Published online: 30.01.

2020

74 Review Article

Oral Anticoagulation in the Elderly and Frail

Rupert M. Bauersachs1,2 Joerg Herold1

1 Department of Vascular Medicine, Klinikum Darmstadt GmbH, Address for correspondence Rupert M. Bauersachs, Department of
Darmstadt, Germany Vascular Medicine, Klinikum Darmstadt GmbH, Grafenstraße 9, 64283
2 Center for Thrombosis and Hemostasis, University Medical Center, Darmstadt, Germany (e-mail: bauersachs@em.uni-frankfurt.de).
Mainz, Germany

Hämostaseologie 2020;40:74–83.

Abstract The proportion of elderly patients will increase substantially over the next decades, and
both atrial fibrillation (AF) and venous thromboembolism (VTE) are more common in

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
the elderly. Age is a risk factor not only for stroke and thromboembolism but also for
bleeding, particularly in frail patients, in whom numerous pathophysiological changes
occur that alter drug kinetics and toxicity of standard doses of oral anticoagulants
(OACs). AF trials showed that the relative benefits of direct OACs (DOACs) also applied
to elderly patients, and due to their higher risk this translates into a higher absolute risk
reduction compared with vitamin K antagonists, suggesting that DOACs are the better
choice. All DOACs—at varying extent—are eliminated via the kidney and it is crucial to
evaluate renal function at initiation and during follow-up, especially for dabigatran. The
fear of falls is a common reason against OAC. However, there is still a benefit with OAC,
particularly with DOACs given the lower risk of intracranial hemorrhage. Polypharmacy
Keywords represents a common challenge, nevertheless DOACs and warfarin were classified as
► anticoagulation beneficial. Nonetheless, attempts should be undertaken to reduce comedication, and
► elderly drug–drug interactions should be assessed. Coadministration of platelet inhibitors
► frail increases bleeding risk and should be avoided. In conclusion, elderly and frail patients
► atrial fibrillation requiring anticoagulation for AF or VTE are at higher risk of adverse outcomes, but also
► venous have a higher absolute benefit from OAC. Important practical aspects to improve
thromboembolism efficacy and safety in this challenging population are summarized in this overview.

Epidemiology
future. In addition, the case fatality rate of VTE is higher in
Elderly Patients and Venous Thromboembolism the elderly, especially in those with cancer, which is more
In Germany the percentage of persons above 60 years will prevalent at higher age. Unfortunately, the risk of major
increase by 39% from 2013 to 2050, when more than one- bleeding during anticoagulation is also age-dependent, con-
third of the population will be above 60 years,1 and the tributing to the vulnerability of this patient group.3 In view of
proportion above 67 years will increase by 33% from 2020 these challenges, it is remarkable that this age group is
to 2060, when more than 21 million persons above that age underrepresented in most of the large VTE trials,5 which is
will live in Germany.2 While the overall incidence of venous an additional hurdle for the translation of evidence-based
thromboembolism (VTE) is approximately 1 to 2 per 1,000 data into clinical practice.
per year, it increases to 1% per year in the elderly.3 Thus, the
great majority of patients treated for VTE are elderly: already Elderly Patients and Atrial Fibrillation
20 years ago more than 70% of VTE patients were above The increase of the elderly population has a similar bearing on
60 years,4 and it is foreseeable that this will increase in the the number of patients suffering from atrial fibrillation (AF)

received © 2020 Georg Thieme Verlag KG DOI https://doi.org/

September 11, 2019 Stuttgart · New York 10.1055/s-0040-1701476.
accepted after revision ISSN 0720-9355.
December 10, 2019
 Anticoagulation in Elderly and Frail Bauersachs, Herold 75

and the subsequent risk of major stroke. The ATRIA study,6 doses of VKA are usually lower in elderly patients, and the
based on a total of 17,974 patients with AF, showed that the recommended standard warfarin starting dose of 5 mg may be
prevalence of AF increases from 0.1% in adults younger than too high for up to 82% of women and 65% of men aged over
55 years to 9.0% in persons of 80 years or older. It was projected 70 years.19,20 Because of the age-related physiological changes,
that the prevalence of AF will increase in the United States to comorbidities, and concomitant medications, the management
more than 5.6 million by the year 2050, with more than 50% of of VKA in elderly patients is challenging and associated with an
affected individuals being 80 years or older. The importance of increased risk of adverse events.21 In addition, older patients
age as a serious risk factor for stroke in AF is reflected in the more often have problems to attend clinics regularly for INR
CHA2DS2-VASc Score,7 where age is counted double. monitoring and to maintain—often complicated—dosing
schedules. Therefore, poor anticoagulation control with the
Frailty classical VKA treatment is likely,21 and there is an increased risk
Frailty is a geriatric syndrome resulting from age-related of bleeding in anticoagulated patients above 65 years.19 None-
cumulative declines across multiple of physiologic systems, theless, the risk for fatal PE is greater than the risk of fatal
with impaired homeostatic reserve and a reduced capacity of bleeding in elderly patients, emphasizing the need to carefully
evaluate the benefit–risk ratio in these patients.16

