Quality of life of elderly people on warfarin for atrial fibrillation

SIR—Atrial fibrillation/flutter (AF) predisposes to ischaemic stroke and is associated with greater disability and mortality than those without AF [1]. The prevalence of AF rises with age and increases sharply in older people [2]. In one study, AF accounted for 10–15% of ischaemic strokes in younger people and nearly 25% in people aged over 80 years [3]. Therapeutic oral anticoagulation with international normalized ratio (INR) of 2–3 is well established both for primary and secondary stroke prevention. This confers an additional 40% risk reduction in recurrent stroke compared with antiplatelet therapy [4]. Furthermore, recent evidence suggests that this may also be associated with reduced stroke severity [5] and overall mortality [6] in patients with ischaemic stroke due to AF. The CHADS2 score, which estimates risk of stroke in patients with AF, has been validated in recently discharged hospital patients aged 65 to 95 years [7]. The adjusted stroke rate per 100 patient-years is estimated from <1% (age <65 years, no risk factors) to 18.2% in patients with a CHADS2 score of six points [8]. Anticoagulation is recommended in high-risk (CHADS2 >2) patients even after conversion of AF to sinus rhythm [9]. Warfarin is underused in older people [8]. Medical practitioners have felt that old age deters anticoagulation independent of severity of stroke and bleeding risk [10, 11]. Elderly patients with AF are often prescribed aspirin, on the presumption that it is safer. However, in BAFTA (Birmingham Atrial Fibrillation Treatment of the Aged) study there was no difference in major bleeding rates between warfarin (INR 2–3) and aspirin (75 mg) [12]. Assessment of quality of life (QL) has received increasing attention as an outcome measure of the subjective sequelae of disease. Measures of QL take into consideration the potential burdens and side effects of any treatment that leads to improvement in specific symptoms but that may negatively affect general well being. Health-related QL is a summarized outcome assessment that is designed to reflect the subjective impairment in general well being caused by individual aspects of pain, and psychological, emotional and physical disturbances. Perceived reduction in QL is an important factor which may influence physicians’ prescription and patients’ use of warfarin therapy. Bleeding complications together with the inconvenience of anticoagulation monitoring are thought to reduce an individual's QL. Conversely, patients may be more averse to the potential consequences of stroke and less bothered by the side effects of antithrombotic treatment than doctors. In an observational study, patients at high risk of stroke placed more value on the avoidance of stroke and less value on the avoidance of bleeding than did physicians [13]. Improving QL is a major goal in the management of chronic disease. QL improves when AF is treated, irrespective of the mode of treatment [14]. It might be anticipated that long-term warfarin therapy affects QL but in a previous study we found no difference between patients with short- and long-term warfarin treatment [15]. We have undertaken a further observational study in the same individuals at the beginning of warfarin therapy for thromboprophylaxis of AF and after 6 months to determine if there is a perceived change in QL with this treatment. This was a sequential study in the same individuals comparing their QL before starting warfarin (taken as control) and 6 months after treatment. The setting was the Leeds anticoagulation clinic. Inclusion criteria were patients aged over 75 years taking warfarin for AF thromboprophylaxis. Patients were excluded if they were on warfarin for other reasons, had received warfarin before or had cognitive impairment as determined by an Abbreviated Mental Test Score of <7/10. QL was measured using the Short Form 12 version 2 (SF12v2) questionnaire [16, 17], an abbreviated form of SF-36, which is well validated in the Western populations. This provides composite QL scores of mental and physical health. Patients completed an SF12v2 questionnaire on starting anticoagulation, and at the same time they were given a blank SF12v2 questionnaire with a stamped addressed envelope and were asked to complete and return it after being on warfarin for 6 months. A telephone reminder to complete the form was made when necessary. The physical component scores (PCS) and the mental component scores (MCS) were calculated from responses using the SF12v2 handbook [18]. These are designed to have a mean population score of 50 with a standard deviation (SD) of 10, and in accordance with other studies, we deemed a difference of 5 to be clinically significant. We proposed a paired t-test for the comparison, and taking the SD of the difference in SF12v2 scores as 14.14(10⁠) [17], we calculated we would require 85 participants to achieve 90% power with a difference of at least 5 points and a significance level of 0.05. In order to allow for ‘drop outs’, we determined we would need a target sample size of 100. This sample size calculation was performed using STATA (StataCorp LP) version 10.0 [19]. We have termed the physical and mental component score at the start of warfarin therapy as PCS and MCS, respectively, and those after being on warfarin for 6 months as PCS6 and MCS6. Matched-pairs t-tests were used to assess changes in scores using STATA software. The study was approved by the East Leeds Research Ethics Committee, and informed written consent was obtained from all participants. One hundred and ten eligible patients filled in the SF12v2 questionnaire at the beginning of induction