Barriers to the Use of Anticoagulation for Nonvalvular Atrial Fibrillation
A Representative Survey of Australian Family Physicians
Background and Purpose— Anticoagulation reduces the risk of stroke in nonvalvular atrial fibrillation yet remains underused. We explored barriers to the use of anticoagulants among Australian family physicians.
Methods— The authors conducted a representative, national survey.
Results— Of the 596 (64.4%) eligible family physicians who participated, 15.8% reported having a patient with nonvalvular atrial fibrillation experience an intracranial hemorrhage with anticoagulation and 45.8% had a patient with known nonvalvular atrial fibrillation experience a stroke without anticoagulation. When presented with a patient at “very high risk” of stroke, only 45.6% of family physicians selected warfarin in the presence of a minor falls risk and 17.1% would anticoagulate if the patient had a treated peptic ulcer. Family physicians with less decisional conflict and longer-standing practices were more likely to endorse anticoagulation.
Conclusion— Strategies to optimize the management of nonvalvular atrial fibrillation should address psychological barriers to using anticoagulation.
Anticoagulation reduces the risk of stroke in nonvalvular atrial fibrillation (NVAF)1 yet remains underused.2 Educational strategies that aim to promote anticoagulant uptake among clinicians such as audit and feedback3 and decision tools4 have not been effective. To assist in developing innovative strategies, we carried out a survey to identify as-yet unexplored psychological barriers to anticoagulant prescribing.
Subjects and Methods
One-thousand randomly selected family physicians (FPs) across Australia were requested to report, through a questionnaire, adverse events in any of their patients with NVAF and their anticipated responsibility for these adverse outcomes (Table 2).5 Eight case scenarios assessed FPs’ self-reported management of NVAF (Table 3). Case 4, describing a patient at “very high risk of stroke,”6 was modified by adding 4 histories to assess the effect of perceived bleeding risk on prescribing (Table 3). Responses to 5 statements (Table 2) were summed to produce a score assessing “decisional conflict.”7 Data collection was completed in July 2006.
FPs who had retired, died, were ill, could not be contacted, or who were on extended leave were ineligible. We received 596 questionnaires from 926 eligible FPs (64.4%). Responders and nonresponders were similar with respect to age, years in practice, sex, and state of practice (Ps=0.16 to 0.52). Respondents were representative of Australian FPs (Table 1).
Although 15.8% of FPs reported having a patient with NVAF experience an intracranial hemorrhage with anticoagulation, 45.8% reported having a patient experience an ischemic stroke without anticoagulation.
Approximately one fifth (17.6%) anticipated feeling most responsible for an intracranial hemorrhage on anticoagulation, whereas 31.5% anticipated feeling most responsible for an ischemic stroke in a patient without anticoagulation.
FPs who anticipated feeling most responsible for an intracranial hemorrhage were more likely to have previously experienced this outcome compared with FPs who anticipated feeling most responsible for a stroke (21.9% versus 11.2%; χ2=6.11; P=0.047). The experience of a stroke was not correlated with FPs’ anticipated responsibility for adverse outcomes (χ2=3.42; P=0.18).
Self-Reported Management of Nonvalvular Atrial Fibrillation
Although most FPs selected warfarin for the 65 year old at “high” risk of stroke (case 2), fewer selected warfarin for the 75 year old at “moderate to high” risk or at “very high” risk of stroke (cases 3 and 4) (Ps<0.001) (Table 3).
Effect of Perceived Bleeding Risk
Although 71.0% of respondents selected warfarin for the 75-year-old patient at “very high” risk of stroke (case 4), 45.6% selected warfarin when this patient was described as having a minor falls risk (case 5). Only 17.1% and 28.9%, respectively, would do so if case 4 had a recent gastrointestinal bleed now on antiulcer treatment (case 6) or frequent nosebleeds (case 7; Table 3).
Twenty-nine percent (29.2) switched their preference for warfarin when case 4 was described as having a “minor falls risk” (case 5). These FPs had been in family practice for fewer years (mean=17.2 years) than FPs who maintained their preference (mean=21.1 years) or who had not initially selected warfarin (mean=18.5 years; P=0.001).
Over half (57.7%) switched their preference from warfarin if case 4 had a gastrointestinal bleed caused by a peptic ulcer (now treated; case 6). FPs who maintained their preference for warfarin had been practicing for longer (P=0.026) and had less decisional conflict than other FPs (P=0.008).
