As the population ages, an increasing number of patients are taking anticoagulant or antiplatelet medications.1-4 Anticoagulant and antiplatelet use increases the risk of traumatic intracranial hemorrhage and trauma-related morbidity and mortality.5-8 This risk is a source of concern for clinicians, with early identification using cranial computed tomographic (CT) imaging being paramount. In addition, further concern exists regarding the risk of intracranial hemorrhage and neurologic deterioration after an initial cranial CT scan with normal results (ie, a delayed traumatic intracranial hemorrhage).9,10 Previous studies11-14 of patients taking anticoagulant or antiplatelet medications who had blunt head trauma have reported delayed intracranial hemorrhage rates from 0.6% to 6%.
Current guidelines recommend immediate neuroimaging for adult patients with head injury taking anticoagulant or antiplatelet medications and in patients 65 years and older.15-17 However, the management beyond the initial cranial CT scanning is variable. Some experts recommend admission for 24 hours of observation and routine repeated cranial CT imaging, while others recommend immediate discharge if the initial CT imaging demonstrates no intracranial hemorrhage.14,18 Actual data on the risk of delayed traumatic intracranial hemorrhage are limited, particularly for patients taking newer medications, such as the direct-acting oral anticoagulants (DOACs) or in patients not taking any anticoagulant or antiplatelet medications. The primary objective of this study was to investigate the overall incidence of delayed traumatic intracranial hemorrhage in older adults with head trauma and specifically across anticoagulant and antiplatelet medications.
This investigation was a countywide prospective observational cohort study conducted at 11 hospitals in northern California. Four hospitals serve as level I or II trauma centers, and 7 are nontrauma centers. Institutional review board approval was obtained at all study sites, and oral informed consent was obtained from all study participants.
Patients 55 years and older with head trauma who were transported to a participating hospital by emergency medical services (EMS) between August 1, 2015, and September 30, 2016, were eligible for inclusion. We excluded patients with penetrating head trauma, those with interfacility transfers, those with intracranial hemorrhage on the initial cranial CT, those who did not undergo cranial CT at their index emergency department (ED) visit, those who declined consent for a follow-up telephone call, and those who were without reliable means for such a call, as well as people who were incarcerated.
The EMS clinicians completed a standardized data collection form that included demographic and clinical variables. The EMS records were linked to hospital records using name, date of birth, and EMS transport dates. A trained research coordinator (1 of us, S.D.G.) abstracted data variables from EMS and hospital electronic medical records (EMRs), including patient demographics, mechanism of injury, anticoagulant or antiplatelet use, laboratory results, cranial CT results, Abbreviated Injury Scale score and Injury Severity Score for hospitalized patients, and ED and hospital disposition. Anticoagulant use included warfarin sodium or DOACs (dabigatran, rivaroxaban, apixaban, and edoxaban). Antiplatelet medications included aspirin, clopidogrel bisulfate, ticlopidine hydrochloride, prasugrel, dipyridamole, cilostazol, and ticagrelor. Isolated head injury was defined as an Abbreviated Injury Scale score less than 3 in all nonhead body regions.19 Cranial CT imaging and hospital admission were conducted at the discretion of the patients’ treating physicians.
Patients were followed up after their index ED visit to assess for the incidence of delayed traumatic intracranial hemorrhage within 14 days. Patients who were admitted to the hospital for 14 days or more underwent EMR review. Patients discharged from the ED or patients who were hospitalized for less than 14 days had a standardized telephone questionnaire completed 14 to 28 days after the index ED visit. Patients or their surrogates were asked sequential questions to determine if an intracranial hemorrhage had been reported on a follow-up cranial CT scan. If an intracranial hemorrhage was documented on a follow-up cranial CT scan, this was confirmed through review of the EMR and cranial CT imaging reports at the imaging site. If the patient was dead at the time of follow-up, we asked surrogates if the patient’s death was from a head injury. We attempted to contact patients or their surrogates up to 6 times, varying the time of the telephone call. If unsuccessful after 6 attempts, patients were considered lost to follow-up. We reviewed the EMR of patients who were dead at the time of follow-up and that of patients who were lost to follow-up to evaluate for return ED visits.
Our primary outcome measure was the incidence of delayed traumatic intracranial hemorrhage on cranial CT within 14 days of the index ED visit in the study without the patient having experienced an additional head injury. We reported the incidence of delayed traumatic intracranial hemorrhage as a proportion with 95% CIs.
Patient characteristics are reported using descriptive statistics. We also performed a sensitivity analysis based on whether the initial presenting hospital was at a trauma center. Normally distributed continuous data are reported as means (SDs), and nonnormally distributed data are reported as medians with 25% to 75% interquartile ranges (IQRs). Data were analyzed using statistical software (Stata, version 14.2; StataCorp).
