INTRODUCTION
Life-threatening hemorrhage—such as traumatic brain injury, hemorrhagic stroke, gastrointestinal bleeding or vascular or visceral injury from blunt or penetrating trauma—is among the most important time-critical conditions treated by prehospital emergency medical services (EMS) [1,2]. To address these emergencies, EMS personnel commonly perform interventions including direct hemorrhage control, tourniquet application, and administration of intravenous fluids and vasopressors. Additionally, select EMS units are also able to treat hemorrhage with blood products such as red blood cells, whole blood, and plasma to stabilize patients [3–5].
Oral anticoagulants (OACs) including warfarin and direct OACs (DOACs) such as dabigatran, rivaroxaban, and apixaban are widely used in the outpatient setting for treating or preventing a range of medical conditions. For example, OACs are used to prevent strokes in patients with atrial fibrillation or those who have had prosthetic heart valve replacement [6]. OACs are also commonly used to treat or prevent thromboembolism [7]. Internationally, the use of oral anticoagulants has been on the rise, with recent estimates indicating an increase from 1.6 million in 2011 to 5.4 million in 2020 [7–12]. A major pitfall of OACs is that their use increases the risk of serious bleeding, which can lead to life-threatening hemorrhages in both trauma and nontrauma scenarios [6]. Given the role of EMS in the emergent treatment of hemorrhage, understanding a patient’s history of OAC use is essential [13–16]. Early identification of this history can guide triage to a trauma center or prompt the administration of medications to reverse anticoagulant effects. However, obtaining this information in the prehospital environment is often challenging.
Despite the critical implications of OAC use, there is a lack of comprehensive understanding regarding the prevalence and characteristics of patients with a history of OAC use receiving EMS care. In this study, we sought to determine the frequency and characteristics of prehospital EMS patients reporting a history of OAC use.
METHODS
Ethics statement
This study was approved by the Ohio State University Office of Responsible Research Practices (No. 2021E0422). The requirement for informed consent was waived because data were deidentified and publicly available.
Study design and setting
We analyzed 2018–2020 data from the ESO Data Collaborative, one of the largest out-of-hospital electronic health record systems in the United States. The software product collects clinical information regarding EMS encounters, including event characteristics, patient demographics, clinical signs and symptoms, interventions, vital signs, and outcomes. The software contains data validation (forcing) options to require or constrain user entries for select data elements. We applied data cleaning methods, including validating medication history against established protocols, checking for inconsistencies in recorded vital signs and demographics, and excluding any incomplete or duplicate records. This rigorous process ensured the reliability and accuracy of the dataset used in the analysis. Current software follows the National EMS Information System (NEMSIS) ver. 3 standard and the associated data definitions [17]. More than 2,000 EMS agencies use the ESO EHR software system. These data have been used in a wide range of research studies [18–21]. In the current study, we focused on individual patient encounters. Patients and the public were not involved in the design, conduct, reporting, or dissemination of the research.
Selection of participants
We included all adult patients (≥18 years) who received “911” EMS care with a reported history of OAC use. EMS providers reported medication history as obtained during clinical care. OACs used included warfarin (Coumadin), dabigatran (Pradaxa), rivaroxaban (Xarelto), and apixaban (Eliquis). While not part of the primary analysis, we also identified instances of reported use of injected anticoagulants, including heparin, enoxaparin (Lovenox), dalteparin (Fragmin), fondaparinux (Arixtra), and edoxaban (Savaysa). We included only medications reported as part of the patient’s medical history, not medications administered during EMS care. We did not study other oral medications that may potentiate bleeding such as antiplatelet agents, aspirin, or clopidogrel (Plavix).
Outcomes
We focused on clinical variables plausibly associated with or influenced by a history of OAC use. Variables of primary interest included patient characteristics (age, sex), incident characteristics (location type, incident type [medical vs. trauma], cardiac arrest, census region). We did not include race or ethnicity in the analysis due to lack of biologic plausibility [22]. We determined initial vital signs as well as select treatments relevant to hemorrhage care (administration of intravenous fluids and blood products or performance of advanced airway management). We determined the primary clinical impression reported by EMS providers; we grouped the impressions according to the International Classification of Diseases, 10th Revision, categories as implemented in the NEMSIS ver. 3 system [17,18]. Within each primary category, we also identified subcategory conditions potentially linked to hemorrhagic emergencies; for example, head injury or altered mental status, which may reflect intracranial hemorrhage. These latter conditions were determined by consensus of the study team.
Statistical analysis
We determined the number and incidence of adult 911 responses with a reported patient history of OAC use. Using univariate logistic regression, we compared patient, incident, and care characteristics between OAC and non-OAC users. We also compared EMS primary clinical impressions between OAC and non-OAC users. We used SAS ver. 9.4 (SAS Institute) for all statistical analyses.
RESULTS
During the study period (2018–2020), among 21.3 million patient care records, there were 17.3 million 911 responses and 16.2 million adult 911 responses. EMS personnel reported a history of patient OAC use for 906,575 events (55.8 per 1,000 events; 95% confidence interval [CI], 55.7–55.9) (Fig. 1). Of the 906,575 responses, the most widely used OACs were apixaban (46.0%), warfarin (34.8%), rivaroxaban (17.5%), and dabigatran (2.5%) (Table 1). Use of other injectable anticoagulants was reported for a smaller number of patients. When stratified by EMS agency, median incidence of OAC use was 54 per 1,000 responses (95% CI, 52–56; range, 0–500).
Compared with other patients, OAC users were more likely to be older and female (Table 2). Patient OAC use history was more likely for calls in nursing, rehabilitation or long-term care facilities, and less likely for calls in public locations. Compared with medical patients, patients with trauma-associated conditions were less likely to report OAC use. However, OAC use was more common among patients with mixed medical/trauma conditions. OAC use was twice as likely among patients presenting with cardiac arrest. OAC use was least likely among EMS events in the Midwest census region.