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
the organism to withstand stress, thus increasing vulnerability
to adverse health outcomes including falls, hospitalization, These considerations often lead physicians to not prescribe
institutionalization, and mortality.8,9 However, also other oral anticoagulant (OAC), which remains underutilized in the
losses of human functioning occur, i.e., psychological or social. elderly,22 even though it has been shown in several analyses that
The actual state of the frail elderly person is dynamic and it can elderly patients benefit from VKA, and even more from direct
be positioned on a continuum between nonfrail and frail.10 Not OAC (DOAC) administration.23 A systematic review and meta-
unexpectedly therefore, different and varying definitions of regression analysis of 10 studies comparing warfarin with no
frailty are being utilized. Most commonly, a phenotypic defi- warfarin and 16 studies comparing warfarin with DOACs was
nition of frailty is being used, based on readily identifiable performed in AF patients above 65 years. Warfarin was found to
physical aspects11; three or more of the following character- be superior to no antithrombotic therapy [relative risk (RR): 0.59
istics support a frailty diagnosis: unintended weight loss, (95% confidence interval [CI]: 0.51–0.76)] and aspirin [RR: 0.44
exhaustion, weakness, slow gait speed, and low physical (0.24–0.64)] for stroke/thromboembolism (TE) prevention.
activity.8 Recent trials on antithrombotic treatment have Warfarin use was associated with a nonsignificant increase in
defined fragile patients as including, but not limited to, elderly risk of major bleeding compared with no antithrombotic thera-
patients (e.g., above 75 years), patients with renal impairment py (RR: 1.26 [0.99–1.52]) or with aspirin (RR: 1.20 [0.91–1.50]),
(glomerular filtration rate [GFR] below 50 mL/min), and those respectively. DOACs were superior to warfarin for stroke/TE
with low body weight (below 50 kg).12 Fragility is associated prevention (hazard ratio [HR]: 0.81 [0.73–0.89]), and DOACs also
with decreased reserve to challenges due to a general decline were associated with a reduced risk of major bleeding compared
in physiologic systems leading to poorer outcomes, complica- with warfarin (HR: 0.87 [0.77–0.97]).
tions, and mortality.12–14 In a recent systematic review and The phase III trials of DOACs in AF included a significant
meta-analysis, approximately 40% of adults with AF over the proportion of patients above 75 years, ranging from 31 to 43%
age of 80 who are admitted to an acute care hospital and representing more than 27,000 elderly patients in whom
are diagnosed as frail.15 Clinical studies as well as data from DOAC treatment was analyzed. A meta-analysis showed no
the prospective RIETE registry indicate that VTE treatment in interaction for age with respect to both safety and efficacy,24
elderly patients and those with renal impairment or low body indicating that the RR difference was not different for elderly
weight is associated with a higher bleeding risk.16,17 patients, while the absolute risk reduction for both thrombotic
and bleeding events was higher in the elderly with DOACs
compared with VKA, translating into a lower number needed to
Challenges of Anticoagulation in the Elderly
treat compared with younger patients.25 The PREFER in AF
and Frail
prospective registry26 with 3,825 patients above 75 years
Elderly Patients showed that the net composite endpoint, including major
Multiple physiological and pathological changes occur in bleeding and ischemic cardiovascular events, occurred in
the elderly including changes in body composition, relative 6.6% per year with DOACs (apixaban, dabigatran, and rivarox-
increase of body fat due to a loss of lean body mass, and aban) and in 9.1% per year with VKA (odds ratio [OR]: 0.71
reductions in total body water, which alter drug kinetics18 [0.51–0.99]), and DOAC therapy was associated with a lower
and the distribution volume of anticoagulation drugs, increas- rate of major bleeding compared with VKA use (OR: 0.58
ing the toxicity of standard doses.19 Due to limited dietary [0.38–0.90]). It was concluded that DOACs are associated
vitamin K intake or reduced absorption of vitamin K, elderly with a better net clinical benefit in elderly AF patients,
patients have a lower ability to synthesize clotting factors.19 For primarily due to lower rates of major bleeding.
clinical management it is important to be aware that elderly In VTE trials also, a reduced risk of major bleeding was
patients show the greatest variability in vitamin K antagonist found with rivaroxaban in patients above 75 years [HR: 0.26
(VKA) dose requirements and take longer time to return to a [0.12–0.56]), CrCl below 50 mL/min (0.21 [0.06–0.73]), or frail
normal international normalized ratio (INR) from either thera- patients (0.27 [0.13–0.54]).14,27,28 Similarly, major bleeding was
peutic or supratherapeutic INR.19 The initial and maintenance reduced with apixaban versus VKA (HR: 0.63 [0.51–0.77]).29

Hämostaseologie Vol. 40 No. 1/2020

76 Anticoagulation in Elderly and Frail Bauersachs, Herold

Also, in the phase III VTE trial with edoxaban, a high efficacy was Vitamin K Antagonists
found in elderly patients.30 In patients with impaired renal function treated with VKA, an
increased risk for bleeding is well known and this is reflected in
Renal Impairment several bleeding scores including the HAS-BLED Score45 or the
There is a well-known association between increasing age and HEMORR2HAGES Score.46 Due to the increased bleeding with
deterioration of renal function13 (►Fig. 1). Criteria for catego- VKA in patients with CKD, an FDA (U.S. Food and Drug
rizing chronic kidney disease (CKD) have been published31–33 Administration) black-box warning was issued for warfarin,
and they are mainly based on equations to estimate the GFR and in Germany phenprocoumon is contraindicated in manifest
from serum creatinine (eGFRcreat) rather than relying on the renal insufficiency. Renal impairment alters binding to plasma
serum creatinine alone. The degree of albuminuria further proteins, volume of distribution, and nonrenal clearance of
defines the outcome of CKD. many drugs, leading either to toxicity or ineffective therapy.47
In AF there is a double-sided relationship with CKD, as AF Although warfarin is hepatically cleared, CKD may impact its
facilitates the progression of CKD, while on the other hand the metabolism due to downregulation of hepatic enzymes.47 Thus,
incidence and prevalence of AF increases with deteriorating patients with CKD require more frequent monitoring to ensure
renal function.34 In the Chronic Renal Insufficiency Cohort therapeutic anticoagulation with VKAs.48 CKD is associated

