Article Text

Download PDFPDF

Association between anticoagulants and mortality and functional outcomes in older patients with major trauma
  1. Nobuhiro Sato1,
  2. Peter Cameron1,2,
  3. Susan Mclellan1,
  4. Ben Beck1,3,
  5. Belinda Gabbe1,4
  1. 1School of Public Health and Preventive Medicine, Monash University, Melbourne, Victoria, Australia
  2. 2Emergency and Trauma Centre, Alfred Health, Melbourne, Victoria, Australia
  3. 3Faculty of Medicine, Laval University, Quebec, Québec, Canada
  4. 4Health Data Research UK, Swansea University, Swansea, West Glamorgan, UK
  1. Correspondence to Dr Nobuhiro Sato, School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC 3004, Australia; nobuhiro.sato{at}monash.edu

Abstract

Background The number of trauma patients taking anticoagulants and antiplatelet agents is increasing as society ages. However, there have been limited and inconsistent reports of the association between anticoagulants and mortality and functional outcomes. This study aimed to quantify the association between anticoagulant/antiplatelet medication at the time of injury and both short-term and longer-term outcomes in older major trauma patients.

Methods This was a population-based registry study using data from the Victorian State Trauma Registry from July 2017 to June 2018. We included patients with major trauma aged 65 years and older. The outcomes of interest were in-hospital mortality, hospital length of stay, intensive care unit length of stay and the Extended Glasgow Outcome Scale (GOS-E) at 6 months after injury. We examined the association between the outcomes and anticoagulants/antiplatelet agents at the time of injury and used multivariable logistic regression models to account for known confounders.

Results There were 1323 older adults eligible for inclusion in the study, of which 249 (18.8%) were taking anticoagulants (n=8 were taking both anticoagulants and antiplatelet agents), 380 (28.7%) were taking antiplatelet agents and 694 (52.5%) were not using either. Any anticoagulant use was associated with higher odds of in-hospital mortality (adjusted OR (AOR), 2.38; 95% CI 1.58 to 3.59) compared with not using anticoagulants. No differences were observed in the GOS-E at 6 months after injury between any anticoagulants use, antiplatelet use and no anticoagulant use (anticoagulant AOR, 0.71; 95% CI 0.48 to 1.05, antiplatelet AOR, 1.02; 95% CI 0.73 to 1.42).

Conclusion Anticoagulant use at the time of injury was associated with higher odds of in-hospital mortality but did not adversely impact functional outcomes at 6 months after injury. These findings demonstrate the importance of seeking an accurate history of anticoagulant use and its indication, as well as the immediate initiation of reversal therapies.

  • trauma
  • research
  • geriatrics
  • major trauma management
  • death/mortality

Data availability statement

Data are available upon reasonable request.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known on this subject

  • Some retrospective cohort studies have shown an association between preinjury warfarin and increased mortality, while other studies of similar design found no association.

  • There is limited evidence linking long-term outcomes with preinjury anticoagulants.

What this study adds

  • We examined the association between anticoagulants/antiplatelet agents at the time of injury and both short-term and longer-term outcomes in older hospitalised major trauma patients using a population-based registry in Victoria, Australia.

  • Our study suggests that anticoagulant use at the time of injury was associated with higher odds of in-hospital mortality, while there was no evidence of association between function at 6 months and anticoagulants.

  • Our findings demonstrate the importance of seeking an accurate history of anticoagulants use and its indication, as well as the immediate initiation of reversal therapies.

Introduction

Anticoagulants and antiplatelet agents are used to treat and prevent thromboembolic events associated with atrial fibrillation, mechanical heart valves and deep venous thrombosis, which are common among older people.1 2 Direct oral anticoagulants (DOACs), such as dabigatran, rivaroxaban and apixaban, offer effective alternatives to warfarin because of safety and increased compliance.3 While oral anticoagulants dramatically reduce the risk of thromboembolic diseases, patients on anticoagulants have an increased bleeding risk, which may result in non-traumatic intracerebral haemorrhage, gastrointestinal haemorrhage and traumatic haemorrhage.4 5

Major trauma has traditionally been viewed as a disease of the young; however, the number of older adults with major trauma is increasing rapidly.6 As a result, the number of trauma patients taking anticoagulants and antiplatelet agents has been steadily increasing.7 A number of retrospective cohort studies have shown an association between preinjury warfarin and increased mortality,7 8 while other studies of similar design found no association.9 10 Overall, while there is evidence showing increased mortality risk following trauma, there is limited evidence linking long-term outcomes with preinjury anticoagulants. Therefore, the aim of this study was to investigate the association between anticoagulants at the time of injury and both short-term and longer-term outcomes in major trauma patients.

