FormalPara Clinician’s capsule

What is known about the topic?

CAEP’s Best Practices Checklist for acute atrial fibrillation/flutter provides a framework for decisions around anticoagulation and disposition/follow-up.

What did this study ask?

How aligned are ED physicians with the anticoagulation and follow-up portions of the checklist?

What did this study find?

This study found that 6.4% of patients had one or more safety issues relating to stroke prevention and follow-up.

Why does this study matter to clinicians?

Identifying areas of discrepancy can guide the development of interventions to improve safety for patients presenting with atrial fibrillation/flutter.

Introduction

Acute atrial fibrillation (AF)/flutter (AFL) is a common emergency department (ED) presentation, with ED practices varying between countries, hospitals and individual providers [1, 2]. In 2021, an updated version of the Canadian Association of Emergency Physicians’ (CAEP) Acute AF/AFL Best Practices Checklist was published, seeking to improve ED management of these conditions [3].

Most patients in Canada who present with acute AF/AFL are managed in the ED and then discharged home, with recent studies showing a discharge rate of over 90% [4]. With many patients receiving a new diagnosis of acute AF/AFL, and some also having limited contact with the medical system, this is a critical moment for both stroke prevention and appropriate follow-up recommendations. While the CAEP Checklist provides recommendations on both topics, it is unclear what the actual physician alignment is with these recommendations and whether deviation from the checklist leads to any change in safety outcomes for patients.

Previous studies have explored reasoning for lack of compliance with stroke prevention guidelines and found an array of patient-related and physician-related factors that may have an impact [5]. With a goal of improving safety for patients presenting with acute AF/AFL to EDs, an assessment of current alignment with guidelines is necessary to identify gaps in care. Thus, our objective was to assess the alignment with and safety of application of the CAEP checklist as it pertains to stroke prevention, disposition, and follow-up recommendations. A complementary paper evaluated alignment with the first two parts of the checklist: initial assessment and rhythm and rate control [6].

Methods

Study design and setting

This was a health records review of all patients presenting to the ED with acute AF/AFL who were treated by an emergency physician and discharged home.

Study setting

Our study setting was two tertiary care academic EDs between January and August 2022. There are a total of 180,000 patient visits to these sites annually. Patients are managed by one of the 95 attending physicians and 55 emergency medicine resident physicians/fellows.

Participants

We included all adults 18 years or older who presented to the ED with acute AF/AFL, who were treated by the ED physician and were discharged home. Acute AF/AFL was defined as in the checklist as being cases of recent-onset AF/AFL with an onset generally less than 48 h but possibly up to seven days [3]. Patients were excluded for the following reasons: initial ECG-recorded heart rate less than 100 beats per minute, patient transferred from another center or directly to an admitted service, patient admitted to hospital or deceased, patient presentation not related to AF/AFL, or if they were ultimately not found to have acute AF/AFL. Our study was approved by the Ottawa Health Science Network Research Ethics Board.

Outcome measures

Our primary outcome was the proportion of patient encounters with one or more identified safety issues pertaining to the “Stroke Prevention” and the “Disposition and Follow-up” sections of the CAEP Checklist. Seven potential safety issues were identified and predetermined by a research team that included resident and attending ED physicians as well as members of the team that formulated the CAEP Checklist. The issues were categorized as either moderate or severe safety concerns. The identified severe issues are as follows: 1. Failure to prescribe anticoagulation when indicated, 2. Prescription of anticoagulation if contraindicated, 3. Prescription of a direct-acting oral anticoagulant (DOAC) if contraindicated, 4. Inappropriate dosing of a DOAC. The potential moderate safety concerns were: 1. Inappropriate prescription of warfarin over a DOAC, 2. Inappropriate prescription of rate or rhythm control medication, 3. Failure to recommend follow-up within 1 week for new warfarin prescription or rate control medication. Secondary outcomes included the proportion of individuals who were recommended follow-up, the number and reason for return ED visits within 30 days, and the number of adverse events within 30 days.

Data collection

Patient records were screened via an ECG database. This database contains all ECGs recorded in the ED that are all over-read by a staff cardiologist. ECGs were extracted from the database if a mention of “atrial fibrillation” or “atrial flutter” was found in either the preliminary machine-generated diagnosis or the formal diagnosis made by the reviewing cardiologist. The patient record was then reviewed to ascertain inclusion and exclusion criteria. For included patients, data were collected on demographics, history and physical exam during ED visit, ED treatment, disposition decisions, outpatient prescriptions, referrals and 30-day outcomes.

Data analysis and sample size

Following initial data collection, each case was assessed according to the seven predetermined potential safety issues. Any disagreement between independent reviewers were discussed with all team members are resolved by consensus. Cases with identified safety issues were then reviewed to assess for a relationship between the presence of safety issues and the occurrence of adverse events. We estimated that at least 300 patient visits were required to adequately depict the incidence of adverse events. Analysis consisted of descriptive statistics with 95% confidence intervals (CI) appropriate for continuous, ordinal, and categorical outcomes.