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
(CRIC),35 nearly one in five participants with CKD had evidence with decreased warfarin maintenance doses and poorer anti-
of AF, a prevalence similar to patients with end-stage renal coagulation stability.49 Warfarin dose adjustments are
disease and two to three times higher than the general required twice as often in patients with CKD compared with
population. Patients with impaired renal function and AF or patients with normal renal function (22 vs. 12% of visits,
after VTE have a significantly increased risk for stroke or VTE respectively), and time in therapeutic range (TTR) is signifi-
recurrence, respectively.36–40 At the same time these patients cantly lower in these patients (62 vs. 74%).34 Patients with CKD
are at an increased risk for bleeding,39,40 with an additional are four times more likely to be overanticoagulated (INR > 4.0),
increase while on OAC.41 increasing the possibility of hemorrhage.48,49 The incidence of
For VTE, renal insufficiency had not been considered a both minor and major bleeding events is significantly increased
classical risk factor for the occurrence of VTE; however, an in patients with severe CKD compared with patients with
analysis of more than 75,000 postoperative patients showed moderate CKD and those with normal renal function.48
a doubling of the VTE incidence in patients with renal
insufficiency.42 More than 60% of all VTE occur in patients Direct Oral Anticoagulants
over 70 years, and up to 25% of patients hospitalized for VTE All four available DOACs are at least partially eliminated via the
have moderate-to-severe renal impairment.43,44 kidney, with dabigatran etexilate having the largest extent of

Fig. 1 Correlation between different stages of renal impairment according to the KDIGO (Kidney Disease: Improving Global Outcomes)
classification27,29 and age in 2,150 vascular patients. (Data on file Klinikum Darmstadt GmbH).

Hämostaseologie Vol. 40 No. 1/2020

 Anticoagulation in Elderly and Frail Bauersachs, Herold 77

renal elimination (80%), less for edoxaban (50%), rivaroxaban warfarin among patients with mild renal impairment (risk
(33%), and apixaban (22%), respectively.13 Therefore, it is crucial ratio: 0.79; 0.68–0.91) and moderate renal impairment (0.80;
to evaluate renal function at the initiation of DOAC antico- 0.69–0.92). Interestingly, no major differences were found in
agulation. As mentioned above, equations to estimate the GFR patients with normal renal function. Additionally, DOACs were
should be utilized, and the CKD-EPI equation estimating the associated with fewer major bleedings among patients
GFR is recommended by the National Kidney Foundation, with normal renal function (0.77; 0.70–0.84), mild (0.86;
because it has been shown to be reliable across the range of 0.77–0.95), and moderate renal impairment (0.73;
CKD stages. In patients on DOACs, renal function needs to be 0.65–0.82), respectively. The authors conclude that DOACs
monitored diligently, at least yearly to detect changes in renal provide a greater clinical benefit than warfarin in patients
function and adapt the dose accordingly. If renal function is with impaired renal function. DOACs are associated with a
impaired (i.e., CrCl < 60 mL/min), a more frequent evaluation is comparatively lower risk of stroke and major bleeding, as well
recommended (e.g., by dividing CrCl by 10 to obtain the renal lower eGFR deterioration over time. This suggests that these
function testing intervals in months).34 In patients with addi- agents are a better choice in renal disease.
tional risk factors (e.g., older age, frailty, multiple comorbid- A Danish registry study found that patients who had their
DOAC dose reduced “only” due to age and frailty without a

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
ities), renal function may be evaluated even more frequently,
especially if on dabigatran. Intercurrent acute illnesses like renal indication for dose reduction had a fivefold increased
infections, acute heart failure, or contrast media exposure may stroke rate with apixaban without significantly reducing
rapidly affect the renal function and should also trigger prompt bleeding risk compared with nonfrail patients (or patient
reevaluation; importantly, patients need to be alerted that they under full therapeutic DOAC dose)51; thus it is important to
should seek contact with their health care provider in such adhere to the Summary of Product Characteristics (SmPC)
situations.34 dose recommendations (see ►Table 1).
There are no randomized controlled trial (RCT) data on the For VTE treatment, the SmPC recommended dose in renal
use of DOACs for stroke prevention in AF patients with severe failure differs from that in AF treatment (►Table 2); while in AF
CKD or on renal replacement therapy, as all landmark DOAC the doses of apixaban and rivaroxaban are reduced for a CrCl
trials essentially excluded patients with a CrCl of <30 mL/min. below 30 and 50 mL/min, respectively, there is no dose reduc-
However, VKAs have also never been prospectively assessed in tion in VTE treatment. This dosing difference between AF and
RCTs in this patient population.34 VTE can be explained with the need for acute anticoagulant
A meta-analysis of RCTs compared VKA with DOACs50 in AF treatment in established acute VTE, versus prevention of blood
patients with normal, mild, or moderate renal function (except formation in AF. Of course, when full-dose anticoagulation is
severe renal impairment: CrCl < 30 mL/min). Five clinical used in renal insufficiency, this treatment should be per-
trials were assessed, involving 72,608 patients. Pooled analysis formed with great caution. For rivaroxaban, a reduction of
indicated that the risk of stroke was lower for DOACs than for the maintenance dose from 20 to 15 mg od (once daily) can be

Table 1 DOAC dose recommendations for patients with atrial fibrillation and renal insufficiency, based on German Summary of
Product Characteristics (SmPC)

Atrial fibrillation
Trade name Drug Dialysis Creatinine clearance (mL/min) > 50
15–29 30–50
Pradaxa Dabigatran Not recommended Contraindicated 150 mg bid (age
80 y or
verapamil: 110 mg bid)
CrCl 30–50 mL/min or
75–80 y, gastritis, elevated
bleeding risk: 110 mg bid or
150 mg bid depending on
thromboembolic and
bleeding risk
Xarelto Rivaroxaban 15 mg od 15 mg od 20 mg od
Eliquis Apixaban 2.5 mg bid 5 mg bid
2.5 mg bid if at least two of:
Age
80 y
60 kg
S Crea
1.5 mg/dL
(133 µmol/L)
Lixiana Edoxaban 30 mg od 60 mg oda

Abbreviations: bid, twice a day; DOAC, direct oral anticoagulant; od, once daily.
a
In patients with a body weight < 60 kg or concomitant use of the P-gp-Inhibitors Ciclosporine, Dronedarone, Erythromycine or Ketoconazole: 30 mg
edoxaban od.