Patients and methods

Study design and participants

This study included all older hospitalised major trauma patients in Victoria, on the Victorian State Trauma Registry (VSTR), with a date of injury from 1 July 2017 to 30 June 2018. Older was defined as aged 65 years or more, consistent with previous research.6 Patients with missing anticoagulant use documentation at the time of injury were excluded.

Setting

The state of Victoria, Australia, has a population of 6.5 million people.11 The Victorian State Trauma System (VSTS) was implemented between 2000 and 2003.12 The VSTS is a centrally coordinated trauma network with one paediatric and two adult hospitals as major trauma services.

Victorian State Trauma Registry

The VSTR is a population-based registry that collects data about all hospitalised major trauma patients in Victoria.13 A patient is classified as major trauma in the VSTR if any of the following criteria are met: death due to injury; an Injury Severity Score (ISS) >12 as determined by the Abbreviated Injury Scale (AIS) (2005 version 2008 update); admission to an intensive care unit (ICU) for more than 24 hours and requiring mechanical ventilation for at least part of their ICU stay; urgent surgery; 20%–29% total body surface area partial or full thickness burns.

Data collection

Demographic factors, cause of injury, injury type and severity, interhospital transfer status, Shock Index (SI), head injury status, GCS score and information about anticoagulants at the time of injury were extracted from the registry. Data collectors reviewed the patients’ medical notes, and if there was documentation of the patient taking anticoagulation or antiplatelet therapy at the time of the injury, the type of medication was entered into the VSTR. As part of quality control, International Classification of Diseases 10th Revision-Australian Modification (ICD-10-AM) codes were checked for code Z92.1 (Personal history of long term (current) use of anticoagulants). Head injury was defined as an AIS severity score greater than two in the AIS head region. The anticoagulants collected by the registry include warfarin, DOACs (rivaroxaban, apixaban and dabigatran), ticagrelor, as well as aspirin, and other platelet function inhibitors (eg, clopidogrel). On the basis of this information, patients were categorised into: (1) warfarin or DOACs user (any anticoagulants group), (2) ticagrelor, clopidogrel and/or aspirin user (antiplatelet group), (3) no anticoagulants. Patients taking both anticoagulants and antiplatelets were assigned to the any anticoagulant group. The comorbid status of the patient was defined using the Charlson Comorbidity Index (CCI), where ICD-10AM codes for comorbid conditions were mapped to the CCI.14

Outcomes measures

To assess the association between anticoagulant use and outcomes, the groups were compared using the following outcomes of interest; in-hospital mortality, hospital length of stay, ICU length of stay and the Extended Glasgow Outcome Scale (GOS-E) at 6 months after injury. The VSTR follows up all survivors to hospital discharge by telephone at 6, 12 and 24 months after injury to collect patient-reported outcomes data.15 The GOS-E categorises patient function into one of eight categories, with upper good recovery representing return to preinjury function.16 The GOS-E is recommended for use in trauma populations.17

Statistical analysis

Continuous data with skewed distributions were summarised using the median and IQR. Categorical data were reported using frequencies and proportions. The Pearson’s χ test or Kruskal-Wallis test was used to explore the association between anticoagulant use and key patient and injury characteristics, and outcomes.

Multivariable analyses were used to assess the association between in-hospital mortality and anticoagulants at the time of injuries using logistic regression models, and ORs with 95% CIs were calculated. Clinically relevant variables demonstrating a significant association with the outcome on univariate testing were entered into a multivariable logistic regression analysis. These variables included age (65–74 years, 75–84 years, ≥85 years), CCI weighting (0, 1, ≥2), cause of injury (traffic injury, low falls (≤1 m), high falls (>1 m), others), ISS, SI (<1.0, ≥1.0),18 head injury (yes, no) and the GCS (3–8, 9–12, 13–15). The association between GOS-E and anticoagulants at the time of injuries was also assessed using multivariable binary logistic regression analysis (good recovery and moderate disability vs severe disability, vegetative state and death), and adjusted for age, sex, CCI weighting, cause of injury, ISS, head injury and GCS. In-hospital deaths were included in the GOS-E model. Serious injury types for patients that died in hospital were assessed. Serious injury was defined as an AIS severity score greater than 2. Data were analysed using SPSS, V.25.0 (IBM Corporation) and a p value <0.05 was considered significant.

Patient and public involvement

Retrospective data from the VSTR database were used in this study, limiting the role of patient and public involvement in design, recruitment and conduct of this study.

Results

Between July 2017 and June 2018, 3645 cases of major trauma were recorded on the VSTR. Of the 1407 older patients, 1323 patients were eligible for inclusion (figure 1).

Figure 1

Patients with major trauma during the study period and patients included in the analysis.

