Retrospective Analysis of LMWH and UFH in Pediatric Trauma Patients: A Comparative Analysis

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 y 2 0 2 0 ( 2 4 9 ) 1 2 1 e1 2 9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Retrospective Analysis of LMWH and UFH in Pediatric Trauma Patients: A Comparative Analysis Michael Hunter Culbert, BS,* Mohammad Hamidi, MD, Muhammad Zeeshan, MD, Kamil Hanna, MD, Andrew Romero, PharmaD, Bellal Joseph, MD, FACS, and Terence O’Keeffe, MB, ChB, MSPH Department of Surgery, University of Arizona College of Medicine, Tucson, Arizona

article info

abstract

Article history:

Background: Chemoprophylaxis with either unfractionated heparin (UFH) or Low-

Received 28 June 2019

Molecular-Weight Heparin (LMWH) are recommended to prevent Venous Thromboembo-

Received in revised form

lism (VTE) after trauma. Experimental work has shown beneficial effects of LMWH in an-

28 September 2019

imal models, but it is unknown if similar effects exist in humans. We hypothesized that

Accepted 3 November 2019

treatment with LMWH is associated with a survival benefit when compared to UFH.

Available online xxx

Methods: We performed a retrospective analysis of our level I trauma center database from January 2009 to June 2018. Pediatric patients (age < 18) were included if they received either

Keywords:

LMWH or UFH during their stay. Outcome measures included mortality, VTE complica-

Trauma

tions, and hospital length of stay (HLOS).

Pediatrics

Results: A total of 354 patients were included. Patients who received LMWH had lower

Mortality

mortality compared to those who received UFH. After multivariate logistic regression,

VTE prophylaxis

LMWH was still independently associated with improved survival. No association was

Low molecular weight heparin

found between LMWH and UFH regarding deep vein thrombosis (DVT) or pulmonary embolism (PE) rates. No association was found between LMWH with HLOS. Conclusions: LMWH was associated with improved survival compared to UFH in our pediatric trauma patients. This was independent of injury severity or VTE complications. Further studies are required to understand better the mechanisms by which LMWH improves survival. Level of Evidence: 3. ª 2019 Elsevier Inc. All rights reserved.

Introduction Venous thromboembolism (VTE) is a serious complication caused by the inappropriate clotting of blood. This can result

in either a deep vein thrombosis (DVT), a clot usually found in the deep veins of the lower extremities, or a pulmonary embolism (PE), a clot found in the major arteries of the lung. VTE complications are clinically significant, and can often result in

We have no conflicts of interest to disclose. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declarations of interest: none. * Corresponding author. Department of Surgery, University of Arizona College of Medicine, 1501 N Campbell Avenue, Tucson, AZ 85724. Tel.: þ1 5206036799; fax: þ1 5206265016. 0022-4804/$ e see front matter ª 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jss.2019.11.019

122

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 y 2 0 2 0 ( 2 4 9 ) 1 2 1 e1 2 9

a fatality, with mortality rates of up to 10%-30% within 30 d of onset.1 Trauma and orthopedic injuries associated with trauma are significant risk factors for developing VTE, irrespective of age.1,2 Although the incidence is significantly higher in adults, and traditionally children are thought to be low-risk for VTE, the rate of VTE in the pediatric population is increasing, highlighting the importance of chemical and ambulatory prophylactic care.3 The current standard of care for VTE prophylaxis in adults is the administration of either low-molecular-weight heparin (LWMH) or unfractionated heparin (UFH).4 In pediatric patients, the standard of care is less clear and can be dependent on the age of the patient and the severity of their injury.5 Current EAST guidelines state that pharmacologic prophylaxis be used only in children over fifteen or postpubertal children under the age of 15 with an ISS greater than 25.5 However, these guidelines are not definitive due to the lack of supportive data and low quality of evidence. While both UFH and LMWH are used regularly in a clinical setting, the lower cost and more predictable dose-response has led to LMWH often being the preferred choice of chemoprophylaxis.6-8 In addition, LMWH has also been shown to demonstrate beneficial effects in vitro and in vivo animal models,9-11 although there is little data to support whether this benefit translates to the pediatric population. While clinical data support the use of LMWH, there could be other mechanisms still to be elucidated that impact the benefit it may have on patient survival. Given this prior research, we planned to investigate if a similar outcome would be seen in pediatric trauma patients that received LMWH compared to UFH. Our hypothesis was that LMWH would confer a survival benefit in pediatric trauma patients that receive it when compared to UFH.