Hämostaseologie Vol. 40 No. 1/2020

78 Anticoagulation in Elderly and Frail Bauersachs, Herold

Table 2 DOAC dose recommendations for patients with venous thromboembolism and renal insufficiency, based on German
Summary of Product Characteristics (SmPC)

Treatment of venous thromboembolism

Trade name Drug Dialysis Creatinine clearance (- > 50
mL/min)
15–29 30–50
Pradaxa Dabigatran Not recommended Contraindicated 150 mg bid (age
80 y or verapamil: 110 mg bid)
CrCl 30–50 mL/min or 75–80 y, gastritis, elevated
bleeding risk: 110 mg bid or 150 mg bid
depending on thromboembolic and bleeding risk
Xarelto Rivaroxaban 15 mg bid for 3 weeks, 15 mg bid for 3 weeks, then 20 mg od
then 20 mg od or 15 mg
od, if the estimated risk of

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
bleeding is higher than the
risk for thromboembolic
recurrence (with caution
for 15–29 mL CrCl)
Eliquis Apixaban 10 mg bid for 1 week, then 5 mg bid (with caution for 15–29 mL CrCl)
Lixiana Edoxaban 30 mg od 60 mg oda

Abbreviations: bid, twice a day; DOAC, direct oral anticoagulant; od, once daily.
a
In patients with a body weight < 60 kg or concomitant use of the P-gp-Inhibitors Ciclosporine, Dronedarone, Erythromycine or Ketoconazole: 30 mg
edoxaban od.

Table 3 Risk score to assess the probability of falls per year53

considered in patients with moderate renal impairment (CrCl:
30–49 mL/min), if the patient’s assessed bleeding risk out-
Previous falls Yes/No
weighs the risk of recurrent VTE12 (►Table 2). This is sup-
ported by observations from the RIETE registry39: depending > 4 medications Yes/No
on the degree of renal insufficiency (CrCl > 60, 30–60, and Psychotropics Yes/No
<30 mL/min), the incidence of fatal pulmonary embolism Low visual acuity Yes/No
within 15 days after diagnosis was 1.0, 2.6, and 6.6%, respec-
Diminished sensation Yes/No
tively, which are significantly higher than the observed rates of
Near tandem stand 10 s Yes/No
fatal bleeding (0.2, 0.3, and 1.2%, respectively).
Alternate step test 10 s Yes/No
Falls Sit to stand < 12 s Yes/No
Several scales have been proposed for the definition of frailty Score 0–1 2–3 4–5 6þ
and functional deficits in elderly persons,34 including factors
Probability of fall per year 7% 13% 27% 49%
that are associated with a high risk of falls.25,52 These include
prior history of falls, lower extremity weakness, poor balance,
cognitive impairment, orthostatic hypotension, use of psycho- the Markov model, while other complications, like intracere-
tropic drugs, severe arthritis, and dizziness. A more formal tool bral hemorrhage (ICH) and—for example—non-CNS hemor-
is summarized in ►Table 3, which had been developed and rhages were 30-fold more common. In patients above
validated in 1,126 older community-living persons.53 In the 90 years, mortality of falls is greatly increased (5.5 vs. 0.9% at
general population above 65 years, there is a 1 to 2% risk of falls <60 years), and the likelihood of home discharge is dramati-
per year, and 5% of falls will result in fracture and hospitaliza- cally decreased (18.2 vs. 73.7%, respectively).56 In this retro-
tion.54 The fear of falls and subsequent subdural hemorrhage spective study of 5,088 patients with ground level falls, both
are often reasons for not to initiate OAC or to discontinue OAC. A aspirin use (1.42 [1.0–2.03]; p ¼ 0.049) and anticoagulation
Markov decision analytic model has claimed that VKA patients with warfarin (OR: 2.24 [1.56–3.21]; p < 0.001) were associat-
would need to fall 295 times to outweigh the benefits of OACs ed with increased overall mortality.56 Urgently, new data are
by the risk of subdural hematoma.55 While this mathematical required to either confirm or reject the 20-year old calculations
analysis has been cited numerously to support the use of OAC in mentioned above55; until then, this reference should be cited
elderly with risks of fall, the flaws of this analysis have not been with caution before drawing firm conclusions for patient
addressed sufficiently: many of the important underlying management in 2020 and beyond.
assumptions for the Markov model were collected more than Practically, patients who fall on OAC should be referred to
40 years ago. Most importantly, the number of 295 falls re- a special service for multifactorial and multidisciplinary
quired to outweigh the benefits of OAC refers to subdural assessment of risk factors and to address remediable pathol-
hematoma, the rarest and not most severe complication in ogy and/or prescribe interventions (e.g., exercise programs;

Hämostaseologie Vol. 40 No. 1/2020

 Anticoagulation in Elderly and Frail Bauersachs, Herold 79

home environmental assessment a.s.o.) that reduce risk of take care of providing medication may lead to a higher
further falls.57 adherence.
Given the lower risk of subdural bleeding and ICH with
DOACs compared with VKA, the “number needed to fall” would Polypharmacy
be beneficial with the use of DOACs. The effect of DOACs versus With increasing age and with several underlying comorbid-
VKA in patients at risk of falling was analyzed specifically in ities, polypharmacy represents an additional challenge for
two trials: prospectively in ENGAGE-AF TIMI 4858 with edox- OAC, particularly in AF. A posthoc analysis of the Aristotle
aban, and retrospectively in the ARISTOTLE study with apix- study analyzed the association between polypharmacy
aban.59 The relative DOAC-treatment benefit in patients with (
5 drugs), comorbidities, and the occurrence of complica-
an increased risk of falling was consistent with those without tions in patients receiving either apixaban or warfarin.61
an increased risk. Because of the larger absolute risk of events Patients were divided into three groups, with either receiv-
in patients at risk for falls, DOACs are associated with a larger ing 0 to 5, 6 to 8, or more than 9 concomitant drugs,
absolute risk reduction compared with VKA. respectively. Patients received a median of six drugs, and
polypharmacy was seen in 76.5% of the 18,201 trial partic-