Patient characteristics

Patient characteristics according to anticoagulants are listed in table 1. There were 249 (18.8%) patients using any anticoagulants, 380 (28.7%) patients using antiplatelets, and 694 (52.5%) patients did not use anticoagulant or antiplatelet agents. Eight patients were taking both anticoagulants and antiplatelets. There were differences between the groups with respect to age, CCI and cause of injury. In addition, a higher proportion of patients in the ‘any anticoagulant’ and antiplatelet group had a head injury. The ISS was similar between the groups. Subdural haematoma was the most common serious injury in patients who died in hospital (table 2).

Table 1

Characteristics of the study population (2017–2018 major trauma) according to anticoagulants (exclude no information of anticoagulants, unknown anticoagulants)

Table 2

Serious injury types (Abbreviated Injury Scale >2) for patients that died in hospital

The association between anticoagulants/antiplatelet agents and outcome

The in-hospital mortality was 31.7% for patients who used anticoagulants (OR 2.73; 95% CI 1.94 to 3.83), 18.4% for patients who used antiplatelet agents (OR 1.33; 95% CI 0.95 to 1.85) and 14.6% for patients who did not use anticoagulants (table 3). The hospital length of stay was similar between the groups. After adjusting for age, CCI, cause of injury, ISS, SI, head injury and GCS for 1191 (90.0%) patients, the odds of in-hospital mortality were higher in the group taking any anticoagulants when compared with the group not taking anticoagulants (adjusted OR (AOR) 2.38; 95% CI, 1.58 to 3.59) (table 4). There was no association between antiplatelet agents and mortality (AOR 1.12; 95% CI 0.74 to 1.71) in our study population.

Table 3

Outcomes according to anticoagulants (exclude no information of anticoagulants, unknown anticoagulants) in older hospitalised major trauma patients

Table 4

Multivariable association of anticoagulants with mortality in older hospitalised major trauma patients (n=1191)

There were 1198 (90.6%) patients with a valid GOS-E score at 6 months, which included 228 (91.6%) in the group taking any anticoagulants, 349 (91.8%) in the group taking antiplatelets and 620 (89.3%) in the group not taking anticoagulants. In the severe disability GOS-E category, the proportion of patients who used any anticoagulant was higher than those who did not use anticoagulants, while the proportion of patients who did not use anticoagulants was higher than those who used any anticoagulant in moderate disability and lower good recovery. After adjusting for age, sex, CCI, cause of injury, ISS, head injury and GCS for 1115 (93.1%) patients in the GOS-E categories, there was no evidence of association between anticoagulant use and the GOS-E (anticoagulant AOR 0.71; 95% CI 0.48 to 1.05, antiplatelet agents AOR 1.02; 95% CI 0.73 to 1.42) in our study population (table 5). There was no evidence of association with GOS-E for patients with high energy injury (traffic injury and high falls) (anticoagulant AOR, 0.66; 95% CI 0.28 to 1.54; antiplatelet AOR 0.97; 95% CI 0.50 to 1.90) in our study population.

Table 5

Multivariable association of anticoagulants with Extended Glasgow Outcome Scale at 6 months after injury (n=1115)

Discussion

In this population-based, registry study, we found an association between in-hospital mortality and anticoagulant use at the time of injury. We found no evidence of association between anticoagulant use at the time of injury and function at 6 months after injury, when early mortality in hospital was included.

Results in context

Any anticoagulant use was associated with higher adjusted odds of in-hospital death in our study. Other retrospective studies have found that preinjury warfarin use did not increase mortality.9 10 However, in the study by Wojcik et al, the study population included less seriously injured patients (median ISS 11.64 for head injury, 7.18 for non-head injury) when compared with our study (median ISS 17)9 and was likely to be underpowered to detect a difference. The study by Kennedy et al did not analyse specific injuries and was also underpowered (61 patients administered warfarin).10

A large American retrospective cohort study and a European retrospective cohort study found that preinjury warfarin use was associated with increased mortality in trauma patients,7 8 which were similar to our results. In the study by Dossett et al, adult patients admitted to trauma centres were included, and 9.3% of warfarin users died compared with only 4.8% of non-users.7 The prior studies only included patients presenting to specialised trauma centres which may introduce selection bias. In contrast, our study used data on all hospitalised major trauma patients in Victoria. In the study by Lecky et al, patients aged 16 years and older and resulting in hospitalisation for >72 hours, intensive care admission, transfer for specialised care or death prior to discharge were included, and 22.4% of warfarin users died compared with 7.1% of non-users at 30 days.8 The mortality was higher in our study (any anticoagulant 31.7% vs no anticoagulant 14.6%), probably because of the older population and inclusion of community hospitals.