of anticoagulation. Fifty-five per cent were women. The mean age of men and women were 80 years and 82 years, respectively. Ninety-four patients returned forms after 6 months. Seven incomplete responses were excluded, and 87 responses were analysed. Of the 16 non-responders, 6 did not wish to fill in the second questionnaire, 3 patients died, 1 suffered a stroke, 2 had to stop warfarin for bleeding complications and 4 patients could not be contacted. The SF12v2 norms for the population mean and standard deviation (SD) for PCS and MCS in over 75 age group were used as baseline QL values [18]. Inspecting the baseline values for QL, we found that our sample of patients with AF had PCS and MCS below (Table 1) those expected for an over 75 age group, and for females this reduction was statistically significant at the 5% level. Comparison of base line PCS and MCS PCS = physical component scores, MCS = mental component scores. Comparison of base line PCS and MCS PCS = physical component scores, MCS = mental component scores. In order to determine if the QL scores were influenced by commencing warfarin therapy, matched-pairs t-tests were used for both the PCS and MCS. The results are shown in Table 2. There were reductions in the mean scores of both PCS and MCS, but neither difference reached statistical significance. Results of matched-pairs t-tests PCS = physical component scores, MCS = mental component scores. Results of matched-pairs t-tests PCS = physical component scores, MCS = mental component scores. In our previous study, we compared the QL of older AF patients those on warfarin for less than a year (mean 0.5 year) with those on warfarin for over a year (mean 5.5 year) [15] and did not find any difference in QL with duration of anticoagulation and recommended a further study assessing QL before and after warfarin treatment in the same individuals. We found that mental QL was affected only by increasing age and that physical QL was affected only by the CHADS2 score, which was unsurprising as CHADS2 is a surrogate marker of disease burden. Indeed, the fact that we found this in our study provides further confirmation of its reliability. A North American Study, Boston Area Anticoagulation Trial for Atrial Fibrillation (BAATAF) similarly showed no change in QL in a younger patient population (mean age 68 years) [20]. In our study, we used SF12v2 to assess health-related QL, as we were unable to find any other specific tool to use in elderly patients with AF. We found that QL remained the same in individuals over a period of 6 months on warfarin. All patients had to attend a hospital-based anticoagulation clinic for the initiation of warfarin therapy. Most patients were then subsequently referred to and monitored by community-based anticoagulation clinics close to their home. Community monitoring reduces patient inconvenience from waiting for transport and the duration of testing as well as reducing the overall cost of anticoagulation monitoring. We assessed how much of the time patients’ physical and emotional problems interfered with their social activities such as visiting relatives and friends whilst on warfarin. We did not find any significant difference after being on warfarin for 6 months. We did not address whether taking warfarin impaired their ability to take holidays. Until recently, vitamin K antagonists have been the only available oral antithrombotic agents, and warfarin is the most commonly used drug in this group. However, it is a drug with some inherent problems including high plasma protein binding, genetic variation in metabolism, interactions with food and drugs, a narrow therapeutic index requiring frequent monitoring and slow onset of action requiring overlap with initial parenteral anticoagulation. Alternative oral agents such as direct thrombin inhibitors are being compared with warfarin for stroke prevention in AF. These newer agents do not require anticoagulation monitoring; however, there is no current test available to check their efficacy, and they must be used with caution in patients with renal impairment which occurs commonly in older people. Newer agents are also expensive and at present, no antidote is available to reverse over-anticoagulation. At least for the time being, warfarin remains the drug of choice for stroke prevention in patients with AF, and patients’ QL, which in turn may affect compliance, remains an important aspect of treatment. This is an observational study in community-dwelling elderly people. The results are not applicable to patients with cognitive impairment, in whom anticoagulation is often avoided. The case for wider but judicious use of anticoagulation for older people with AF is strengthened further by this study. Our study excluded elderly patients who had previous complications arising from warfarin treatment, and those in care homes. As suggested by a recent Australian study, we need a better way to identify the risk of bleeding in frail elderly people [21]. A further study with repeat assessments in the same individuals after being on warfarin for an even longer period of time might be helpful and it would be valuable to include other activities which might affect QL such as the ability to take holidays. Warfarin treatment per se does not affect either the physical or mental QL of older patients with AF over 6 months period. Physicians and their patients can be reassured that concerns regarding deterioration in QL need not play a significant role in deciding on warfarin use. This study was supported by grant 10/2004 from the British Geriatrics Society. The results of this study have been accepted as an abstract for the autumn meeting of the BGS 2008.