[Skip to Content]
[Skip to Content Landing]
Figure.  CONSORT Diagram
CONSORT Diagram

SDM indicates shared decision-making.

Table 1.  Patient Characteristics
Patient Characteristics
Table 2.  Clinician Characteristics
Clinician Characteristics
Table 3.  Participant-Reported Quality of Shared Decision-making
Participant-Reported Quality of Shared Decision-making
Table 4.  Observed Encounter Outcomes
Observed Encounter Outcomes
1.
Morillo  CA, Banerjee  A, Perel  P, Wood  D, Jouven  X.  Atrial fibrillation: the current epidemic.   J Geriatr Cardiol. 2017;14(3):195-203.PubMedGoogle Scholar
2.
Colilla  S, Crow  A, Petkun  W, Singer  DE, Simon  T, Liu  X.  Estimates of current and future incidence and prevalence of atrial fibrillation in the US adult population.   Am J Cardiol. 2013;112(8):1142-1147. doi:10.1016/j.amjcard.2013.05.063 PubMedGoogle ScholarCrossref
3.
Tawfik  A, Bielecki  JM, Krahn  M,  et al.  Systematic review and network meta-analysis of stroke prevention treatments in patients with atrial fibrillation.   Clin Pharmacol. 2016;8:93-107. doi:10.2147/CPAA.S105165 PubMedGoogle Scholar
4.
Gallagher  AM, Rietbrock  S, Plumb  J, van Staa  TP.  Initiation and persistence of warfarin or aspirin in patients with chronic atrial fibrillation in general practice: do the appropriate patients receive stroke prophylaxis?   J Thromb Haemost. 2008;6(9):1500-1506. doi:10.1111/j.1538-7836.2008.03059.x PubMedGoogle ScholarCrossref
5.
Ogilvie  IM, Newton  N, Welner  SA, Cowell  W, Lip  GYH. Underuse of oral anticoagulants in atrial fibrillation: a systematic review. Am J Med. 2010;123(7):638-645.
6.
Yao  X, Abraham  NS, Alexander  GC,  et al.  Effect of adherence to oral anticoagulants on risk of stroke and major bleeding among patients with atrial fibrillation.   J Am Heart Assoc. 2016;5(2):e003074. doi:10.1161/JAHA.115.003074 PubMedGoogle Scholar
7.
Hernandez  I, He  M, Chen  N, Brooks  MM, Saba  S, Gellad  WF.  Trajectories of oral anticoagulation adherence among Medicare beneficiaries newly diagnosed with atrial fibrillation.   J Am Heart Assoc. 2019;8(12):e011427. doi:10.1161/JAHA.118.011427 PubMedGoogle Scholar
8.
January  CT, Wann  LS, Alpert  JS,  et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines.  2014 AHA/ACC/HRS guideline 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 Heart Rhythm Society.   J Am Coll Cardiol. 2014;64(21):e1-e76. doi:10.1016/j.jacc.2014.03.022 PubMedGoogle ScholarCrossref
9.
O’Neill  ES, Grande  SW, Sherman  A, Elwyn  G, Coylewright  M.  Availability of patient decision aids for stroke prevention in atrial fibrillation: a systematic review.   Am Heart J. 2017;191:1-11. doi:10.1016/j.ahj.2017.05.014 PubMedGoogle ScholarCrossref
10.
Stacey  D, Legare  F, Lewis  K,  et al.  Decision aids for people facing health treatment or screening decisions.   Cochrane Database Syst Rev. 2017;4(4):CD001431. doi:10.1002/14651858.CD001431.pub5 PubMedGoogle Scholar
11.
Clarkesmith  DE, Lip  GYH, Lane  DA.  Patients’ experiences of atrial fibrillation and non–vitamin K antagonist oral anticoagulants (NOACs), and their educational needs: a qualitative study.   Thromb Res. 2017;153:19-27. doi:10.1016/j.thromres.2017.03.002 PubMedGoogle ScholarCrossref
12.
Thomson  RG, Eccles  MP, Steen  IN,  et al.  A patient decision aid to support shared decision-making on anti-thrombotic treatment of patients with atrial fibrillation: randomised controlled trial.   Qual Saf Health Care. 2007;16(3):216-223. doi:10.1136/qshc.2006.018481 PubMedGoogle ScholarCrossref
13.
Eckman  MH, Costea  A, Attari  M,  et al.  Shared decision-making tool for thromboprophylaxis in atrial fibrillation—a feasibility study.   Am Heart J. 2018;199:13-21. doi:10.1016/j.ahj.2018.01.003 PubMedGoogle ScholarCrossref
14.
Fatima  S, Holbrook  A, Schulman  S, Park  S, Troyan  S, Curnew  G.  Development and validation of a decision aid for choosing among antithrombotic agents for atrial fibrillation.   Thromb Res. 2016;145:143-148. doi:10.1016/j.thromres.2016.06.015 PubMedGoogle ScholarCrossref
15.
Saposnik  G, Joundi  RA.  Visual aid tool to improve decision making in anticoagulation for stroke prevention.   J Stroke Cerebrovasc Dis. 2016;25(10):2380-2385. doi:10.1016/j.jstrokecerebrovasdis.2016.05.037 PubMedGoogle ScholarCrossref
16.
Eckman  MH, Wise  RE, Naylor  K,  et al.  Developing an atrial fibrillation guideline support tool (AFGuST) for shared decision making.   Curr Med Res Opin. 2015;31(4):603-614. doi:10.1185/03007995.2015.1019608 PubMedGoogle ScholarCrossref
17.
Kunneman  M, Branda  ME, Noseworthy  PA,  et al.  Shared decision making for stroke prevention in atrial fibrillation: study protocol for a randomized controlled trial.   Trials. 2017;18(1):443. doi:10.1186/s13063-017-2178-y PubMedGoogle ScholarCrossref
18.
Zeballos-Palacios  CL, Hargraves  IG, Noseworthy  PA,  et al; Shared Decision Making for Atrial Fibrillation (SDM4AFib) Trial Investigators.  Developing a conversation aid to support shared decision making: reflections on designing anticoagulation choice.   Mayo Clin Proc. 2019;94(4):686-696. doi:10.1016/j.mayocp.2018.08.030 PubMedGoogle ScholarCrossref
19.
Montori  VM, Kunneman  M, Brito  JP.  Shared decision making and improving health care: the answer is not in.   JAMA. 2017;318(7):617-618. doi:10.1001/jama.2017.10168 PubMedGoogle ScholarCrossref
20.
Mayo Clinic. Anticoagulation choice decision aid. Accessed June 8, 2020. https://anticoagulationdecisionaid.mayoclinic.org/
21.
Lip  GYH, Nieuwlaat  R, Pisters  R, Lane  DA, Crijns  HJGM.  Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor–based approach: the Euro Heart Survey on Atrial Fibrillation.   Chest. 2010;137(2):263-272. doi:10.1378/chest.09-1584 PubMedGoogle ScholarCrossref
22.
Pisters  R, Lane  DA, Nieuwlaat  R, de Vos  CB, Crijns  HJGM, Lip  GYH.  A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey.   Chest. 2010;138(5):1093-1100. doi:10.1378/chest.10-0134 PubMedGoogle ScholarCrossref
23.
Chew  LD, Bradley  KA, Boyko  EJ.  Brief questions to identify patients with inadequate health literacy.   Fam Med. 2004;36(8):588-594.PubMedGoogle Scholar
24.
Fagerlin  A, Zikmund-Fisher  BJ, Ubel  PA, Jankovic  A, Derry  HA, Smith  DM.  Measuring numeracy without a math test: development of the Subjective Numeracy Scale.   Med Decis Making. 2007;27(5):672-680. doi:10.1177/0272989X07304449 PubMedGoogle ScholarCrossref
25.
Consumer Assessment of Healthcare Providers and Systems. CAHPS research on survey design and administration. Agency for Healthcare Research and Quality. Updated January 2020. Accessed April 21, 2017. https://www.ahrq.gov/cahps/surveys-guidance/survey-methods-research/index.html
26.
O’Connor  AM.  Validation of a decisional conflict scale.   Med Decis Making. 1995;15(1):25-30. doi:10.1177/0272989X9501500105 PubMedGoogle ScholarCrossref
27.
Elwyn  G, Hutchings  H, Edwards  A,  et al.  The OPTION scale: measuring the extent that clinicians involve patients in decision-making tasks.   Health Expect. 2005;8(1):34-42. doi:10.1111/j.1369-7625.2004.00311.x PubMedGoogle ScholarCrossref
28.
Diggle  PJ, Heagerty P, Liang  K-Y, Zeger  SL.  Analysis of Longitudinal Data. Oxford University Press; 1994. Oxford Statistical Science Series; vol. 25.
29.
Rubin  DB.  Multiple Imputation for Nonresponse in Surveys. John Wiley & Sons; 1987. doi:10.1002/9780470316696
30.
Thissen  D, Steinberg  L, Kuang  D.  Quick and easy implementation of the Benjamini-Hochberg procedure for controlling the false positive rate in multiple comparisons.   J Educ Behav Stat. 2002;27(1):77-83. doi:10.3102/10769986027001077 Google ScholarCrossref
31.
Chew  LD, Griffin  JM, Partin  MR,  et al.  Validation of screening questions for limited health literacy in a large VA outpatient population.   J Gen Intern Med. 2008;23(5):561-566. doi:10.1007/s11606-008-0520-5 PubMedGoogle ScholarCrossref
32.
Weymiller  AJ, Montori  VM, Jones  LA,  et al.  Helping patients with type 2 diabetes mellitus make treatment decisions: statin choice randomized trial.   Arch Intern Med. 2007;167(10):1076-1082. doi:10.1001/archinte.167.10.1076 PubMedGoogle ScholarCrossref
33.
Mullan  RJ, Montori  VM, Shah  ND,  et al.  The diabetes mellitus medication choice decision aid: a randomized trial.   Arch Intern Med. 2009;169(17):1560-1568. doi:10.1001/archinternmed.2009.293 PubMedGoogle ScholarCrossref
34.
LeBlanc  A, Herrin  J, Williams  MD,  et al.  Shared decision making for antidepressants in primary care: a cluster randomized trial.   JAMA Intern Med. 2015;175(11):1761-1770. doi:10.1001/jamainternmed.2015.5214 PubMedGoogle ScholarCrossref
35.
Peterson  ED.  Machine learning, predictive analytics, and clinical practice: can the past inform the present?   JAMA. 2019; 322(23):2283-2284. doi:10.1001/jama.2019.17831 PubMedGoogle Scholar
36.
Beaser  AD, Cifu  AS.  Management of patients with atrial fibrillation.   JAMA. 2019;321(11):1100-1101. doi:10.1001/jama.2019.1264 PubMedGoogle ScholarCrossref
37.
Decision-Making and Choices to Inform Dialogue and Empower AFib Patients (DECIDE) Center. Project summary. Patient-Centered Outcomes Research Institute. Updated February 20, 2020. Accessed January 27, 2020. https://www.pcori.org/research-results/2018/decision-making-and-choices-inform-dialogue-and-empower-afib-patients-decide
38.
Kunneman  M, Gionfriddo  MR, Toloza  FJK,  et al.  Humanistic communication in the evaluation of shared decision making: a systematic review.   Patient Educ Couns. 2019;102(3):452-466. doi:10.1016/j.pec.2018.11.003 PubMedGoogle ScholarCrossref
39.
Hargraves  IG, Montori  VM, Brito  JP,  et al.  Purposeful SDM: a problem-based approach to caring for patients with shared decision making.   Patient Educ Couns. 2019;102(10):1786-1792. doi:10.1016/j.pec.2019.07.020 PubMedGoogle ScholarCrossref
40.
Couet  N, Desroches  S, Robitaille  H,  et al.  Assessments of the extent to which health-care providers involve patients in decision making: a systematic review of studies using the OPTION instrument.   Health Expect. 2015;18(4):542-561. doi:10.1111/hex.12054 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 4,340
    Citations 0
    Original Investigation
    July 20, 2020