Antiplatelet agents showed no evidence of association with mortality in our study population. The results of this study support the findings of prior studies. Single-institution trauma database analyses documented preinjury warfarin was an independent risk factor for mortality, while antiplatelet agents did not affect mortality.19 One small retrospective study did find an association between antiplatelet agents and increased mortality for older patients with traumatic intracranial haemorrhage20; however, it did not adjust for other risk factors.

In our study, subdural haematoma was the most common serious injury in patients who died in hospital. In particular, the proportion of head injuries increased with anticoagulant use. Consistent with our findings, previous studies have described preinjury warfarin use associated with higher risk of traumatic intracranial haemorrhage and mortality.21 22

After accounting for early mortality in hospital, no evidence of association between anticoagulant and antiplatelet use and function was found at 6 months in our study population. This result differs from previous studies.23 24 One small study demonstrated older adults with preinjury warfarin or clopidogrel use and mild traumatic intracranial haemorrhage may be at increased risk for unfavourable GOS-E at 6 months compared with similar patients without preinjury use of these medications.23 However, patients in the preinjury anticoagulant or antiplatelet cohort were older than those in the control cohort and this was not adjusted for. Another small retrospective study showed preinjury warfarin was a risk factor for worse GOS-E at 6 months.24 The study included only patients with moderate to severe traumatic brain injury. Our study included all major trauma patients and the proportion of patients with head injury was less (~50%), which might have influenced the result. Our study suggests that anticoagulants impact on the survival, but there might be no impact on longer term outcomes even when early mortality in hospital was included. Further research is required to explore the association between anticoagulant and longer term outcome.

The proportion of patients taking anticoagulants was high in our study. Of all older hospitalised major trauma patients, 18.8% were using anticoagulants. This compares with estimated warfarin use of 1%–2% of adults and 8% of those aged over 80 years in the developed world.25 26 The proportion of warfarin users increased from 2.3% (2002) to 4.0% (2006) in all patients and from 7.3% (2002) to 12.8% (2006) in patients older than 65 years in the National Trauma Databank in the USA.7 The reason for the high proportion of patients taking anticoagulants might be that warfarin use is increasing over time—and our study included older hospitalised major trauma patients from 2017 to 2018, a decade on from previous studies. The benefits of anticoagulants or antiplatelet agents for preventing cerebrovascular and cardiovascular diseases are well established.1 2 Our findings highlight the importance of seeking an accurate history of anticoagulant use and its indication, as well as the immediate initiation of reversal therapies. In addition, emergency physicians might be better able to predict the prognosis of the patients with a history of anticoagulant use.

Limitations

Our study has some limitations. First, we could not assess the degree of anticoagulation (ie, prothrombin time, international normalised ratio), dosage or compliance with anticoagulant prescriptions at the time of injury. Second, we did not divide into warfarin and DOACs in any anticoagulant group. Third, we did not include prehospital deaths. Fourth, our study may not have been sufficiently powered to detect a difference in GOS-E. Finally, this is an observational study, therefore causality cannot be inferred. Future studies should measure the degree and compliance of anticoagulation and optimisation of reversal of anticoagulants.

Conclusion

We observed a significant association between in-hospital mortality with anticoagulants, while there was no evidence of association between function at 6 months and anticoagulants with our study population. Antiplatelet agents showed no evidence of association with mortality or functional outcome. This study demonstrates the importance of seeking an accurate history of anticoagulant use and indication, as well as the immediate initiation of reversal therapies. Further larger prospective studies are needed to better understand the effect of anticoagulants on trauma patients and to optimise clinical pathways for management.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

The VSTR has ethical approval from the Department of Health and Human Services (DHHS) Human Research Ethics Committee (HREC), 138 trauma-receiving hospitals in Victoria and the Monash University HREC.

Acknowledgments

The Victorian State Trauma Outcome Registry and Monitoring group is thanked for the provision of VSTR data.

References

Footnotes

  • Handling editor Ellen J Weber

  • Twitter @Nobu_2Sato, @DrBenBeck

  • Contributors NS conceived the study. PC, BB, SM and BG supervised the conduct of the trial and data collection. NS and SM managed the data, including quality control. PC and BG provided statistical advice on the study design and analysed the data. NS and PC chaired the data oversight committee. NS, PC, BB and BG drafted the manuscript, and all authors contributed substantially to its revision. NS takes responsibility for the paper as a whole. All authors read and approved the final manuscript.

  • Funding BB was supported by an Australian Research Council Discovery Early Career Researcher Award Fellowship (DE180100825). PC was supported by a Medical Research Futures Fund (MRFF) Practitioner Fellowship. BG was supported by an Australian Research Council Future Fellowship (FT170100048). The Victorian State Trauma Registry is a Department of Health, State Government of Victoria and Transport Accident Commission funded project.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.