Results

A total of 2242 ECGs were screened for inclusion and 358 patients met inclusion criteria (Appendix 1 Flow Diagram). Details of patient demographics and ED course can be found in Table 1. At discharge, one third of patients (33.5%) were prescribed a new medication, with 99 individuals (27.7%) receiving a new anticoagulant prescription and 54 patients (15.1%) starting a new rate or rhythm control agent. Most patients (71.2%) were recommended follow-up. Thirty days following discharge, 24% of included patients had one or more ED return visits. Acute AF/AFL was the reason for return in most (73.3%) cases. Adverse events within 30 days included one stroke, one major bleeding event, and one minor bleeding event. Given that data was obtained from clinician documentation, there were rare instances of missing data of low significance. For example, in 24/358 patients, the timing since onset of symptoms was not evident in the patient chart and was documented as “unknown”.

Table 1 Patient demographics, ED course, prescriptions and referrals, 30-day follow-up, and safety criteria outcomes (n = 358)
Full size table

The primary outcome of safety concern was identified in 6.4% (95% CI 4.3–9.5) of encounters, representing 28 safety issues in 23 individuals. The severe safety issues identified were: failure to prescribe anticoagulation when indicated (n = 6) and inappropriate dosing of a DOAC (n = 2). Moderate issues identified included: inappropriate prescription of rate or rhythm control medication (n = 9) and failure to recommend appropriately timed follow-up for new warfarin or rate control medication (n = 11). All nine cases with inappropriate rate/rhythm control prescriptions were patients who had converted to sinus rhythm but were still prescribed a beta blocker or calcium channel blocker on discharge. Where a safety concern was identified; a patient prescribed a subtherapeutic dose of a DOAC subsequently suffered a stroke.

Discussion

Interpretation

There was overall very good alignment with the 2021 CAEP Acute AF/AFL Best Practices Checklist, with one or more safety issues being identified in 6.4% of cases. The most common safety issues were the omission of indicated anticoagulation, the inappropriate prescription of rate control agents, and a failure to ensure appropriate follow-up after prescribing these medications. We found that about a quarter of patients included in our study returned to the ED within 30 days, with AF/AFL being their most frequent reason for returning. Within 30 days of discharge, there were very few adverse events.

Previous studies

While no study to our knowledge has assessed ED physician alignment with the CAEP checklist, previous studies have supported the initiation of anticoagulation for stroke prevention directly in the ED. It is well established that when indicated, anticoagulation significantly decreases stroke risk [7]. Atzema et al. demonstrated a substantially higher oral anticoagulant use at six months post ED visit when this was prescribed during the ED visit as opposed to later by another provider [8]. The importance of prompt initiation of anticoagulation is emphasized in the checklist and in national guidelines [3, 9].

A complementary paper from our institution evaluated alignment with recommendations for initial assessment as well as rate and rhythm control, but did not assess stroke prevention and disposition [6].

Strengths and limitations

This is the first paper to evaluate ED physician alignment with the CAEP checklist for stroke prevention and disposition. A major strength of this study was the large number of included patients as we included all patients with acute AF/AFL over a prolonged period of eight months, allowing for a valuable assessment of current practices at the studied institutions. A limitation, given this study’s health records design, is our reliance on chart documentation, which may lead to overestimated misalignment with the checklist. For instance, shared decision-making discussions and informal follow-up recommendations may not always be detailed in a patient’s chart. Of the six patients who were not prescribed anticoagulation when indicated, one was recommended a DOAC; however, it was documented that the patient had refused. It is not possible to know if a similar situation occurred for other patients included in our study. Second, while the checklist is well-known at the study institutions via faculty-wide Grand Rounds and emails, access to a mobile app, and high visibility posters around the ED, it is not possible to know whether physicians were consistently using it or if their treatment decisions were based on their usual practice.

Clinical implications

Our study showed overall high physician alignment with the CAEP checklist at two ED sites. There remain important areas where safety can be further improved within the stroke prevention and follow-up portions of AF/AFL management. Specifically, efforts can be made to verify dosing for DOAC prescriptions and to avoid rate control medications in those who have converted to sinus rhythm. This can be accomplished by frequent familiarization with guidelines. We therefore encourage ED physicians to refer to the CAEP checklist or the free smartphone application (CAEP Atrial Fibrillation Guide on iOS/Android) [3].

Research implications

Our improved understanding of the areas where checklist deviation is most common can guide future research and initiatives to improve these aspects of care. Previous studies have shown that quality improvement initiatives utilizing algorithmic care pathways for patients presenting to EDs with acute AF/AFL can lead to better compliance with stroke prevention and follow-up guidelines [10]. Our study may prompt similar initiatives both locally and nationally to improve ED management of Acute AF/AFL.

Conclusion

In conclusion, our study demonstrated very good overall alignment with CAEP’s Checklist regarding stroke prevention and disposition/follow-up. There remain opportunities to further improve care with respect to recommendation of anticoagulation and reducing inappropriate prescriptions of rate or rhythm control medications.