- Email: [email protected]
Lower Extremity Duplex Ultrasound Screening Protocol for Moderate- and High-Risk Trauma Patients
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Lower Extremity Duplex Ultrasound Screening Protocol for Moderate and High Risk Trauma Patients Grace E. Martin, MD,1 Amanda Pugh, MD,1 Susan G. Williams, MD, Dennis Hanseman, PhD, Vanessa Nomellini, MD, PhD, Amy T. Makley, MD, Timothy A. Pritts, MD, PhD, and Michael D. Goodman, MD* Department of Surgery, University of Cincinnati, Cincinnati, Ohio
article info
abstract
Article history:
Background: Deep vein thrombosis (DVT) remains a significant cause of morbidity after
Received 21 June 2018
injury. Lower extremity duplex ultrasound screening (LEDUS) is designed to identify early,
Received in revised form
asymptomatic DVTs in moderate and high risk patients. We sought to describe when
26 August 2018
thrombus is detected and identify which trauma patients benefit from LEDUS.
Accepted 2 October 2018
Materials and methods: A retrospective review was conducted on trauma patients who were
Available online xxx
moderate or high risk for venous thromboembolism based on risk assessment profile (RAP)
Keywords:
We defined moderate venous thromboembolism risk as an RAP score of 5-9 and high risk as
Lower extremity duplex ultrasound
an RAP score of 10. Demographics, injury characteristics, and chemoprophylaxis type and
scoring. Patients with RAP scores 5 underwent LEDUS on hospital Day 4 and then weekly.
screening
timing were analyzed.
Deep vein thrombosis
Results: A total of 579 trauma patients underwent a total of 820 ultrasounds in 1 y. Eighty-
Venous thromboembolic event
eight acute DVTs were identified. There was only one progression of a below- to above-the-
Clot propagation
knee DVT. Patients with RAP scores 10 had significantly higher rates of DVTs compared with patients with lower RAP scores in addition to longer lengths of stay and time to DVT prophylaxis. Moderate- and high-risk patients had similar rates of pulmonary embolism. Two-thirds of all DVTs were diagnosed on the first screening examination. The rate of DVTs in patients with RAP scores 7-9 was 15.4% compared with 6.1% of those with RAP scores of 5-6. Conclusions: LEDUS allows for early identification of asymptomatic DVTs. Moderate-risk patients with RAP scores of 7 should be considered for LEDUS, given higher rates of DVT. ª 2018 Elsevier Inc. All rights reserved.
The postal address for all the previously mentioned authors is the same and listed in the following: University of Cincinnati, Department of Surgery, 231 Albert Sabin Way, ML 0558, Cincinnati, OH 45267. * Corresponding author. Department of Surgery, Section of General Surgery, University of Cincinnati, 231 Albert Sabin Way, ML 0558, Cincinnati, OH 45267. Tel.: þ1 513 558-5661; fax: þ1 513 558-3474. E-mail address: [email protected] (M.D. Goodman). 1 These authors contributed equally to this study and article. 0022-4804/$ e see front matter ª 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2018.10.010
martin et al dvt screening in trauma patients
Introduction Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), remains a significant cause of morbidity and mortality in hospitalized patients after traumatic injury.1 After a major trauma, patients often experience some combination of vascular injury, prolonged immobilization, and decreased fibrinolysis.2 Given these multiple risk factors, the trauma population remains at high risk for the development of VTE. The incidence of DVT in severely injured patients can be as high as 50% in the absence of DVT prophylaxis and has been shown to reach 44% even in those who receive chemoprophylaxis.3,4 Although the initial diagnosis of VTE during hospitalization can incur significant morbidity, associated long-term complications, including recurrent VTE, postthrombotic syndrome, and pulmonary hypertension, can lead to further complication, requirements for long-term anticoagulation, and increased health care costs.5 Therefore, it remains critical to identify and treat VTE early to avoid both short- and long-term consequences. Lower extremity duplex ultrasound screening (LEDUS) has been shown to be useful in the identification of asymptomatic DVTs in the trauma patient population.6-8 However, the use of LEDUS for screening asymptomatic moderate- and low-risk trauma patients remains controversial, with some studies suggesting that LEDUS identifies more DVTs without preventing PEs.9,10 Since the validation of the risk assessment profile (RAP) score in the trauma population as a risk assessment tool for VTE, the scoring system has been used to categorize patients as low, moderate, or high risk for VTE.11,12 Although high-risk patients commonly undergo LEDUS, it remains unclear which moderate risk patients may benefit from screening. Currently, clinicians use the RAP score in conjunction with several other patient factors including Injury Severity Score (ISS), duration of immobilization, and associated comorbidities to determine which injured patients are at elevated risk for VTE and would therefore benefit from LEDUS protocols. There are several unanswered questions regarding routine LEDUS including RAP score cutoffs and timing of screening. As a result, screening practices are highly dissimilar among institutions. A lack of agreement in LEDUS protocols has led to mixed results when studying the value of LEDUS in this population.13 In the present study, we sought to characterize risk factors for DVT in the trauma population and identify the timing and location of DVT occurrence. We hypothesized that the RAP score would help differentiate which trauma patients at moderate risk for VTE may benefit from LEDUS.
Methods Study population This study was undertaken at the University of Cincinnati Medical Center (UCMC) and was approved by the Institutional Review Board of the University of Cincinnati. UCMC is an American College of Surgeonseverified level I trauma center
281
and maintains an independent registry of those patients evaluated after a traumatic injury. The UCMC trauma registry was retrospectively queried for patients admitted to the Trauma Surgery service between July 2014 and June 2015. Patient demographics, admission vital signs, regional Abbreviated Injury Scale (AIS) scores, ISS, RAP scores, time at which ultrasounds were performed, administration of DVT chemoprophylaxis, and hospital length of stay (LOS) were collected. Any data not available from the institutional trauma registry were supplemented with additional review of the electronic medical record. Vascular laboratory reports for initial and subsequent weekly ultrasounds were reviewed individually. LEDUS examinations evaluated the deep veins of the lower extremities from the inguinal ligament to the ankle. Information on the presence, location (above- versus below-the-knee), clot acuity, and progression of DVT was collected. Popliteal DVTs at any level were considered to be above-the-knee DVTs. Chronic clots were defined by the attending vascular surgeon who interpreted the duplex ultrasound examination. CT angiography was used to diagnose patients with symptomatic PE and was ordered based on provider clinical judgment. PE was defined as any filling defect of the pulmonary arterial tree to the subsegmental level. Multivariate logistic regression analysis was undertaken to identify associated risk factors of those patients diagnosed with DVT. Patient gender, weight, mechanism of injury, ISS, RAP score, extremity AIS, and days to DVT prophylaxis were included in the multivariate analysis as these have been shown to be predictive of VTE. Based on institutional protocol and data presented in the following, patients with RAP scores 10 were considered high risk for VTE, whereas those with RAP scores between 5 and 9 were considered moderate risk for VTE. Patients in these moderate and high risk for VTE groups were compared with respect to time to DVT prophylaxis, LOS, incidence of acute DVT, and incidence of PE.
Lower extremity duplex ultrasound screening program Patients admitted to the Trauma Service at UCMC receive DVT prophylaxis and LEDUS based on RAP scoring that occurs within 24 h of admission (Table 1). Any patient diagnosed with an above-the-knee DVT received therapeutic anticoagulation if no contraindications existed. If patients were not candidates for therapeutic anticoagulation, an inferior vena cava filter was placed. Patients at moderate or high risk for VTE underwent LEDUS on hospital Day 4 if otherwise asymptomatic and received 30 mg of subcutaneous enoxaparin every 12 h for DVT prophylaxis. Patients with RAP scores of 5 and weights <125 kg received 30 mg enoxaparin twice daily, whereas those weighing >125 kg received a starting dose of 40 mg enoxaparin twice daily. Serum anti-Xa levels were drawn before the fourth dose of enoxaparin, and the subsequent enoxaparin dose was increased by 10 mg if the anti-Xa level was undetectable. Moderate and high risk patients continued to undergo LEDUS weekly for the remainder of their hospital stay. Low risk patients with RAP scores <5 did not undergo LEDUS and but did receive DVT prophylaxis in the form of 5000 units of subcutaneous unfractionated heparin every 8 h during their
282
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
undertaken. Those variables significant at P < 0.2 on bivariate analysis were entered in a stepwise logistic regression model. Statistical significance was considered for P values < 0.05.
