LAUSR

This editorial comments on the article by Mailhot et al. Oral anticoagulants (OACs) reduce the risk of stroke for people with atrial fibrillation (AF), and OACs are recommended in evidence-based guidelines for most people with AF. The net clinical benefit of OACs compared with no treatment or aspirin is clear for most people with AF, apart from those at lowest stroke risk. The common risk factors for incident AF are also risk factors for stroke and bleeding in AF. There are well-validated risk scores to assess stroke and bleeding risks in AF, and an independent PatientCentered Outcomes Research Institute (PCORI) systematic review and evidence appraisal found that of commonly used AF risk scores, the CHADS2, CHA2DS2 −VASc and HAS-BLED scores were the best validated scores for use in clinical practice. However, stroke risk in people with AF is increased by aging and incident risk factors, so should be regularly repeated, given that risk is dynamic and not a static “one-off” assessment. Similarly, bleeding risk is dynamic and should be determined for all people with AF before (and after) commencing OAC treatment. Indeed, regular reassessment using the HASBLED score is associated with mitigation of modifiable bleeding risk factors, reduced bleeding risk, and an increase in OAC use. Importantly, a high bleeding risk score should not be used as the sole reason to not initiate anticoagulation. Instead, people should be appropriately treated with OAC, monitored and also reassessed to determine any changes in risks over time. There is no exception in current guidelines for initiating anticoagulation for AF in people with frailty or cognitive impairment. The risk of stroke without treatment is often of greater consequence than bleeding risk by prescribing OACs for people with AF, including older people, people with cognitive impairment, and/or those with a history of falls or frailty. Thus, people with AF, frailty, and/or cognitive impairment should be appropriately assessed for stroke and bleeding risk and treated and monitored accordingly. Consideration should be given as to whether a caregiver is available to assist with anticoagulant adherence for people with cognitive impairment and dementia. However, there are issues in the current evidence base as people with cognitive impairment or dementia are often excluded from research studies and clinical trials. In this issue of the Journal of the American Geriatrics Society, Mailhot and colleagues determine the independent and concurrent prevalence of cognitive impairment and frailty in a cohort of over 1200 people aged 65 and older with nonvalvular AF in the United States. All participants had a CHA2DS2-VASc score ≥2 and had no contraindications to the use of OACs. The Fried Frailty Scale was used to assess frailty, and the Montreal Cognitive Assessment (MoCA) was used to assess cognitive impairment, with a defined cutpoint of 23 to categorize cognitive impairment with the MoCA. The Anti-Clot Treatment Scale (ACTS) was used to assess patient satisfaction of OACs. The authors reported that almost one half of the study participants had frailty, cognitive impairment, or both; approximately 5% had frailty only, 34% had cognitive impairment only, and 9% had both frailty and cognitive impairment. The majority of participants (85%) were receiving OACs, and frailty or cognitive impairment did not associate with OAC prescribing. This finding indicates that prescribers for the participants of this study were mostly following current guidelines and not withholding oral anticoagulation due to frailty status or cognitive impairment. For the remaining 15%, the reasons for why OACs are not prescribed or whether the participants have ever received OACs are not reported. The analysis of ACTS scores in the study showed people with cognitive impairment, but not frailty or both cognitive impairment and frailty, reported low perceived benefit of OACs compared with people with no impairment, but treatment burden did not significantly differ between the groups. This was partially explained by adjusting for other factors known to associate with treatment adherence such as social support and levels of education but remained statistically significant after adjustments (odds ratio for low benefit of OACs for people with cognitive impairment and frailty vs. people with no impairment: 1.87; 95% confidence intervals: 1.08, 3.27). Consideration DOI: 10.1111/jgs.17142 Journal of the American Geriatrics Society