Impact of minor bleeds on confidence in anticoagulation therapy, adherence to treatment and quality of life in patients using a non-vitamin K antagonist oral anticoagulant for atrial fibrillation

OBJECTIVE: In a previous study on the reasons for discontinuation of novel oral anticoagulation therapy (NOAC) in patients with atrial fibrillation, we showed that minor bleeding was the second most important reason for NOAC discontinuation. This finding suggest that the impact of minor bleeds on the patient's perspective on NOACs cannot be ignored, especially those minor bleeds for which the patient searched medical care. We therefore performed a study in which we explored the impact of minor bleed (clinically relevant non-major bleeds) on patient confidence in therapy, adherence to treatment and quality of life in AF patients using NOAC's.

METHODS: Details on NOAC therapy, level of confidence, adherence and quality of life were assessed using a semi-structured telephone interview. Questions related to annoyance, concern and trust were scored on a scale of 1-10, with score 10 representing the highest level of impact or trust. For the adherence to treatment before and after a minor bleed the Medication Adherence Rating Scale (MARS-5) questionnaire was used. The total MARS-5 score ranges from 5 to 25; score 5 is indicative of poor adherence and 25 suggest perfect adherence. Furthermore, the impact (restrictions) of a bleed on physical activities and daily life was assessed using a VAS score ranging from 1 to 10; score 10 representing the highest level of restriction.

RESULTS: A total of 142 patients were included. During NOAC treatment, minor bleeds were reported by 87 patients (61%) of whom 16 (11%) suffered from a CRNMB. All patients that suffered from a CRNMB contacted their treating physician and 13 received treatment. The most frequently reported CRNMBs were epistaxis (n = 6), a bleed after injury (n = 3), rectal bleed (n = 2) and an eye bleed (n = 2). With regard to the impact of a CRNMB, the median level of annoyance was 8 (min-max 2-10)) and the level of concern 4 (min-max 1-10). The MARS-5 score for adherence to treatment was only marginally influenced. Also the level of trust remained high after the CRNMB. Out of 12 patients, only 2 patients reported a reduction of 1 and 8 points, respectively. Ten patients scored no change and one patient scored a 1 point increase in the level of trust. The mean reduction in trust was 0.7 (95%CI -0.8 to 2.2). With regard to the reported restrictions on physical activities, 8 out of 12 patients reported no (score 1) or only marginal impact (score 2), 2 patients reported a moderate impact (score 5 and 6) and 2 patients reported a high impact score 7 (mean score 2.9 (95%CI 1.3-4.5). For the daily life impact question, 5 out of 9 patients reported no or only marginal impact, 3 patients reported a moderate impact (scores 4 and 5) and 1 patient reported maximal impact (score 10) (mean score 3.2 (95%CI 0.9-5.5)).

CONCLUSION: In our study the level of trust, annoyance and concern were not significantly impacted by the CRNMB, nor the adherence to treatment and impact on daily life and physical activities. However, on an individual basis, there were patients that reported a high impact. We hope that future data on impact of this type of minor bleeds will help us identify and guide suboptimaly adherent NOAC patients in shared decision manner.