Table 1 e Risk assessment profile score. RAP score Underlying conditions Obese (BMI 30 kg/m2)
2
Malignancy
2
Abnormal coagulation factors
2
History of thromboembolism
3
Iatrogenic factors Central femoral line > 24 h
2
Four or more transfusions during first 24 h
2
Surgical procedures > 2 h
2
Repair or ligation of major venous injury
3
Injury-related factors AIS >2 for the chest
2
AIS >2 for the abdomen
2
Spinal fractures
2
AIS >2 for the head
3
Coma (GCS <8 for >4 h)
3
Severe lower extremity fracture*
4
Pelvic fracture
4
Spinal cord injury with paraplegia or quadriplegia
4
Age 40-59
2
60-74
3
75
4
BMI ¼ body mass index; GCS ¼ Glasgow Coma Scale. * Severe lower extremity fracture ¼ femur fracture, open tibia fracture (including pilon), and open bimalleolar or trimalleolar fracture.
hospital stay. Only those patients with RAP scores performed within 24 h of arrival were included for analysis. Patients with symptoms of lower extremity DVT or PE that prompted duplex ultrasonography were excluded from this study. In addition, any patient who did not undergo LEDUS within the first 4 d of hospital admission was excluded (Fig. 1). Of the 133 patients excluded from analysis, 65.4% were discharged by the fourth day of their hospital admission. The other 34.6% of these patients had a median age of 71 y, LOS of 7 d, and RAP score of 8. The most common reasons these patients did not undergo LEDUS were that they were already therapeutically anticoagulated or had preexisting conditions for which therapeutic anticoagulation would have been contraindicated or deemed unsafe by the attending trauma surgeon.
Statistical analysis All statistical analyses were performed using the statistical software package Prism 7 (GraphPad, La Jolla, CA). Continuous data are reported as a median with interquartile range with comparisons performed using a ManneWhitney test of significance for nonnormally distributed data. Discrete and categorical data are reported as percentages and rates with comparisons performed using chi-square statistics. To identify independent predictors of presence of lower extremity DVT, clot propagation, and PE, bivariate analysis was first
Results A total of 820 LEDUS examinations were performed on 579 patients included in the study. Most patients were observed to have a blunt mechanism of injury and were between the ages of 30 and 70 y (Table 2). At least one DVT was identified in 77 patients, and a total of 88 DVTs were identified. Below-theknee DVTs accounted for 53 of the DVTs (60.2%), and abovethe-knee DVTs accounted for 29 of the DVTs (33.0%). There were six instances of concomitant unilateral above-the-knee and below-the-knee DVTs. Of note, all six of these DVTs had above-the-knee and below-the-knee components noted on the same scan, eliminating the ability to determine which component originated first. Only one of the 53 patients (1.9%) with acute below-the-knee DVT experienced propagation to an above-the-knee DVT. This particular patient was an 87-yold woman with an RAP score of 11. There were nine chronic clots identified on LEDUS, with the majority (77.8%) located above-the-knee. PE was identified in 12 patients (2.1%). Of those with PE, 50% of these patients had no DVT previously identified on screening or diagnostic ultrasound. Three patients (6.4%) with below-the-knee DVTs experienced PE. Of these three patients, one was moderate risk, and two were high risk for VTE based on RAP scores. Another three patients developed PE concomitantly with unilateral above- and below-the-knee DVT. All three patients with concomitant above- and below-the-knee DVTs had RAP scores between 7 and 8. Most DVTs were identified on the first LEDUS examination performed within 4 d of admission (61.4%) or on the second LEDUS examination 1 wk later (26.1%; Table 3). No significant differences were observed with respect to timing of LEDUS examination in identification of either below-the-knee or above-the-knee clots. Only two DVTs (both above-the-knee) were identified in patients who received four or five LEDUS examinations. For the single patient in which clot propagation was observed, the below-the-knee component was identified on the first LEDUS examination with the above-the-knee component identified on the second LEDUS examination. Multivariate analysis comparing variables of patients diagnosed with DVT to those without DVT revealed that weight (odds ratio [OR] 1.02, confidence interval [CI] 1.01-1.03; P < 0.01), RAP score (OR 1.15, CI 1.04-1.26; P < 0.01), and lower extremity AIS score (OR 3.53, CI 1.22-1.26; P < 0.05) were significant predictors of DVT (Table 4).
Outcomes of moderate versus high risk for VTE patients Of the 579 patients included, 377 patients were identified as moderate risk for VTE with another 202 patients identified as high risk for VTE. There were no significant differences in age, weight, or mechanism of injury between the moderate- and high-risk groups (Table 2). High risk patients had significantly higher ISS scores, increased lower extremity AIS scores, and longer LOS compared with moderate risk patients. High risk
martin et al dvt screening in trauma patients
283
Fig. 1 e Consort diagram of patients included in study.
patients were also observed to have significantly longer time to initiation of DVT chemoprophylaxis compared with moderate risk patients. In addition, high risk patients had a higher incidence of acute above-the-knee DVT and acute below-the-knee DVT compared with those at moderate risk. There were no significant differences in below-the-knee DVT clot propagation or in the incidence of PE between the two groups.
Relationship of RAP score to VTE events Analysis of RAP score in relationship to the presence of DVT was undertaken (Fig. 2). An increasing trend was observed in RAP score and the rate of any kind of DVT (P ¼ 0.0002), rate of above-the-knee DVT (P ¼ 0.006), and rate of below-the-knee DVT (P ¼ 0.001). No trend was observed between RAP score and rate of PE. Approximately 20% of patients with RAP scores of 11 were observed to have any kind of acute DVT with at least 5% of these patients developing acute above-the-knee DVT. Below-the-knee DVTs were present in more than 10% of patients with RAP scores of 10. Moderate risk patients with RAP scores 5-9 were just as likely to experience PE as high risk patients, despite lower rates of DVT. The rate of any kind of DVT in patients with RAP scores 7-9 was >10%, whereas only 3.0% and 6.9% of those with RAP scores of 5 and 6, respectively, developed a DVT. Similarly, rates of below-theknee DVT were equivalent for patients with RAP scores of 79 and elevated compared with those patients with RAP scores of 5 and 6.
Discussion Hospitalized trauma patients continue to be at higher than average risk for venous thromboembolic events including DVT and PE.3,4 Given the potential long-term complications and health care costs that VTE can incur, many centers across the country now use LEDUS to identify patients with asymptomatic DVT. These screening protocols vary from center to center, which makes studying the consequences of these programs difficult and fraught with error. To date, it remains unclear which patients at moderate risk for VTE will benefit most from LEDUS. Using the RAP score, we found that patients with RAP scores of 5 and 6 had an incidence of DVT of 3% and 6.9%, respectively, whereas patients with RAP score of 7-9 had an incidence of DVT of >10%. These data have allowed our institution to better stratify the moderate risk patients who are at increased risk for VTE and may benefit from LEDUS screening. The RAP score was originally designed by a panel of experts to assess risk of VTE in trauma patients.14 Since the original publication of these “Greenfield Guidelines”, Gearhart et al. validated the RAP score in this population, confirming that patients with RAP scores of 5 are three times more likely to experience VTE compared with patients with RAP scores of <5. However, the authors found that the only independent variable associated with the development of DVT was the transfusion of four or more units of blood in the initial 24 h.12 In their study, no DVTs developed in any patients with complex lower extremity fractures or with chest trauma. Our
284
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
Table 2 e Demographics and rate of VTE of patients with RAP score >5. Study cohort characteristics
All patients
RAP score 5-9
10
n ¼ 579
n ¼ 377
n ¼ 202
Age (y)
49 (31-66)
50 (31-65)
47 (30-71.3)
Weight (kg)
82 (69.8-97)
81 (68-97)
83 (70.3-97)
81.9
80.4
84.7
Blunt mechanism (%) RAP score (n)
8 (6-10)
7 (6-8)
17 (13-26)
17 (10-22)
Highest AIS score (n)
3 (3-4)
3 (3-3)
3 (3-4)*
Lower extremity AIS score (n)
3 (2-3)
2 (2-3)
3 (2-3)*
ISS (n)
LOS (d)
8 (5-14)
7 (5-11.5)
Time to DVT prophylaxis (d)
1 (1-2)
1 (1-2)
12 (10-13)* 22 (17-33.3)*
11 (7-19)* 2 (1-2)y
Patients with acute DVT (%)
77 (13.3)
34 (9.0)
43 (21.3)*
Above-the-knee (%)
28 (4.8)
10 (2.7)
18 (8.9)y
Below-the-knee (%)
47 (8.6)
19 (5.0)
28 (13.9)y
Above- and belowthe-knee (%)
6 (1.0)
5 (1.3)
1 (0.5)
Total acute DVTs (n)
88
37
51y
Above-the-knee (n)
29
10
19y
Below-the-knee (n)
53
22
31y
Above- and belowthe-knee (n)
6
5
1
1 (1.9)
0 (0)
1 (3.2)
12 (2.1)
5 (1.3)
7 (3.5)
0 (0)
0 (0)
0 (0)
Acute below-theknee DVT clot propagation (%) Patients with pulmonary embolism (%) Associated with above-the-knee DVT (%) Associated with below-the-knee DVT (%)
3 (25.0)
1 (20)
Associated with above- and below-the-knee DVT (%)
3 (25.0)
3 (60)
No DVT identified
6 (50%)
1 (20%)
2 (28.6)
0 (0)
5 (71.4%)z
*
P < 0.0001 compared with RAP score 5-9. y P < 0.01 compared with RAP score 5-9. z P < 0.05 compared with RAP score 5-9.