    Assessment of Shared Decision-making for Stroke Prevention in Patients With Atrial Fibrillation: A Randomized Clinical Trial

    Author Affiliations
    • 1Knowledge and Evaluation Research Unit, Mayo Clinic, Rochester, Minnesota
    • 2Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands
    • 3Department of Biostatistics and Informatics, Colorado School of Public Health, Anschutz Medical Campus, University of Colorado Denver, Aurora
    • 4Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, Mayo Clinic, Rochester, Minnesota
    • 5Division of General Internal Medicine, Hennepin Health, Minneapolis, Minnesota
    • 6Thrombosis Clinic and Anticoagulation Services, Park Nicollet Health Services, St Louis Park, Minnesota
    • 7Division of Cardiology, Department of Medicine, University of Mississippi Medical Center, Jackson
    • 8Department of Internal Medicine, Division of Cardiovascular Disease, University of Alabama at Birmingham, Birmingham
    • 9Department of Emergency Medicine, University of Alabama at Birmingham, Birmingham
    • 10Department of Psychiatry and Psychology, Mayo Clinic, Rochester, Minnesota
    • 11Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota
    • 12Heart Rhythm Services, Department of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
    JAMA Intern Med. Published online July 20, 2020. doi:10.1001/jamainternmed.2020.2908
    Key Points

    Question  Does use of the Anticoagulation Choice Shared Decision Making encounter tool affect the quality of shared decision-making and anticoagulant treatment selection in patients with atrial fibrillation who are at risk of experiencing stroke?

    Findings  In this randomized clinical trial of 922 patients with atrial fibrillation and 151 clinicians, use of the Anticoagulation Choice Shared Decision Making encounter tool resulted in several improvements in markers of shared decision-making quality and clinician satisfaction without changing anticoagulant treatment rates or encounter length.

    Meaning  The results indicate that use of a tool for shared decision-making in the clinical encounter contributes to the care of patients with atrial fibrillation who are considering anticoagulant treatment.

    Abstract

    Importance  Shared decision-making (SDM) about anticoagulant treatment in patients with atrial fibrillation (AF) is widely recommended but its effectiveness is unclear.

    Objective  To assess the extent to which the use of an SDM tool affects the quality of SDM and anticoagulant treatment decisions in at-risk patients with AF.