(mechanism of injury, injury severity score [ISS], and body regions abbreviated injury scale score (AIS) [head-AIS, thoraxAIS, abdomen-AIS, extremity-AIS]); emergency department (ED) vitals (Glasgow Coma Scale [GCS], systolic blood pressure [SBP], heart rate [HR], and respiratory rate [RR]); type of operative intervention (craniotomy, craniectomy, laparotomy, laparoscopy, colectomy, thoracotomy, cardiac procedure, open reduction internal fixation [tibia and fibula, tarsals and metatarsals, femur, ankle, carpals and metacarpals, humerus, facial bones, radius and ulna], vertebral operative intervention); ventilation days; ICU length of stay (LOS); type of thromboprophylactic agent used (LMWH versus UFH); complications (DVT, PE, superficial surgical site infection [SSSI], organ space surgical site infection (SI), deep surgical site infection (SSI), acute renal failure [ARF], sepsis, and mortality); hospital characteristics (hospital LOS, ICU length of stay, and hospital charges). The definitions of our variables were defined based on the descriptions from the Trauma Quality Improvement Database (TQIP). AIS is defined as the following abbreviated definition: “severity codes that reflect the patient’s injuries.” The Scale is ranked from 1 to 6, with 1 being classified as a “minor injury” and 6 being classified as “maximum injury, virtually unsurvivable.” SSI is defined as the following abbreviated TQIP guidelines: “involves deep soft tissues of the incision (e.g., fascial and muscle layers) and has purulent drainage or organisms identified from the incision and has either a fever >38 C, localized pain or tenderness, or other signs of infection.” Organ space SI is defined as the following abbreviated TQIP guidelines: “infection involving any part of the body deeper than the fascial/muscle layers that is opened or manipulated during the operative procedure and has either purulent drainage from a drain that is placed into the organ/ space, organisms identified from organ space, or an abscess or other evidence of infection involving the organ/space.”

Methods

Statistical analysis

Study design, population, and outcomes

Data are reported as proportions for categorical variables (%), as the means with standard deviation for continuous parametric variables (mean  SD), and as medians with interquartile range (median [IQR]) for continuous nonparametric variables. We performed the student’s t-test and the ManneWhitney U-test to explore the differences between parametric and nonparametric continuous variables in the two groups. The Pearson Chi-square (X2) test was used to compare the categorical variables between the two groups. Alpha was set at 5%, and a P-value <0.05 was considered statistically significant. We performed univariate logistic and linear regression analyses to assess the association between each variable and the outcomes. The variables with the significant association on univariate analysis (i.e., P-value < 0.2) were then used in a multivariate regression model. On multivariate logistic and linear regression analyses, the association was considered significant if the P-value was <0.05. Multivariate regression was performed to control for possible confounders, including age, gender, race, mechanism of injury, ISS, GCS, ED SBP, ED HR, hospital and ICU LOS, ventilation days, and type of operative intervention. For comparing survival between the two

We performed a retrospective analysis of our level I trauma center pediatric database from January 5, 2009, to June 30, 2018. We included all pediatric trauma patients who received thromboprophylaxis with either LMWH or UFH. Inclusion criteria were patients under 18 years of age and have received either LMWH or UFH for thromboprophylaxis during their hospital stay. Patients were excluded if they were dead on arrival or had medication crossover during their stay. Patients were stratified into two groups; LMWH versus UFH. Our primary outcome measure was survival. Secondary outcome measures were VTE complications, hospital length of stay, and hospital charges. Appropriate Institutional Review Board (IRB) review and approval were obtained for the study. Informed consent was not required from the patients due to the retrospective nature of the analysis.

Variables analyzed The following variables were evaluated for each patient: demographics (age, gender, race, ethnicity); injury parameters

123

culbert et al  lmwh and ufh in pediatric trauma patients

Table 1 e Patient characteristics. UFH (n ¼ 237)

LMWH (n ¼ 117)

P-value

Age, y, mean  SD

11.4  5.9

15.4  2.4

<0.01

Gender, male, n (%)

147 (62.0)

71 (60.7)

0.80

Race, White, n (%)

199 (84.0)

104 (88.9)

0.21

Ethnicity, Hispanic/Latino, n (%)

123 (51.9)

64 (54.7)

0.62

200 (84.4)

99 (84.6)

0.95

Variable

Mechanism of injury, n (%) Blunt Penetrating Other

8 (3.4)

2 (1.7)

29 (12.2)

16 (13.7)

Injury characteristics, median [IQR] 17 [10-28]

11 [9-19]

<0.01

Head-AIS

4 [3-5]

2 [2-3]

<0.01

Thorax-AIS

3 [2-3]

2 [2-3]

0.25

Abdomen-AIS

3 [2-4]

2 [2-3]

0.02

Extremity-AIS

2 [2-3]

3 [2-3]

0.01

GCS

14 [3-15]

15 [14-15]

<0.01

ED SBP, mm Hg

123  27

131  19

<0.01

ED HR, BPM

114  36

100  23

<0.01

21  7

20  5

<0.01

144 (60.8)

84 (71.8)

0 (0)