multivariate analysis similarly did not find an association with either of these AIS scores; only the extremity AIS score was associated with DVT. Furthermore, in this study, the RAP score failed to appropriately triage several patients identified as moderate risk that ultimately developed PE. Previous authors have hypothesized that use of inflammatory markers
may also aid in predicting which patients go onto developing VTE.11,15 Future studies designed to assess an association between inflammatory markers, such as C-reactive protein level, white blood cell count (in the absence of infection), or serum microvesicles, and risk of VTE may help to improve the sensitivity and specificity of the RAP score. Clot propagation with subsequent embolization is a potential concern for physicians taking care of patients with below-the-knee DVT. Previous studies have cited below-theknee DVT clot propagation rates as ranging from 0% to 35.7%.16-19 However, few studies have investigated below-theknee DVT clot propagation in the trauma population. Olsen et al. reported a higher rate of 8% of below-the-knee clot propagation; however, it is unclear how they defined “moderate” and “high” risk for VTE. In addition, a portion of the patients included underwent twice-weekly LEDUS examinations, which may have resulted in a higher rate of detection of clot propagation.20 Iskander et al. have also reported a higher rate of below-the-knee clot propagation of 12.8%; however, the patients with clot propagation had a median ISS of 35, which is considerably higher than the patients in our high risk group (median ISS ¼ 22) or our moderate risk group (median ¼ 17).16 Using a risk stratification tool similar to the RAP score, Sharpe et al. demonstrated a much lower incidence (4.7%) of clot propagation in high and very high risk trauma patients.21 This rate of propagation is similar to the observed rate in our high-risk patients (3.2%), defined as having RAP scores of 10. No patients classified as moderate risk for VTE based on RAP scoring of 5-9 experienced below-the-knee clot propagation in our study. Given these low rates of clot propagation, patients at our center with isolated below-the-knee DVT do not receive therapeutic anticoagulation. In addition, because treatment is not initiated in these moderate-risk patients, subsequent weekly LEDUS examinations do not include below-the-knee visualization. If a patient is found to have a below-the-knee DVT, the patient will continue to undergo LEDUS in the nonaffected lower extremity. Not surprisingly, the time to DVT prophylaxis is longer in patients with higher RAP scores. As demonstrated by the significantly higher ISS and AIS scores of the high risk group to the moderate risk group, these high risk patients often have more severe and hemorrhagic injuries complicating the balance of bleeding and clotting. Initiation of DVT chemoprophylaxis in this high-risk group is delayed by the presence of recent solid organ or brain injury, ongoing bleeding, coagulopathy, and/or need for operative intervention. It remains unclear whether these patients are experiencing higher VTE rates because of longer periods without chemoprophylaxis or whether this is because of the higher incidence of risk factors included in the RAP score. It is likely the combination of these two factors that leads to the observed increase in incidence of VTE events, making it even more important that these highrisk patients undergo early routine LEDUS examinations. DVTs are known to occur early after trauma.4,18-20 Here, we have confirmed that nearly two-thirds of posttraumatic asymptomatic DVTs occur within 4 d of injury. In a porcine model of trauma and hemorrhagic shock, animals developed hypercoagulability within 4 h of resuscitation. This hypercoagulable state persisted up to 72 h after injury with animals demonstrating increased clot firmness and increased platelet-
285
martin et al dvt screening in trauma patients
Table 3 e Timing of DVT identification. LEDUS examination
DVTs identified (n)
Below-theknee DVT (n)
Above-theknee DVT (n)
Below- and above-theknee DVT (n)
1st
54
36
14
4
2nd
23
14
8
1
3rd
9
3
5
1
4th
1
0
1
0
5th
1
0
1
0
fibrin interaction compared with animals that only received shock and resuscitation.22 Similarly, Gary et al. found in trauma patients evaluated on admission with rapid thrombelastography that despite more hypocoagulable R-times, patients with elevated maximum amplitudes (suggesting platelet hyperaggregability) demonstrated significantly higher rates of VTE compared with patients with normal or low maximum amplitudes.23 Future studies to further delineate how early DVTs form is warranted. Perhaps, LEDUS examinations should be undertaken even earlier in high risk patients as hypercoagulability can occur within hours of traumatic injury and resuscitation. True PEs, defined as those that arise and embolize from extremity veins, are less common in trauma patients as the majority appear to arise de novo as thrombi in the pulmonary vasculature after injurydespecially in those patients with chest trauma.24-27 However, previous studies have
demonstrated that LEDUS protocols decrease rates of PE.28 In our cohort, 50% of patients with PE had an associated concomitant DVT. Interestingly, 80% of moderate risk patients who developed PE had concomitant DVT compared with only 28.6% of high risk patients with PE. The high risk patients with PE may have had a higher median chest AIS scores compared with moderate risk patients (3 versus 2, high risk versus moderate risk, P ¼ 0.2), although this did not reach statistical significance because of sample size. However, given the small number of PEs observed in our cohort, we cannot comment on the likelihood of true thromboembolic events because of DVTs. Based on our data demonstrating the incidence of VTE, patients with RAP scores 7-9 should be considered for LEDUS. Although this recommendation does not provide a single RAP score at which risk of VTE becomes absolutely necessary to screen for DVTs, we have shown that higher RAP scores even
Table 4 e Logistic regression models for VTE. All patients Variable
Univariate
Multivariate
OR
95% CI
P value
Female
0.60
0.34-1.04
0.07
Weight
1.02
1.01-1.03
<0.01
ISS
1.04
1.02-1.06
<0.01
RAP
1.18
1.10-1.27
<0.01
AIS cervical
1.73
1.03-2.91
0.04
AIS external
4.05
1.48-11.1
<0.01
Days to prophylaxis
1.21
1.05-1.39
<0.01
Female
0.41
0.17-1.03
0.06
Weight
1.02
1.01-1.03
<0.01
ISS
1.05
1.01-1.09
0.02
Blunt trauma
0.56
0.25-1.21
0.14
RAP
1.31
0.98-1.76
0.07
AIS cervical
2.10
1.01-4.36
0.05
1.58
0.87-2.88
0.14
1.86-105.3
0.01
1.31
1.04-1.64
0.02
Weight
1.02
1.00-1.03
0.02
AIS lumbar
4.03
0.62-26.4
0.15
OR
95% CI
P value
1.02
1.01-1.03
<0.01
1.15
1.04-1.26
<0.01
3.53
1.22-1.26
0.02
1.02
1.01-1.04
<0.01
1.06
1.02-1.10
<0.01
1.02
1.00-1.03
0.02
RAP score 5-9
AIS chest AIS external Days to prophylaxis
14.0
RAP score 10
286
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
Fig. 2 e Association between RAP score and DVT occurrence.