    Design, Setting, and Participants  This encounter-randomized trial recruited patients with nonvalvular AF who were considering starting or reviewing anticoagulant treatment and their clinicians at academic, community, and safety-net medical centers between January 30, 2017 and June 27, 2019. Encounters were randomized to either the standard care arm or care that included the use of an SDM tool (intervention arm). Data were analyzed from August 1 to November 30, 2019.

    Interventions  Standard care or care using the Anticoagulation Choice Shared Decision Making tool (which presents individualized risk estimates and compares anticoagulant treatment options across issues of importance to patients) during the clinical encounter.

    Main Outcomes and Measures  Quality of SDM (which included quality of communication, patient knowledge about AF and anticoagulant treatment, accuracy of patient estimates of their own stroke risk [within 30% of their estimate], decisional conflict, and satisfaction), decisions made during the encounter, duration of the encounter, and clinician involvement of patients in the SDM process.

    Results  The clinical trial enrolled 922 patients (559 men [60.6%]; mean [SD] age, 71 [11] years) and 244 clinicians. A total of 463 patients were randomized to the intervention arm and 459 patients to the standard care arm. Participants in both arms reported high communication quality, high knowledge, and low decisional conflict, demonstrated low accuracy in their risk perception, and would similarly recommend the approach used in their encounter. Clinicians were significantly more satisfied after intervention encounters (400 of 453 encounters [88.3%] vs 277 of 448 encounters [61.8%]; adjusted relative risk, 1.49; 95% CI, 1.42-1.53). A total of 747 of 873 patients (85.6%) chose to start or continue receiving an anticoagulant medication. Patient involvement in decision-making (as assessed through video recordings of the encounters using the Observing Patient Involvement in Decision Making 12-item scale) scores were significantly higher in the intervention arm (mean [SD] score, 33.0 [10.8] points vs 29.1 [13.1] points, respectively; adjusted mean difference, 4.2 points; 95% CI, 2.8-5.6 points). No significant between-arm difference was found in encounter duration (mean [SD] duration, 32 [16] minutes in the intervention arm vs 31 [17] minutes in the standard care arm; adjusted mean between-arm difference, 1.1; 95% CI, −0.3 to 2.5 minutes).

    Conclusion and Relevance  The use of an SDM encounter tool improved several measures of SDM quality and clinician satisfaction, with no significant effect on treatment decisions or encounter duration. These results help to calibrate expectations about the value of implementing SDM tools in the care of patients with AF.

    Trial Registration  ClinicalTrials.gov Identifier: NCT02905032

    Introduction

    Atrial fibrillation (AF) is the most common cardiac arrhythmia observed in clinical practice, with more than 5 million people experiencing AF in the US alone.1,2 Atrial fibrillation is associated with increased stroke and systemic embolism rates and increased morbidity and mortality.1 Anticoagulant treatment reduces the risk of stroke by approximately 65% in patients with nonvalvular AF.3 Almost one-half of patients at risk of experiencing stroke do not start, and a similar proportion do not continue, to receive anticoagulant treatment and experience preventable strokes.4-7 This gap in care is likely multifactorial. It may reflect misunderstandings about the risk of stroke or the association of that risk with anticoagulant treatment (warfarin or direct oral anticoagulant [DOAC] medications), or it may result from concerns about bleeding, activity, diet and drug interactions, anticoagulant treatment reversal, out-of-pocket costs, or the need for periodic monitoring. Some patients may not be able to use anticoagulant medications safely and consistently.

    In 2014, 3 major cardiovascular organizations formulated guidelines and issued a class 1 recommendation for the use of shared decision-making (SDM) to individualize the anticoagulant treatment of patients with nonvalvular AF who are at risk of experiencing stroke.8 To implement this recommendation, several tools to facilitate SDM among patients with AF have been developed.9-13 However, most of these tools have not been rigorously evaluated, omit DOAC medications, present outdated data, do not directly support the patient-clinician conversation, or do not address practical considerations that are important to the success of ongoing safe anticoagulant treatment, such as leisure activities, diet, travel, and out-of-pocket costs.9,11,14-16

    To address these limitations and support at-risk patients with AF and their clinicians in making decisions about anticoagulant treatment, we developed the Anticoagulation Choice Shared Decision Making tool.17,18 The aim of the current study was to assess the extent to which the use of the Anticoagulation Choice Shared Decision Making tool affects the quality of SDM and anticoagulant treatment decisions in patients with AF who are at risk of experiencing stroke.

    Methods
    Design, Setting, and Participants

    This encounter-level multicenter randomized clinical trial compared the use of standard care during the clinical encounter with the use of the Anticoagulation Choice Shared Decision Making tool (which presents individualized risk estimates and compares anticoagulant treatment options across issues of importance to patients) during the clinical encounter to examine the effects of the 2 approaches on SDM and clinical outcomes. This report addresses the outcome data collected during and immediately following the index clinical encounter. The institutional review boards at the coordinating center (Mayo Clinic) and other participating sites (Hennepin Health, Park Nicollet Health Partners, the University of Alabama at Birmingham, and the University of Mississippi Medical Center) approved the study procedures, and the study protocol for the clinical trial was published previously.17 The trial protocol is available in Supplement 1. Written informed consent was obtained from all participants. The study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline for randomized clinical trials.

    The clinical trial took place in emergency and inpatient hospital departments and outpatient safety-net, primary care, and cardiology clinics at US academic medical centers. Participant recruitment began at an academic medical center (Mayo Clinic), a suburban group practice (Park Nicollet Health Partners), and an urban safety-net health system (Hennepin Health) in Minnesota in January 2017. The University of Alabama at Birmingham and the University of Mississippi Medical Center joined the study in December 2018.

    All clinicians at the participating sites who regularly had conversations about anticoagulant treatment with patients with AF were eligible for participation. Participating clinicians provided written informed consent before enrolling patients. Adult patients (aged ≥18 years) were eligible for participation if they were able to read and understand the informed consent document, had a diagnosis of nonvalvular AF, and were at high risk of experiencing a thromboembolic event. The risk of a thromboembolic event was measured using the CHA2DS2-VASc score (congestive heart failure, hypertension, age ≥75 years, diabetes, previous stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, and sex category; score range, 0-9, with higher scores indicating higher risk); a CHA2DS2-VASc score of 1 or more for men and 2 or more for women indicated high risk.

    Patients were classified into 1 of 2 cohorts: start and review. The start cohort comprised patients who were new to the receipt of anticoagulant treatment (ie, had not received anticoagulant treatment within the past 6 months). During the clinical trial, the practice of prescribing a DOAC medication in the emergency department and referring the patient to an outpatient clinic for full discussion of anticoagulant treatment led us to modify the protocol by including in the start cohort all patients who had been prescribed an anticoagulant medication within 10 days of the clinical trial encounter but were otherwise new to the receipt of anticoagulant treatment. The review cohort comprised patients who were currently receiving ongoing anticoagulant medication or who had received anticoagulant medication within the past 6 months.

