Appropriate Antithrombotic Management for Older Adults Living with Dementia

Abstract

Antithrombotic medication is recommended for individuals who have a history of atrial fibrillation, venous thromboembolism, acute coronary events, or post-percutaneous coronary intervention. The purpose of this study was to describe the use of antithrombotics among older adults living with dementia at hospital admission and 1-month follow-up, treatments prescribed, and associated complications. The sample included 404 hospitalized older adults living with dementia, the majority of whom were White females, mean age in years of 82 (SD = 8). On admission, 69% of the patients were on at least one antithrombotic medication, and at 1-month post-discharge, this decreased to 64%. At 1-month post-discharge, the percentage of individuals on two or more antithrombotics decreased from admission at 34% to 14%. On admission, 11 (4%) of patients were admitted with adverse events from antithrombotics, and at 1-month post-discharge 5 (2%), patients were readmitted due to adverse events from antithrombotics. Given the risks and benefits of antithrombotic use among older adults living with dementia, a shared decision-making approach with patients and caregivers is recommended. This approach is the best way to help patients achieve their individual goals of care.

1. Introduction

There are a number of reasons antithrombotics are recommended, including when there is a diagnosis of atrial fibrillation for the prevention of stroke [1], for patients with post-acute coronary syndrome or post-percutaneous coronary intervention [2], or for treatment and prevention of venous thromboembolism [3,4]. The benefits of antithrombotic use varied depending on the reason for treatment. Treating atrial fibrillation in all adults, for example, has been noted to reduce stroke risk by almost two-thirds [5]. For those with atrial fibrillation, the CHA₂DS₂-VA score is the currently recommended validated tool to predict the risk of stroke and guide recommendations as it is more sensitive to those who are 75 years of age or older [1]. Scoring considers age, history of congestive heart failure, hypertension, prior stroke, trans-ischemic attach or thromboembolism, vascular disease, and diabetes mellitus. The risk of stroke is based on a score of ≥2, and if the score is 1, clinical judgment should be used to determine the pros and cons of thrombolytics. In a recent study [6] with a sample of 8294 patients with atrial fibrillation, it was noted that all of the individuals 65 years of age and older, based on use of the CHA₂DS₂-VASc score (which also includes gender as a risk factor), would have been started on anticoagulation. Thus, there are concerns that the CHA₂DS₂-VASc, which is similar to the CHA₂DS₂-VA, score may result in overprescribing thrombolytics to older adults [7,8].

For individuals with atrial fibrillation who have had a recent percutaneous coronary intervention, the recommendation is for triple therapy, which is defined as one antithrombotic (e.g., apixaban, rivaroxaban, warfarin, dabigatran, and enoxaparin) and two antiplatelet agents (e.g., aspirin and a P2Y₁₂ inhibitor such as clopidogrel, ticagrelor, prasugrel) for a short duration (depending on the risk of bleeding for the participant, this is usually just for one month) [9]. This is then followed by combination therapy with an oral antithrombotic plus a P2Y₁₂ inhibitor for four to six weeks or up to 12 months, again depending on the risk level for the patient [10,11]. After 12 months in patients with atrial fibrillation and stable coronary artery disease not requiring any further intervention, the guidelines recommend monotherapy with an oral antithrombotic, specifically an anticoagulant [10,11,12].

For patients with peripheral artery disease, the consensus recommendation [13] is that asymptomatic patients should not be treated. Patients with ischemia should be given the option for platelet anti-aggregation with either low-dose aspirin or a P2Y₁₂ inhibitor. A combination of aspirin with a low dose of a direct oral anticoagulant (DOAC), specifically rivaroxaban, could be considered depending on the patient’s risk factors for use of antithrombotic [14]. Patients at high risk of cardiogenic embolization should be considered for long-term antithrombotic treatment with a DOAC. These recommendations, however, are based on studies where the mean age of the participants was generally in the 70’s (some with small numbers of individuals > 80 years of age), and the participants were competent and able to self-consent [14].

For individuals who have experienced a venous thromboembolism (VTE), there are guidelines for initiation of treatment and recommendations for the use of long-term treatment [3,4]. If there is a known cause (e.g., surgery) for the VTE, treatment can be stopped generally after three to six months. For those in whom the VTE was spontaneous, and no clear cause was identified, long-term treatment may be an option if the individual is not at high risk for bleeding. In a recent study comparing treatment of patients who had a VTE, it was noted that there was no difference in recurrence rates of VTE among those who did versus did not receive continued antithrombotics [15]. More bleeding occurred, however, among those who were on prolonged antithrombotic treatment, with bleeding being most common for those on a standard-dose DOAC compared to low-dose DOACs.