ISS

Emergency department (ED) vitals, mean  SD

ED RR Operative intervention, n (%) Craniotomy, n (%)

0.04

2 (1.7)

0.32

Craniectomy, n (%)

7 (3)

0 (0)

0.45

Laparotomy, n (%)

19 (8)

5 (4.3)

Laparoscopy, n (%)

9 (4)

1 (0.9)

0.21

Colectomy, n (%)

6 (2.5)

6 (5)

0.07

Thoracotomy, n (%)

3 (1.3)

1 (0.9)

0.73

Cardiac procedure, n (%)

3 (1.3)

0 (0)

0.22

Ventilation days, median [IQR]

0.11

<0.01

5 [2-10]

2 [1-4]

Vertebral operative intervention, n (%)

2 (0.8)

2 (1.8)

0.52

Open reduction internal fixation, n (%)

25 (11)

27 (23)

<0.01

7 (6)

<0.01

Tibia and fibula ORIF, n (%)

2 (0.8)

Tarsals and metatarsals ORIF, n (%)

1 (0.4)

Femur ORIF, n (%)

4 (8.5)

0 (0.0)

0.48

10 (8.5)

<0.01

Ankle ORIF, n (%)

0 (0.0)

1 (0.9)

0.15

Carpals and metacarpals ORIF, n (%)

0 (0.0)

1 (0.9)

0.15

Humerus ORIF, n (%) Facial bones ORIF, n (%) Radius and ulna ORIF, n (%)

4 (1.7) 12 (5.1) 2 (0.8)

groups, Kaplan Meier survival curves were plotted and compared using the log-rank test. For further evaluating the outcomes across age groups, a subanalysis was performed by breaking down the patient population into three groups (ages 0 through 9; age 10 through 14; age 15 through 17).

Results From the 9 years of available data and an average annual trauma volume of 200-300 pediatric patients, a total of 354

0 (0.0) 8 (6.8) 0 (0.0)

0.16 50 0.32

pediatric trauma patients were identified (Table 1) shows their demographics and injury characteristics). The sample was stratified based on the thromboprophylaxis agent used: 237 UFH and 117 LMWH. Three patients had medication crossover over during their stay and were not included in the study population. Mean age was 12.7  5.4 years, 61.6% were male and 86% were white. The majority of patients (84%) sustained blunt trauma compared to (13%) penetrating and (3%) burns, with a mean ISS of 17 [9-26]. Most patients were normotensive with an SBP of 126  22 mmHg but tachycardic with a pulse of 109  28 beats per minute (BPM). Level of consciousness and

124

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 y 2 0 2 0 ( 2 4 9 ) 1 2 1 e1 2 9

Table 2 e Outcome measures. UFH (n ¼ 237)

LMWH (n ¼ 117)

P-value

DVT

4 (1.7)

0 (0)

0.15

PE

2 (0.8)

0 (0)

0.32

SSSI

1 (0.4)

2 (1.7)

0.21

Organ space SI

1 (0.4)

0 (0)

0.48

Deep SSI

1 (0.4)

0 (0)

0.48

ARF

3 (1.3)

1 (0.9)

0.73

Variable Complications, n (%)

Sepsis

1 (0.4)

0 (0)

0.48

24 (10.1)

1 (0.9)

0.001

Hospital LOS, d, median [IQR]

8 [4-17]

4 [2-8]

<0.01

ICU LOS, d, median [IQR]

5 [3-11]

3 [1-5]

<0.01

174,297  223,714

94,485  102,720

<0.01

Mortality Hospital characteristics

Hospital charges, $, mean  SD

neurological status was generally preserved with a median GCS of 14 [5-15]. 64% of patients underwent an operative intervention following admission. In comparison to patients who received UFH, those who received LMWH were older (15.4  2.4 versus 11.4  5.9; P < 0.01), had a higher GCS on presentation (15 [14-15] versus 14 [3-15]; P < 0.01), higher ED SBP (131  19 versus 123  27; P ¼ 0.008), less ventilation days (2 [1-4] versus 5 [2-10]; P < 0.01) and were more likely to undergo an operative intervention (71.8% versus 60.8%; P ¼ 0.04), specifically open reduction internal fixation (23% versus 11%; P < 0.01). Patients who received LMWH sustained injuries of lower severity (11 [9-19] versus 17 [10-28]; P < 0.01) especially in the head (2 [2-3] versus 4 [3-5]; P < 0.01) and abdominal region (2 [2-3] versus 3 [2-4]; P ¼ 0.02). No significant differences were found between the two groups regarding gender (P ¼ 0.8), race (P ¼ 0.21), ethnicity (P ¼ 0.62), mechanism of injury (P ¼ 0.95), and operative intervention type ([craniotomy (P ¼ 0.32), [craniectomy (P ¼ 0.34)], [laparotomy (P ¼ 0.11], [laparoscopy