    Randomization and Intervention

    Encounters were randomized on a 1:1 ratio to either standard care or care that included use of the SDM tool, which allowed clinicians to participate in both study arms. The randomization algorithm (generated within the Remote Data Capture [REDCap] software system; Vanderbilt University), which was built by the clinical trial statistician (M.E.B.), used a stratified block randomization with blocks of random size. The clinical trial was stratified by medical center, cohort (start vs review), and stroke risk (CHA2DS2-VASc score of 1 for men and 2 for women vs >1 for men and >2 for women).

    In the standard care arm, clinical encounters were conducted according to the clinicians’ usual approach. In the intervention arm, clinicians were asked to use the Anticoagulation Choice Shared Decision Making tool in their encounters. This tool is a freely available online conversation aid that is designed for use within the encounter.19,20 The tool calculates the patient’s risk of stroke using the CHA2DS2-VASc score21 and provides the patient’s individualized risk of experiencing stroke at 1 year or 5 years, with and without anticoagulant treatment, using natural frequency expressions (eg, “out of 100 people like you”) and 100-person pictographs that illustrate the proportion of people experiencing nondisabling strokes, disabling or fatal strokes, or no such events. The tool then supports the comparison of available anticoagulant treatment options (ie, warfarin and DOAC medications) across patient-important issues, such as how to use the medications, the need for periodic monitoring, the reversibility of anticoagulant treatment, the estimated out-of-pocket costs, and the association of lifestyle or medical factors with the risk of bleeding (using the HAS-BLED [hypertension, abnormal kidney or liver function, stroke, bleeding, labile international normalized ratio, elderly age (>65 years), and drug or alcohol use] estimator; score range, 0-9, with higher scores indicating higher risk22). The tool offers a patient report and tailored text that can be copied into the clinical note to document the conversation and the decision. Participating clinicians at each site completed a training session with a study coordinator, including an overview of the Anticoagulation Choice Shared Decision Making tool and a video tutorial about its intended use.

    Outcomes

    Clinicians completed a baseline survey at enrollment, and both patients and clinicians completed a survey immediately after the clinical encounter (eMethods in Supplement 2). The survey captured patients’ sociodemographic characteristics, health literacy (measured by a series of screening questions, with inadequate health literacy defined as a patient self-report of being “not at all” or “a little bit” confident in filling out medical forms without assistance),23 and subjective numeracy (measured by the Subjective Numeracy Scale; score range, 1-6, with higher scores indicating higher subjective numeracy).24 With the participant’s written consent, the encounter was recorded (either audiovisual or audio only).

    Participant-Reported Outcomes

    The primary outcome was the quality of SDM, a multidimensional concept that requires high-quality communication, effective knowledge transfer to the patient, agreement between the patient and the clinician on the course of action selected at the end of the encounter, and satisfaction with the decision-making process. The Consumer Assessment of Healthcare Providers and Systems Clinician and Group Survey was used to assess the quality of communication.25 Each item was coded as yes (definitely or somewhat) or no. Six questions about AF and anticoagulant treatment were used to assess knowledge transfer. To assess the accuracy of patients’ estimations of their own stroke risk, we asked patients to provide the number of people like them (out of 100 people) who they perceived could be expected to have a stroke within the next year. We considered a correct response any answer that was within either 10% (strict threshold) or 30% (liberal threshold) of the respondent’s actual CHA2DS2-VASc risk score. Comparison of the patient’s and clinician’s reported course of action was used to assess decision concordance.

    Decisional satisfaction was assessed using the Decisional Conflict Scale (score range, 0-100, with higher scores indicating greater decisional conflict), which reflected the degree of uncertainty about the choice.26 Participants indicated, on a 7-point Likert scale (with higher scores indicating stronger recommendation), the extent to which they would recommend the approach used in the encounter to other patients and clinicians. Clinicians indicated, on a 5-point Likert scale (with higher scores indicating greater satisfaction), the extent to which they were satisfied with their conversation with the patient. Each question was converted to a binary response of strongly recommend (6-7 points) or completely satisfied (4-5 points), respectively.

    Observed Encounter Outcomes

    After training and documentation of reliability, reviewers from the study team, working independently and in duplicate, used the Observing Patient Involvement in Decision Making 12-item (OPTION12) scale (score range, 0-100, with 0 indicating minimal behavior and 100 indicating maximal behavior) to code clinicians’ behavior to involve patients in decision-making.27 Interrater reliability (using the Lin concordance correlation coefficient [CCC]) was verified at baseline, at 33% of encounter recordings, and at 66% of encounter recordings (Lin CCC range, 0.84%-0.96%).

    Similarly, reviewers coded user fidelity (ie, use of the conversation aid as intended) using an ad hoc scale with adequate reliability (ie, κ interrater agreement of 0.8-1.0 across items). On-site study coordinators captured the duration (measured in minutes) of the full encounter. When unavailable, the duration documented in the recording of the encounter was used.

    Statistical Analysis

    The study was conducted and analyzed according to the intention-to-treat principle, which included all encounters in the arm to which they were randomly assigned. The primary analysis was conducted at the encounter level using mixed-effects models that were adjusted by arm, cohort (start vs review), and stroke risk (CHA2DS2-VASc score of 1 for men and 2 for women vs >1 for men and >2 for women), with the random effect of clinic and clinician.28 The 15 participants who correctly answered 3 or fewer knowledge questions were grouped into 1 category, with the remaining participants grouped into 3 categories based on the exact number of correct answers (4, 5, or 6 answers). Knowledge was modeled as correct responses (successes) out of 6 questions as a mixed-effects logistic regression analysis with a family of binomial distribution, adjusted by arm, cohort, and stroke risk. Assumptions for all models were verified, with no deviations found.

    The clinical trial recruitment goal was 1000 patient encounters (500 encounters per arm) based on previously reported sample size estimations.17 As conducted, the clinical trial produced estimates of between-arm differences with a margin of error (one-half of the 95% CI) of 1.9 (out of 100) for the Decisional Conflict Scale, 1.4 (out of 100) for the OPTION12 scale, and 1.4 minutes for the encounter duration. For binary outcomes, the margin of error ranged from 3% to 9%, with patient knowledge estimates having a margin of error of 14%.

    Missing outcome data owing to the nonreturn of surveys or incomplete survey responses occurred in 2% to 7% of encounters across outcomes. According to the statistical plan, outcomes associated with encounter recordings were not imputed because we considered it inappropriate to assume patients or clinicians chose not to be recorded at random. Multiple imputation was conducted with data that were treated as missing at random,29 data with 5 imputations, data with relative efficiency ranging from 98% to 99%, and data within range values.

    The exploration of heterogeneity of treatment effect included all SDM outcomes using the χ2 test statistic of the differences in the log likelihood to test interactions by type of clinic (academic, community, or safety net), by cohort (start or review), by stroke risk (CHA2DS2-VASc score of >1 for men or >2 for women), and by numeracy (mean score on the Subjective Numeracy Scale of ≤4 points or >4 points, with >4 points considered adequate numeracy). The Benjamini-Hochberg method was used to account for multiple comparisons.30 All tests were 2-sided and unpaired, and data analysis was conducted using Stata statistical software, version 15 (StataCorp LLC). Data were analyzed from August 1 to November 30, 2019.