2. Risk to Antithrombotic Use

There is a long list of risk factors that are associated with taking an antithrombotic as noted in

Table 1. Factors Associated with Increased Risk of Bleeding with Use of Antithrombotics.

Factors Associated with Increased Risk of Bleeding
Age > 65 years
A history of prior bleeding
A history of cancer, particularly metastatic disease
Renal impairment or failure
Liver impairment or failure
Thrombocytopenia
Anemia
Poor anticoagulant control/medication adherence
Comorbidity and reduced functional capacity
Frequent falls
Alcohol abuse
Use of multiple antithrombotics

[16]. The major risk considered is bleeding, generally referred to as major bleeding or clinically relevant nonmajor bleeding (e.g., requiring hospitalization or emergency room evaluation). Older adults particularly are at increased risk of subdural hematomas due to brain shrinkage and impact of the brain against the skull when there is any head trauma [17]. Bleeding risks are individually based and influenced by underlying disease and other medications [18,19]. There are also non-hemorrhagic risk factors associated with antithrombotic use, although these are less common and vary by type of treatment (

Table 2. Nonhemorrhagic Side Effects of Antithrombotics.

Aspirin	Vitamin K Antagonists Nonhemorrhagic Risk	DOACs Nonhemorrhagic Risks	P2Y12 Inhibitor
Allergic reactions including hives; difficult breathing; swelling of the face, lips, tongue, or throat. Ringing in the ears Confusion Hallucinations Tachypnea Seizures Nausea/vomiting, or Stomach pain Fever	Skin necrosis Limb gangrene Purple toe syndrome Possible vascular calcification, nephropathy, and osteoporosis.	Dyspepsia Renal impairment Dizziness Headache Dyspnea Abdominal pain	Allergic response: skin rash, itchiness, flushing, or shortness of breath Thrombotic thrombocytopenic purpura (TTP) Diarrhea Muscle pain

).

3. The Benefits and Risks of Antithrombotic Use Among Older Adults Living with Dementia

Although there are clear benefits to the use of antithrombotics in older adults for the prevention of stroke, myocardial infarction, or an embolic event, these need to be considered against the significant risks. Further, the benefits may only be effective with certain treatment approaches (e.g., a decrease in myocardial infarction is only effective with rivaroxaban 2.5 mg and dual pathway inhibition) [20]. The cost/benefit of the use of antithrombotics has not been well studied in older adults living with dementia. These individuals commonly experience transitions of care and changes in their primary health care providers and are unable to provide information as to why antithrombotic use has been initiated. Through these transitions, antithrombotics may be continued or discontinued without a strong rationale to do so.

To obtain a sense of the current use of antithrombotics among older adults living with dementia during care transitions, we evaluated the use of these medications from admission to acute care to one-month post-discharge. Specifically, the percentage of individuals on antithrombotics, the number of antithrombotics prescribed, and evidence of associated bleeding with treatment on admission and at one-month post-discharge were evaluated. The findings from this study can be used to guide approaches for determining antithrombotic treatment among older adults living with dementia.

4. Methods

4.1. Design

This was a descriptive study using baseline data and one-month follow-up data from the study Testing the Implementation of Function Focused Care in Hospitalized Older Adults Living with Dementia (R01AG065338; Clinicaltrials.gov NCT04235374). The study was reviewed and approved by the University of Maryland Institutional Review Board. To participate, hospitals had to have a unit dedicated to medical patients. Patients were eligible to participate if they were 55 years of age or older, were admitted with a medical diagnosis, screened positive for dementia based on a score of ≤20 on the Saint Louis University Mental Status Examination [21] completed by the participant, and completion of the following three measures based on input from the legally authorized representative: a score of >2 on the AD8 Dementia Screening Interview [22] based on input from the legally authorized representative; a score of 0.5 to 2.0 on the Clinical Dementia Rating Scale [23]; and a score of ≥9 on the Functional Activities Questionnaire to differentiate between dementia and mild cognitive impairment [24]. Patients were excluded if they were enrolled in Hospice, on the unit for more than 48 h, had no contact noted to provide follow-up information, required surgery, or had psychiatric or neurological problems that caused cognitive changes not due to dementia such as a brain tumor. All potentially eligible patients were approached about the study and completed an Evaluation to Sign Consent. The Evaluation to Sign Consent includes 5 items that ask potential participants such things as what they would be expected to do if participating in the study, and what to do if they wanted to stop participating. If the individual did not pass the Evaluation to Sign Consent, assent was obtained to contact their legally authorized representative and proceed with the consenting process. A total of 6471 patients were screened with ineligibility being due to having COVID-19, discharged or not available to be approached, the patient was non-communicative and could not even assent, not being eligible based on criteria, or refusals by patient or legally authorized representative. A total of 404 patients were recruited.