(P ¼ 0.21)], [colectomy (P ¼ 0.07)], [thoracotomy (P ¼ 0.73)], [cardiac procedure (P ¼ 0.22)], [vertebral operative intervention (P ¼ 0.52)]). Table 2 demonstrates the difference in outcomes between the two groups. Patients who received LMWH had a lower mortality rate (0.9% versus 10.1%, P ¼ 0.001) compared to those who received UFH. After multivariable logistic regression analysis, adjusting for the following variables (age, gender), injury parameters (ISS, body regions AIS), admission parameters (GCS, ED SBP), ventilation days, ICU LOS, the need and type of operative intervention (craniotomy, craniectomy, laparotomy, laparoscopy, colectomy, thoracotomy, cardiac procedure, open reduction internal fixation, vertebral operative intervention), LMWH was an independent predictor of survival in comparison to UFH (OR 1.11 [1.05-1.2]; P ¼ 0.04). After separating the population into three different age groups (age 0 through 9; age 10 through 14; age 15 through 17) and assessing for survival based on prophylaxis

Table 3 e Univariate and multivariate logistic regression analysis of survival. Variable

Univariate analysis OR [95% CI]

Multivariate analysis P-value

OR [95% CI]

P-value

Age

1.12 [1.04-1.18]

0.004

1.09 [0.94-1.23]

Gender, male

0.88 [0.39-2.18]

0.82

-

-

Race, white

0.80 [0.13-2.87]

0.92

-

-

-

0.25

Mechanism of injury, blunt

1.70 [0.58-3.74]

0.53

ISS

0.84 [0.91-0.97]

<0.01

0.95 [0.90-0.99]

0.03

-

GCS

1.42 [1.18-1.47]

<0.01

1.31 [1.15-1.49]

<0.01

ED SBP

1.22 [1.01-1.04]

<0.01

1.01 [0.99-1.04]

0.08

ED HR

0.89 [0.98-1.01]

0.41

-

-

Hospital LOS

1.19 [1.01-1.18]

0.03

1.25 [1.10-1.42]

<0.01

ICU LOS

1.74 [0.97-1.81]

0.25

-

Operative intervention

5.26 [2.13-12.9]

<0.01

1.41 [0.41-4.80]

0.57

LMWH

3.04 [1.74-5.67]

0.01

1.11 [1.05-1.20]

0.04

-

125

culbert et al  lmwh and ufh in pediatric trauma patients

Table 4 e Univariate and multivariate linear regression analysis of hospital length of stay. Variable

Univariate analysis

Multivariate analysis

B [95% CI]

P-value

B [95% CI]

Age

0.66 [0.89 to 0.42]

<0.01

0.17 [0.33 to 0.01]

0.031

Gender, male

2.00 [4.73 to 0.72]

0.15

0.27 [1.82 to 1.28]

0.73

Race, white

0.63 [3.14 to 4.41]

0.74

e

Mechanism of injury, blunt

0.11 [3.54 to 3.77]

0.95

e

ISS

0.41 [0.31 to 0.52]

<0.01

0.01 [0.06 to 0.08]

0.81

GCS

0.79 [1.04 to 0.54]

<0.01

0.02 [0.18 to 0.14]

0.79

ED SBP

0.06 [0.11 to 0.01]

0.006 [0.36 to 0.02]

0.67

ED HR

0.09 [0.05 to 0.13]

<0.01

0.005 [0.02 to 0.03]

ICU LOS

1.07 [1.00 to 1.13]

<0.01

1.02 [0.94 to 1.08]

Operative intervention

6.16 [3.45 to 8.86]

<0.01

2.85 [1.13 to 4.57]

0.001

Heparin versus LMWH

6.37 [3.63 to 9.12]

<0.01

0.39 [1.53 to 2.32]