    A data and safety monitoring board met before study initiation to approve the board’s charter and met biannually thereafter. The board monitored study conduct, data quality, and safety signals, although no interim efficacy analyses were planned or conducted. On review of the results, the board released the data for publication.

    Results
    Participant Characteristics

    Study recruitment occurred from January 30, 2017, to June 27, 2019. Patients and clinicians consented to participate in 942 of the 1827 eligible encounters (52%). A total of 922 patients and 244 clinicians were enrolled and included in the analyses. Among patients, 463 individuals (291 men [62.9%]; mean [SD] age, 71 [11] years) were randomized to the intervention arm, and 459 individuals (268 men [58.4%]; mean [SD] age, 71 [10] years) were randomized to the standard care arm. Among clinicians, 112 of 222 individuals (50.5%) were women, with a mean (SD) age of 43 (12) years (Table 1, Table 2, and Figure).

    All patient factors were balanced across arms. Because no significant interactions were found across any of the planned subgroup analyses (eTable 1 in Supplement 2), including no differential effects by cohort (start vs review) or clinic, the clinical trial results are presented for the whole cohort.

    Participant-Reported Outcomes

    Almost all patients in both arms reported that the clinician showed respect (426 of 428 patients [99.5%] in the intervention arm and 427 of 427 patients [100%] in the standard care arm), listened carefully (428 of 430 patients [99.5%] in the intervention arm and 427 of 427 patients [100%] in the standard care arm), and used terms that were easy to understand (431 of 432 patients [99.8%] in the intervention arm and 422 of 425 patients [99.3%] in the standard care arm) during the encounter. Patients in both arms (345 of 445 patients [77.5%] in the intervention arm and 315 of 433 patients [72.7%] in the standard care arm; P = .15) correctly answered most questions (5 or 6 correct responses of 6 total questions) about anticoagulant treatment for AF (Table 3).

    No significant difference was observed between study arms with regard to patients’ accuracy of their own perceived risk of stroke using both the strict threshold (30 of 445 patients [6.7%] in the intervention arm vs 24 of 434 patients [5.1%] in the standard care arm; adjusted relative risk [aRR], 1.4; 95% CI, 0.8-2.2) and the liberal threshold (49 of 445 patients [11.0%] in the intervention arm vs 40 of 434 patients [9.2%] in the standard care arm; aRR, 1.3; 95% CI, 0.8-1.8). Decisional conflict was low (Decisional Conflict Scale unadjusted mean [SD] score, 16.6 [14.4] points in the intervention arm and 17.9 [14.9] points in the standard care arm), and patient-clinician concordance about treatment selection was high in both arms (381 of 434 patients [87.8%] in the intervention arm vs 369 of 424 patients [87.0%] in the standard care arm), with no significant between-arm differences (aRR, 1.0; 95% CI, 0.9-1.1).

    Patients would similarly recommend the communication approach used during the clinical encounter across clinical trial arms (390 of 429 encounters [90.9%] in the intervention arm and 378 of 425 encounters [88.9%] in the standard care arm). More clinicians were satisfied with the encounter in the intervention arm (400 of 453 encounters [88.3%]) compared with the standard care arm (277 of 448 encounters [61.8%]; aRR, 1.49; 95% CI, 1.42-1.53), and they were more likely to recommend using the SDM tool (396 of 453 encounters [87.4%]) than the standard care approach (199 of 448 encounters [44.4%]; aRR, 2.1; 95% CI, 2.0-2.2) to their colleagues. These results indicated good performance relative to the use of a per-protocol analyses or multiple imputation analyses for missing data rather than an intention-to-treat analysis (eTable 2 in Supplement 2).

    Observed Encounter Outcomes

    Clinician involvement of patients in decision-making about anticoagulant treatment was significantly greater in the intervention arm compared with the standard care arm (OPTION12 mean [SD] score, 33.0 [10.8] points vs 29.1 [13.1] points, respectively; adjusted mean between-arm difference, 4.2 points; 95% CI, 2.8-5.6 points) (Table 4). Clinicians used the SDM tool with high fidelity (mean [SD] score, 5.6 [1.4] points of 7.0 possible points). However, the conversation focused first on the issue (eg, risk of bleeding, need for monitoring, or costs) identified by the patient as the highest priority in only 53 of 419 encounters [12.7%] in which the SDM tool was used. No significant difference was found in the duration of encounters between the intervention and standard care arms (mean [SD] duration, 32 [16] minutes vs 31 [17] minutes, respectively; adjusted mean between-arm difference, 1.1-minute; 95% CI, −0.3 to 2.5 minutes).

    Overall, a median of 2 patients (interquartile range [IQR], 1-6 patients; range, 1-76 patients) were enrolled per clinician, with a median of 1 patient (range, 1-38 patients) per clinician in the standard care arm and 2 patients (range, 1-44 patients) per clinician in the intervention arm. Of 151 clinicians who had encounters with 1 or more patients enrolled in the study, 68 clinicians enrolled patients in both arms of the clinical trial (median, 7 patients per clinician; IQR, 3-14 patients per clinician). Minimal data were found to indicate clinicians’ use of the SDM tool in the standard care arm or contamination of SDM behaviors between arms (data not shown).

    Discussion
    Main Findings

    This encounter-level multicenter randomized clinical trial found that adding an SDM tool to standard care during clinical encounters with patients with AF improved several aspects of SDM quality without significantly affecting anticoagulant treatment decisions or lengthening the duration of the encounters. Clinicians who used the SDM tool were significantly more likely to engage patients in SDM and to be more satisfied with the encounters in which they used the SDM tool.

    To date, 3 clinical trials have tested the effect of using SDM tools to facilitate real-life decisions about anticoagulant treatment in patients with AF; those clinical trials yielded inconsistent results with regard to patient knowledge, decisional conflict, and anticoagulant treatment choices.9 Compared with those studies, the present clinical trial combined the evaluation of an SDM tool that supported both patients and clinicians in deciding how to prevent strokes (including the option to receive DOAC medications), rather than supporting patients alone, with an assessment of the intervention’s effectiveness through the use of recorded encounters.

    Results from the present clinical trial are consistent with previous SDM clinical trials conducted by the Knowledge and Evaluation Research Unit at Mayo Clinic,32-34 with the exception of the lack of significant improvements in patient knowledge or decisional conflict (which was nearly optimal at baseline). Compared with the findings of 105 randomized clinical trials of SDM interventions that were included in a 2017 Cochrane systematic review,10 the present clinical trial yielded similar results, with the exception of no significant improvements in patient knowledge or decisional conflict (both were reported to improve in the review) and no significant change in encounter duration (reported to lengthen by 3 minutes in the review). Overall, the present study is, to our knowledge, one of the largest SDM clinical trials conducted and one of the few to intervene in the clinical encounter and directly observe SDM behaviors, fidelity of use, and contamination.

    Implications

    This clinical trial demonstrates a feasible and acceptable approach to implementing an SDM tool to guide anticoagulant treatment discussions during clinical encounters with patients with AF in diverse practice settings, and it provides credible estimates of the efficacy of the SDM approach. This approach also offers a way to bring risk assessment (in this case, through use of the CHA2DS2-VASc score) into the patient-centered decision-making process.35 During the clinical trial, the SDM recommendation for encounters with patients with AF shifted from discussing warfarin and DOAC medications as treatment choices to discussing only DOAC medications as treatment choices.36 This change may have reduced the scope of SDM for decisions about how to treat patients for anticoagulation and may have limited SDM application to the decision of whether to treat patients for anticoagulation. Further research may be necessary to understand the extent to which current anticoagulant treatment decisions are inappropriate across clinical and patient factors.