4.2. Procedures, Measures, and Data Analysis

Data were collected by research evaluators based on chart review within the first 48 h of admission to the hospital unit and via telephone follow-up at one-month post-discharge from the patient’s caregiver/legally authorized representative. In addition, any adverse events including falls, emergency room transfers, hospital admissions, and death were obtained during the telephone follow-up. Demographic and descriptive information was also obtained from electronic health records and included the reason for admission, age, gender, race, and comorbidities. Cognitive status was obtained using the Saint Louis Mental Status Exam (SLUMS) [21]. Descriptive analyses using SPSS 28.0 including means, standard deviations, frequencies, and percentages were performed to describe the sample, medication use, and adverse events.

5. Results

The first 404 participants in the parent study were included in this analysis. The mean age in years of the participants was 82 (SD = 8) and the majority were female (252, 62%) and White (282, 70%) older adults. The mean SLUMS was 7.7 (SD = 6.0), which indicated moderate to severe dementia, and they had an average of 3 (SD = 2) comorbidities. These individuals were from 10 hospitals. Comorbidities of the participants are shown in

Table 3. Description of Comorbidities Among the Sample (N = 404).

Comorbidity	Frequency	Percentage
Myocardial infarction Yes No	14 385	3 97
Congestive Heart Failure Yes No	97 302	24 76
Peripheral Vascular Disease Yes No	41 358	10 90
Dementia Yes No	192 207	48 52
Cerebral Vascular Disease Yes No	52 347	13 87
Chronic Pulmonary Disease Yes No	42 357	10 90
Gastrointestinal Ulcers Yes No	7 391	2 98
Liver Disease Yes No	13 394	4 96
Hemiplegia Yes No	7 392	2 98
Renal Disease Yes No	129 270	32 68
Diabetes Yes No	135 264	34 66
Cancer/Tumor Yes No	51 348	13 87
Leukemia Yes No	5 394	1 99
Lymphoma Yes No	6 393	1 99
Metastatic Cancer Yes No	16 383	4 96

. Admission diagnoses varied significantly across the sample including such things as abdominal pain, abnormal labs, a jaw fracture, a change in function, anemia, change in cognition, cystitis, diarrhea, congestive heart failure, pneumonia or other infections, renal failure, respiratory failure, fever, weakness, among others. Discharge location also varied with 188 (47%) living in home settings (with a caregiver), 25 (6%) living in assisted living, 181 (45%) in nursing homes or subacute settings, 5 (1%) died, 3 (1%) went to acute rehabilitation settings, and 2 (1%) were unknown locations. At the time of admission, there were 124 (31%) individuals who were not on any antithrombotic treatment with the remaining 280 (69%) on at least one antithrombotic medication. As shown in

Table 4. Admission Anticoagulant or Antiplatelet Medications (280 Individuals on Treatment).

Admission Single Antithrombotic
Medication	Number of Individuals (%)
Apixaban	42 (18%)
Clopidogrel	7 (3%)
Aspirin	69 (29%)
Rivaroxaban	9 (4%)
Dabigatran	2 (1%)
Enoxaparin sodium injection	20 (8%)
Heparin injection	30 (13%)
Warfarin	8 (3%)
Ticagrelor	0 (0%)
Admission Two Antithrombotics (an antiplatelet and an anticoagulant)
Clopidogrel and Heparin injection	6 (3%)
Clopidogrel and Enoxaparin injection	2 (1)
Clopidogrel and Apixaban	4 (2%)
Clopidogrel and Rivaroxaban	1 (1%)
Aspirin and Heparin injection	24 (10%)
Aspirin and Enoxaparin injection	17 (7%)
Aspirin and Apixaban	7 (3%)
Aspirin and Warfarin	3 (1%)
Aspirin and Rivaroxaban	3 (1%)
Clopidogrel and Aspirin	10 (4%)
Aspirin and Ticagrelo	2 (1%)
Anticoagulant and Anticoagulant
Rivaroxaban and Enoxaparin injection	1 (1%)
Heparin injection and Enoxaparin injection	1 (1%)
Heparin injection and Apixaban	1 (1%)
Enoxaparin injection and warfarin	1 (1%)
Admission Three Antithrombotics
Clopidogrel, Heparin, and Enoxaparin injection	1 (1%)
Aspirin, Apixaban, and Enoxaparin injection	1 (1%)
Aspirin, Enoxaparin, and Heparin injection	1 (1%)
Aspirin, Clopidogrel, and Heparin injection	1 (1%)
Heparin injection, Aspirin, Clopidogrel	2 (1%)
Enoxaparin injection, Clopidogrel, Aspirin	1 (1%
Aspirin, Clopidogrel, and Enoxaparin injection	2 (1%)
Aspirin, Clopidogrel, Rivaroxaban	1 (1%)

, there were 187 patients (67%) out of the 280 treated with one antithrombotic medication. Specifically, 83 individuals (30%) were on a combination of two medications with 67 (24%) on an antiplatelet and anticoagulant, 12 (4%) on two antiplatelet medications, and 4 (1%) on two anticoagulants. There were 10 (4%) individuals out of the 280 being treated who were on three antithrombotics. Of the 280 individuals on antithrombotics on admission, only 13 individuals (5%) had a clear reason documented in the medical record for treatment (atrial fibrillation, a venous thrombus embolism, or acute cardiac event based on admission diagnoses or comorbidities). Among the 280 participants on antithrombotics, there were 12 individuals (4%) who were admitted due to an antithrombotic-associated complication (evidence of bleeding based on a head trauma, evidence of blood in stool, or anemia). Only 4 (33%) out of the 12 individuals admitted with bleeding-related events were on antithrombotics.

At one-month post-hospital, discharge data were obtained on 340 (84%) out of the 404 participants. Among the missing 64 individuals, 25 (39%) died, 14 (22%) were unreachable, and the others were pending follow-up (25, 39%). As shown in

Table 5. Month 1 Antithrombic Therapy (N = 340).

Month 1 Single Anticoagulant or Antiplatelet Medication
Medication	Number of Individuals (%)
Apixaban	44 (13%)
Clopidogrel	15 (4%)
Aspirin	99 (29%)
Rivaroxaban	13 (4%)
Dabigatran	2 (1$)
Enoxaparin sodium injection	1 (1%)
Heparin injection	2 (1%)
Warfarin	10 (3%)
Ticagrelor	0 (0%)
Medication	Number of Individuals
Month 1 Double Antithrombotic Therapy
Clopidogrel and Heparin injection	0 (0%)
Clopidogrel and Enoxaparin injection	0 (0%)
Clopidogrel and Apixaban	4 (1%)
Clopidogrel and Rivaroxaban	1 (1%)
Aspirin and Heparin injection	0 (0%)
Aspirin and Enoxaparin injection	2 (1%)
Aspirin and Apixaban	7 (2%)
Aspirin and Warfarin	2 (1%)
Aspirin and Rivaroxaban	2 (1%)
Clopidogrel and Aspirin	10 (3%)
Aspirin and Ticagrelor	1 (1%)
Rivaroxaban and Enoxaparin injection	0 (0%)
Heparin injection and Enoxaparin injection	0 (0%)
Heparin injection and Apixaban	0 (0%)
Enoxaparin injection and Warfarin	0 (0%)
Month 1 Three Antithrombotic Therapies
Medication	Number of Individuals
Aspirin, Enoxaparin injection, Warfarin	1 (1%)

, of the 340 individuals with follow-up data, there were 219 (64%) individuals on antithrombotic therapy and 121 (36%) on no antithrombotic therapy. Of the 219, 188 (86%) were on one antithrombotic therapy, 30 (13%) were on a combination of antithrombotic therapies, and one individual (1%) was still on three antithrombotic therapies, two antiplatelet medications, and one anticoagulant. At one-month post-discharge, there were 5 (1%) individuals who had complications from antithrombotic therapy. As with admission complications associated with antithrombotics, complications at 1 month were based on evidence of gastrointestinal bleeding and head trauma. Only 2 (40%) out of the 5 individuals were on antithrombotics.

Over time, there was a decrease in the percentage of older adults living with dementia who were on antithrombotics (

Table 6. Percentage of Older Adults Living with Dementia on Antithrombotics and Potentially Associated Complications Between Hospital Admission and One-Month Post-Discharge.

