. NSQIP collects large patient numbers from over
201
650 hospitals nationwide, which helps make findings generalizable to most physicians
202
across the country considering instituting an outpatient TJA protocol. The NSQIP
203
database is stronger than data derived from billing codes in other administrative databases.
204
Data collectors are certified each year, and random audits have found rates of discrepancy
205
less around 2% within each site.[17] One such limitation is in the availability of baseline
AC C
199
ACCEPTED MANUSCRIPT VTE following Revision THA
10
comorbidities. Information regarding history of DVT and PE was unavailable. In the
207
current paper “cardiac disease” was characterized by a history of percutaneous coronary
208
intervention. This excludes a variety of clinically significant cardiac diseases that have
209
been shown to effect outcomes.
RI PT
206
The NSQIP database only includes complications, reoperation and readmission
211
information for 30-days, which may under-report complications when compared to 90-
212
day follow up. Additionally, the database lacks orthopaedic specific complications
213
including dislocation, fracture, and poor functional outcome scores. The database did not
214
track which VTE prophylaxis was administered to each patient. As some data suggests a
215
trend toward less aggressive anticoagulation,[4,7] the variability in chemoprophylaxis
216
could affect VTE rates. Weight-bearing status following revision surgery was also not
217
reported. Patients with restricted weight bearing may be at risk for higher VTE rates
218
because of prolonged immobilization. There is no differentiation regarding diagnostic
219
modalities for DVT in the NSQIP database, therefore only clinically symptomatic DVTs
220
and DVTs identified by ultrasound were included in the same group. In the absence of
221
routine ultrasound screening, asymptomatic DVTs were likely overwhelmingly
222
underreported. While we tracked types of revisions by CPT code, there could still be
223
variability in the surgical procedures ranging from a simple head and liner exchange to a
224
more complex both component revision requiring an extended trochanteric osteotomy.
225
Caution should also be made when interpreting certain differences between the primary
226
and revision groups such as age, which is statistically different but not clinically different.
227
Our study found a higher prevalence of DVT in revision versus primary THA, but
228
no difference in the rate of PE. While a higher prevalence of DVT existed in revision
AC C
EP
TE D
M AN U
SC
210
ACCEPTED MANUSCRIPT VTE following Revision THA
11
THA, the most significant finding of the present study is that revision THA is not an
230
independent risk factor for DVT and PE. These findings agree with previous finding by
231
Wakabayashi et al, which concluded that the prevalence of DVT was higher in the
232
revision THA than for the primary cases, however following a multivariate analysis,
233
revision THA was not an independent risk factor.[11] These findings suggest that
234
confounding variables exist such as increased age, comorbidities, and increased operative
235
time that are associated with revision surgery. While revision THA is not itself an
236
independent risk factor, many of these confounding variables associated with revision
237
THA were found to increase the risk of DVT, and must be considered when selecting,
238
counseling and treating this patient population.
239
arthroplasty or higher risk primary arthroplasty patients may have chosen more
240
aggressive chemoprophylaxis, which could explain the comparable VTE rates. We also
241
did not find any difference in DVT or PE rate based on type of revision surgery (both
242
component, acetabular only, femoral only). There is a lack of consensuses in the current
243
literature regarding optimal DVT prophylaxis following THA.[1,2,5–7,9–11] While the
244
optimal prophylaxis for revision surgery is not clear, one study by Deirmengian et al did
245
not find a difference in incidence of VTE between aspirin and warfarin following
246
revision THA.[4] Patients undergoing revision THA without further risk factors for VTE
247
may benefit from less aggressive anticoagulation without being at increased risk for DVT
248
and PE.
SC
M AN U
EP
TE D
Surgeons also treating revision
AC C
249
RI PT
229
The present study found that independent risk factors for DVT were age over 70
250
years, general anesthesia, arthroplasty procedure for infection, operating time over three
251
hours, and ASA IV or greater. Furthermore, independent risk factors for PE included age
ACCEPTED MANUSCRIPT VTE following Revision THA
12
over 70 years, and operating time over three hours. These results are consistent with
253
previous studies. One such national study by Dua et al found that patients with diagnosed
254
DVT following primary TJA were older, more likely to be African American and more
255
likely to have significant comorbidities including congestive heart failure, peripheral
256
artery disease, and end-stage renal disease.[1] Another previous study found an
257
association with anesthesia time with an increased risk of DVT.[5]
RI PT
252
We present an easy to use risk score for VTE events following revision THA. An
259
optimal cutoff has not been established and will depend on each patient and surgeon’s
260
tolerance for risk.