    Whether a more selective implementation approach could yield larger effects remains unclear and deserves examination. Such an examination may need to focus on patients who may find it difficult to decide whether or how to use anticoagulant treatments, such as patients with low to intermediate stroke risk, patients who have experienced difficulty maintaining therapeutic international normalized ratio levels, or patients who find DOAC medications unaffordable. Some explorations are currently occurring in an ongoing multicenter randomized clinical trial funded by the American Heart Association and the Patient Centered Outcomes Research Institute that compares standard care, an SDM encounter tool, a patient decision aid, or a combination of these options.37

    The findings of this clinical trial contribute to the discussion of the recommendations for SDM in published guidelines, which inspired this clinical trial, and the mandated use of SDM by payers as a requisite for reimbursement. Both practices assume that clinicians and health care systems can implement forms of SDM that are capable of responding effectively to the problematic situation for which patients seek care.38 The finding that the patient’s highest priority led the discussion in only a limited number of SDM encounters challenges this assumption. There is a clinical and ethical need for patients and clinicians to work together to form plans of care. How best to do so remains to be determined. Furthermore, it remains necessary to develop ways of identifying which patients, decisions, encounters, and clinicians need more support to enact which form of SDM.39 The results of the present study suggest it may be premature to proceed with a wholesale implementation of SDM tools that is sustained by mandates or financial incentives.

    Strengths and Limitations

    This study had several strengths. Some aspects of the conduct of the clinical trial contribute to the credibility of its findings. The implementation of allocation concealment and adherence to the intention-to-treat principle in the conduct and analyses of the study mitigated the intrusion of bias. Video review demonstrated that, in most cases, the clinicians used the intervention correctly, and no substantial contamination occurred. This clinical trial, one of the largest clinical trials of SDM to date,10 has yielded precise estimates of effect.

    The study also had several limitations. Some features of the clinical trial may have contributed to an overestimation of the effect of the SDM intervention. Selection bias could have been introduced when enrolled clinicians chose not to enroll an eligible patient encounter into the clinical trial. In addition, bias may have affected the unblinded assessment of recorded encounters and the scoring of those encounters using the OPTION12 scale. Some features of the clinical trial may alternately have produced underestimations of the effect of the SDM intervention. Based on the results found in the standard care arm, participants in the present study were particularly competent at implementing SDM compared with those in previous studies.40 In addition, patients who were already receiving anticoagulant treatment (ie, those enrolled in the review cohort), who comprised most of the clinical trial participants, may have been generally satisfied with their current anticoagulant treatment regimen or may have had no difficulty deciding on their best course of action (ie, they already had high levels of knowledge and low levels of decisional conflict at baseline). Both factors may have limited the potential contribution of the SDM tool, although no significant treatment-outcome interactions by clinic or cohort (start vs review) were found.

    Conclusions

    The use of an encounter tool to foster and support SDM resulted in improvements in several aspects of SDM quality and clinician satisfaction, with no significant effect on treatment decisions or encounter duration. These results question the view that implementing SDM tools for anticoagulant treatment can improve care for patients with AF.

    Back to top
    Article Information

    Accepted for Publication: May 27, 2020.

    Published Online: July 20, 2020. doi:10.1001/jamainternmed.2020.2908

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Kunneman M et al. JAMA Internal Medicine.

    Corresponding Author: Victor M. Montori, MD, MSc, Knowledge and Evaluation Research Unit, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (montori.victor@mayo.edu).

    Author Contributions: Ms Branda and Dr Montori had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Kunneman, Branda, Hargraves, Sivly, Burnett, Suzuki, Linzer, Brand-McCarthy, Brito, Noseworthy, Montori.

    Acquisition, analysis, or interpretation of data: Kunneman, Branda, Sivly, Lee, Gorr, Suzuki, Jackson, Hess, Linzer, Brito, Montori.

    Drafting of the manuscript: Kunneman, Hargraves, Sivly, Jackson, Brand-McCarthy, Noseworthy, Montori.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Branda, Lee, Montori.

    Obtained funding: Montori.

    Administrative, technical, or material support: Hargraves, Sivly, Gorr, Burnett, Suzuki, Hess, Brand-McCarthy, Noseworthy, Montori.

    Supervision: Hargraves, Gorr, Suzuki, Jackson, Hess, Brand-McCarthy, Brito, Noseworthy, Montori.

    Conflict of Interest Disclosures: Dr Kunneman reported receiving grants from the National Heart, Lung, and Blood Institute during the conduct of the study. Ms Branda reported receiving grants from the National Heart, Lung, and Blood Institute during the conduct of the study. Dr Hargraves reported receiving grants from the National Institutes of Health during the conduct of the study. Ms Sivly reported receiving grants from Mayo Clinic during the conduct of the study and outside the submitted work. Dr Gorr reported receiving grants from the National Institutes of Health during the conduct of the study. Dr Burnett reported receiving grants and personal fees from the Mayo Clinic and the National Institutes of Health during the conduct of the study and personal fees from the Mayo Clinic outside the submitted work. Dr Jackson reported receiving grants from the Mayo Clinic during the conduct of the study and research funding from Amgen and the National Institutes of Health outside the submitted work. Dr Hess reported receiving grants from the Patient-Centered Outcomes Research Institute outside the submitted work. Dr Linzer reported receiving grants from the National Institutes of Health during the conduct of the study and grants from the American College of Physicians, the American Medical Association, and the Institute for Healthcare Improvement outside the submitted work. Dr Brito reported being the medical director of the Shared Decision Making National Resource Center at the Mayo Clinic. Dr Montori reported receiving grants from the National Heart, Lung, and Blood Institute during the conduct of the study and serving as board chair of The Patient Revolution outside the submitted work. No other disclosures were reported.

    Funding/Support: The clinical trial was funded by grant RO1 HL131535-01 from the National Heart, Lung, and Blood Institute of the National Institutes of Health.