within the moderate risk definition are associated with increased VTE risk. Clinicians with different practice patterns and biases may choose different RAP scores at which to perform LEDUS examinations; the decision ultimately depends on the risk of VTE that an individual provider is willing to accept. This often becomes problematic for physicians as VTEs are now used as a marker of quality by several health care agencies.29 Increased VTE rates are often portrayed as suboptimal care and may bias clinicians against aggressive DVT screening. There are several limitations to this study. First, this was a retrospective study at a single institution, and extrapolation of our results to other hospitals could be limited. Second, the rates of DVT and PE were not known for patients with RAP scores <5, with the assumption that these rates are low. Although unlikely based on previous RAP score studies, these patients could have similar risk of DVT and PE as the moderate risk patients in our study.13 Third, this study did not take into account rates of VTE on readmission or postdischarge and may not have fully accounted for all DVTs in the discharged trauma population. Fourth, many patients developed VTE by the time of the initial LEDUS. It remains unclear as to the optimal time to screen trauma patients for VTE and LEDUS examinations that occur earlier in patients’ hospital course may be necessary. Fifth, the data analyzed are from 3 y ago and may not accurately reflect the rates of DVT that our center currently experiences. The protocol for LEDUS in trauma patients changed in late 2015, and our group is working to characterize a protocol that is safe and cost-effective. In addition, given the retrospective nature of the study, the number and timing of missed chemoprophylaxis doses were not reliably identified or included in statistical analyses. Finally, as not all trauma centers use LEDUS examinations, and only a proportion of those that screen do so based on the RAP score, these results may not be broadly applicable to all
trauma centers. Although we realize that nonselective LEDUS protocols are cost-prohibitive for all trauma patients in all centers, the information provided by this study may guide physicians in the selection of patients at high and moderate risk for VTE who would most greatly benefit from screening and more aggressive chemoprophylaxis. The RAP score continues to be useful in the identification of moderate- and high-risk patients who require chemoprophylaxis and LEDUS. Patients at moderate risk for VTE have lower DVT rates but similar PE rates compared with RAPstratified high risk patients; therefore, the decision to perform LEDUS examinations should be considered for all moderate risk patients with RAP scores 7-9. Serial examinations need not be undertaken routinely as the rate of clot propagation in our study was low. Finally, nearly two-thirds of acute DVT were identified within 4 d of injury, confirming that DVT occurs and can be detected early after trauma.
Acknowledgment Authors’ contributions: G.E.M., T.A.P., J.B.E., and M.D.G. contributed to the design of the study. G.E.M., A.P., S.W., and K.B. contributed to the acquisition of data. A.P., G.E.M., D.H., V.N., A.T.M., and M.D.G. contributed to data analysis. A.P. and G.E.M. drafted the article. All authors contributed to critical review of the article and have agreed to the final version of the article. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosure None.
martin et al dvt screening in trauma patients
references
1. Jacobs BN, Cain-Nielsen AH, Jakubus JL, et al. Unfractionated heparin versus low-molecular-weight heparin for venous thromboembolism prophylaxis in trauma. J Trauma Acute Care Surg. 2017;83:151e158. 2. Alhassan S, Pelinescu A, Gandhi V, Naddour M, Singh AC, Bihler E. Clinical presentation and risk factors of venous thromboembolic disease. Crit Care Nurs Q. 2017;40:201e209. 3. Geerts WH, Jay RM, Code KI, et al. A comparison of low-dose heparin with low-molecular-weight heparin as prophylaxis against venous thromboembolism after major trauma. N Engl J Med. 1996;335:701e707. 4. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP. A prospective study of venous thromboembolism after major trauma. N Engl J Med. 1994;331:1601e1606. 5. Dobesh PP. Economic burden of venous thromboembolism in hospitalized patients. Pharmacotherapy. 2009;29:943e953. 6. Cipolle MD, Wojcik R, Seislove E, Wasser TE, Pasquale MD. The role of surveillance duplex scanning in preventing venous thromboembolism in trauma patients. J Trauma. 2002;52:453e462. 7. Bandle J, Shackford SR, Kahl JE, et al. The value of lowerextremity duplex surveillance to detect deep vein thrombosis in trauma patients. J Trauma Acute Care Surg. 2013;74:575e580. 8. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L. Prospective ultrasound evaluation of venous thrombosis in high-risk trauma patients. J Trauma. 1993;35:405e408. 9. Spain DA, Richardson JD, Polk Jr HC, Bergamini TM, Wilson MA, Miller FB. Venous thromboembolism in the highrisk trauma patient: do risks justify aggressive screening and prophylaxis? J Trauma. 1997;42:463e469. 10. Dietch ZC, Edwards BL, Thames M, Shah PM, Williams MD, Sawyer RG. Rate of lower-extremity ultrasonography in trauma patients in associated with rate of deep venous thrombosis but not pulmonary embolism. Surgery. 2015;158:379e385. 11. Hegsted D, Gritsiouk Y, Schlesinger P, Gardiner S, Gubler KD. Utility of risk assessment profile for risk stratification of venous thrombotic events for trauma patients. Am J Surg. 2013;205:517e520. 12. Gearhart MM, Luchette FA, Proctor MC, et al. The risk assessment profile score identifies trauma patients at risk for deep vein thrombosis. Surgery. 2000;128:631e640. 13. Haut ER, Chang DC, Pierce CA, et al. Predictors of posttraumatic deep vein thrombosis (DVT): hospital practice versus patient factors-an analysis of the National Trauma Databank (NTDB). J Trauma. 2009;66:994e999. 14. Greenfield LJ, Proctor MC, Rodriguez JL, Luchette FA, Cipolle MD, Cho J. Posttrauma thromboembolism prophylaxis. J Trauma. 1997;42:100e103.
287
15. Kim Y, Abplanalp WA, Jung AD, et al. Endocytosis of red blood cell microparticles by pulmonary endothelial cells is mediated by Rab5. Shock. 2018;49:288e294. 16. Iskander GA, Nelson RS, Morehouse DL, Tenguist JE, Szlabick RE. Incidence and propagation of infrageniculate deep venous thrombosis in trauma patients. J Trauma. 2006;61:695e700. 17. Masuda EM, Kistner RL, Musikasinthorn C, Liquido F, Geling O, He Q. The controversy of managing calf vein thrombosis. J Vasc Surg. 2012;55:550e561. 18. Horner D, Hogg K, Body R, Nash MJ, Baglin T, MackwayJones K. The anticoagulation of calf thrombosis (ACT) project: results from the randomized controlled external pilot trial. Chest. 2014;146:1468e1477. 19. Garry J, Duke A, Labropoulos N. Systematic review of the complications following isolated calf deep vein thrombosis. Br J Surg. 2016;103:789e796. 20. Olsen EF, Zander AL, Van Gent JM, et al. Below-knee deep vein thrombosis: an opportunity to prevent pulmonary embolism? J Trauma Acute Care Surg. 2014;77:459e463. 21. Sharpe RP, Gupta R, Gracias VH, et al. Incidence and natural history of below-knee deep venous thrombosis in high-risk trauma patients. J Trauma. 2002;53:1048e1052. 22. Mulier KE, Greenberg JG, Beilman GJ. Hypercoagulability in porcine hemorrhagic shock is present early after trauma and resuscitation. J Surg Res. 2012;174:e31ee35. 23. Gary JL, Schneider PS, Galpin M, et al. Can thromboelastography predict venous thromboembolic events in patients with severe extremity trauma? J Orthop Trauma. 2016;30:294e298. 24. Knudson MM, Ikossi DG, Khaw L, Morabito D, Speetzen LS. Thromboembolism after trauma: an analysis of 1602 episodes from the American College of Surgeons National Trauma Databank. Ann Surg. 2004;240:490e496. 25. Brakenridge SC, Henley SS, Kashner TM, et al. Comparing clinical predictors of deep venous thrombosis versus pulmonary embolus after severe injury: a new paradigm for posttraumatic venous thromboembolism? J Trauma Acute Care Surg. 2013;74:1231e1237. 26. Velmahos GC, Spaniolas K, Tabbara M, et al. Pulmonary embolism and deep venous thrombosis in trauma: are they related? Arch Surg. 2009;144:928e932. 27. Knudson MM, Gomez D, Haas B, Cohen MJ, Nathens AB. Three thousand seven hundred thirty-eight posttraumatic pulmonary emboli: a new look at an old disease. Ann Surg. 2011;254:625e632. 28. Allen CJ, Murray CR, Meizoso JP, et al. Surveillance and early management of deep vein thrombosis decreases rate of pulmonary embolism in high risk trauma patients. J Am Coll Surg. 2016;222:65e72. 29. National Consensus Standards for the prevention and care of venous thromboembolism. National Quality Forum project. Available at https://www.jointcommission.org/venous_ thromboembolism/. Accessed April 17, 2018.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Lower Extremity Duplex Ultrasound Screening Protocol for Moderate and High Risk Trauma Patients Grace E. Martin, MD,1 Amanda Pugh, MD,1 Susan G. Williams, MD, Dennis Hanseman, PhD, Vanessa Nomellini, MD, PhD, Amy T. Makley, MD, Timothy A. Pritts, MD, PhD, and Michael D. Goodman, MD* Department of Surgery, University of Cincinnati, Cincinnati, Ohio
article info
abstract
Article history:
Background: Deep vein thrombosis (DVT) remains a significant cause of morbidity after
Received 21 June 2018
injury. Lower extremity duplex ultrasound screening (LEDUS) is designed to identify early,
Received in revised form
asymptomatic DVTs in moderate and high risk patients. We sought to describe when
26 August 2018
thrombus is detected and identify which trauma patients benefit from LEDUS.