OAC use history was most common among patients with a reported clinical impression involving cardiovascular, infectious, or respiratory conditions (Table 3). Specific primary clinical impressions strongly associated with OAC use history included cardiovascular conditions, gastrointestinal bleeding, infections, traumatic brain injury, hemorrhage, intracranial hemorrhage, strokes (including transient ischemic attacks and stroke-like syndromes), respiratory emergencies, and other bleeding conditions.
DISCUSSION
A history of OAC use is a critical risk factor in life-threatening hemorrhagic and neurologic emergencies. Our study offers important information regarding the epidemiology of OAC users receiving EMS care. In this national series, 1 of every 18 adult EMS responses involved a patient with history of OAC use. In some EMS agencies, the prevalence of OAC use was even higher. As expected, OAC use was common in certain patients with high-risk conditions such as cardiovascular and respiratory diseases as well as central neurologic emergencies. OAC use was also strongly associated with EMS responses to calls from nursing, rehabilitation, and long-term facilities. These observations highlight the prevalence of OAC use among EMS patients and are particularly important given the increasing prevalence of OAC use internationally [7–9,11,12].
Knowledge of a patient’s history of OAC use has extremely important implications for EMS care. OAC use may directly amplify the severity and risk of critical conditions, such as hemorrhagic stroke, traumatic brain injury, or gastrointestinal or extremity bleeding [7,13]. In cases of acute, life-threatening bleeding, awareness of a patient’s OAC use history may prompt hospitals to utilize rescue therapies, such as blood products (e.g., plasma or red blood cells), procoagulants (e.g., tranexamic acid, vitamin K, prothrombin complex concentrate), or DOAC-specific reversal agents (e.g., andexanet and idarucizumab) [7,23]. Because of the potential for rapid deterioration of patients with life-threatening hemorrhage, even the early accurate identification of OAC use history can have important downstream benefits [14]. The identification of OAC use history may prompt earlier notification to receiving hospitals or transportation to specialty care centers; for example, direct transportation to a trauma center [14]. Nishijima et al. [2] found that the addition of OAC use to trauma triaging criteria improved the sensitivity for detection of intracranial hemorrhage and death. Novel technologies such as mobile stroke units may also be leveraged to accelerate identification or exclusion of intracranial hemorrhage in high-risk OAC users [15,16]. In certain settings, prehospital administration of procoagulants or OAC reversal agents may be feasible and appropriate.
However, prehospital identification of a patient’s OAC use history may be difficult. In a retrospective analysis of 2,110 EMS responses for older persons (age ≥55 years) with a head injury, Nishijima et al. [2] found that EMS personnel correctly identified OAC use history in only 28% of patients. These findings are not surprising given that many EMS patients are unconscious or critically ill and that many older patients have cognitive impairment and may not accurately report medication history. Additionally, under certain prehospital environment conditions, EMS personnel may overlook medication history.
These findings underscore the need for alternate strategies for prehospital identification of OAC use history. Our study provides a foundational step in this direction by identifying patient and care factors that are associated with OAC use. This information is clinically significant; for example, unconscious patients who match the risk profile for OAC use may benefit from immediate triage to trauma centers or tertiary care facilities equipped to manage OAC-related hemorrhages. Furthermore, early administration of hemostatic agents (e.g., tranexamic acid) or OAC reversal agents (e.g., vitamin K, 4-factor prothrombin complex concentrate [4F-PCC], andexanet alfa, and idarucizumab) may be warranted in high-risk patients who match the risk profile for OAC use [24,25].
Limitations
While the ESO data set is nationally scaled, its representativeness across all regions remains unclear. Although the data include information from over 2,000 EMS agencies, it may not fully capture the diversity of care practices, which could affect the generalizability of our findings. Additionally, OAC use was determined by EMS personnel reports, which may have led to an underestimation of its prevalence. EMS providers likely relied on clinical data sources, such as patient and family reports, but did not have direct access to medication records or pharmacy databases. Also, determination of OAC history may have been difficult in patients who were unconscious or had altered mentation.
Our analysis identified instances of OAC use but did not explore the reasons for its use or the specific dosages involved. While the focus of this study was on patterns of OAC use and associations with certain conditions, we did not assess outcomes, which are influenced by the underlying disease process and require risk adjustment that is not standardized for linked EMS-hospital data. We also did not examine use of parenteral anticoagulants or other agents, like antiplatelet medications, that can also potentiate bleeding. Future research should address these limitations by including detailed medication histories, investigating the reasons for OAC use, and exploring patient outcomes.
Finally, we did not assess geographic variations in OAC use, though such mapping could aid in the strategic deployment of resources like mobile stroke units, procoagulant medications, and novel OAC reversal agents. Our observation of a larger proportion of EMS calls involving OAC users in the Northeast than in other regions suggests regional differences that warrant further investigation. Additionally, our reliance on data not specifically designed to address these research questions introduces challenges, including potential misclassification bias, unmeasured confounding factors, and missing data. These issues may limit the generalizability and reliability of our findings, and future studies should overcome these limitations to improve our understanding of the characteristics of OAC users in prehospital settings, potentially improving their management and outcomes.
Conclusions
In this national series, 1 in 18 adult EMS encounters involved patients using OACs. While OAC use was common in high-risk populations, injuries were less frequently reported. These findings underscore the critical need for heightened awareness and preparedness within EMS systems regarding OAC use, as these systems contain important epidemiologic information that can be used to improve protocols and enhance patient care for this vulnerable population.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