    Role of the Funder/Sponsor: The funding source had no influence on the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Shared Decision Making for Atrial Fibrillation (SDM4AFib) Trial Investigators: steering committee: Principal investigator: Victor Montori, MD, MSc; study statistician: Megan E. Branda, MS; coinvestigators: Juan P. Brito, MD, MSc, Marleen Kunneman, PhD, and Ian Hargraves, PhD; study coordinator: Angela L. Sivly, CCRP; study manager: Kirsten Fleming, BA; site principal investigators: Bruce Burnett, MD (Park Nicolette Health Partners, Minneapolis, Minnesota), Mark Linzer, MD, and Haeshik Gorr, MD (Hennepin Health, Minneapolis, Minnesota), Elizabeth Jackson, MD, and Erik Hess, MD (University of Alabama at Birmingham), Takeki Suzuki, MD, MPH, PhD, and James Hamilton IV, MD (University of Mississippi Medical Center, Jackson), and Peter A. Noseworthy, MD (Mayo Clinic, Rochester, Minnesota). Site teams (in alphabetical order): Hennepin Health: Haeshik Gorr, MD, Alexander Haffke, BS, Mark Linzer, MD, Jule Muegge, MPH, Sara Poplau, BA, Benjamin Simpson, BS, Miamoua Vang, BS, and Mike Wambua, MS. Mayo Clinic: Joel Anderson, MPH, Emma Behnken, BA, Fernanda Bellolio, MD, Juan P. Brito, MD, MSc, Renee Cabalka, BS, Michael Ferrara, MS, Kirsten Fleming, BA, Rachel Giblon, MS, Ian Hargraves, PhD, Jonathan Inselman, MS, Marleen Kunneman, PhD, Annie LeBlanc, PhD, Alexander Lee, BS, Victor Montori, MD, MSc, Peter A. Noseworthy, MD, Marc Olive, BA, Paige Organick, BA, Nilay Shah, PhD, Angela Sivly, CCRP, Gabriela Spencer-Bonilla, MD, Amy Stier, CNP, Anjali Thota, BA, Henry Ting, MD, Derek Vanmeter, and Claudia Zeballos-Palacios, MD. Park Nicollet Health Partners: Carol Abullarade, RN-BC, CACP, Bruce Burnett, MD, Lisa Harvey, RD, and Shelly Keune, BSN, RN. University of Alabama at Birmingham: Elizabeth Jackson, MD, Erik Hess, MD, MSc, Timothy Smith, CRNP, and Shannon Stephens, EMTP. University of Mississippi Medical Center: Bryan Barksdale, MA, James Hamilton IV, MD, Theresa Hickey, Roma Peters, FNP, Memrie Price, BSc, Takeki Suzuki, MD, MPH, PhD, Connie Watson, BA, and Douglas Wolfe, MD. Data safety and monitoring board: Gordon Guyatt, MD (chair), Brian Haynes, MD, and George Tomlinson, PhD. Expert advisory panel: Paul Daniels, MD, Bernard Gersh, MD, Erik Hess, MD, MSc, Thomas Jaeger, MD, Robert McBane, MD, and Peter A. Noseworthy, MD (chair).

    Data Sharing Statement: See Supplement 3.

    Additional Contributions: The investigators thank all of the patients, caregivers, clinicians, study coordinators, patient and expert advisors, and members of the data monitoring board who kindly and enthusiastically made this clinical trial possible.