Accepted 2 October 2018
Materials and methods: A retrospective review was conducted on trauma patients who were
Available online xxx
moderate or high risk for venous thromboembolism based on risk assessment profile (RAP)
Keywords:
We defined moderate venous thromboembolism risk as an RAP score of 5-9 and high risk as
Lower extremity duplex ultrasound
an RAP score of 10. Demographics, injury characteristics, and chemoprophylaxis type and
scoring. Patients with RAP scores 5 underwent LEDUS on hospital Day 4 and then weekly.
screening
timing were analyzed.
Deep vein thrombosis
Results: A total of 579 trauma patients underwent a total of 820 ultrasounds in 1 y. Eighty-
Venous thromboembolic event
eight acute DVTs were identified. There was only one progression of a below- to above-the-
Clot propagation
knee DVT. Patients with RAP scores 10 had significantly higher rates of DVTs compared with patients with lower RAP scores in addition to longer lengths of stay and time to DVT prophylaxis. Moderate- and high-risk patients had similar rates of pulmonary embolism. Two-thirds of all DVTs were diagnosed on the first screening examination. The rate of DVTs in patients with RAP scores 7-9 was 15.4% compared with 6.1% of those with RAP scores of 5-6. Conclusions: LEDUS allows for early identification of asymptomatic DVTs. Moderate-risk patients with RAP scores of 7 should be considered for LEDUS, given higher rates of DVT. ª 2018 Elsevier Inc. All rights reserved.
The postal address for all the previously mentioned authors is the same and listed in the following: University of Cincinnati, Department of Surgery, 231 Albert Sabin Way, ML 0558, Cincinnati, OH 45267. * Corresponding author. Department of Surgery, Section of General Surgery, University of Cincinnati, 231 Albert Sabin Way, ML 0558, Cincinnati, OH 45267. Tel.: þ1 513 558-5661; fax: þ1 513 558-3474. E-mail address: [email protected] (M.D. Goodman). 1 These authors contributed equally to this study and article. 0022-4804/$ e see front matter ª 2018 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2018.10.010
martin et al dvt screening in trauma patients
Introduction Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), remains a significant cause of morbidity and mortality in hospitalized patients after traumatic injury.1 After a major trauma, patients often experience some combination of vascular injury, prolonged immobilization, and decreased fibrinolysis.2 Given these multiple risk factors, the trauma population remains at high risk for the development of VTE. The incidence of DVT in severely injured patients can be as high as 50% in the absence of DVT prophylaxis and has been shown to reach 44% even in those who receive chemoprophylaxis.3,4 Although the initial diagnosis of VTE during hospitalization can incur significant morbidity, associated long-term complications, including recurrent VTE, postthrombotic syndrome, and pulmonary hypertension, can lead to further complication, requirements for long-term anticoagulation, and increased health care costs.5 Therefore, it remains critical to identify and treat VTE early to avoid both short- and long-term consequences. Lower extremity duplex ultrasound screening (LEDUS) has been shown to be useful in the identification of asymptomatic DVTs in the trauma patient population.6-8 However, the use of LEDUS for screening asymptomatic moderate- and low-risk trauma patients remains controversial, with some studies suggesting that LEDUS identifies more DVTs without preventing PEs.9,10 Since the validation of the risk assessment profile (RAP) score in the trauma population as a risk assessment tool for VTE, the scoring system has been used to categorize patients as low, moderate, or high risk for VTE.11,12 Although high-risk patients commonly undergo LEDUS, it remains unclear which moderate risk patients may benefit from screening. Currently, clinicians use the RAP score in conjunction with several other patient factors including Injury Severity Score (ISS), duration of immobilization, and associated comorbidities to determine which injured patients are at elevated risk for VTE and would therefore benefit from LEDUS protocols. There are several unanswered questions regarding routine LEDUS including RAP score cutoffs and timing of screening. As a result, screening practices are highly dissimilar among institutions. A lack of agreement in LEDUS protocols has led to mixed results when studying the value of LEDUS in this population.13 In the present study, we sought to characterize risk factors for DVT in the trauma population and identify the timing and location of DVT occurrence. We hypothesized that the RAP score would help differentiate which trauma patients at moderate risk for VTE may benefit from LEDUS.
Methods Study population This study was undertaken at the University of Cincinnati Medical Center (UCMC) and was approved by the Institutional Review Board of the University of Cincinnati. UCMC is an American College of Surgeonseverified level I trauma center
281
and maintains an independent registry of those patients evaluated after a traumatic injury. The UCMC trauma registry was retrospectively queried for patients admitted to the Trauma Surgery service between July 2014 and June 2015. Patient demographics, admission vital signs, regional Abbreviated Injury Scale (AIS) scores, ISS, RAP scores, time at which ultrasounds were performed, administration of DVT chemoprophylaxis, and hospital length of stay (LOS) were collected. Any data not available from the institutional trauma registry were supplemented with additional review of the electronic medical record. Vascular laboratory reports for initial and subsequent weekly ultrasounds were reviewed individually. LEDUS examinations evaluated the deep veins of the lower extremities from the inguinal ligament to the ankle. Information on the presence, location (above- versus below-the-knee), clot acuity, and progression of DVT was collected. Popliteal DVTs at any level were considered to be above-the-knee DVTs. Chronic clots were defined by the attending vascular surgeon who interpreted the duplex ultrasound examination. CT angiography was used to diagnose patients with symptomatic PE and was ordered based on provider clinical judgment. PE was defined as any filling defect of the pulmonary arterial tree to the subsegmental level. Multivariate logistic regression analysis was undertaken to identify associated risk factors of those patients diagnosed with DVT. Patient gender, weight, mechanism of injury, ISS, RAP score, extremity AIS, and days to DVT prophylaxis were included in the multivariate analysis as these have been shown to be predictive of VTE. Based on institutional protocol and data presented in the following, patients with RAP scores 10 were considered high risk for VTE, whereas those with RAP scores between 5 and 9 were considered moderate risk for VTE. Patients in these moderate and high risk for VTE groups were compared with respect to time to DVT prophylaxis, LOS, incidence of acute DVT, and incidence of PE.
Lower extremity duplex ultrasound screening program Patients admitted to the Trauma Service at UCMC receive DVT prophylaxis and LEDUS based on RAP scoring that occurs within 24 h of admission (Table 1). Any patient diagnosed with an above-the-knee DVT received therapeutic anticoagulation if no contraindications existed. If patients were not candidates for therapeutic anticoagulation, an inferior vena cava filter was placed. Patients at moderate or high risk for VTE underwent LEDUS on hospital Day 4 if otherwise asymptomatic and received 30 mg of subcutaneous enoxaparin every 12 h for DVT prophylaxis. Patients with RAP scores of 5 and weights <125 kg received 30 mg enoxaparin twice daily, whereas those weighing >125 kg received a starting dose of 40 mg enoxaparin twice daily. Serum anti-Xa levels were drawn before the fourth dose of enoxaparin, and the subsequent enoxaparin dose was increased by 10 mg if the anti-Xa level was undetectable. Moderate and high risk patients continued to undergo LEDUS weekly for the remainder of their hospital stay. Low risk patients with RAP scores <5 did not undergo LEDUS and but did receive DVT prophylaxis in the form of 5000 units of subcutaneous unfractionated heparin every 8 h during their
282
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
undertaken. Those variables significant at P < 0.2 on bivariate analysis were entered in a stepwise logistic regression model. Statistical significance was considered for P values < 0.05.