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Dementia and Anticoagulation ipants. More drugs were being used in elderly patients,
Dementia is common in older age groups, and AF itself is a women, and in the United States. The number of comorbid-
risk factor for dementia, and there is evidence that use of OAC ities increased with the number of drugs, including those
may reduce the risk of dementia in AF by approximately one- that can interact with warfarin or apixaban. There was a
third.60 significant increase of mortality as well as stroke/TE rate
Dementia poses unique challenges for decision making, with the number of concomitant drugs61 (►Fig. 2). However,
choice of treatment, and managing drug adherence. Impor- the RR reduction of stroke/TE with the use of a DOAC
tantly, dementia should not be a viewed as a general remained consistent regardless of the number of concomi-
contraindication to OAC, especially if well managed (see tant drugs. This analysis shows the magnitude of the problem
below). Dementia patients should have a careful assessment of OAC for stroke prevention: three quarters of the patients
of their ability to understand and make a treatment deci- have five or more drugs, associated with increased comor-
sion regarding OAC. Where capacity is deficient, it is ratio- bidities, more drug–drug interactions, higher mortality, and
nal for the physician to recommend treatment on the basis higher rates of thromboembolic or bleeding complications.
of the “best medical interest” principle, ideally including Yet, use of DOACs was more effective and at least as safe as
family members. warfarin in these patients.61
Adherence to OAC intake is very important in dementia. For patients above 65 years, the appropriateness of antico-
Once daily medications, weekly tablet boxes, reminders, and agulant drugs was reviewed based on the Fit-fOR-The-Aged
packing may be helpful. Paradoxically, the fact that others (FORTA) classification.62–65 In a structured comprehensive

Fig. 2 Occurrence of stroke/TE, major bleeding and mortality, and number of concomitant drugs. Analysis of the ARISTOTLE-Study, which
compared warfarin with apixaban in patients with atrial fibrillation. 61 TE, thromboembolism.

Hämostaseologie Vol. 40 No. 1/2020

80 Anticoagulation in Elderly and Frail Bauersachs, Herold

review of RCTs and summaries of individual SmPC, the result- reduce time on triple therapy,67 and administration of proton-
ing evidence was discussed. Decisions on age appropriateness pump inhibitors should be considered.34
were made using a Delphi process. Even though over 24,000
patients above 75 years were studied for warfarin, only two Practical Considerations
studies reported on frailty, falls, and dementia. Apixaban was The above-mentioned specific aspects in elderly and frail
classified as highly beneficial, and dabigatran, high-dose edox- patients illustrate that the decision on the type and dose of
aban, and warfarin were classified beneficial. Phenprocoumon, OAC is complex and is influenced by many clinical factors,
acenocoumarol, and fluindione were questionable, mainly which have to be considered both at the initiation of OAC and
because of lack of data. In conclusion, DOACs and warfarin during regular follow-up visits. An example for a checklist to
were classified as beneficial or very beneficial in older patients. be thoroughly and regularly reviewed in elderly and frail
Pharmacodynamic interactions are as important and should patients is provided in ►Table 4.
always be assessed. Coadministration of platelet inhibitors In patients with AF, age also represents a risk factor for
(including off-the-counter nonsteroidal anti-inflammatory major bleeding, as indicated in the HAS-BLED Score.45
drugs [NSAIDs] not listed in medication plan) increases Importantly however, the HAS-BLED Score should not serve

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
the risk of bleeding, particularly in patients with underlying as a reason not to anticoagulate, as these patients commonly
gastrointestinal lesions. In combination with NSAIDs or alone, also have a very high risk for stroke; rather, attention should
selective serotonin reuptake inhibitors, in particular escitalo- be paid to try to correct and minimize modifiable bleeding
pram, are associated with increased risk of major bleeding.66 risk factors in these patients while on oral anticoagulation.7
Thus, these combinations should be carefully balanced against Implantation of a left atrial appendage (LAA) occluder or
their potential benefit. The indication for coadministration of surgical occlusion may be an option instead of long-term
DOACs with dual-antiplatelet drugs requires active measures to anticoagulation.7 Unfortunately, at present there is no

Table 4 Checklist of clinical parameters to be assessed at initiation of OAC and during regular follow-up visits

Assess thrombotic risk (CHA2DS2-VASc Score or risk for VTE recurrence)

Identify individual bleeding risk. A high bleeding risk per se should not result in the decision not to anticoagulate, rather to
correct and minimize modifiable risks
Check renal function, global coagulation tests, liver function parameters, and CBC: in the case of anemia try to identify
underlying pathologies
Check for contraindications for DOACs (e.g., mechanical heart valves, hemodialysis, DAPT)
Choose best suitable anticoagulant and appropriate dose; od vs. bid?
Check whether the patient qualifies for specific dose reduction according to SmPC?
Check whether DDI are expected, including over-the-counter drugs, and review number and significance of DDI (see ►Table 3 in
Steffel et al34)
Try to reduce the number of comedications (e.g., FORTA List [62–65]), particularly those with platelet inhibition properties
Use alternative DOAC with less potential for DDI, if possible
Evaluate need for protein pump inhibitors
Obtain informed consent and tailor patient education to the elderly patient and repeat as needed; consider use of additional
teaching and adherence tools like videos, graphical material a.s.o; involve partner and family
Hand out anticoagulation card
Evaluate indication for left atrial appendage occluder in case of contraindication to OAC
Organize follow-up:
Determine follow-up intervals for re-checking of renal function (see above), and schedule follow-up with particular focus on side
effects, renal function, hemoglobin, liver function tests, and repeated education; early initial follow-up may be required
Check appropriate intake and dose at follow-up visit
Check for potential adherence problems and evaluate whether the patient would benefit from a switch from VKA to DOACs or
vice versa
Explore options for dose reduction or discontinuation in VTE secondary prevention
Consider in selected special cases drug level measurement at trough (for accumulation, safety) or peak (for compliance,
efficacy): diluted thrombin time (for dabigatran) or anti-Xa activity measurements (for Xa-inhibitors)
Assess the risk for falls (►Table 3) and intervene to reduce risk of further falls; frailty and risk of falling should not generally be a
reason not to be anticoagulated

Abbreviations: CBC, complete blood count; DAPT, dual antiplatelet therapy; DDI, drug–drug interactions; DOAC, direct oral anticoagulant; OAC, oral
anticoagulant; SmPC, German Summary of Product Characteristics; VKA, vitamin K antagonist; VTE, venous thromboembolism.

