Abstract
Objective To compare patient-reported limitations, concerns, and burdens in those receiving and not receiving warfarin for thromboprophylaxis in atrial fibrillation (AF).
Methods We conducted a cross-sectional survey study of patients with AF receiving thromboprophylaxis for stroke prevention. Patients were administered the validated Anti-Clot Treatment Scale (ACTS). Mean scores of patients receiving and not receiving warfarin were compared for each ACTS item, and for the Burden and Benefit subscales.
Results From July 2010 to August 2011, 80 patients with AF were administered the survey, with 65 patients receiving a regimen containing warfarin and 15 patients not receiving a regimen containing warfarin. Six of the 17 individual questions depicting patient- perceived limitations in physical activity due to bleeding, limitations on diet, feelings of inconvenience of occasional aspects of thromboprophylaxis therapy, and frustration, and burden had less favorable scores in the warfarin-managed patients compared with the patients not receiving warfarin (P < 0.05 for all). Mean ACTS Burden scores were more favorable in the no-warfarin group (44.5 ± 6.4) compared with the warfarin group (39.8 ± 8.0; P = 0.003). No difference was seen between the 2 groups on the ACTS Benefits score (11.1 ± 3.4 vs 10.4 ± 3.7; P = 0.38).
Conclusion Patients with AF receiving warfarin may have less favorable feelings regarding thromboprophylaxis versus those receiving non-warfarin thromboprophylaxis. Patients report having more limitations and having greater feelings of burden on warfarin.
Multiple randomized controlled trials have demonstrated that the use of long-term anticlot therapy with warfarin reduces the relative risk of stroke in atrial fibrillation (AF) by approximately 65%.1,2Unfortunately, there are significant disadvantages to warfarin therapy, including complex dosing regimens, inconvenient regular blood monitoring, dangerous food and drug interactions, and as high as a 3% annual risk of major bleeding. Antiplatelet agents (aspirin alone or with clopidogrel) have been found to be somewhat efficacious in lowering stroke risk in patients with AF,1and newer oral anticoagulants including rivaroxaban, dabigatran, and apixaban bring the promise of at least similar efficacy to warfarin but with enhanced safety and ease of use.1Although it might be assumed that these disadvantages associated with warfarin would intuitively reduce patient satisfaction versus other choices, this has not been evaluated adequately to date. Given the long-standing experience with warfarin and its relatively low cost versus other antithrombotic choices, research in this area is needed. As such, we report the results of our comparison of patient-reported feelings of limitation, concern, and burden when receiving a warfarin- and non–warfarin-containing thromboprophylaxis regimen.
MATERIALS AND METHODS
We conducted a cross-sectional survey study of patients with AF from 3 arrhythmia clinics associated with a large urban teaching hospital receiving thromboprophylaxis for stroke prevention. The study was institutional review board approved, and all patients provided written informed consent. Eligible patients had to have a diagnosis of AF, be receiving thromboprophylaxis to prevent stroke (warfarin, dabigatran, and antiplatelet or combination therapy), and live in the community (not living in an assisted living or a skilled nursing facility). Eligible patients were asked to participate at check-in to a scheduled arrhythmia clinic follow-up visit. Patients self-administered the validated Anti-Clot Treatment Scale (ACTS).3,4The ACTS was chosen because it has been recently used in 2 large randomized controlled trials4,5and is included as an outcome in both the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation and the Global Anticoagulant Registry in the FIELD prospective, longitudinal AF patient registries.6,7The scale contains 17 questions about the impact of thromboprophylaxis treatment scored on a 5-point Likert scale (1, “not at all”, to 5, “extremely”). For the first 13 questions, higher scores represented less favorable perceptions of thromboprophylaxis treatment. For questions 14 to 17, higher scores represented more favorable perceptions. The ACTS questions were also separated into 2 domains, “Burden” and “Benefit”, each with their own scores. The Burden score was calculated as the sum of questions 1 to 12 subtracted from 60, with a higher score suggesting greater satisfaction with treatment. The Benefit subscale was calculated as the sum of questions 14 to 16, again with a higher score suggesting greater satisfaction. The 2 domains also showed good internal consistency with Cronbach alpha (a measure of the reliability of a psychometric test) ranging from 0.90 to 0.93 (scores ≥0.70 are typically considered desirable). Questions 13 and 17 were intended to stand alone and assess patients’ global perception of negative and positive aspects of anticoagulation.
Mean scores with accompanying standard deviations and medians with minimum and maximum values of patients receiving and not receiving warfarin were compared for each of the 17 items and for the Burden and Benefit subscales. Mann-Whitney U tests were used, with P ⩽ 0.05 considered significant. Analysis was performed using SPSS version 17.0 (SPSS Inc, Chicago, IL).