261
anticoagulation have been shown to be effective in primary THA and TKA.13 Our study
262
identifies patients that are at-risk for DVT and PE that may benefit from stronger
263
anticoagulation. The primary limitation of the current VTE risk score presented is that it
264
does not account for differences in VTE prophylaxis protocols in the patients studied.
265
Historically surgeons have treated revisions as high risk cases, and therefore may have
266
prophylactically treated such patients with more aggressive forms of DVT prophylaxis.
267
One notable risk factor missing from the scoring system is history of DVT or PE, as this
268
data was not available in the NSQIP database. Despite this limitation the the present
269
authors conclude that while aggressive anticoagulation is associated with increased
270
adverse events following revision THA, surgeons should consider stratifying their
271
revision THA patients when considering pharmacoprophylaxis against VTE.
272
prospective studies are needed to validate the VTE risk score and subsequent treatment
273
recommendations.
274
SC
258
AC C
EP
TE D
M AN U
Stratification protocols to prevent complications associated with
Further
ACCEPTED MANUSCRIPT VTE following Revision THA 275
13
Conclusion
276
Revision THA alone is not an independent risk factor for DVT and PE when
277
compared to primary hip arthroplasty.
278
operating time > 3 hours, and age > 70 years are at higher risk for VTE events. Surgeons
279
should weigh the risks and benefits of more aggressive anticoagulation in these high-risk
280
patients. Further study is needed to determine the effectiveness of risk stratification
281
protocols for thromboprophylaxis following revision THA.
282
References
[1]
RI PT
SC
M AN U
283 284
Patients undergoing a THA for infection,
Dua A, Desai SS, Lee CJ, Heller JA. National Trends in Deep Vein Thrombosis
285
following Total Knee and Total Hip Replacement in the United States. Ann Vasc
286
Surg 2016. doi:10.1016/j.avsg.2016.05.110. [2]
Wong KL, Daguman R, Lim KH, Shen L, Lingaraj K. Incidence of deep vein
TE D
287 288
thrombosis following total hip arthroplasty: a Doppler ultrasonographic study. J
289
Orthop Surg (Hong Kong) 2011;19:50–3. [3]
Memtsoudis SG, Besculides MC, Gaber L, Liu S, González Della Valle A. Risk
EP
290
factors for pulmonary embolism after hip and knee arthroplasty: A population-
292
based study. Int Orthop 2009;33:1739–45. doi:10.1007/s00264-008-0659-z.
293 294 295
AC C
291
[4]
Can Be Used as Prophylaxis for Prevention of Venous Thromboembolism Following Revision Hip and Knee Arthroplasty. J Arthroplasty 2016;31:11–4.
296 297
Deirmengian GK, Heller S, Smith EB, Maltenfort M, Chen AF, Parvizi J. Aspirin
doi:10.1016/j.arth.2016.03.031. [5]
Jaffer AK, Barsoum WK, Krebs V, Hurbanek JG, Morra N, Brotman DJ. Duration
ACCEPTED MANUSCRIPT VTE following Revision THA
14
298
of anesthesia and venous thromboembolism after hip and knee arthroplasty. Mayo
299
Clin Proc 2005;80:732–8. doi:10.1016/S0025-6196(11)61526-7.
300
[6]
Nam D, Nunley RM, Johnson SR, Keeney JA, Clohisy JC, Barrack RL. Thromboembolism Prophylaxis in Hip Arthroplasty: Routine and High Risk
302
Patients. J Arthroplasty 2015;30:2299–303. doi:10.1016/j.arth.2015.06.045.
303
[7]
RI PT
301
Forster R, Stewart M. Anticoagulants (extended duration) for prevention of venous thromboembolism following total hip or knee replacement or hip fracture repair.
305
Cochrane Database Syst Rev 2016. doi:10.1002/14651858.CD004179.pub2. [8]
Eikelboom JW, Karthikeyan G, Fagel N, Hirsh J. American Association of
M AN U
306
SC
304
Orthopedic Surgeons and American College of Chest Physicians guidelines for
308
venous thromboembolism prevention in hip and knee arthroplasty differ: what are
309
the implications for clinicians and patients? Chest 2009;135:513–20.
310
doi:10.1378/chest.08-2655.
311
[9]
TE D
307
Kester BS, Merkow RP, Ju MH, Peabody TD, Bentrem DJ, Ko CY, et al. Effect of post-discharge venous thromboembolism on hospital quality comparisons
313
following hip and knee arthroplasty. J Bone Joint Surg Am 2014;96:1476–84.
314
doi:10.2106/JBJS.M.01248.
316 317 318 319 320
[10]
Nam D, Nunley RM, Johnson SR, Keeney JA, Clohisy JC, Barrack RL. The
AC C
315
EP
312
Effectiveness of a Risk Stratification Protocol for Thromboembolism Prophylaxis
After Hip and Knee Arthroplasty. J Arthroplasty 2016;31:1299–306. doi:10.1016/j.arth.2015.12.007.
