Artificial intelligence MacHIne learning for the detection and treatment of atrial fibrillation guidelines in the emergency department setting (AIM HIGHER): Assessing a machine learning clinical decision support tool to detect and treat non-valvular atrial fibrillation in the emergency department.

OBJECTIVE: Advanced machine learning technology provides an opportunity to improve clinical electrocardiogram (ECG) interpretation, allowing non-cardiology clinicians to initiate care for atrial fibrillation (AF). The Lucia Atrial Fibrillation Application (Lucia App) photographs the ECG to determine rhythm detection, calculates CHA2DS2-VASc and HAS-BLED scores, and then provides guideline-recommended anticoagulation. Our purpose was to determine the rate of accurate AF identification and appropriate anticoagulation recommendations in emergency department (ED) patients ultimately diagnosed with AF. METHODS: We performed a single-center, observational retrospective chart review in an urban California ED, with an annual census of 70,000 patients. A convenience sample of hospitalized patients with AF as a primary or secondary discharge diagnosis were evaluated for accurate ED AF diagnosis and ED anticoagulation rates. This was done by comparing the Lucia App against a gold standard board-certified cardiologist diagnosis and using the American College of Emergency Physicians AF anticoagulation guidelines. RESULTS: Two hundred and ninety seven patients were enrolled from January 2016 until December 2019. The median age was 79 years and 44.1% were female. Compared to the gold standard diagnosis, the Lucia App detected AF in 98.3% of the cases. Physicians recommended guideline-consistent anticoagulation therapy in 78.5% versus 98.3% for the Lucia App. Of the patients with indications for anticoagulation and discharged from the ED, only 25.0% were started at discharge. CONCLUSION: Use of a cloud-based ECG identification tool can allow non-cardiologists to achieve similar rates of AF identification as board-certified cardiologists and achieve higher rates of guideline-recommended anticoagulation therapy in the ED.

Identification and early anticoagulation in patients with atrial fibrillation in the emergency department.

BACKGROUND: Emergency departments (ED) in the United States see more than half a million atrial fibrillation visits a year, however guideline recommended anticoagulation is prescribed in <55% of eligible patients. OBJECTIVE: The purpose of this study was to measure guideline recommended anticoagulation prescribing in patients with nonvalvular atrial fibrillation (NVAF) presenting to the ED, with the goal of closing any treatment gap established. METHODS: We conducted an observational, prospective cohort study in consecutive patients presenting to the ED with a diagnosis of NVAF. CHA2DS2-VASc and HAS-BLED scores were calculated and used as predefined criteria to establish guideline-based oral anticoagulation compliance in comparing routine care (baseline cohort) versus a multidisciplinary team approach. Transition of Care (TOC) services and follow-up were also provided in the multidisciplinary cohort. The primary endpoint was to compare the proportion of patients on guideline based oral anticoagulant (OAC) therapy at admission and discharge between the groups. RESULTS: In the Baseline Cohort (BC) (n = 99), 62.3% of patients with a moderate-high risk of stroke (CHA2DS2-VASc score ≥ 2) were discharged on guideline-based OAC therapy versus 87.8% in the Multidisciplinary Team Cohort (MTC) (n = 131), a 25.5% overall improvement for appropriate anticoagulation (p-value <.001, 95% CI (0.14-0.37)). CONCLUSIONS: A multidisciplinary team approach with TOC services for the identification and early intervention of NVAF patients at risk of stroke in the ED can significantly improve the percentage of moderate to high-risk patients that are discharged home with guideline based OAC.

The recidivism characteristics of an emergency department observation unit.

STUDY OBJECTIVE: We describe the recidivism characteristics of an adult emergency department (ED) observation unit population and determine whether rates differ according to demographic or clinical features. METHODS: This prospective observational cohort study of a protocol-driven ED observation unit reviewed all discharged ED observation unit patients who returned within 14 days of discharge for an unscheduled ED visit or direct inpatient admission to the study hospital, or a proximate affiliated hospital, during 6 consecutive months. Age, sex, initial ED observation unit diagnosis, ED observation unit length of stay, and return visit disposition were determined from hospital databases and confirmed by chart review. All return visits were classified as related or unrelated to the index visit. RESULTS: There were 55,727 ED visits, with 4,348 patients admitted to the ED observation unit, of whom 80.7% (3,509) were discharged. Patients with or without a return visit were similar in terms of age (56.9 years [standard deviation (SD) 19.5 years]), percentage of male patients (40.0%), or initial ED observation unit length of stay (15.0 hours [SD 6.0 hours]). Of discharged ED observation unit patients, 375 (10.8%) had a return visit, of which 277 (7.9%) were related. Of return visits, 86.3% of patients had only 1 return visit, 11.6% had 2, and 2.1% had 3 or more; 4.2% of returns occurred at an affiliated hospital. Time to first return visit was clustered within the first week for related visits, with a mean time to return of 4.5 days (SD 3.9 days). On return visit, 40.2% of patients were treated and discharged from the ED, 36.2% were treated in the ED and admitted, 14.4% were treated in the ED and then the ED observation unit and discharged home, 12.3% were directly admitted to the hospital, and 2.5% were treated in the ED and then the ED observation unit and admitted. Among common conditions, related return visit rates were highest for headache (16.1%), back pain (13.8%), and abdominal pain (12.7%) and lowest for chest pain (3.6%). As a group, therapeutic protocols, and specifically painful conditions, had significantly higher related return visit rates than diagnostic protocols (10.8% versus 5.1%). CONCLUSION: Patients who return after an ED observation unit visit are similar to patients who do not return in terms of age, sex, or initial length of stay. However, ED observation unit recidivism rates do differ according to observation category, with painful conditions showing the highest recidivism rates.