Hämostaseologie Vol. 40 No. 1/2020

 Anticoagulation in Elderly and Frail Bauersachs, Herold 81

evidence from RCTs for LAA occlusion after OAC bleeding tion developed in collaboration with EACTS. Eur Heart J 2016;37
under OAC, e.g., compared with continued DOAC treatment (38):2893–2962
in this challenging clinical situation. Additional important 8 Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al; European Working
Group on Sarcopenia in Older People. Sarcopenia: European
and valuable practical aspects for the treatment of AF with
consensus on definition and diagnosis: report of the European
DOACs are summarized in the European Heart Rhythm Working Group on Sarcopenia in Older People. Age Ageing 2010;
Association Practical Guide.34 39(04):412–423
9 Lacas A, Rockwood K. Frailty in primary care: a review of its
conceptualization and implications for practice. BMC Med 2012;
Conclusions 10:4
10 Gobbens RJ, Luijkx KG, Wijnen-Sponselee MT, Schols JM. Towards
Elderly and frail patients requiring anticoagulation for AF or
an integral conceptual model of frailty. J Nutr Health Aging 2010;
VTE are at a higher risk of adverse outcomes, but—at the same 14(03):175–181
time—have a higher absolute benefit from OAC. Altered 11 Fried LP, Tangen CM, Walston J, et al; Cardiovascular Health
responses to drugs due to age-related physical decline, renal Study Collaborative Research Group. Frailty in older adults:
impairment, or low body weight have to be taken into evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001;
56(03):M146–M156

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
account and may lead to suboptimal anticoagulation. Even
12 Bauersachs RM. Managing venous thromboembolism with novel
though DOACs are superior to VKA because they exhibit oral anticoagulants in the elderly and other high-risk patient
predictable pharmacokinetics, eliminating the need for rou- groups. Eur J Intern Med 2014;25(07):600–606
tine coagulation monitoring and dose adjustment, with 13 Bauersachs RM. Use of anticoagulants in elderly patients. Thromb
fewer drug–drug interactions, extra attention and regular Res 2012;129(02):107–115
reviews are required to ensure safe and effective anticoagu- 14 Prins MH, Lensing AW, Bauersachs R, et al; EINSTEIN Investiga-
tors. Oral rivaroxaban versus standard therapy for the treatment
lation in elderly and frail patients.
of symptomatic venous thromboembolism: a pooled analysis of
the EINSTEIN-DVT and PE randomized studies. Thromb J 2013;11
Conflicts of Interest (01):21
J.H. has received honoraria for advisory boards, and/or 15 Oqab Z, Pournazari P, Sheldon RS. What is the impact of frailty on
travel support from LEO Pharma, Bayer, Bristol-Myers prescription of anticoagulation in elderly patients with atrial
fibrillation? A systematic review and meta-analysis. J Atr Fibril-
Squibb, and Pfizer.
lation 2018;10(06):1870
R.M.B. has acted as a principal investigator in anticoagu-
16 López-Jiménez L, Montero M, González-Fajardo JA, et al; RIETE
lation studies by Bayer, Bristol-Myers Squibb, Daiichi Investigators. Venous thromboembolism in very elderly patients:
Sankyo, LEO, and Pfizer, and received honoraria for advi- findings from a prospective registry (RIETE). Haematologica
sory boards or lectures from Bayer, Bristol-Myers Squibb, 2006;91(08):1046–1051
Daiichi Sankyo, LEO, and Pfizer. 17 Trujillo-Santos J, Schellong S, Falga C, et al; RIETE Investigators.
Low-molecular-weight or unfractionated heparin in venous
thromboembolism: the influence of renal function. Am J Med
2013;126(05):425.e1–434.e1
References 18 Mangoni AA, Jackson SH. Age-related changes in pharmacokinet-
1 Bevölkerung in absoluten Zahlen, Anteile der Altersgruppen in ics and pharmacodynamics: basic principles and practical appli-
Prozent, 1960 bis 2050 Statistisches Bundesamt: Lange Reihen: cations. Br J Clin Pharmacol 2004;57(01):6–14
Bevölkerung nach Altersgruppen, 13. koordinierte Bevölkerungs- 19 Garcia D, Regan S, Crowther M, Hughes RA, Hylek EM. Warfarin
vorausberechnung: Bevölkerung Deutschlands bis 2060: Bundes- maintenance dosing patterns in clinical practice: implications for
zentrale für Politische Bildung; [Available from: Available at: safer anticoagulation in the elderly population. Chest 2005;127
https://www.bpb.de/nachschlagen/zahlen-und-fakten/soziale-sit- (06):2049–2056
uation-in-deutschland/61541/altersstruktur 20 Samama MM. Use of low-molecular-weight heparins and new
2 Ergebnisse der 14. koordinierten Bevölkerungsvorausberechnung: anticoagulants in elderly patients with renal impairment. Drugs
Statistisches Bundesamt. Available at: https://www.destatis.de/DE/ Aging 2011;28(03):177–193
Themen/Gesellschaft-Umwelt/Bevoelkerung/Bevoelkerungsvoraus 21 Silverstein RL, Bauer KA, Cushman M, Esmon CT, Ershler WB,
berechnung/Tabellen/variante-1-2-3-altersgruppen.html;jsessio- Tracy RP. Venous thrombosis in the elderly: more questions than
nid¼452C6718CB5A674FB003FFF71486C2B9.internet712 answers. Blood 2007;110(09):3097–3101
3 Rosendaal FR, VAN Hylckama Vlieg A, Doggen CJ. Venous throm- 22 Fohtung RB, Novak E, Rich MW. Effect of new oral anticoagulants
bosis in the elderly. J Thromb Haemost 2007;5(Suppl 1):310–317 on prescribing practices for atrial fibrillation in older adults. J Am
4 Naess IA, Christiansen SC, Romundstad P, Cannegieter SC, Rose- Geriatr Soc 2017;65(11):2405–2412
ndaal FR, Hammerstrøm J. Incidence and mortality of venous 23 Bai Y, Guo SD, Deng H, et al. Effectiveness and safety of oral
thrombosis: a population-based study. J Thromb Haemost 2007;5 anticoagulants in older patients with atrial fibrillation: a system-
(04):692–699 atic review and meta-regression analysis. Age Ageing 2018;47
5 van Es N, Coppens M, Schulman S, Middeldorp S, Büller HR. Direct (01):9–17
oral anticoagulants compared with vitamin K antagonists for 24 Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the
acute venous thromboembolism: evidence from phase 3 trials. efficacy and safety of new oral anticoagulants with warfarin in
Blood 2014;124(12):1968–1975 patients with atrial fibrillation: a meta-analysis of randomised
6 Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial trials. Lancet 2014;383(9921):955–962
fibrillation in adults: national implications for rhythm management 25 Kato ET, Giugliano RP, Ruff CT, et al. Efficacy and safety of
and stroke prevention: the AnTicoagulation and Risk Factors in edoxaban in elderly patients with atrial fibrillation in the ENGAGE
Atrial Fibrillation (ATRIA) Study. JAMA 2001;285(18):2370–2375 AF-TIMI 48 trial. J Am Heart Assoc 2016;5(05):e003432
7 Kirchhof P, Benussi S, Kotecha D, et al; ESC Scientific Document 26 Patti G, Pecen L, Lucerna M, et al. Net clinical benefit of non-
Group. 2016 ESC Guidelines for the management of atrial fibrilla- vitamin K antagonist vs vitamin K antagonist anticoagulants in