[11]
Wakabayashi H, Hasegawa M, Niimi R, Sudo A. Clinical analysis of preoperative deep vein thrombosis risk factors in patients undergoing total hip arthroplasty.
ACCEPTED MANUSCRIPT VTE following Revision THA 321 322
15
Thromb Res 2015;136:855–8. doi:10.1016/j.thromres.2015.06.021. [12]
Burnett RSJ, Clohisy JC, Wright RW, McDonald DJ, Shively RA, Givens SA, et al. Failure of the American College of Chest Physicians-1A Protocol for Lovenox
324
in Clinical Outcomes for Thromboembolic Prophylaxis. J Arthroplasty
325
2007;22:317–24. doi:10.1016/j.arth.2007.01.007. [13]
327 328
Mont MA, Jacobs JJ. Preventing Venous Thromboembolic Disease in Patients Undergoing Elective Hip and Knee Arthroplasty Guideline. vol. 19. 2011.
[14]
SC
326
RI PT
323
Barrack RL, Burnett RSJ. Deep Vein Thrombosis Prophylaxis: Protecting the Patient or the Surgeon? Semin Arthroplasty 2008;19:109–11.
330
doi:10.1053/j.sart.2007.12.013.
331
[15]
332
M AN U
329
American College of Surgeons. ACS National Surgical Quality Improvement Program (ACS NSQIP) n.d.
[16]
American College of Surgeons. NSQIP participant use data file 2014.
334
[17]
American College of Surgeons. Data collection, analysis, and reporting n.d.
EP AC C
335
TE D
333
ACCEPTED MANUSCRIPT
Figure 1. A DVT risk score predicting probability of having a DVT within 30 days of total hip arthroplasty (p<0.001)
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. A PE risk score predicting probability of having a PE within 30 days of total hip arthroplasty (p<0.001)
ACCEPTED MANUSCRIPT
EP
AC C
M AN U
SC
59034 (79) 5002 (7) 1060 (1.4) 7566 (10) 4865 (6.5) 1663 (2.2) 1038 (1.4) 8691 (12) 1840 (3) 2490 (3) 4618 (3) 407 (0.5) 80 (0.1) 2130 (2.9) 3440 (5)
RI PT
Number 64.9 (11.7) 29.8 (7.0) 33058 (44)
TE D
Demographic Data (N=74,405) Mean age (years) (SD) Mean BMI (kg/m^2) (SD) Male Gender (%) Ethnicity (%) White African American Asian Revision Arthroplasty (%) Both components Acetabulum only Femur only Diabetes Mellitus (%) Smoking History (%) Malnourished (%) Kidney Disease (%) Cardiac Disease (%) History of Stroke (%) Kidney Disease (%) Low Preoperative Hematocrit (%) ASA Classification (%) I II III IV V Mean Operative Time (minutes) (SD) Operative Time > 3 hours (%) Return to Operating Room (%) Readmission within 30 days (%) Mortality (%) Any Complication (%) Deep Venous Thrombosis (%) Pulmonary Embolism (%) Mean time to PE (days) (SD) Mean time to DVT (SD)
3039 (4) 39752 (53) 29751 (40) 1773 (2) 5 (<0.1) 99.8 (49.9) 4195 (6) 268 (0.4) 452 (0.6) 220 (0.3) 14200 (19) 324 (0.4) 208 (0.3) 13.2 (8.5) 9.1 (9.1)
Table 1. Descriptive statistics and demographic data on the entire patient population
ACCEPTED MANUSCRIPT
Revision THA N=7,566 65.8 (12.5) 29.1 (8.0) 3383 (45)
P value
53007 (79) 4502 (7) 978 (1) 7706 (12) 37261 (56) 1574 (2) 1762 (3) 1813 (3) 344 (0.5) 64 (0.1) 1813 (3) 2467 (4)
6027 (80) 500 (7) 82 (1) 985 (13) 5464 (72) 266 (4) 728 (10) 317 (4) 63 (0.8) 16 (0.2) 317 (4) 973 (13)
0.028
2887 (4) 36585 (55) 25909 (39) 1373 (2) 4 (<0.1) 94.4 (42) 2233 (3) 189 (0.3) 356 (0.5) 166 (0.2) 11411 (17) 278 (0.4) 180 (0.2) 9.2 (9.1) 12.9 (8.5)
AC C
EP
TE D
<0.001 <0.001 0.601
RI PT
SC
Mean age (years) (SD) Mean BMI (kg/m^2) (SD) Male Gender (%) Ethnicity (%) White African American Asian Diabetes Mellitus (%) General Anesthesia (%) Smoking History (%) Malnourished (%) Kidney Disease (%) Cardiac Disease (%) History of Stroke (%) Kidney Disease (%) Low Preoperative Hematocrit (%) ASA Classification (%) I II III IV V Mean Operative Time (minutes) (SD) Operative Time > 3 hours (%) Return to Operating Room (%) Readmission within 30 days (%) Mortality (%) Any Complication (%) Deep Venous Thrombosis (%) Pulmonary Embolism (%) Mean time to PE (days) (SD) Mean time to DVT (days) (SD)
Primary THA N=66,839 64.7 (11.6) 29.8 (6.9) 29675 (44)
M AN U
Demographic Data
152 (2) 3167 (420 3842 (51) 400 (5) 1 (<0.1) 147.0 (79) 1962 (26) 79 (1) 96 (1) 54 (0.7) 2789 (37) 46 (0.6) 28 (0.3) 8.6 (9.3) 14.8 (7.9)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.004 <0.001 <0.001 <0.001
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.016 0.116 0.215 0.890
Table 2. Comparison of patient characteristics between the primary THA and revision THA groups.
