Effectiveness and Safety of Direct Oral Anticoagulants versus Vitamin K Antagonists for People Aged 75 Years and over with Atrial Fibrillation: A Systematic Review and Meta-Analyses of Observational Studies

Abstract

Older people, are underrepresented in randomised controlled trials of direct oral anticoagulants (DOACs) for stroke prevention in atrial fibrillation (AF). The aim of this study was to combine data from observational studies to provide evidence for the treatment of people aged ≥75 years. Medline, Embase, Scopus and Web of Science were searched. The primary effectiveness outcome was ischaemic stroke. Safety outcomes were major bleeding, intracranial haemorrhage, gastrointestinal bleeding, myocardial infarction, and mortality. Twenty-two studies were eligible for inclusion. Two studies related specifically to people ≥75 years but were excluded from meta-analysis due to low quality; all data in the meta-analyses were from subgroups. The pooled risk estimate of ischaemic stroke was slightly lower for DOACs. There was no significant difference in major bleeding, mortality, or myocardial infarction. Risk of intracranial haemorrhage was 44% lower with DOACs, but risk of GI bleeding was 46% higher. Our results suggest that DOACs may be preferable for the majority of older patients with AF, provided they are not at significant risk of a GI bleed. However, these results are based entirely on data from subgroup analyses so should be interpreted cautiously. There is a need for adequately powered research in this patient group.

Keywords: aged; anticoagulants; atrial fibrillation; hemorrhage; stroke.

Figure 1

Study Selection.

Figure 2

Meta-analysis of observational studies on ischaemic stroke stratified by direct oral anticoagulant (DOAC), then grouped by age band and DOAC dose. New users = no previous use of VKA, and switchers = previous use of VKA prior to starting DOAC. Effect sizes reported are hazard ratios. Sex is male and female unless otherwise stated.

Figure 3

Meta-analysis of observational studies on composite effectiveness outcomes stratified by DOAC, then grouped by age band and DOAC dose. HS = haemorrhagic stroke, ICH = intracranial haemorrhage, IS = ischaemic stroke, RI = retinal infarct, SE = systemic embolism, TIA = transient ischaemic attack, and US = unspecified stroke. Effect sizes reported are hazard ratio. Sex is male and female unless otherwise stated.

Figure 4

Meta-analysis of observational studies on composite safety outcomes stratified by DOAC, then grouped by age band and DOAC dose. CRNMB = clinically-relevant non-major bleeding, GI = gastrointestinal bleeding, HS = haemorrhagic stroke, IC = intracranial haemorrhage, and MB = major bleeding. Effect sizes reported are hazard ratio. Sex is male and female unless otherwise stated.

Figure 5

Meta-analysis of observational studies on major bleeding stratified by DOAC, then grouped by age band and DOAC dose. Effect sizes reported are hazard ratios, except where ^ = sub-hazard ratio, and * = incident rate ratio. Sex is male and female unless otherwise stated.

Figure 6

Meta-analysis of observational studies on mortality stratified by DOAC, then grouped by age band and DOAC dose. Effect sizes reported are hazard ratios, except where ^ = sub-hazard ratio. Sex is male and female unless otherwise stated.

Figure 7

Meta-analysis of observational studies on myocardial infarction stratified by DOAC, then grouped by age band and DOAC dose. New users = no previous use of VKA, and switchers = previous use of VKA prior to starting DOAC. Effect sizes reported are hazard ratios. Sex is male and female unless otherwise stated.

Figure 8

Meta-analysis of observational studies on intracranial haemorrhage stratified by DOAC, then grouped by age band and DOAC dose. New users = no previous use of VKA, and switchers = previous use of VKA prior to starting DOAC. Effect sizes reported are hazard ratios. Sex is male and female unless otherwise stated.

Figure 9

Meta-analysis of observational studies on gastrointestinal bleeding stratified by DOAC, then grouped by age band and DOAC dose. New users = no previous use of VKA, and switchers = previous use of VKA prior to starting DOAC. Effect sizes reported are hazard ratios, except where * = incident rate ratio. Sex is male and female unless otherwise stated.