Table 1 e Risk assessment profile score. RAP score Underlying conditions Obese (BMI 30 kg/m2)
2
Malignancy
2
Abnormal coagulation factors
2
History of thromboembolism
3
Iatrogenic factors Central femoral line > 24 h
2
Four or more transfusions during first 24 h
2
Surgical procedures > 2 h
2
Repair or ligation of major venous injury
3
Injury-related factors AIS >2 for the chest
2
AIS >2 for the abdomen
2
Spinal fractures
2
AIS >2 for the head
3
Coma (GCS <8 for >4 h)
3
Severe lower extremity fracture*
4
Pelvic fracture
4
Spinal cord injury with paraplegia or quadriplegia
4
Age 40-59
2
60-74
3
75
4
BMI ¼ body mass index; GCS ¼ Glasgow Coma Scale. * Severe lower extremity fracture ¼ femur fracture, open tibia fracture (including pilon), and open bimalleolar or trimalleolar fracture.
hospital stay. Only those patients with RAP scores performed within 24 h of arrival were included for analysis. Patients with symptoms of lower extremity DVT or PE that prompted duplex ultrasonography were excluded from this study. In addition, any patient who did not undergo LEDUS within the first 4 d of hospital admission was excluded (Fig. 1). Of the 133 patients excluded from analysis, 65.4% were discharged by the fourth day of their hospital admission. The other 34.6% of these patients had a median age of 71 y, LOS of 7 d, and RAP score of 8. The most common reasons these patients did not undergo LEDUS were that they were already therapeutically anticoagulated or had preexisting conditions for which therapeutic anticoagulation would have been contraindicated or deemed unsafe by the attending trauma surgeon.
Statistical analysis All statistical analyses were performed using the statistical software package Prism 7 (GraphPad, La Jolla, CA). Continuous data are reported as a median with interquartile range with comparisons performed using a ManneWhitney test of significance for nonnormally distributed data. Discrete and categorical data are reported as percentages and rates with comparisons performed using chi-square statistics. To identify independent predictors of presence of lower extremity DVT, clot propagation, and PE, bivariate analysis was first
Results A total of 820 LEDUS examinations were performed on 579 patients included in the study. Most patients were observed to have a blunt mechanism of injury and were between the ages of 30 and 70 y (Table 2). At least one DVT was identified in 77 patients, and a total of 88 DVTs were identified. Below-theknee DVTs accounted for 53 of the DVTs (60.2%), and abovethe-knee DVTs accounted for 29 of the DVTs (33.0%). There were six instances of concomitant unilateral above-the-knee and below-the-knee DVTs. Of note, all six of these DVTs had above-the-knee and below-the-knee components noted on the same scan, eliminating the ability to determine which component originated first. Only one of the 53 patients (1.9%) with acute below-the-knee DVT experienced propagation to an above-the-knee DVT. This particular patient was an 87-yold woman with an RAP score of 11. There were nine chronic clots identified on LEDUS, with the majority (77.8%) located above-the-knee. PE was identified in 12 patients (2.1%). Of those with PE, 50% of these patients had no DVT previously identified on screening or diagnostic ultrasound. Three patients (6.4%) with below-the-knee DVTs experienced PE. Of these three patients, one was moderate risk, and two were high risk for VTE based on RAP scores. Another three patients developed PE concomitantly with unilateral above- and below-the-knee DVT. All three patients with concomitant above- and below-the-knee DVTs had RAP scores between 7 and 8. Most DVTs were identified on the first LEDUS examination performed within 4 d of admission (61.4%) or on the second LEDUS examination 1 wk later (26.1%; Table 3). No significant differences were observed with respect to timing of LEDUS examination in identification of either below-the-knee or above-the-knee clots. Only two DVTs (both above-the-knee) were identified in patients who received four or five LEDUS examinations. For the single patient in which clot propagation was observed, the below-the-knee component was identified on the first LEDUS examination with the above-the-knee component identified on the second LEDUS examination. Multivariate analysis comparing variables of patients diagnosed with DVT to those without DVT revealed that weight (odds ratio [OR] 1.02, confidence interval [CI] 1.01-1.03; P < 0.01), RAP score (OR 1.15, CI 1.04-1.26; P < 0.01), and lower extremity AIS score (OR 3.53, CI 1.22-1.26; P < 0.05) were significant predictors of DVT (Table 4).
Outcomes of moderate versus high risk for VTE patients Of the 579 patients included, 377 patients were identified as moderate risk for VTE with another 202 patients identified as high risk for VTE. There were no significant differences in age, weight, or mechanism of injury between the moderate- and high-risk groups (Table 2). High risk patients had significantly higher ISS scores, increased lower extremity AIS scores, and longer LOS compared with moderate risk patients. High risk
martin et al dvt screening in trauma patients
283
Fig. 1 e Consort diagram of patients included in study.
patients were also observed to have significantly longer time to initiation of DVT chemoprophylaxis compared with moderate risk patients. In addition, high risk patients had a higher incidence of acute above-the-knee DVT and acute below-the-knee DVT compared with those at moderate risk. There were no significant differences in below-the-knee DVT clot propagation or in the incidence of PE between the two groups.
Relationship of RAP score to VTE events Analysis of RAP score in relationship to the presence of DVT was undertaken (Fig. 2). An increasing trend was observed in RAP score and the rate of any kind of DVT (P ¼ 0.0002), rate of above-the-knee DVT (P ¼ 0.006), and rate of below-the-knee DVT (P ¼ 0.001). No trend was observed between RAP score and rate of PE. Approximately 20% of patients with RAP scores of 11 were observed to have any kind of acute DVT with at least 5% of these patients developing acute above-the-knee DVT. Below-the-knee DVTs were present in more than 10% of patients with RAP scores of 10. Moderate risk patients with RAP scores 5-9 were just as likely to experience PE as high risk patients, despite lower rates of DVT. The rate of any kind of DVT in patients with RAP scores 7-9 was >10%, whereas only 3.0% and 6.9% of those with RAP scores of 5 and 6, respectively, developed a DVT. Similarly, rates of below-theknee DVT were equivalent for patients with RAP scores of 79 and elevated compared with those patients with RAP scores of 5 and 6.
Discussion Hospitalized trauma patients continue to be at higher than average risk for venous thromboembolic events including DVT and PE.3,4 Given the potential long-term complications and health care costs that VTE can incur, many centers across the country now use LEDUS to identify patients with asymptomatic DVT. These screening protocols vary from center to center, which makes studying the consequences of these programs difficult and fraught with error. To date, it remains unclear which patients at moderate risk for VTE will benefit most from LEDUS. Using the RAP score, we found that patients with RAP scores of 5 and 6 had an incidence of DVT of 3% and 6.9%, respectively, whereas patients with RAP score of 7-9 had an incidence of DVT of >10%. These data have allowed our institution to better stratify the moderate risk patients who are at increased risk for VTE and may benefit from LEDUS screening. The RAP score was originally designed by a panel of experts to assess risk of VTE in trauma patients.14 Since the original publication of these “Greenfield Guidelines”, Gearhart et al. validated the RAP score in this population, confirming that patients with RAP scores of 5 are three times more likely to experience VTE compared with patients with RAP scores of <5. However, the authors found that the only independent variable associated with the development of DVT was the transfusion of four or more units of blood in the initial 24 h.12 In their study, no DVTs developed in any patients with complex lower extremity fractures or with chest trauma. Our
284
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
Table 2 e Demographics and rate of VTE of patients with RAP score >5. Study cohort characteristics
All patients
RAP score 5-9
10
n ¼ 579
n ¼ 377
n ¼ 202
Age (y)
49 (31-66)
50 (31-65)
47 (30-71.3)
Weight (kg)
82 (69.8-97)
81 (68-97)
83 (70.3-97)
81.9
80.4
84.7
Blunt mechanism (%) RAP score (n)
8 (6-10)
7 (6-8)
17 (13-26)
17 (10-22)
Highest AIS score (n)
3 (3-4)
3 (3-3)
3 (3-4)*
Lower extremity AIS score (n)
3 (2-3)
2 (2-3)
3 (2-3)*
ISS (n)
LOS (d)
8 (5-14)
7 (5-11.5)
Time to DVT prophylaxis (d)
1 (1-2)
1 (1-2)
12 (10-13)* 22 (17-33.3)*
11 (7-19)* 2 (1-2)y
Patients with acute DVT (%)
77 (13.3)
34 (9.0)
43 (21.3)*
Above-the-knee (%)
28 (4.8)
10 (2.7)
18 (8.9)y
Below-the-knee (%)
47 (8.6)
19 (5.0)
28 (13.9)y
Above- and belowthe-knee (%)
6 (1.0)
5 (1.3)
1 (0.5)
Total acute DVTs (n)
88
37
51y
Above-the-knee (n)
29
10
19y
Below-the-knee (n)
53
22
31y
Above- and belowthe-knee (n)
6
5
1
1 (1.9)
0 (0)
1 (3.2)
12 (2.1)
5 (1.3)
7 (3.5)
0 (0)
0 (0)
0 (0)
Acute below-theknee DVT clot propagation (%) Patients with pulmonary embolism (%) Associated with above-the-knee DVT (%) Associated with below-the-knee DVT (%)
3 (25.0)
1 (20)
Associated with above- and below-the-knee DVT (%)
3 (25.0)
3 (60)
No DVT identified
6 (50%)
1 (20%)
2 (28.6)
0 (0)
5 (71.4%)z
*
P < 0.0001 compared with RAP score 5-9. y P < 0.01 compared with RAP score 5-9. z P < 0.05 compared with RAP score 5-9.