	Admission (N = 404)		One-Month Post-Discharge (N = 340)
	On Anticoagulation N (%)	Not On Antithrombotic (%)	On Anticoagulation N (%)	Not On Anticoagulation N (%)
Any Anticoagulant	280 (69%)	124 (31%)	205 (60%)	115 (40%)
1 anticoagulant	186 (66%)	-	176 (86%)	-
2 antithrombotic	84 (30%)	-	28 (13%)	-
3 antithrombotic	10 (4%)	-	1 (1%)	-
Adverse events associated with antithrombotic therapy	11 (4%)	-	5 (2%)	-

). For those still being treated, there was an increase in the number and percentage of individuals on a single antithrombotic treatment and a decrease in the percentage of individuals on two and three treatments.

6. Discussion

This was a descriptive study intending to review the current use of antithrombics in a hospitalized sample of older adults living with dementia. Our descriptive findings showed that the rates of antithrombotic use in our current study of older hospitalized patients living with dementia were similar to prior research evaluating all adults across different sites of care (e.g., home, nursing home, hospital), with rates ranging from 48% to 83% [25,26]. In our sample, there was a slight decrease in antithrombotic use over one-month post-discharge, but this was still at 60%. There was also a decrease noted in the percentage of individuals on multiple drugs. We do not know the rationale for the use of antithrombotics and whether or not it is appropriate among these individuals as there was no provider documentation around medication prescribing. Guidelines regarding use continue to evolve and ongoing work is needed to carefully monitor the appropriate use of these medications.

The impact of antithrombotic use in this study noted that 3% of the study participants were admitted to the hospital with major bleeding-associated events and 1% had major bleeding-associated events at one-month post-discharge. This was consistent with an older 2006 study that reported the major risk of bleeding for individuals on antithrombotics ranged from 1 to 3% [27]. In a recent 2023 study, however, the rate of bleeding was higher at 10% [28]. Our low rates may be due to incorrect admission or readmission diagnoses and reasons for admission may not have been accurately documented on admission and were based on input from a legally authorized representative at 1 month. Further, there were many situations in which the legally authorized representative did not know the reason for readmission.

6.1. Making Treatment Decisions

The rates of known adverse events associated with antithrombotic use were low in this study. Further, we do not know if antithrombotic treatment prevented a cardiovascular event. Given the known benefits [29,30] as well as the known risks associated with the use of these medications [16,17,31,32], it is challenging to know what treatment recommendations to make when working with older adults living with dementia. Benefits may be particularly limited for nursing home residents living with dementia as prior research noted that treatment with antithrombotics for atrial fibrillation provided only a modest benefit for risk of death with a promise of 76 days of longer life and no protection against stroke [31]. Risks of bleeding increase with the use of dual or triple antithrombotic therapy [30]. It is therefore critically important to establish if and when use of antithrombotics is needed. This requires obtaining information from prior providers to determine the rationale for use (e.g., dates of cardiovascular procedures or events, history of atrial fibrillation).

6.2. Study Strengths and Limitations

A strength of this study was the large sample of older adults living with dementia from across 10 hospitals. This was a secondary data analysis using data from the first 404 participants in the parent study; thus, detailed data such as the reasons providers had individuals on antithrombotics and the reasons for continuing them at discharge or discontinuing them were not obtained. Likewise, the one-month follow-up data for adverse events were based on the recall of the legally authorized representatives and not hospital records and documentation of the actual cause of the readmission and if it was related to anticoagulant use. Further, we do not know the number of events that may have been prevented if individuals were on antithrombotics (e.g., strokes, blood clots).

6.3. Conclusions

The findings from this study indicate that there is still a high rate of antithrombotic use among older adults living with dementia. We noted there are some risks for use based on causes for hospital admission and readmission during the first-month post-discharge. Guidance for antithrombotic use among older adults, particularly those living with dementia and those greater than 90 years of age, is vague, and the recommendation suggests the use of “careful consideration” when making treatment decisions [30]. In situations in which there are not clear data to guide practice such as what we have seen with regard to some immunizations [33], a shared decision-making and a person-centered approach may be the most appropriate way to proceed. Person-centered care is focused on engaging the individual or their legally authorized representative in shared decision-making around a care-related topic. Helping older adults living with dementia and their legally authorized representatives understand the benefits and risks associated with antithrombotic use and identifying their goals of care can help these individuals establish what risk they are willing to take. Documentation of this discussion and ongoing evaluation of the use of antithrombotics should also be conducted. With increased opportunities for treatment in terms of number and types of drugs available, our understanding of the risks and benefits associated with antithrombotics will develop as new data are accrued. For now, the person-centered shared decision-making approach may be the safest and most appropriate way to address antithrombotic use among older adults living with dementia.