Just under half (46.6%) switched their preference from warfarin when case 4 was described as having frequent nose bleeds (case 7).
Only 8.1% of FPs maintained a preference for warfarin when case 4 had a history of a hypertensive intracranial hemorrhage (case 8). Compared with other FPs, these FPs were more likely to have had a patient experience an intracranial hemorrhage on anticoagulation (P=0.045), to feel most responsible for a stroke (P=0.054), and to be members of Australia’s peak organization for family practitioners (P=0.013).
Experience of bleeding events and fear of bleeding appear to influence prescribing. Furthermore, the experience of an intracranial hemorrhage in a patient with NVAF with anticoagulation appeared to condition FPs to feeling responsible for this outcome. However, the more common experience of a stroke did not affect their sense of responsibility. Clinicians may have more “chagrin” over harm arising from so-called acts of commission because they are inculcated to first “do no harm.”8 They may therefore abandon potentially harmful therapies, even if the benefits outweigh the risks.8 Our findings are consistent with a recent study showing that clinicians are less likely to prescribe anticoagulants for NVAF if any one of their patients experiences a serious bleeding event yet are no more likely to prescribe warfarin if a patient with NVAF experiences a stroke.9
For a patient at “very high risk” of stroke, the majority would substitute warfarin with another treatment if the patient has nose bleeds, treated peptic ulcers, and a minor falls risk, although these bleeding risks are preventable and outweighed by the benefits of anticoagulation.6,10 FPs with longer-standing practices were more likely to maintain a preference for warfarin, indicating that experience desensitizes FPs to fears of anticoagulation. Reduced decisional conflict may also increase anticoagulant prescribing. Anticoagulation for patients with a history of an intracranial hemorrhage remains controversial11 and was not endorsed by the majority of FPs. However, the minority maintaining a preference for warfarin who did so felt more responsible for a stroke and did so despite having patients experience an intracranial hemorrhage.
Our results may only generalize to FPs and the usual caveats about self-reported data apply. Our results commend the implementation of interventions that will reduce fears about bleeding, promote a rational appraisal of the risk and benefits of anticoagulation, and resolve decisional conflict.
We express our gratitude to Ms Tracey Coles for administering advance letters to FPs and for her diligent data entry. We also thank Ms Maria-Jose Velasco for her diligent double data entry of all questionnaires to maximize data quality. A posthumous acknowledgement is extended to Dr Ian W. Black for being involved in the formation of this study. The University of New South Wales, Human Research Ethics Committee approved the study.
Sources of Funding
M.G. is supported by a postdoctoral Public Health Training Fellowship from the National Health and Medical Research Council of Australia (Reg Key #300616). M.G. is named as the principal investigator on an Early Career Researcher grant awarded from the Faculty of Medicine, The University of New South Wales; all other authors were associate investigators.
- Received May 29, 2007.
- Accepted June 7, 2007.
Weir CJ, Lees KR, MacWalter RS, Muir KW, Wallesch CW, McLelland EV, Hendry A. Cluster-randomized, controlled trial of computer-based decision support for selecting long-term anti-thrombotic therapy after acute ischaemic stroke. Quality Journal in Medicine. 2003; 96: 143–153.
Poses RM, De Saintonge DM, McClish DK, Smith WR, Huber EC, Clemo FL, Schmitt BP, Alexander-Forti D, Racht EM, Colenda CC 3rd, Centor RM. An international comparison of physicians’ judgments of outcome rates of cardiac procedures and attitudes toward risk, uncertainty, justifiability, and regret. Med Decis Making. 1998; 18: 131–140.
ACC/AHA/ESC 2006 Guidelines for the Management of Patients with Atrial Fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients with Atrial Fibrillation): developed in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society. Circulation. 2006; 114: e257–354.
Dolan JG. A method for evaluating health care providers’ decision making: the Provider Decision Process Assessment Instrument. Med Decis Making. 1999; 19: 38–41.
Aberegg SK, Arkes H, Terry PB. Failure to adopt beneficial therapies caused by bias in medical evidence evaluation. Med Decis Making. 2006; 26: 575–582.
Choudhry NK, Anderson GM, Laupacis A, Ross-Degnan D, Normand SL, Soumerai SB. Impact of adverse events on prescribing warfarin in patients with atrial fibrillation: matched pair analysis. BMJ. 2006; 332: 141–145.