RESULTS
Patients were recruited between July 2010 and August 2011. During this time, 103 patients were approached before an office visit at one of the arrhythmia clinics, of which 23 refused participation. Characteristics of the 80 AF respondents are in Table 1. Of them, 65 respondents were receiving a regimen containing warfarin (n = 39, warfarin; and n = 26, warfarin plus antiplatelet agent/s) and 15 were not (n= 11, aspirin; n = 2, aspirin plus clopidogrel; and n = 2, dabigatran). Patients receiving warfarin spent 63% of the time in the therapeutic international normalized ratio range of 2 to 3 (calculated using the Rosendaal method).8Of the 15 patients not receiving a regimen containing warfarin, 4 patients had a CHADS2 (congestive heart failure, hypertension, advanced age, diabetes and prior stroke/transient ischemic attack) score of 0 to 1, 4 patients were 80 years or older (3 patients were 85 years or older), 2 patients were receiving dabigatran, and 2 were receiving dual antiplatelet therapy (owing to the presence of coronary stents).
Patients’ and Thromboprophylaxis Treatment Characteristics
Six of the 17 individual questions (1, 5, 7, and 11–13) had higher mean scores in the patients receiving warfarin compared with the patients not receiving warfarin (P < 0.05 for all; Table2). No difference was seen between the 2 groups with regard to the ACTS Benefits score (11.1 ± 3.4 vs 10.4 ± 3.7; P = 0.38). The mean ACTS Burden scores were higher (more favorable) in the no-warfarin group (44.5 ± 6.4) compared with the warfarin group (39.8 ± 8.0; P = 0.003). Within the warfarin group, no difference in the ACTS Burden or Benefits scores were observed when comparing those receiving or not receiving aspirin (P > 0.05 for both). However, the ACTS Burden scores were significantly associated with time spent in the therapeutic international normalized ratio range (a small and positive correlation; Pearson r = 0.248; P = 0.046).
Responses to the ACTS According to Warfarin Use
DISCUSSION
Our assessment of the patients with AF receiving thromboprophylaxis suggests that those taking warfarin may have less favorable feelings about their treatment compared to other available agents. The patients’ responses suggested that the respondents believe they suffer significant burden owing to receiving warfarin but receive no incremental benefit compared to other thromboprophylaxis agents. Specifically, the patients perceived they had limitations on both the physical activity they could engage in and regarding their diet. The patients taking warfarin also reported feeling more hassled and frustrated by their thromboprophylaxis.
Currently, there is no published guidance as to what constitutes a minimally important clinical difference on the ACTS scale. However, many researchers have suggested that a change of 0.5 SD units or more on a patient-reported outcomes measure often estimates this value.9The SD on the ACTS Burden score in our overall population was 7.9, suggesting a minimally important clinical difference of approximately 4 points on the ACTS burden domain. Because we observed a 4.7-point difference between the warfarin and no-warfarin groups in our analysis, it seems reasonable to suggest that the difference is not only statistically significant but also clinically relevant.
Of note, 2 recent randomized trials10,11evaluating the use of thromboprophylaxis agents in patients with AF deemed “unsuitable” for warfarin suggest that up to 38% of patients refuse to use warfarin. A number of patient-preference studies demonstrated that if the decision was left to patients, far fewer patients would opt to receive warfarin compared to what is suggested by national treatment guidelines.12–14Countless observational studies have found that only a fraction of patients with AF indicated to receive warfarin are actually receiving it.12,13The results of these studies may be explained by our trial. In the patients who were taking warfarin therapy, many for prolonged periods of time, they felt burdened by the lifestyle limitations and the variable nature of the regimen and blood monitoring versus non–warfarin-based regimens.
It has been suggested that greater patient involvement in health care decision-making processes may lead to increased satisfaction with the decision-making process, improved adherence to treatment modalities, and ultimately, prolongation of quality-adjusted survival. Consequently, at least one set of national AF clinical practice guidelines14has recommended that patient preferences be incorporated into clinical decision making and selection of optimal thromboprophylaxis for individual patients. Factors such as efficacy, safety, cost, and convenience matter to a different extent in different individuals; and extrapolating the values of the caregiver onto the patient may not be as optimal as providing information to the patient and engaging in shared decision making.
There are limitations to this study that should be considered when evaluating our results. We enrolled only a small population into this exploratory study. The small sample size and large proportion of patients taking warfarin may have resulted in a failure to show significant differences on individual survey questions when one truly exists (owing to underpowering). Also of import, the patients with AF were recruited from arrhythmia clinics affiliated with a single urban teaching hospital in the northeastern United States. All patients had health care insurance, the average patient CHADS2 score was 2.1, and the mean number of thromboprophylaxis agents used by patients was 1.4. Because these characteristics may not be representative of all patients with AF, caution should be used in generalizing our results to other practices or settings. Finally, we would have liked to compare warfarin patients’ responses to ACTS with those receiving other agents individually; however, there were too few receiving any one specific alternative thromboprophylaxis regimen. Most of the patients not receiving warfarin were taking aspirin alone, with only a handful receiving dabigatran or clopidogrel plus aspirin. This is likely a result of these other regimens being new to the US market and only recently being included in national AF treatment guidelines.1As such, a study like ours should be repeated in the future on a larger scale when there is greater permeation of other antithrombotic agents in clinical use. The CHADS2 score was very similar in the 2 groups, and we do not believe that the underlying burden of stroke risk would explain our findings and is a strength of our study.
CONCLUSION
Patients with AF receiving warfarin may have less favorable feelings about their thromboprophylaxis compared to non–warfarin-based regimens. Patients may perceive greater burden owing to their warfarin, while not perceiving any incremental benefit, versus non–warfarin-based regimens.