Hämostaseologie Vol. 40 No. 1/2020

82 Anticoagulation in Elderly and Frail Bauersachs, Herold

elderly patients with atrial fibrillation. Am J Med 2019;132(06): Force for the Management of Atrial Fibrillation of the European
749.e5–757.e5 Society of Cardiology (ESC). Eur Heart J 2010;31(19):2369–2429
27 Bauersachs R, Berkowitz SD, Brenner B, et al; EINSTEIN Investi- 46 Gage BF, Yan Y, Milligan PE, et al. Clinical classification schemes for
gators. Oral rivaroxaban for symptomatic venous thromboembo- predicting hemorrhage: results from the National Registry of
lism. N Engl J Med 2010;363(26):2499–2510 Atrial Fibrillation (NRAF). Am Heart J 2006;151(03):713–719
28 Bauersachs RM, Lensing AW, Prins MH, et al. Rivaroxaban versus 47 Dreisbach AW, Lertora JJ. The effect of chronic renal failure on
enoxaparin/vitamin K antagonist therapy in patients with venous drug metabolism and transport. Expert Opin Drug Metab Toxicol
thromboembolism and renal impairment. Thromb J 2014;12:25 2008;4(08):1065–1074
29 Sharma M, Cornelius VR, Patel JP, Davies JG, Molokhia M. Efficacy 48 Limdi NA, Beasley TM, Baird MF, et al. Kidney function influences
and harms of direct oral anticoagulants in the elderly for stroke warfarin responsiveness and hemorrhagic complications. J Am
prevention in atrial fibrillation and secondary prevention of Soc Nephrol 2009;20(04):912–921
venous thromboembolism: systematic review and meta-analysis. 49 Kleinow ME, Garwood CL, Clemente JL, Whittaker P. Effect of
Circulation 2015;132(03):194–204 chronic kidney disease on warfarin management in a pharmacist-
30 Büller HR, Décousus H, Grosso MA, et al; Hokusai-VTE Investigators. managed anticoagulation clinic. J Manag Care Pharm 2011;17
Edoxaban versus warfarin for the treatment of symptomatic venous (07):523–530
thromboembolism. N Engl J Med 2013;369(15):1406–1415 50 Zou R, Tao J, Shi W, et al. Meta-analysis of safety and efficacy for
31 Group IGOCW. KDIGO 2012 clinical practice guideline for the direct oral anticoagulation treatment of non-valvular atrial fibril-

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
evaluation and management of chronic kidney disease. Kidney Int lation in relation to renal function. Thromb Res 2017;160:41–50
Suppl 2013;3:1–150 51 Yao X, Shah ND, Sangaralingham LR, Gersh BJ, Noseworthy PA.
32 Erratum Regarding. Erratum regarding “KDOQI US commentary Non-vitamin K antagonist oral anticoagulant dosing in patients
on the 2012 KDIGO clinical practice guideline for glomerulone- with atrial fibrillation and renal dysfunction. J Am Coll Cardiol
phritis” (Am J Kidney Dis. 2013;62[3]:403-441). Am J Kidney Dis 2017;69(23):2779–2790
2017;69(03):485 52 Donzé J, Clair C, Hug B, et al. Risk of falls and major bleeds in
33 Sarnak MJ, Bloom R, Muntner P, et al. KDOQI US commentary on patients on oral anticoagulation therapy. Am J Med 2012;125(08):
the 2013 KDIGO clinical practice guideline for lipid management 773–778
in CKD. Am J Kidney Dis 2015;65(03):354–366 53 Tiedemann A, Lord SR, Sherrington C. The development and
34 Steffel J, Verhamme P, Potpara TS, et al; ESC Scientific Document validation of a brief performance-based fall risk assessment
Group. The 2018 European Heart Rhythm Association Practical tool for use in primary care. J Gerontol A Biol Sci Med Sci 2010;
Guide on the use of non-vitamin K antagonist oral anticoagulants in 65(08):896–903
patients with atrial fibrillation. Eur Heart J 2018;39(16):1330–1393 54 Rubenstein LZ. Falls in older people: epidemiology, risk factors and
35 Soliman EZ, Prineas RJ, Go AS, et al; Chronic Renal Insufficiency strategies for prevention. Age Ageing 2006;35(Suppl 2):ii37–ii41
Cohort (CRIC) Study Group. Chronic kidney disease and prevalent 55 Man-Son-Hing M, Nichol G, Lau A, Laupacis A. Choosing antith-
atrial fibrillation: the Chronic Renal Insufficiency Cohort (CRIC). rombotic therapy for elderly patients with atrial fibrillation who
Am Heart J 2010;159(06):1102–1107 are at risk for falls. Arch Intern Med 1999;159(07):677–685
36 Lee M, Saver JL, Chang KH, Liao HW, Chang SC, Ovbiagele B. Low 56 Bhattacharya B, Maung A, Schuster K, Davis KA. The older they are
glomerular filtration rate and risk of stroke: meta-analysis. BMJ the harder they fall: injury patterns and outcomes by age after
2010;341:c4249 ground level falls. Injury 2016;47(09):1955–1959
37 Marinigh R, Lip GY, Fiotti N, Giansante C, Lane DA. Age as a risk 57 Tricco AC, Thomas SM, Veroniki AA, et al. Comparisons of inter-
factor for stroke in atrial fibrillation patients: implications for ventions for preventing falls in older adults: a systematic review
thromboprophylaxis. J Am Coll Cardiol 2010;56(11):827–837 and meta-analysis. JAMA 2017;318(17):1687–1699
38 Marinigh R, Lane DA, Lip GY. Severe renal impairment and stroke 58 Steffel J, Giugliano RP, Braunwald E, et al. Edoxaban versus
prevention in atrial fibrillation: implications for thromboprophy- warfarin in atrial fibrillation patients at risk of falling: ENGAGE
laxis and bleeding risk. J Am Coll Cardiol 2011;57(12):1339–1348 AF-TIMI 48 analysis. J Am Coll Cardiol 2016;68(11):1169–1178
39 Monreal M, Falgá C, Valle R, et al; RIETE Investigators. Venous 59 Rao MP, Vinereanu D, Wojdyla DM, et al; Apixaban for Reduction
thromboembolism in patients with renal insufficiency: findings in Stroke Other Thromboembolic Events in Atrial Fibrillation
from the RIETE Registry. Am J Med 2006;119(12):1073–1079 (ARISTOTLE) Investigators. Clinical outcomes and history of fall
40 Olesen JB, Lip GY, Kamper AL, et al. Stroke and bleeding in atrial in patients with atrial fibrillation treated with oral anticoagula-
fibrillation with chronic kidney disease. N Engl J Med 2012;367 tion: insights from the ARISTOTLE trial. Am J Med 2018;131(03):
(07):625–635 269–275.e2
41 Lip GY, Frison L, Halperin JL, Lane DA. Comparative validation of a 60 Friberg L, Andersson T, Rosenqvist M. Less dementia and stroke in
novel risk score for predicting bleeding risk in anticoagulated low-risk patients with atrial fibrillation taking oral anticoagula-
patients with atrial fibrillation: the HAS-BLED (hypertension, tion. Eur Heart J 2019;40(28):2327–2335
abnormal renal/liver function, stroke, bleeding history or predis- 61 Jaspers Focks J, Brouwer MA, Wojdyla DM, et al. Polypharmacy
position, labile INR, elderly, drugs/alcohol concomitantly) score. and effects of apixaban versus warfarin in patients with atrial
J Am Coll Cardiol 2011;57(02):173–180 fibrillation: post hoc analysis of the ARISTOTLE trial. BMJ 2016;
42 Gangireddy C, Rectenwald JR, Upchurch GR, et al. Risk factors and 353:i2868
clinical impact of postoperative symptomatic venous thrombo- 62 Wehling M, Collins R, Gil VM, et al. Appropriateness of oral
embolism. J Vasc Surg 2007;45(02):335–341, discussion 341–342 anticoagulants for the long-term treatment of atrial fibrillation
43 Engbers MJ, van Hylckama Vlieg A, Rosendaal FR. Venous throm- in older people: results of an evidence-based review and interna-
bosis in the elderly: incidence, risk factors and risk groups. tional consensus validation process (OAC-FORTA 2016). Drugs
J Thromb Haemost 2010;8(10):2105–2112 Aging 2017;34(07):499–507
44 Cook LM, Kahn SR, Goodwin J, Kovacs MJ. Frequency of renal 63 Pazan F, Weiss C, Wehling M; FORTA. The FORTA (Fit fOR The Aged)
impairment, advanced age, obesity and cancer in venous throm- list 2015: update of a validated clinical tool for improved pharma-
boembolism patients in clinical practice. J Thromb Haemost 2007; cotherapy in the elderly. Drugs Aging 2016;33(06):447–449
5(05):937–941 64 Pazan F, Weiss C, Wehling M; FORTA. Correction to: the EURO-
45 Camm AJ, Kirchhof P, Lip GY, et al; European Heart Rhythm FORTA (Fit fOR The Aged) list: international consensus validation
Association; European Association for Cardio-Thoracic Surgery. of a clinical tool for improved drug treatment in older people.
Guidelines for the management of atrial fibrillation: the Task Drugs Aging 2018;35(07):677

Hämostaseologie Vol. 40 No. 1/2020

 Anticoagulation in Elderly and Frail Bauersachs, Herold 83

65 Pazan F, Weiss C, Wehling M; FORTA. The EURO-FORTA (Fit fOR 67 Roffi M, Patrono C, Collet JP, et al; ESC Scientific Document Group.
The Aged) list: international consensus validation of a clinical tool 2015 ESC guidelines for the management of acute coronary syn-
for improved drug treatment in older people. Drugs Aging 2018; dromes in patients presenting without persistent ST-segment
35(01):61–71 elevation: task force for the management of acute coronary syn-
66 Samuel NG, Seifert CF. Risk of bleeding in patients on full-dose dromes in patients presenting without persistent ST-segment
enoxaparin with venous thromboembolism and selective serotonin elevation of the European Society of Cardiology (ESC). Eur Heart J
reuptake inhibitors. Ann Pharmacother 2017;51(03):226–231 2016;37(03):267–315

This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.

Hämostaseologie Vol. 40 No. 1/2020