A total of 1356 patients with head injury were identified during the study period. We excluded 497 patients, 144 of whom had no reliable means of follow-up, 130 of whom did not consent for follow-up, 128 of whom did not have a cranial CT scan obtained during the initial ED evaluation or hospitalization, and 95 of whom had intracranial hemorrhage on the initial cranial CT imaging, leaving 859 patients in the study cohort (Figure). The cohort had a median age of 75 years (IQR, 64-85 years), and 389 (45.3%) were male. A total of 343 patients (39.9%) were taking an anticoagulant or antiplatelet medication. Patient characteristics are listed in Table 1.
Three of the 859 patients (0.3%; 95% CI, 0.1%-1.0%) had a delayed traumatic intracranial hemorrhage. Of the 3 patients, 1 of 75 patients (1.3%; 95% CI, 0.0%-7.2%) who were taking warfarin alone and 2 of 516 patients (0.4%; 95% CI, 0.1%-1.4%) who were not taking any anticoagulant or antiplatelet medications had a delayed traumatic intracranial hemorrhage. No patients taking aspirin, other antiplatelet medications, DOACs, or concomitant medications had a delayed traumatic intracranial hemorrhage. The specific characteristics of the 3 patients with a delayed traumatic intracranial hemorrhage are listed in Table 2. There was a similar incidence of delayed traumatic intracranial hemorrhage in patients initially seen at a trauma center (2 of 514 [0.4%]; 95% CI, 0.0%-1.4%) compared with those initially seen at a nontrauma center (1 of 345 [0.3%]; 95% CI, 0.0%-1.6%).
At the time of telephone follow-up, 30 patients (3.5%; 95% CI, 2.3%-4.9%) had died, and 39 patients (4.5%; 95% CI, 3.2%-6.2%) were lost to follow-up. This includes 8 patients taking warfarin alone (5 had died, and 3 were lost to follow-up), 7 patients taking DOAC medications alone (6 had died, and 1 was lost to follow-up), 11 patients taking aspirin alone (7 had died, and 4 were lost to follow-up), 3 patients taking other antiplatelet medications (0 had died, and 3 were lost to follow-up), 3 patients taking concomitant anticoagulant and antiplatelet medications (1 had died, and 2 were lost to follow-up), and 37 patients taking no anticoagulant or antiplatelet medications (11 had died, and 26 were lost to follow-up). None of the patients who had died at the time of follow-up had demonstrated any repeated ED visits or hospitalizations suggestive of a delayed traumatic intracranial hemorrhage.
Unlike prior studies11-13 that evaluated the incidence of delayed traumatic intracranial hemorrhage in patients taking anticoagulant or antiplatelet medications, this study included patients taking a broad range of anticoagulant and antiplatelet medications, as well as patients not taking any of these medications. We found that less than 1% of older adults with head injuries with an initial cranial CT scan with normal results have a delayed traumatic intracranial hemorrhage within 14 days of the index ED visit. This risk was also low across specific anticoagulation and antiplatelet medications. Our findings are consistent with those of prior studies11-14 that focused on patients taking anticoagulant and antiplatelet medications.
Two of the 3 patients with a delayed traumatic intracranial hemorrhage were not taking any anticoagulant or antiplatelet medications. It is also notable that in 2 of the 3 patients the delayed traumatic intracranial hemorrhage occurred 3 days and 5 days after the initial cranial CT scan. Therefore, it is likely that even with 24-hour observation and repeated cranial CT imaging a delayed traumatic intracranial hemorrhage might not have been identified in these patients. Based on our results and assuming the higher end of the 95% CI and a standard 24-hour observation period, 115 patients taking anticoagulant or antiplatelet medications and 93 patients not taking anticoagulant or antiplatelet medications would need to be admitted for repeated imaging to detect one delayed traumatic intracranial hemorrhage.
This study has limitations. It is possible that patients who were dead at the time of follow-up or who were lost to follow-up had delayed traumatic intracranial hemorrhages that were undiagnosed. However, for patients who were dead at the time of follow-up, we ascertained from family or caregivers that none of these patients had additional hospital visits for symptoms consistent with a delayed traumatic intracranial hemorrhage. We also only included patients transported via EMS. This was done as a means to reliably identify patients with blunt head trauma prospectively. The inclusion of only EMS-transported patients likely resulted in a population who had more significant mechanisms of injury and higher injury severity compared with those who would self-present to the ED. This would be expected to overestimate the risk for a delayed traumatic intracranial hemorrhage, further strengthening our findings. Finally, the small number of patients in each anticoagulant and antiplatelet group resulted in wide 95% CIs for specific anticoagulant or antiplatelet medications.
The risk of delayed traumatic intracranial hemorrhage is low in older adults after blunt head trauma regardless of their specific anticoagulant or antiplatelet medication. A delayed traumatic intracranial hemorrhage may also occur in patients not taking any anticoagulant or antiplatelet medications. A standard 24-hour observation period with repeated imaging would not have detected 2 of the 3 delayed traumatic intracranial hemorrhage cases in our study. This highlights the importance of clinical judgment regarding the severity of trauma, additional injuries, and ability to monitor the patient for deterioration when making decisions about admission for older patients after blunt head trauma.