    References
    1.
    Morillo  CA, Banerjee  A, Perel  P, Wood  D, Jouven  X.  Atrial fibrillation: the current epidemic.   J Geriatr Cardiol. 2017;14(3):195-203.PubMedGoogle Scholar
    2.
    Colilla  S, Crow  A, Petkun  W, Singer  DE, Simon  T, Liu  X.  Estimates of current and future incidence and prevalence of atrial fibrillation in the US adult population.   Am J Cardiol. 2013;112(8):1142-1147. doi:10.1016/j.amjcard.2013.05.063 PubMedGoogle ScholarCrossref
    3.
    Tawfik  A, Bielecki  JM, Krahn  M,  et al.  Systematic review and network meta-analysis of stroke prevention treatments in patients with atrial fibrillation.   Clin Pharmacol. 2016;8:93-107. doi:10.2147/CPAA.S105165 PubMedGoogle Scholar
    4.
    Gallagher  AM, Rietbrock  S, Plumb  J, van Staa  TP.  Initiation and persistence of warfarin or aspirin in patients with chronic atrial fibrillation in general practice: do the appropriate patients receive stroke prophylaxis?   J Thromb Haemost. 2008;6(9):1500-1506. doi:10.1111/j.1538-7836.2008.03059.x PubMedGoogle ScholarCrossref
    5.
    Ogilvie  IM, Newton  N, Welner  SA, Cowell  W, Lip  GYH. Underuse of oral anticoagulants in atrial fibrillation: a systematic review. Am J Med. 2010;123(7):638-645.
    6.
    Yao  X, Abraham  NS, Alexander  GC,  et al.  Effect of adherence to oral anticoagulants on risk of stroke and major bleeding among patients with atrial fibrillation.   J Am Heart Assoc. 2016;5(2):e003074. doi:10.1161/JAHA.115.003074 PubMedGoogle Scholar
    7.
    Hernandez  I, He  M, Chen  N, Brooks  MM, Saba  S, Gellad  WF.  Trajectories of oral anticoagulation adherence among Medicare beneficiaries newly diagnosed with atrial fibrillation.   J Am Heart Assoc. 2019;8(12):e011427. doi:10.1161/JAHA.118.011427 PubMedGoogle Scholar
    8.
    January  CT, Wann  LS, Alpert  JS,  et al; American College of Cardiology/American Heart Association Task Force on Practice Guidelines.  2014 AHA/ACC/HRS guideline 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 Heart Rhythm Society.   J Am Coll Cardiol. 2014;64(21):e1-e76. doi:10.1016/j.jacc.2014.03.022 PubMedGoogle ScholarCrossref
    9.
    O’Neill  ES, Grande  SW, Sherman  A, Elwyn  G, Coylewright  M.  Availability of patient decision aids for stroke prevention in atrial fibrillation: a systematic review.   Am Heart J. 2017;191:1-11. doi:10.1016/j.ahj.2017.05.014 PubMedGoogle ScholarCrossref
    10.
    Stacey  D, Legare  F, Lewis  K,  et al.  Decision aids for people facing health treatment or screening decisions.   Cochrane Database Syst Rev. 2017;4(4):CD001431. doi:10.1002/14651858.CD001431.pub5 PubMedGoogle Scholar
    11.
    Clarkesmith  DE, Lip  GYH, Lane  DA.  Patients’ experiences of atrial fibrillation and non–vitamin K antagonist oral anticoagulants (NOACs), and their educational needs: a qualitative study.   Thromb Res. 2017;153:19-27. doi:10.1016/j.thromres.2017.03.002 PubMedGoogle ScholarCrossref
    12.
    Thomson  RG, Eccles  MP, Steen  IN,  et al.  A patient decision aid to support shared decision-making on anti-thrombotic treatment of patients with atrial fibrillation: randomised controlled trial.   Qual Saf Health Care. 2007;16(3):216-223. doi:10.1136/qshc.2006.018481 PubMedGoogle ScholarCrossref
    13.
    Eckman  MH, Costea  A, Attari  M,  et al.  Shared decision-making tool for thromboprophylaxis in atrial fibrillation—a feasibility study.   Am Heart J. 2018;199:13-21. doi:10.1016/j.ahj.2018.01.003 PubMedGoogle ScholarCrossref
    14.
    Fatima  S, Holbrook  A, Schulman  S, Park  S, Troyan  S, Curnew  G.  Development and validation of a decision aid for choosing among antithrombotic agents for atrial fibrillation.   Thromb Res. 2016;145:143-148. doi:10.1016/j.thromres.2016.06.015 PubMedGoogle ScholarCrossref
    15.
    Saposnik  G, Joundi  RA.  Visual aid tool to improve decision making in anticoagulation for stroke prevention.   J Stroke Cerebrovasc Dis. 2016;25(10):2380-2385. doi:10.1016/j.jstrokecerebrovasdis.2016.05.037 PubMedGoogle ScholarCrossref
    16.
    Eckman  MH, Wise  RE, Naylor  K,  et al.  Developing an atrial fibrillation guideline support tool (AFGuST) for shared decision making.   Curr Med Res Opin. 2015;31(4):603-614. doi:10.1185/03007995.2015.1019608 PubMedGoogle ScholarCrossref
    17.
    Kunneman  M, Branda  ME, Noseworthy  PA,  et al.  Shared decision making for stroke prevention in atrial fibrillation: study protocol for a randomized controlled trial.   Trials. 2017;18(1):443. doi:10.1186/s13063-017-2178-y PubMedGoogle ScholarCrossref
    18.
    Zeballos-Palacios  CL, Hargraves  IG, Noseworthy  PA,  et al; Shared Decision Making for Atrial Fibrillation (SDM4AFib) Trial Investigators.  Developing a conversation aid to support shared decision making: reflections on designing anticoagulation choice.   Mayo Clin Proc. 2019;94(4):686-696. doi:10.1016/j.mayocp.2018.08.030 PubMedGoogle ScholarCrossref
    19.
    Montori  VM, Kunneman  M, Brito  JP.  Shared decision making and improving health care: the answer is not in.   JAMA. 2017;318(7):617-618. doi:10.1001/jama.2017.10168 PubMedGoogle ScholarCrossref
    20.
    Mayo Clinic. Anticoagulation choice decision aid. Accessed June 8, 2020. https://anticoagulationdecisionaid.mayoclinic.org/
    21.
    Lip  GYH, Nieuwlaat  R, Pisters  R, Lane  DA, Crijns  HJGM.  Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor–based approach: the Euro Heart Survey on Atrial Fibrillation.   Chest. 2010;137(2):263-272. doi:10.1378/chest.09-1584 PubMedGoogle ScholarCrossref
    22.
    Pisters  R, Lane  DA, Nieuwlaat  R, de Vos  CB, Crijns  HJGM, Lip  GYH.  A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey.   Chest. 2010;138(5):1093-1100. doi:10.1378/chest.10-0134 PubMedGoogle ScholarCrossref
    23.
    Chew  LD, Bradley  KA, Boyko  EJ.  Brief questions to identify patients with inadequate health literacy.   Fam Med. 2004;36(8):588-594.PubMedGoogle Scholar
    24.
    Fagerlin  A, Zikmund-Fisher  BJ, Ubel  PA, Jankovic  A, Derry  HA, Smith  DM.  Measuring numeracy without a math test: development of the Subjective Numeracy Scale.   Med Decis Making. 2007;27(5):672-680. doi:10.1177/0272989X07304449 PubMedGoogle ScholarCrossref
    25.
    Consumer Assessment of Healthcare Providers and Systems. CAHPS research on survey design and administration. Agency for Healthcare Research and Quality. Updated January 2020. Accessed April 21, 2017. https://www.ahrq.gov/cahps/surveys-guidance/survey-methods-research/index.html
    26.
    O’Connor  AM.  Validation of a decisional conflict scale.   Med Decis Making. 1995;15(1):25-30. doi:10.1177/0272989X9501500105 PubMedGoogle ScholarCrossref
    27.
    Elwyn  G, Hutchings  H, Edwards  A,  et al.  The OPTION scale: measuring the extent that clinicians involve patients in decision-making tasks.   Health Expect. 2005;8(1):34-42. doi:10.1111/j.1369-7625.2004.00311.x PubMedGoogle ScholarCrossref
    28.
    Diggle  PJ, Heagerty P, Liang  K-Y, Zeger  SL.  Analysis of Longitudinal Data. Oxford University Press; 1994. Oxford Statistical Science Series; vol. 25.
    29.
    Rubin  DB.  Multiple Imputation for Nonresponse in Surveys. John Wiley & Sons; 1987. doi:10.1002/9780470316696
    30.
    Thissen  D, Steinberg  L, Kuang  D.  Quick and easy implementation of the Benjamini-Hochberg procedure for controlling the false positive rate in multiple comparisons.   J Educ Behav Stat. 2002;27(1):77-83. doi:10.3102/10769986027001077 Google ScholarCrossref
    31.
    Chew  LD, Griffin  JM, Partin  MR,  et al.  Validation of screening questions for limited health literacy in a large VA outpatient population.   J Gen Intern Med. 2008;23(5):561-566. doi:10.1007/s11606-008-0520-5 PubMedGoogle ScholarCrossref
    32.
    Weymiller  AJ, Montori  VM, Jones  LA,  et al.  Helping patients with type 2 diabetes mellitus make treatment decisions: statin choice randomized trial.   Arch Intern Med. 2007;167(10):1076-1082. doi:10.1001/archinte.167.10.1076 PubMedGoogle ScholarCrossref
    33.
    Mullan  RJ, Montori  VM, Shah  ND,  et al.  The diabetes mellitus medication choice decision aid: a randomized trial.   Arch Intern Med. 2009;169(17):1560-1568. doi:10.1001/archinternmed.2009.293 PubMedGoogle ScholarCrossref
    34.
    LeBlanc  A, Herrin  J, Williams  MD,  et al.  Shared decision making for antidepressants in primary care: a cluster randomized trial.   JAMA Intern Med. 2015;175(11):1761-1770. doi:10.1001/jamainternmed.2015.5214 PubMedGoogle ScholarCrossref
    35.
    Peterson  ED.  Machine learning, predictive analytics, and clinical practice: can the past inform the present?   JAMA. 2019; 322(23):2283-2284. doi:10.1001/jama.2019.17831 PubMedGoogle Scholar
    36.
    Beaser  AD, Cifu  AS.  Management of patients with atrial fibrillation.   JAMA. 2019;321(11):1100-1101. doi:10.1001/jama.2019.1264 PubMedGoogle ScholarCrossref
    37.
    Decision-Making and Choices to Inform Dialogue and Empower AFib Patients (DECIDE) Center. Project summary. Patient-Centered Outcomes Research Institute. Updated February 20, 2020. Accessed January 27, 2020. https://www.pcori.org/research-results/2018/decision-making-and-choices-inform-dialogue-and-empower-afib-patients-decide
    38.
    Kunneman  M, Gionfriddo  MR, Toloza  FJK,  et al.  Humanistic communication in the evaluation of shared decision making: a systematic review.   Patient Educ Couns. 2019;102(3):452-466. doi:10.1016/j.pec.2018.11.003 PubMedGoogle ScholarCrossref
    39.
    Hargraves  IG, Montori  VM, Brito  JP,  et al.  Purposeful SDM: a problem-based approach to caring for patients with shared decision making.   Patient Educ Couns. 2019;102(10):1786-1792. doi:10.1016/j.pec.2019.07.020 PubMedGoogle ScholarCrossref
    40.
    Couet  N, Desroches  S, Robitaille  H,  et al.  Assessments of the extent to which health-care providers involve patients in decision making: a systematic review of studies using the OPTION instrument.   Health Expect. 2015;18(4):542-561. doi:10.1111/hex.12054 PubMedGoogle ScholarCrossref
    ×