0.68

administration, patients age 10 through 14 were less likely to survive if they received Heparin compared to LWMH (0.0% versus 10.1%, P ¼ 0.04). Patients age 9 showed clinically significant improvement in survival if they received LWHW; however, it was not statistically significant (0.0% versus 16.9%, P ¼ 0.37). Patients age 15 through 17 showed no difference in survival if they received LMWH or UFH. GCS was also a significant predictor for survival (OR 1.31 [1.15-1.49]; P < 0.01). On the other hand, ISS (OR 0.95 [0.9-0.99]; P ¼ 0.03) was an independent predictor of increased mortality. Table 3 demonstrates the results of univariate and multivariate regression analyses. There were no significant differences between the two groups regarding rates of all in-hospital complications, such as DVT (0% versus 1.7%; P ¼ 0.15), PE (0% versus 0.8%; P ¼ 0.32), SSSI (1.7% versus 0.4%; P ¼ 0.21), ARF (0.9% versus 1.3%; P ¼ 0.73), and sepsis (0% versus 0.4%; P ¼ 0.48) (Table 3). Patients in the LMWH group had significantly shorter hospital LOS (4 [2-8] versus 8 [4-17]; P < 0.01), ICU LOS (3 [1-5] versus 5 [311]; P < 0.01) and lower mean hospital charges (P < 0.01). Although UFH was a significant predictor of longer hospital LOS on univariate analysis, after adjustment for confounding variables, the choice of thromboprophylaxis agent was no longer a significant predictor of hospital LOS (b ¼ 0.39 [1.53 to 2.32]; P ¼ 0.68). Significant predictors of longer hospital LOS were operative intervention (b ¼ 2.85 [1.13-4.57]; P ¼ 0.001) while younger age was a predictor of a shorter hospital LOS (b ¼ 0.17 [0.33 to 0.01]; P ¼ 0.031). Factors that did not have an independent effect on hospital LOS were gender (P ¼ 0.73), ISS (P ¼ 0.81), GCS (P ¼ 0.79), and ED SBP (P ¼ 0.67). Table 4 demonstrates the results of linear regression analysis. Figure demonstrates Kaplan Meier survival curves for the two groups. Survival in the LMWH group was significantly higher compared to the UFH group (P ¼ 0.014).

Discussion These data supported our hypothesis, indicating that LMWH independently improved survival in this group of moderate to

0.03

P-value

-

0.69 <0.01

severely injured pediatric trauma patients when compared to UFH. This apparent advantage was not related to or confounded by the reduction in VTE associated morbidity and mortality, sepsis, SSSI, and ARF complication rates. This survival benefit could be explained by the different pharmacokinetic and pharmacodynamic profiles of the medications. While LMWH is derived from UFH, there are notable differences between how each medication gains its anticoagulant activity. UFH is indirect and stems from the binding of UFH to antithrombin (AT).12 This binding produces rapid inhibition of thrombin and other clotting factors, reducing the ability of blood to coagulate effectively.12 UFH also binds to plasma proteins and cells, giving it an unpredictable pharmacological profile in vivo so that patients have more variation in anticoagulation state.7 LMWH, prepared by depolymerization of UFH, is one-third the molecular weight and has less affinity to proteins and cells, giving it a more predictable dose-response profile, increasing its specificity, and increasing its halflife.7,13,14 Additionally, UFH has been shown to be more difficult to maintain at therapeutic levels for the pediatric population.15 This is because primarily UFH is renally cleared at a higher rate in pediatric patients.16 Second, the presentation of VTE complications in children are often not immediately recognized in a clinical setting, or they present with atypical symptoms.16,17 This could potentially increase the rate at which UFH is cleared by the body from the already progressed clotting pathology.16,18 LMWH also has a lower incidence of heparin-induced thrombocytopenia (HIT) and probably reduced risk of osteoporosis, further contributing to the benefit to the patient.19,20 Animal studies that have investigated the varying effects of LMWH could also explain why our LMWH patient population had such a profound mortality benefit. A study published in the CNS Drug Reviews by Stutzmann et al.11 found that enoxaparin, an LMWH, exhibited neuroprotective effects to rats that suffered from cerebral ischemia, brain edema, or traumatic brain injuries (TBI). In the TBI model, they report that enoxaparin promoted significant memory improvement in a dose-dependent fashion. They also showed LMWH was more effective at preventing microthrombus and blood

126

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 y 2 0 2 0 ( 2 4 9 ) 1 2 1 e1 2 9

Fig e Kalpan meier survival curve analysis. (Color version of figure is available online.)

coagulation with less risk of hemorrhage for the same antiXa activity when compared to UFH. Enoxaparin and UFH were both found to reduce cerebral inflammation; however, enoxaparin had a lower potential for hemorrhage and longer halflife, suggesting it is more beneficial and safer to use for its possible neuroprotective effects.11 While this neuroprotective benefit of enoxaparin has been replicated in humans in a randomized clinical trial, the small patient population of the study limits the accuracy of the results.21 In regards to our study, the majority of our patient population with a worse head-AIS score received UFH rather than LWMH, suggesting that UFH may not supply this same benefit. Both UFH and LMWH have also been shown to demonstrate antiinflammatory properties.22,23 This is significant from the perspective of trauma, as the inflammation from trauma, especially when prolonged, has been associated with a worse overall prognosis.24 One animal study observed LMWH to markedly reduce the antiinflammatory response with a similar mechanism of potent nonsteroidal anti-inflammatory drugs (NSAIDs), while UFH did not.25 Additionally, a different study by Downing et al.23 suggests that the mechanism for LMWH’s anti-inflammatory effects is completely separate from its anticoagulant properties, further demonstrating that LMWH may have an unperceived underlying benefit for those who receive it for their VTE prophylaxis. In regards to our study population, specifically, a reduction in systemic inflammation may significantly contribute to the perceived mortality benefit seen in the LMWH trauma patients.

Furthermore, the currently available evidence also suggests that LMWH provides a benefit to patients through a reduction in circulating Von Willebrand Factor (vWF). vWF, a glycoprotein released by endothelial cells after primary hemostasis, is important in mediating adhesion of platelets and the binding of clotting factors.26,27 UFH has been shown to release more vWF when compared to LMWH.28-30 However, an increased serum level of vWF has been associated with higher rates of revascularization, myocardial infarction (MI), and mortality.31 In children specifically, elevated serum vWF has been shown to be related to vascular damage.32 There is some evidence that LMWHs, particularly enoxaparin, lowers serum vWF; however, conclusions from this study should be tempered, as it was performed with a small sample size.29 Although patients with more severe injuries received UFH, both LMWH and UFH demonstrated no difference in VTE associated morbidity or mortality in our study population. Current literature investigating chemical prophylaxis in pediatric patients is very limited, and there is no clear evidence as to which medication is better at preventing VTE complications. Available evidence investigating the VTE rate in adults typically states that LMWH has a higher efficacy in VTE prevention.33,34 Jacobs et al.35 found that LMWH was better at reducing the rate of VTE, PE, DVT, and mortality in trauma patients. However, it should be noted that a different study that targeted the nontrauma patients found that LMWH and UFH are comparable, with no treatment providing a significant reduction in VTE complication

culbert et al  lmwh and ufh in pediatric trauma patients

rate.36 While our study observed no statistical difference in overall VTE rate between those that received UFH or LWMH, we did observe mortality differences between different age groups. UFH was associated with a higher likelihood of mortality for patients aged 10 to 14. Patients below age 10 demonstrated clinically significant findings of increased mortality associated with UFH use. Interestingly, post-pubertal patients between the ages of 15 and 17 demonstrated no difference in mortality rate between LMWH and UFH use. This may suggest that LMWH is more beneficial in younger patient populations versus those with adult or near-adult physiology. Their is a lack of literature in chemical VTE prophylaxis for children with few guidelines or algorithms regarding VTE screening or VTE prophylaxis protocol. Our institution had no definitive guidelines for VTE prophylaxis in place for the years data was collected. Prophylaxis administration was largely done on an independent case-by-case basis with a multi-disciplinary approach. It was typically considered unnecessary to administer prophylaxis if the patient was under 13, post-pubertal or if they were ambulatory, which is recommended by EAST guidelines. Patients with TBI or other neurological injuries were placed on prophylaxis according to neurosurgical recommendations. For these patients, prophylaxis did not start until 24 h after a stable neurological exam and head imaging. For patients with isolated spinal injuries, LMWH was typically the preferred anticoagulant; however, UFH is the more preferred choice of prophylaxis for patients with head injuries due to its reversibility in light of a significant bleeding event. VTE prophylaxis was administered on a symptomatic basis and was not routinely screened for in our population, based upon EAST guidelines for VTE screening in pediatric populations.5 We recommend that multicenter prospective studies be performed to determine which medication is more effective at preventing VTE complications in this population specifically. Patients who received UFH were more likely to stay in the hospital for twice as long when compared to those that received LMWH. However, this can easily be attributed to the fact that UFH recipients sustained more serious injuries and had a lower GCS on admission. Upon multivariate analysis and after controlling for confounding variables, UFH did not have an impact on hospital LOS. Admission GCS also did not impact hospital LOS; however, it was an independent predictor of survival. It should be noted that results investigating LOS do not account for insurance or other complications that develop during the patient’s care that could prolong LOS. Hospital charges were also lower in patients that received LMWH, but again this is likely explained by the fact that patients who received UFH had a longer hospital LOS and more severe injuries. However, several studies investigating patient cost versus VTE prophylaxis found LMWH, while more expensive upfront, reduces the cost of care through fewer complications.25,33 While our results support the overall lower hospital costs of LMWH, it cannot be drawn conclusively due to the lack of VTE incident and other controlled variables. We recommend explicit cost studies be performed to determine which medication treatment has a greater cost benefit for patients. The limitations of our study stem from its retrospective nature performed in a single level one trauma center (not