ACCEPTED MANUSCRIPT
RI PT
95% Confidence Interval 0.665-1.051 0.731-1.675 0.564-1.232 1.157-1.829 1.030-1.654 0.864-1.502 0.749-1.440 0.759-2.581 0.225-3.805 0.995-2.660 0.905-2.315 2.180-7.951 1.179-2.553 0.889-2.080 1.056-2.848
SC
Odds Ratio 0.836 1.107 0.833 1.455 1.305 1.139 1.039 1.400 0.925 1.626 1.447 4.163 1.735 1.360 1.734
M AN U
Risk Factor Male Gender African American Ethnicity Revision Arthroplasty Age > 70 years General Anesthesia BMI > 35 kg/m2 Diabetes Mellitus Smoking History Cardiac Disease Kidney Disease Malnutrition Infection Operating Time > 3 hours Low Preoperative Hematocrit ASA IV or greater
P value 0.125 0.632 0.360 0.001 0.028 0.355 0.819 0.281 0.913 0.053 0.123 <0.001 0.005 0.156 0.030
AC C
EP
TE D
Table 3. Multivariate logistic regression analysis identifying independent risk factors for deep venous thrombosis within 30 days of hip arthroplasty.
ACCEPTED MANUSCRIPT
95% Confidence Interval 0.702-1.253 0.988-2.534 0.630-1.616 1.271-2.277 0.887-1.606 0.686-1.435 0.806-1.817 0.592-3.022 0.450-2.137 0.785-2.767 0.379-4.337 1.124-3.026 0.817-2.484 0.312-1.934
RI PT
Odds Ratio 0.938 1.582 1.009 1.701 1.193 0.993 1.210 1.337 0.981 1.474 1.281 1.845 1.424 0.777
SC
Risk Factor Male Gender African American Ethnicity Revision Arthroplasty Age > 70 years General Anesthesia BMI > 35 kg/m2 Diabetes Mellitus Smoking History Kidney Disease Malnutrition Infection Operating Time > 3 hours Low Preoperative Hematocrit ASA IV or greater
AC C
EP
TE D
M AN U
Table 4. Multivariate logistic regression analysis identifying independent risk factors for pulmonary embolism within 30 days of hip arthroplasty.
P value 0.665 0.056 0.971 <0.001 0.244 0.968 0.357 0.485 0.961 0.227 0.690 0.015 0.213 0.587
ACCEPTED MANUSCRIPT
Odds Ratio 1.452 1.314 1.651 1.571 3.818 1.668 1.782 1.527
95% Confidence Interval 1.162-1.816 1.038-1.664 1.021-2.672 1.003-2.461 2.132-6.837 1.155-2.407 1.089-2.915 1.396-1.671
P value 0.001 0.023 0.041 0.048 <0.001 0.006 0.021 <0.001
PE Risk Score 1 1 1 3
Odds Ratio 1.744 1.657 2.040 1.802
95% Confidence Interval 1.314-2.315 1.038-2.646 1.294-3.216 1.448-2.241
P value <0.001 0.034 0.002 <0.001
RI PT
Risk Factor for PE Age > 70 years African American Ethnicity Operating Time > 3 hours TOTAL
DVT Risk Score 1 1 1 1 3 1 1 9
SC
Risk Factor for DVT Age > 70 years General Anesthesia Kidney Disease Malnutrition Infection Operating Time > 3 hours ASA IV or greater TOTAL
AC C
EP
TE D
M AN U
Table 5. Forward, stepwise multivariate logistic regression analysis identifying simple risk score for VTE events after total hip arthroplasty.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT