Abstract
Aim
It has been widely reported that pre‐injury use of anticoagulant and antiplatelet agents can affect traumatic brain injury and the associated risk of mortality, however, the effect of these agents on non‐head injury site‐related blood loss remains unclear. Therefore, we investigated the effects of pre‐injury anticoagulant and antiplatelet agents on the transfusion amount and the need for massive transfusion in elderly patients with severe trauma.
Methods
We retrospectively reviewed a cohort of elderly patients with severe trauma (age, ≥65 years; Injury Severity Score, ≥16) between September 2006 and March 2014. The selected patients were subsequently divided into patients who were: only taking warfarin, aspirin, or clopidogrel; taking various combinations of these agents; and a control group who were not taking any of these agents.
Results
During the study period, 67 patients (20%) were taking anticoagulant and antiplatelet agents and 272 patients were included in the control group. Among these patients, 10 were receiving only warfarin, 28 were receiving only aspirin, 14 were receiving only clopidogrel, and 13 were receiving various combinations of these medications. The amount of red cell concentrate and need for massive transfusions were only significantly increased in the warfarin group (P < 0.05). Furthermore, the warfarin group had a significantly higher risk of needing a massive transfusion with multivariate logistic regression analysis (odds ratio, 5.03; 95% confidence interval, 1.25–20.20; P < 0.05).
Conclusion
Patients who were receiving only warfarin before their injury had an increased risk of bleeding due to non‐head injuries.
Keywords: Anticoagulant agent, antiplatelet agent, blood loss, elderly, trauma
Introduction
In recent years, the number of elderly individuals and their life expectancy have been increasing in most developed countries.1 As a result, an increasing number of elderly patients with severe trauma are being admitted to emergency departments. Unfortunately, elderly patients have higher trauma‐related mortality rates relative to younger patients, due to physiological changes that cause a decline in their baseline functions.2, 3, 4, 5 Therefore, early aggressive resuscitation and careful monitoring may be warranted in a large number of elderly patients with severe trauma.
In this context, anticoagulant or antiplatelet agents are commonly prescribed to prevent and treat thromboembolic events, and many elderly patients with trauma receive these treatments. However, the number of elderly patients using these agents before their injury is also increasing.6, 7 It has also been widely reported that pre‐injury use of anticoagulant and antiplatelet agents can affect traumatic brain injury and the associated risk of mortality,8, 9, 10, 11, 12 although few reports have examined the effects of bleeding at non‐head injury sites.6, 7, 13, 14 Furthermore, there are reports regarding the associated risk of mortality in these cases, although very few studies have evaluated the associated blood loss. Thus, the effect of these agents on non‐head injury site‐related blood loss remains unclear. Therefore, we investigated the effects of pre‐injury anticoagulant and antiplatelet agents on the transfusion amount and the need for massive transfusion in elderly patients with severe trauma.
Methods
Study design and patient selection
We retrospectively reviewed a cohort of elderly patients with severe trauma (age, ≥65 years; Injury Severity Score [ISS], ≥16) who were admitted to the Kochi Health Sciences Center (Kochi, Japan) between September 2006 and March 2014. Cases of isolated head injury and cases that were pronounced dead on arrival were excluded. The selected patients were subsequently divided into: those who were only taking warfarin, aspirin, or clopidogrel; those who were taking various combinations of these agents; and a control group who were not taking any of these agents. This study's design was reviewed and approved by the Ethics Committee of the Medical Faculty of the Kochi Health Sciences Center.
Data collection
The patients' data were collected from our institution's electronic patient database or the patients' charts. The demographic characteristics that we evaluated included age, sex, mechanism of injury, and reason for anticoagulant or antiplatelet treatment. The clinical characteristics that we evaluated included shock index (SI) and Glasgow Coma Scale (GCS) on arrival, ISS, international normalized ratio (INR) on arrival, the amount of red cell concentrate (RCC), fresh frozen plasma (FFP), and platelet concentrate (PC) that were administered during the initial 24 h after admission, massive transfusions (≥10 RCC units during the initial 24 h after admission, or early death due to massive bleeding), and mortality due to hemorrhage. Survival or death was assessed during a 28‐day follow‐up period. In our institution, trauma patients with major blood loss are managed based on a poor response to initial fluid resuscitation or suspected active hemorrhage, and group O RCC and group AB FFP are used until the patient's blood type can be determined. Once the patient's blood type is determined, transfusions are carried out with a target ratio of 1:1:1 for RCC : FFP : PC.
Statistical analysis
The transfusion amounts and number of massive transfusions were compared between the control group and the different medication groups. All continuous data were presented as median (interquartile range); categorical variables were presented as number (percentage). Intergroup differences for continuous variables were evaluated using the Mann–Whitney U‐test. Intergroup differences for categorical variables were evaluated using the χ2‐test, the χ2‐test with Yates' correction for continuity, or Fisher's exact test (as double the one‐tailed exact probability). To adjust for potential confounding factors, multivariate logistic regression analyses were carried out, which included age, sex (male), SI, GCS, and ISS as potentially confounding factors. All continuous variables were converted into dichotomous variables using clinically relevant cut‐off points (age, ≥80 years; SI, ≥1; GCS, ≤8; and ISS, ≥25). To identify the risk factors that were independently associated with the need for a massive transfusion, we constructed a stepwise logistic regression model and included risk factors that had a P‐value of <0.2 in the bivariate analysis. The results for the regression model were presented as odds ratios and 95% confidence intervals. Differences with a two‐tailed P‐value of <0.05 were considered statistically significant, and all statistical analyses were carried out using spss software (version 22; SPSS Inc., Chicago, IL, USA).
Results
Patient composition
A total of 1,168 severe trauma patients (ISS, ≥16; age, ≥16 years) had previously been enrolled into a large observational study over a 7‐year and 7‐month period. During the study period, 339 of these patients met our inclusion criteria and were included in our analyses. Among these 339 patients, 67 patients (20%) were receiving anticoagulant and antiplatelet agents, and 272 patients were included in the control group. Among the patients who were receiving anticoagulant and antiplatelet agents, 10 were receiving only warfarin, 28 were receiving only aspirin, 14 were receiving only clopidogrel, and the remaining 13 were receiving various combinations of these medications (Fig. 1).
Figure 1.
Elderly patients (age, ≥65 years) with severe trauma (Injury Severity Score [ISS], ≥16) who participated in this study. Control group, patients who were not taking any anticoagulant or antiplatelet agents.
Baseline characteristics
When we compared the different groups, no differences were observed in their mechanism of injury, SI and GCS on arrival, or ISS and mortality due to hemorrhage. However, significant differences were observed in their age and INR, with the anticoagulant and antiplatelet agent group having a significantly higher INR, because it is notably affected by warfarin. Although the amount of transfusion and need for massive transfusions tended to be higher in the group that was receiving anticoagulant and antiplatelet agents, the difference was not statistically significant (Table 1). Cerebrovascular accidents were the most common reason for anticoagulant use (21 cases, 31%), followed by acute coronary syndrome (19 cases, 28%), atrial fibrillation (9 cases, 13%), and prosthetic heart valves (2 cases, 3%) (Table 2).
Table 1.
Baseline characteristics of elderly patients with severe trauma
| Pre‐injury anticoagulant or antiplatelet agents (n = 67) | Control (n = 272) | P‐value | |
|---|---|---|---|
| Age, years (range) | 79 (74–83) | 76 (70–80) | <0.01 |
| Age ≥80 years, n (%) | 31 (46) | 81 (30) | 0.12 |
| Gender, (male), n (%) | 46 (69) | 171 (63) | 0.38 |
| Mechanism of injury, n (%) | 0.35 | ||
| Traffic accident | 22 (33) | 122 (45) | |
| Fall | 28 (42) | 91 (33) | |
| Work‐related injury | 13 (19) | 47 (17) | |
| Other | 2 (3) | 12 (5) | |
| Blunt trauma, n (%) | 65 (97) | 272 (100) | |
| SI (range) | 0.65 (0.50–0.83) | 0.62 (0.50–0.82) | 0.59 |
| SI ≥ 1, n (%) | 11 (16) | 38 (14) | 0.61 |
| GCS (range) | 15 (14–15) | 14 (13–15) | 0.13 |
| GCS ≤ 8, n (%) | 6 (9) | 31 (11) | 0.57 |
| Injury site, AIS ≥ 3, overlap, n (%) | |||
| Head | 18 (27) | 97 (36) | 0.17 |
| Face | 5 (7) | 23 (8) | 0.79 |
| Chest | 31 (46) | 135 (50) | 0.62 |
| Abdomen | 4 (6) | 29 (11) | 0.25 |
| Spine | 27 (40) | 113 (42) | 0.85 |
| Pelvic | 9 (13) | 37 (14) | 0.97 |
| Extremity | 17 (25) | 52 (19) | 0.25 |
| ISS (range) | 21 (17–27) | 25 (17–34) | 0.14 |
| ISS ≥ 25, n (%) | 30 (45) | 138 (51) | 0.38 |
| INR | 1.08 (1.01–1.40) | 1.04 (0.98–1.14) | <0.01 |
| Emergency procedure, overlap, n (%) | |||
| Thoracotomy | 0 (0) | 6 (2) | 0.48 |
| Laparotomy | 2 (3) | 6 (2) | 0.94 |
| TAE | 4 (6) | 28 (10) | 0.39 |
| Fracture fixation | 6 (9) | 31 (11) | 0.57 |
| Transfusion, units | |||
| RCC | 0 (0–10) | 0 (0–6) | 0.10 |
| FFP | 0 (0–9) | 0 (0–4) | 0.25 |
| PC | 0 (0–0) | 0 (0–0) | 0.99 |
| Massive transfusion requirement, n (%) | 18 (27) | 47 (17) | 0.07 |
| Mortality due to hemorrhage, n (%) | 1 (1.5) | 5 (1.8) | 0.85 |
Values are given as median (interquartile range) or number (%). AIS, abbreviated injury scale; FFP, fresh frozen plasma; GCS, Glasgow Coma Scale; INR, international normalized ratio; ISS, injury severity score; PC, platelet concentration; RCC, red cell concentration; SI, shock index; TAE, transcatheter arterial embolization.
Table 2.
Reasons for anticoagulant or antiplatelet treatment
| Condition | n (%) |
|---|---|
| Cerebrovascular accident | 21 (31) |
| Acute coronary syndrome | 19 (28) |
| Atrial fibrillation | 9 (13) |
| Prosthetic heart valves | 2 (3) |
| Revascularization procedures | 1 (2) |
| Arteriosclerosis obliterans | 1 (2) |
| Various combinations of thromboembolic complications | 4 (6) |
| Other | 1 (2) |
| Unknown | 9 (13) |
Comparing controls and individual agents
When we compared the control group to the individual agent groups, each individual agent group (aspirin, warfarin, and clopidogrel) was significantly older. The INR was significantly higher in the warfarin and combination groups, because warfarin strongly affects the INR. However, the amount of RCC and need for massive transfusions was only significantly increased in the warfarin group. Eight patients (80%) in the warfarin group were receiving reversal agents, such as vitamin K or FFP (Table 3).
Table 3.
Comparison between controls and patients treated with warfarin, aspirin, clopidogrel or various combination before severe injury
| Control | Warfarin | Aspirin | Clopidogrel | Various combination | |||||
|---|---|---|---|---|---|---|---|---|---|
| (n = 272) | (n = 10) | P‐value | (n = 28) | P‐value | (n = 14) | P‐value | (n = 13) | P‐value | |
| Age, years (range) | 76 (70–80) | 82 (77–84) | <0.05 | 74 (70–83) | 0.92 | 80 (78–85) | <0.05 | 81 (79–82) | <0.01 |
| Gender, male, n (%) | 171 (63) | 7 (70) | 0.90 | 21 (75) | 0.20 | 9 (60) | 0.86 | 9 (69) | 0.86 |
| SI | 0.62 (0.50–0.82) | 0.78 (0.55–0.89) | 0.24 | 0.70 (0.53–0.91) | 0.17 | 0.57 (0.49–0.70) | 0.26 | 0.58 (0.46–0.80) | 0.53 |
| GCS | 14 (13–15) | 15 (14–15) | 0.27 | 15 (14–15) | 0.11 | 15 (14–15) | 0.67 | 14 (14–15) | 0.97 |
| ISS | 25 (17–34) | 25 (17–36) | 0.84 | 20 (17–26) | 0.06 | 25 (17–27) | 0.43 | 26 (17–37) | 0.80 |
| INR | 1.04 (0.98–1.14) | 1.77 (1.66–2.14) | <0.001 | 1.05 (0.98–1.12) | 0.96 | 1.08 (1.01–1.10) | 0.85 | 1.13 (1.03–1.86) | <0.05 |
| Transfusion, (units) | |||||||||
| RCC | 0 (0–6) | 9 (1–13) | <0.05 | 0 (0–9) | 0.26 | 0 (0–5) | 0.82 | 0 (0–6) | 0.68 |
| FFP | 0 (0–4) | 9 (5–15) | <0.01 | 0 (0–10) | 0.76 | 0 (0–0) | 0.48 | 0 (0–2) | 0.96 |
| PC | 0 (0–0) | 0 (0–18) | 0.14 | 0 (0–0) | 0.98 | 0 (0–0) | 0.39 | 0 (0–0) | 0.74 |
| Massive transfusion, n (%) | 47 (17) | 5 (50) | <0.05 | 7 (25) | 0.31 | 3 (21) | 0.97 | 2 (15) | 0.84 |
| Mortality due to hemorrhage, n (%) | 5 (1.8) | 0 (0) | 0.43 | 1 (4) | 0.93 | 0 (0) | 0.59 | 0 (0) | 0.56 |
Values are given as median (interquartile range) or number (%). FFP, fresh frozen plasma; GCS, Glasgow Coma Scale; INR, international normalized ratio; ISS, injury severity score; PC, platelet concentration; RCC, red cell concentration; SI, shock index.
Multivariate logistic regression analysis of prognostic factors for massive transfusion
Patients who were receiving only warfarin and had an SI of ≥1, a GCS of ≤8, and an ISS of ≥25 had a significantly higher risk of needing a massive transfusion (odds ratio, 5.03; 95% confidence interval, 1.25–20.20; P < 0.05). However, no increased risk of needing a massive transfusion was observed for patients who were receiving only aspirin, only clopidogrel, or various combinations of these agents (Table 4).
Table 4.
Multivariate logistic regression analysis of prognostic factors for massive transfusion
| Unadjusted | Adjusted | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P‐value | OR | 95% CI | P‐value | |
| Warfarin | 4.54 | 1.27–16.20 | <0.05 | 5.03 | 1.25–20.20 | <0.05 |
| Aspirin | 1.47 | 0.60–3.63 | 0.55 | 2.26 | 0.82–6.26 | 0.12 |
| Clopidogrel | 1.17 | 0.32–4.33 | 0.91 | 1.41 | 0.34–5.78 | 0.63 |
| Various combination | 0.77 | 0.17–3.55 | 0.98 | 0.62 | 0.12–3.31 | 0.58 |
| Gender, male | 0.85 | 0.48–1.46 | 0.62 | 1.09 | 0.58–2.03 | 0.79 |
| Age ≥80 years | 1.61 | 0.92–2.82 | 0.12 | 1.19 | 0.64–2.21 | 0.59 |
| SI ≥ 1 | 3.98 | 2.06–7.68 | <0.001 | 2.84 | 1.37–5.86 | <0.01 |
| GCS ≤ 8 | 3.49 | 1.69–7.21 | <0.001 | 2.30 | 1.04–5.08 | <0.05 |
| ISS ≥ 25 | 4.22 | 2.26–7.90 | <0.001 | 3.53 | 1.79–6.97 | <0.001 |
CI, confidence interval; GCS, Glasgow Coma Scale; ISS, injury severity score; OR, odds ratio; SI, shock index.
Discussion
This study investigated the effect of pre‐injury anticoagulant and antiplatelet agent use on blood loss at non‐head injury sites among elderly patients with severe trauma. Although the amount of RCC and need for massive transfusions tended to be higher in the group that was receiving anticoagulant and antiplatelet agents, this increase was not statistically significant. However, a significantly higher amount of RCC and need for massive transfusions was observed among patients who were receiving only warfarin. Furthermore, we found that only warfarin was an independent risk factor for massive transfusions.
The effects of pre‐injury anticoagulant and antiplatelet agent use on traumatic brain injury and mortality have been widely reported, although few reports have examined these effects for non‐head injuries.6, 7, 13, 14 Furthermore, there are reports regarding the associated risk of mortality,6, 13, 14 although very few studies have evaluated the associated blood loss.7 In one report regarding non‐head injuries, Dossett et al.6 reported that warfarin use was associated with a significantly higher risk of death among all trauma patients, although the causes of the death are not clear. However, Wojcik et al.13 reported that warfarin did not significantly affect the risk of mortality among patients with non‐head injuries, and Bonville et al.14 reported that warfarin significantly affected the risk of mortality among patients with non‐head injuries. Unfortunately, none of these studies evaluated the associated blood loss. Furthermore, Ferraris et al.7 reported that post‐trauma bleeding was not significantly associated with pre‐injury antiplatelet drug use among all trauma patients, although the methods for evaluating bleeding were not clear in that study. Thus, the effect of pre‐injury anticoagulant and antiplatelet agent use on blood loss remains unknown.
Many studies have evaluated the risk of bleeding in cases of orthopedic surgery (hip fracture) and found that anticoagulant and antiplatelet agent use did not substantially increase the risk of bleeding complications.15, 16, 17 In contrast, other studies have reported that pre‐injury anticoagulant and antiplatelet agent use had an adverse effect on the outcome (mortality) in cases of traumatic brain injury. Furthermore, several of these studies have specifically reported that warfarin increased the risk of mortality,8, 9, 11, 14, 18 and have also suggested that the effect of warfarin should be rapidly reversed with vitamin K or FFP if intracranial bleeding is detected.11, 18, 19 However, only a few studies have reported that antiplatelet agent use increased the risk of mortality in cases of traumatic brain injury.12 In this study, patients who were receiving only warfarin before their injury had an increased risk of bleeding due to non‐head injuries. Therefore, rapidly providing treatment to reverse the effect of warfarin may be warranted in select patients with pre‐injury use of only warfarin, even if the case is not complicated with traumatic brain injury.
The main limitation of the present study was its retrospective design and small sample size of patients with pre‐injury anticoagulant and antiplatelet agent use, which was not large enough to carry out a valid statistical analysis. Second, the time to transfusion initiation after arrival and the amount of transfusion may have varied, depending on the expertise of the emergency physicians who were involved in primary care at that time. Therefore, it is possible that the transfusions were inappropriate in some patients, as it is unlikely that all transfusions were performed correctly. Third, we did not consider the effect of the injury site on the amount of bleeding, or the effect of coagulopathy due to traumatic brain injury.20 Finally, it is possible that patients who were receiving pre‐injury anticoagulant and antiplatelet agents had a greater prevalence of preexisting or more severe disease, which may have confounded our comparisons of these two groups.
Conclusion
Patients who were receiving only warfarin before their injury had an increased risk of bleeding due to non‐head injury sites. Furthermore, only warfarin was an independent risk factor for needing a massive transfusion. Therefore, rapidly providing treatment to reverse the effects of warfarin may be warranted in select patients with pre‐injury use of only warfarin, even if the case is not complicated with traumatic brain injury.
Conflict of Interest
None.
References
- 1. Christensen K, Doblhammer G, Rau R, Vaupel JW. Ageing populations: The challenges ahead. Lancet 2009; 374: 1196–1208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Aschkenasy MT, Rothenhaus TC. Trauma and falls in the elderly. Emerg. Med. Clin. North Am. 2006; 24: 413–432. [DOI] [PubMed] [Google Scholar]
- 3. Callaway DW, Wolf R. Geriatric trauma. Emerg. Med. Clin. North Am. 2007; 25: 837–860. [DOI] [PubMed] [Google Scholar]
- 4. Heffernan DS, Thakkar RK, Monaghan SF, Ravindran R, Adams CA Jr, Kozloff MS. Normal presenting vital signs are unreliable in geriatric blunt trauma victims. J. Trauma 2010; 69: 813–820. [DOI] [PubMed] [Google Scholar]
- 5. Giannoudis PV, Harwood PJ, Court‐Brown C, Pape HC. Severe and multiple trauma in older patients; incidence and mortality. Injury 2009; 40: 362–367. [DOI] [PubMed] [Google Scholar]
- 6. Dossett LA, Riesel JN, Griffin MR, Cotton BA. Prevalence and implications of pre‐injury warfarin use: An analysis of the National Trauma Databank. Arch. Surg. 2011; 146: 565–570. [DOI] [PubMed] [Google Scholar]
- 7. Ferraris VA, Bernard AC, Hyde B. The impact of antiplatelet drugs on trauma outcomes. J. Trauma Acute Care Surg. 2012; 73: 492–497. [DOI] [PubMed] [Google Scholar]
- 8. Franko J, Kish KJ, O'Connell BG, Subramanian S, Yuschak JV. Advanced age and pre‐injury warfarin anticoagulation increase the risk of mortality after head trauma. J. Trauma 2006; 61: 107–110. [DOI] [PubMed] [Google Scholar]
- 9. Grandhi R, Duane TM, Dechert T et al Anticoagulation and the elderly head trauma patient. Am. Surg. 2008; 74: 802–805. [DOI] [PubMed] [Google Scholar]
- 10. Ott MM, Eriksson E, Vanderkolk W, Christianson D, Davis A, Scholten D. Antiplatelet and anticoagulation therapies do not increase mortality in the absence of traumatic brain injury. J. Trauma 2010; 68: 560–563. [DOI] [PubMed] [Google Scholar]
- 11. Howard JL 2nd, Cipolle MD, Horvat SA et al Preinjury warfarin worsens outcome in elderly patients who fall from standing. J. Trauma 2009; 66: 1518–1522. [DOI] [PubMed] [Google Scholar]
- 12. Batchelor JS, Grayson A. A meta‐analysis to determine the effect of pre‐injury antiplatelet agents on mortality in patients with blunt head trauma. Br. J. Neurosurg. 2013; 27: 12–18. [DOI] [PubMed] [Google Scholar]
- 13. Wojcik R, Cipolle MD, Seislove E, Wasser TE, Pasquale MD. Preinjury warfarin does not impact outcome in trauma patients. J. Trauma 2001; 51: 1147–1151. [DOI] [PubMed] [Google Scholar]
- 14. Bonville DJ, Ata A, Jahraus CB et al Impact of pre‐injury warfarin and antiplatelet agents on outcomes of trauma patients. Surgery 2011; 150: 861–868. [DOI] [PubMed] [Google Scholar]
- 15. Nydick JA, Farrell ED, Marcantonio AJ, Hume EL, Marburger R, Ostrum RF. The use of clopidogrel (Plavix) in patients undergoing nonelective orthopaedic surgery. J. Orthop. Trauma 2010; 24: 383–386. [DOI] [PubMed] [Google Scholar]
- 16. Chechik O, Thein R, Fichman G, Haim A, Tov TB, Steinberg EL. The effect of clopidogrel and aspirin on blood loss in hip fracture surgery. Injury 2011; 42: 1277–1282. [DOI] [PubMed] [Google Scholar]
- 17. Collinge CA, Kelly KC, Little B, Weaver T, Schuster RD. The effects of clopidogrel (Plavix) and other oral anticoagulants on early hip fracture surgery. J. Orthop. Trauma 2012; 26: 568–573. [DOI] [PubMed] [Google Scholar]
- 18. Mina AA, Bair HA, Howells GA, Bendick PJ. Complications of pre‐injury warfarin use in the trauma patient. J. Trauma 2003; 54: 842–847. [DOI] [PubMed] [Google Scholar]
- 19. Ivascu FA, Howells GA, Junn FS, Bair HA, Bendick PJ, Janczyk RJ. Rapid warfarin reversal in anticoagulated patients with traumatic intracranial hemorrhage reduces hemorrhage progression and mortality. J. Trauma 2005; 59: 1131–1137. [DOI] [PubMed] [Google Scholar]
- 20. Cap AP, Spinella PC. Severity of head injury is associated with increased risk of coagulopathy in combat casualties. J. Trauma 2011; 71: S78–81. [DOI] [PubMed] [Google Scholar]