127

specifically designated for pediatric trauma until 2018). The study was designed to reduce bias as much as possible within the scope of the utilized database using multivariable logistic regression analysis and adjusting for all measurable confounding factors. With this in mind, there still can be unmeasurable variables, and we cannot account for these residual confounding factors. This is a universal limitation within our study that cannot be adjusted for. Additionally, while the data set and results reflect our hospital, this limits the sample size and the ability to extrapolate conclusions nationally. We are currently working on a similar project looking at patients from the TQIP, which may help strengthen our conclusions. The low incidence of VTE complications, while not unusual for a pediatric population, limits the conclusions that can be drawn, and a larger sample size would provide more reliable results. Our dataset did not support a true control group with patients that received no prophylaxis, and we were unable to provide additional analysis on comparing a control group with our patient population. This limits our ability to determine if prophylaxis is increasing mortality in certain patient groups. Information on when prophylaxis started and the number of doses received was unavailable; however, this does not impact the fact that LMWH provided a survival benefit given that each group had no difference in VTE complication rate. While admission GCS and head-AIS were available in our database, we were unable to determine if one form of prophylaxis improved neurological outcomes due to limitations in our data set and lack of data recording any changes in GCS. We recommend a prospective, multicenter study be performed to assess if LMWH provides a significant improvement in neurological outcomes. We were unable to obtain the exact cause of death, limiting the findings of fatal complications that one form of prophylaxis may have been associated with. For operative intervention, we were unable to control for surgeon variety in each group. Last, we were unable to obtain data on why one form of prophylaxis was used over another, as our hospital is still developing a definitive standard of care protocol for VTE prophylaxis in pediatric trauma patients. This limits the application of the results toward medication effectiveness for specific indications.

Conclusion Our results show an apparent association between the use of LMWH and survival in pediatric trauma patients, although the exact mechanism of this beneficial effect remains to be elucidated, especially in light of the fact that LMWH and UFH were comparable in rates of VTE complication and hospital LOS. Further prospective studies should be done to investigate and confirm the observed potential benefits of LMWH in this group of trauma patients.

Acknowledgment Author contributions: Michael Culbert was responsible for literature review and primary manuscript writing.

128

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 y 2 0 2 0 ( 2 4 9 ) 1 2 1 e1 2 9

Mohammad Hamidi, Muhammad Zeeshan, and Kamil Hanna were responsible for statistical analysis, interpretation of data, statistical support, and primary manuscript writing. Andrew Romero was responsible for data retrieval. Bellal Joseph was responsible for project oversight. Terence O’Keeffe was responsible for hypothesis generation, analysis of data, manuscript revision, and project oversight.

Disclosure The authors reported no proprietary or commercial interest in any product mentioned or concept discussed in this article.

references

1. Beckman MG, Hooper WC, Critchley SE, Ortel TL. Venous thromboembolism: a public health concern. Am J Prev Med. 2010;38(4 Suppl l):S495eS501. 2. Heit JA. The epidemiology of venous thromboembolism in the community: implications for prevention and management. J Thromb Thrombolysis. 2006;21:23e29. 3. Mahajerin A, Croteau SE. Epidemiology and risk assessment of pediatric venous thromboembolism. Front Pediatr. 2017;5:68. 4. Laryea J, Champagne B. Venous thromboembolism prophylaxis. Clin Colon Rectal Surg. 2013;26:153e159. 5. Mahajerin A, Petty JK, Hanson SJ, et al. Prophylaxis against venous thromboembolism in pediatric trauma: a practice management guideline from the Eastern association for the Surgery of trauma and the pediatric trauma Society. J Trauma Acute Care Surg. 2017;82:627e636. 6. Pineo G, Lin J, Stern L, Subrahmanian T, Annemans L. Economic impact of enoxaparin versus unfractionated heparin for venous thromboembolism prophylaxis in patients with acute ischemic stroke: a hospital perspective of the PREVAIL trial. J Hosp Med. 2012;7:176e182. 7. Hirsh J, Bauer KA, Donati MB, Gould M, Samama MM, Weitz JI. Parenteral anticoagulants: American College of chest Physicians evidence-based clinical practice guidelines (8th Edition). Chest. 2008;133(6 Suppl l):141Se159S. 8. Hatam N, Bahmei J, Keshavarz K, Feiz F, Sedghi R, BorhaniHaghighi A. Cost-effectiveness analysis of the unfractionated heparin versus low-molecular-weight heparin in hospitalized patients with stroke due to atrial fibrillation in Shiraz, South of Iran. J Vasc Interv Neurol. 2017;9:6e12. 9. Carter CJ, Kelton JG, Hirsh J, Cerskus A, Santos AV, Gent M. The relationship between the hemorrhagic and antithrombotic properties of low molecular weight heparin in rabbits. Blood. 1982;59:1239e1245. 10. Ritter EF, Cronan JC, Rudner AM, Serafin D, Klitzman B. Improved microsurgical anastomotic patency with low molecular weight heparin. J Reconstr Microsurg. 1998;14:331e336. 11. Stutzmann JM, Mary V, Wahl F, Grosjean-Piot O, Uzan A, Pratt J. Neuroprotective profile of enoxaparin, a low molecular weight heparin, in in vivo models of cerebral ischemia or traumatic brain injury in rats: a review. CNS Drug Rev. 2002;8:1e30. 12. Harter K, Levine M, Henderson SO. Anticoagulation drug therapy: a review. West J Emerg Med. 2015;16:11e17. 13. Dabbous MK, Sakr FR, Malaeb DN. Anticoagulant therapy in pediatrics. J Basic Clin Pharm. 2014;5:27e33.