multivariate analysis similarly did not find an association with either of these AIS scores; only the extremity AIS score was associated with DVT. Furthermore, in this study, the RAP score failed to appropriately triage several patients identified as moderate risk that ultimately developed PE. Previous authors have hypothesized that use of inflammatory markers
may also aid in predicting which patients go onto developing VTE.11,15 Future studies designed to assess an association between inflammatory markers, such as C-reactive protein level, white blood cell count (in the absence of infection), or serum microvesicles, and risk of VTE may help to improve the sensitivity and specificity of the RAP score. Clot propagation with subsequent embolization is a potential concern for physicians taking care of patients with below-the-knee DVT. Previous studies have cited below-theknee DVT clot propagation rates as ranging from 0% to 35.7%.16-19 However, few studies have investigated below-theknee DVT clot propagation in the trauma population. Olsen et al. reported a higher rate of 8% of below-the-knee clot propagation; however, it is unclear how they defined “moderate” and “high” risk for VTE. In addition, a portion of the patients included underwent twice-weekly LEDUS examinations, which may have resulted in a higher rate of detection of clot propagation.20 Iskander et al. have also reported a higher rate of below-the-knee clot propagation of 12.8%; however, the patients with clot propagation had a median ISS of 35, which is considerably higher than the patients in our high risk group (median ISS ¼ 22) or our moderate risk group (median ¼ 17).16 Using a risk stratification tool similar to the RAP score, Sharpe et al. demonstrated a much lower incidence (4.7%) of clot propagation in high and very high risk trauma patients.21 This rate of propagation is similar to the observed rate in our high-risk patients (3.2%), defined as having RAP scores of 10. No patients classified as moderate risk for VTE based on RAP scoring of 5-9 experienced below-the-knee clot propagation in our study. Given these low rates of clot propagation, patients at our center with isolated below-the-knee DVT do not receive therapeutic anticoagulation. In addition, because treatment is not initiated in these moderate-risk patients, subsequent weekly LEDUS examinations do not include below-the-knee visualization. If a patient is found to have a below-the-knee DVT, the patient will continue to undergo LEDUS in the nonaffected lower extremity. Not surprisingly, the time to DVT prophylaxis is longer in patients with higher RAP scores. As demonstrated by the significantly higher ISS and AIS scores of the high risk group to the moderate risk group, these high risk patients often have more severe and hemorrhagic injuries complicating the balance of bleeding and clotting. Initiation of DVT chemoprophylaxis in this high-risk group is delayed by the presence of recent solid organ or brain injury, ongoing bleeding, coagulopathy, and/or need for operative intervention. It remains unclear whether these patients are experiencing higher VTE rates because of longer periods without chemoprophylaxis or whether this is because of the higher incidence of risk factors included in the RAP score. It is likely the combination of these two factors that leads to the observed increase in incidence of VTE events, making it even more important that these highrisk patients undergo early routine LEDUS examinations. DVTs are known to occur early after trauma.4,18-20 Here, we have confirmed that nearly two-thirds of posttraumatic asymptomatic DVTs occur within 4 d of injury. In a porcine model of trauma and hemorrhagic shock, animals developed hypercoagulability within 4 h of resuscitation. This hypercoagulable state persisted up to 72 h after injury with animals demonstrating increased clot firmness and increased platelet-
285
martin et al dvt screening in trauma patients
Table 3 e Timing of DVT identification. LEDUS examination
DVTs identified (n)
Below-theknee DVT (n)
Above-theknee DVT (n)
Below- and above-theknee DVT (n)
1st
54
36
14
4
2nd
23
14
8
1
3rd
9
3
5
1
4th
1
0
1
0
5th
1
0
1
0
fibrin interaction compared with animals that only received shock and resuscitation.22 Similarly, Gary et al. found in trauma patients evaluated on admission with rapid thrombelastography that despite more hypocoagulable R-times, patients with elevated maximum amplitudes (suggesting platelet hyperaggregability) demonstrated significantly higher rates of VTE compared with patients with normal or low maximum amplitudes.23 Future studies to further delineate how early DVTs form is warranted. Perhaps, LEDUS examinations should be undertaken even earlier in high risk patients as hypercoagulability can occur within hours of traumatic injury and resuscitation. True PEs, defined as those that arise and embolize from extremity veins, are less common in trauma patients as the majority appear to arise de novo as thrombi in the pulmonary vasculature after injurydespecially in those patients with chest trauma.24-27 However, previous studies have
demonstrated that LEDUS protocols decrease rates of PE.28 In our cohort, 50% of patients with PE had an associated concomitant DVT. Interestingly, 80% of moderate risk patients who developed PE had concomitant DVT compared with only 28.6% of high risk patients with PE. The high risk patients with PE may have had a higher median chest AIS scores compared with moderate risk patients (3 versus 2, high risk versus moderate risk, P ¼ 0.2), although this did not reach statistical significance because of sample size. However, given the small number of PEs observed in our cohort, we cannot comment on the likelihood of true thromboembolic events because of DVTs. Based on our data demonstrating the incidence of VTE, patients with RAP scores 7-9 should be considered for LEDUS. Although this recommendation does not provide a single RAP score at which risk of VTE becomes absolutely necessary to screen for DVTs, we have shown that higher RAP scores even
Table 4 e Logistic regression models for VTE. All patients Variable
Univariate
Multivariate
OR
95% CI
P value
Female
0.60
0.34-1.04
0.07
Weight
1.02
1.01-1.03
<0.01
ISS
1.04
1.02-1.06
<0.01
RAP
1.18
1.10-1.27
<0.01
AIS cervical
1.73
1.03-2.91
0.04
AIS external
4.05
1.48-11.1
<0.01
Days to prophylaxis
1.21
1.05-1.39
<0.01
Female
0.41
0.17-1.03
0.06
Weight
1.02
1.01-1.03
<0.01
ISS
1.05
1.01-1.09
0.02
Blunt trauma
0.56
0.25-1.21
0.14
RAP
1.31
0.98-1.76
0.07
AIS cervical
2.10
1.01-4.36
0.05
1.58
0.87-2.88
0.14
1.86-105.3
0.01
1.31
1.04-1.64
0.02
Weight
1.02
1.00-1.03
0.02
AIS lumbar
4.03
0.62-26.4
0.15
OR
95% CI
P value
1.02
1.01-1.03
<0.01
1.15
1.04-1.26
<0.01
3.53
1.22-1.26
0.02
1.02
1.01-1.04
<0.01
1.06
1.02-1.10
<0.01
1.02
1.00-1.03
0.02
RAP score 5-9
AIS chest AIS external Days to prophylaxis
14.0
RAP score 10
286
j o u r n a l o f s u r g i c a l r e s e a r c h m a r c h 2 0 1 9 ( 2 3 5 ) 2 8 0 e2 8 7
Fig. 2 e Association between RAP score and DVT occurrence.
within the moderate risk definition are associated with increased VTE risk. Clinicians with different practice patterns and biases may choose different RAP scores at which to perform LEDUS examinations; the decision ultimately depends on the risk of VTE that an individual provider is willing to accept. This often becomes problematic for physicians as VTEs are now used as a marker of quality by several health care agencies.29 Increased VTE rates are often portrayed as suboptimal care and may bias clinicians against aggressive DVT screening. There are several limitations to this study. First, this was a retrospective study at a single institution, and extrapolation of our results to other hospitals could be limited. Second, the rates of DVT and PE were not known for patients with RAP scores <5, with the assumption that these rates are low. Although unlikely based on previous RAP score studies, these patients could have similar risk of DVT and PE as the moderate risk patients in our study.13 Third, this study did not take into account rates of VTE on readmission or postdischarge and may not have fully accounted for all DVTs in the discharged trauma population. Fourth, many patients developed VTE by the time of the initial LEDUS. It remains unclear as to the optimal time to screen trauma patients for VTE and LEDUS examinations that occur earlier in patients’ hospital course may be necessary. Fifth, the data analyzed are from 3 y ago and may not accurately reflect the rates of DVT that our center currently experiences. The protocol for LEDUS in trauma patients changed in late 2015, and our group is working to characterize a protocol that is safe and cost-effective. In addition, given the retrospective nature of the study, the number and timing of missed chemoprophylaxis doses were not reliably identified or included in statistical analyses. Finally, as not all trauma centers use LEDUS examinations, and only a proportion of those that screen do so based on the RAP score, these results may not be broadly applicable to all
trauma centers. Although we realize that nonselective LEDUS protocols are cost-prohibitive for all trauma patients in all centers, the information provided by this study may guide physicians in the selection of patients at high and moderate risk for VTE who would most greatly benefit from screening and more aggressive chemoprophylaxis. The RAP score continues to be useful in the identification of moderate- and high-risk patients who require chemoprophylaxis and LEDUS. Patients at moderate risk for VTE have lower DVT rates but similar PE rates compared with RAPstratified high risk patients; therefore, the decision to perform LEDUS examinations should be considered for all moderate risk patients with RAP scores 7-9. Serial examinations need not be undertaken routinely as the rate of clot propagation in our study was low. Finally, nearly two-thirds of acute DVT were identified within 4 d of injury, confirming that DVT occurs and can be detected early after trauma.
Acknowledgment Authors’ contributions: G.E.M., T.A.P., J.B.E., and M.D.G. contributed to the design of the study. G.E.M., A.P., S.W., and K.B. contributed to the acquisition of data. A.P., G.E.M., D.H., V.N., A.T.M., and M.D.G. contributed to data analysis. A.P. and G.E.M. drafted the article. All authors contributed to critical review of the article and have agreed to the final version of the article. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Disclosure None.
martin et al dvt screening in trauma patients
references
1. Jacobs BN, Cain-Nielsen AH, Jakubus JL, et al. Unfractionated heparin versus low-molecular-weight heparin for venous thromboembolism prophylaxis in trauma. J Trauma Acute Care Surg. 2017;83:151e158. 2. Alhassan S, Pelinescu A, Gandhi V, Naddour M, Singh AC, Bihler E. Clinical presentation and risk factors of venous thromboembolic disease. Crit Care Nurs Q. 2017;40:201e209. 3. Geerts WH, Jay RM, Code KI, et al. A comparison of low-dose heparin with low-molecular-weight heparin as prophylaxis against venous thromboembolism after major trauma. N Engl J Med. 1996;335:701e707. 4. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP. A prospective study of venous thromboembolism after major trauma. N Engl J Med. 1994;331:1601e1606. 5. Dobesh PP. Economic burden of venous thromboembolism in hospitalized patients. Pharmacotherapy. 2009;29:943e953. 6. Cipolle MD, Wojcik R, Seislove E, Wasser TE, Pasquale MD. The role of surveillance duplex scanning in preventing venous thromboembolism in trauma patients. J Trauma. 2002;52:453e462. 7. Bandle J, Shackford SR, Kahl JE, et al. The value of lowerextremity duplex surveillance to detect deep vein thrombosis in trauma patients. J Trauma Acute Care Surg. 2013;74:575e580. 8. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L. Prospective ultrasound evaluation of venous thrombosis in high-risk trauma patients. J Trauma. 1993;35:405e408. 9. Spain DA, Richardson JD, Polk Jr HC, Bergamini TM, Wilson MA, Miller FB. Venous thromboembolism in the highrisk trauma patient: do risks justify aggressive screening and prophylaxis? J Trauma. 1997;42:463e469. 10. Dietch ZC, Edwards BL, Thames M, Shah PM, Williams MD, Sawyer RG. Rate of lower-extremity ultrasonography in trauma patients in associated with rate of deep venous thrombosis but not pulmonary embolism. Surgery. 2015;158:379e385. 11. Hegsted D, Gritsiouk Y, Schlesinger P, Gardiner S, Gubler KD. Utility of risk assessment profile for risk stratification of venous thrombotic events for trauma patients. Am J Surg. 2013;205:517e520. 12. Gearhart MM, Luchette FA, Proctor MC, et al. The risk assessment profile score identifies trauma patients at risk for deep vein thrombosis. Surgery. 2000;128:631e640. 13. Haut ER, Chang DC, Pierce CA, et al. Predictors of posttraumatic deep vein thrombosis (DVT): hospital practice versus patient factors-an analysis of the National Trauma Databank (NTDB). J Trauma. 2009;66:994e999. 14. Greenfield LJ, Proctor MC, Rodriguez JL, Luchette FA, Cipolle MD, Cho J. Posttrauma thromboembolism prophylaxis. J Trauma. 1997;42:100e103.
287
15. Kim Y, Abplanalp WA, Jung AD, et al. Endocytosis of red blood cell microparticles by pulmonary endothelial cells is mediated by Rab5. Shock. 2018;49:288e294. 16. Iskander GA, Nelson RS, Morehouse DL, Tenguist JE, Szlabick RE. Incidence and propagation of infrageniculate deep venous thrombosis in trauma patients. J Trauma. 2006;61:695e700. 17. Masuda EM, Kistner RL, Musikasinthorn C, Liquido F, Geling O, He Q. The controversy of managing calf vein thrombosis. J Vasc Surg. 2012;55:550e561. 18. Horner D, Hogg K, Body R, Nash MJ, Baglin T, MackwayJones K. The anticoagulation of calf thrombosis (ACT) project: results from the randomized controlled external pilot trial. Chest. 2014;146:1468e1477. 19. Garry J, Duke A, Labropoulos N. Systematic review of the complications following isolated calf deep vein thrombosis. Br J Surg. 2016;103:789e796. 20. Olsen EF, Zander AL, Van Gent JM, et al. Below-knee deep vein thrombosis: an opportunity to prevent pulmonary embolism? J Trauma Acute Care Surg. 2014;77:459e463. 21. Sharpe RP, Gupta R, Gracias VH, et al. Incidence and natural history of below-knee deep venous thrombosis in high-risk trauma patients. J Trauma. 2002;53:1048e1052. 22. Mulier KE, Greenberg JG, Beilman GJ. Hypercoagulability in porcine hemorrhagic shock is present early after trauma and resuscitation. J Surg Res. 2012;174:e31ee35. 23. Gary JL, Schneider PS, Galpin M, et al. Can thromboelastography predict venous thromboembolic events in patients with severe extremity trauma? J Orthop Trauma. 2016;30:294e298. 24. Knudson MM, Ikossi DG, Khaw L, Morabito D, Speetzen LS. Thromboembolism after trauma: an analysis of 1602 episodes from the American College of Surgeons National Trauma Databank. Ann Surg. 2004;240:490e496. 25. Brakenridge SC, Henley SS, Kashner TM, et al. Comparing clinical predictors of deep venous thrombosis versus pulmonary embolus after severe injury: a new paradigm for posttraumatic venous thromboembolism? J Trauma Acute Care Surg. 2013;74:1231e1237. 26. Velmahos GC, Spaniolas K, Tabbara M, et al. Pulmonary embolism and deep venous thrombosis in trauma: are they related? Arch Surg. 2009;144:928e932. 27. Knudson MM, Gomez D, Haas B, Cohen MJ, Nathens AB. Three thousand seven hundred thirty-eight posttraumatic pulmonary emboli: a new look at an old disease. Ann Surg. 2011;254:625e632. 28. Allen CJ, Murray CR, Meizoso JP, et al. Surveillance and early management of deep vein thrombosis decreases rate of pulmonary embolism in high risk trauma patients. J Am Coll Surg. 2016;222:65e72. 29. National Consensus Standards for the prevention and care of venous thromboembolism. National Quality Forum project. Available at https://www.jointcommission.org/venous_ thromboembolism/. Accessed April 17, 2018.