14. Hirsh J. Low molecular weight heparin. Thromb Haemost. 1993;70:204e207. 15. Andrew M, Marzinotto V, Massicotte P, et al. Heparin therapy in pediatric patients: a prospective cohort study. Pediatr Res. 1994;35:78e83. 16. McDonald MM, Jacobson LJ, Hay WW, Hathaway WE. Heparin clearance in the newborn. Pediatr Res. 1981;15:1015e1018. 17. Male C, Kuhle S, Mitchell L. Diagnosis of venous thromboembolism in children. Semin Thromb Hemost. 2003;29:377e390. 18. Hirsh J, van Aken WG, et al. Heparin kinetics in venous thrombosis and pulmonary embolism. Circulation. 1976;53:691e695. 19. Dix D, Andrew M, Marzinotto V, et al. The use of low molecular weight heparin in pediatric patients: a prospective cohort study. J Pediatr. 2000;136:439e445. 20. Monagle P, Chan AKC, Goldenberg NA, et al. Antithrombotic therapy in neonates and children: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 Suppl l):e737See801S. 21. Baharvahdat H, Ganjeifar B, Etemadrezaie H, Farajirad M, Zabihyan S, Mowla A. Enoxaparin in the treatment of severe traumatic brain injury: a randomized clinical trial. Surg Neurol Int. 2019;10:10. 22. Young E. The anti-inflammatory effects of heparin and related compounds. Thromb Res. 2008;122:743e752. 23. Downing LJ, Strieter RM, Kadell AM, Wilke CA, Greenfield LJ, Wakefield TW. Low-dose low-molecular-weight heparin is anti-inflammatory during venous thrombosis. J Vasc Surg. 1998;28:848e854. 24. Pierce A, Pittet JF. Inflammatory response to trauma: implications for coagulation and resuscitation. Curr Opin Anaesthesiol. 2014;27:246e252. 25. Ceccarelli M, Bani D, Cinci L, et al. Anti-inflammatory effects of low molecular weight heparin derivative in a rat model of carrageenan-induced pleurisy. J Cell Mol Med. 2009;13:2704e2712. 26. Sadler JE. Biochemistry and genetics of von Willebrand factor. Annu Rev Biochem. 1998;67:395e424. 27. Wagner DD. Cell biology of von Willebrand factor. Annu Rev Cell Biol. 1990;6:217e246. 28. Montalescot G, Collet JP, Lison L, et al. Effects of various anticoagulant treatments on von Willebrand factor release in unstable angina. J Am Coll Cardiol. 2000;36:110e114. 29. Montalescot G, Bal-dit-Sollier C, Chibedi D, et al. Comparison of effects on markers of blood cell activation of enoxaparin, dalteparin, and unfractionated heparin in patients with unstable angina pectoris or non-ST-segment elevation acute myocardial infarction (the ARMADA study). Am J Cardiol. 2003;91:925e930. 30. Ho¨dl R, Huber K, Kraxner W, et al. Comparison of effects of dalteparin and enoxaparin on hemostatic parameters and von Willebrand factor in patients with unstable angina pectoris or noneST- segment elevation acute myocardial infarction. Am J Cardiol. 2002;89:589e592. 31. Sonneveld MA, Franco OH, Ikram MA, et al. Von Willebrand factor, ADAMTS13, and the risk of mortality: the Rotterdam study. Arterioscler Thromb Vasc Biol. 2016;36:2446e2451. 32. Cetın N, Sav NM, Karabel D, Yildirim A, Yıldız B. Serum albumin and von Willebrand factor: possible markers for early detection of vascular damage in children undergoing peritoneal dialysis. Clin Invest Med. 2016;39:E111eE119. 33. Fowler RA, Mittmann N, Geerts W, et al. Cost-effectiveness of dalteparin vs unfractionated heparin for the prevention of

culbert et al  lmwh and ufh in pediatric trauma patients

venous thromboembolism in critically ill patients. JAMA. 2014;312:2135e2145. 34. Aujesky D, Smith KJ, Cornuz J, Roberts MS. Cost-effectiveness of low-molecular-weight heparin for treatment of pulmonary embolism. Chest. 2005;128:1601e1610. 35. Jacobs BN, Cain-Nielsen AH, Jakubus JL, et al. Unfractionated heparin versus low-molecular-weight heparin for venous

129

thromboembolism prophylaxis in trauma. J Trauma Acute Care Surg. 2017;83:151e158. 36. Rothberg MB, Pekow PS, Lahti M, Lindenauer PK. Comparative effectiveness of low-molecular-weight heparin versus unfractionated heparin for thromboembolism prophylaxis for medical patients. J Hosp Med. 2012;7:457e463.