Hot Off the Press: A Novel Algorithm to Decrease Unnecessary Hospitalizations in Patients with Atrial Fibrillation

Atrial fibrillation (AFib) is the most common dysrhythmia encountered in the emergency department (ED).1 However, the management of AFib is incredibly variable.2, 3 In the United States, most patients presenting to the ED in AFib are admitted to hospital, accounting for a large proportion of the costs associated with this condition.4-6 However, in other countries, such as Canada, the vast majority of AFib patients are managed as outpatients.2 Inpatient care is more expensive, but allows for a more rapid workup. It is unclear which patients require admission for AFib and which can be managed safely as outpatients. The objective of this study was to determine whether the implementation of a new algorithm for patients with a primary diagnosis of AFib could safely decrease the hospital admission rate. This is a retrospective cohort study with a before-and-after design, comparing the admission rate for patients with AFib during a 1-year period before the initiation of a new management algorithm to a 1-year period after the algorithm was introduced. Physicians were first instructed to determine whether the AFib was secondary to another important diagnosis, such as sepsis or pulmonary embolism. (These patients were admitted to hospital.) Among the patients with primary AFib, physicians assessed for four high-risk criteria: hemodynamic instability, acute heart failure, syncope, and acute coronary syndrome. Patients with no high-risk criteria were managed with either rate or rhythm control and then discharged home with close outpatient follow-up. The admission rate for AFib patients fell from 80% in the before period to 67% after, with no change in ED return visits. There are a number of limitations inherent to the before-and-after study design. Any observational study is at risk of unseen confounders. There is a general trend toward cutting costs and decreasing admissions in medicine, which could explain some of the decrease in admissions seen in this study. Similarly, the Hawthorne effect—where individuals change their behavior because they know they are being observed—could also have contributed to the observed changes. It may be difficult to duplicate the results of this study, because we are not provided with details about how clinicians were educated about and encouraged to use this new algorithm. Furthermore, the results were reported without confidence intervals, making it difficult to assess the precision of the numbers. Perhaps the biggest issue with this study will be external validity. The management of AFib varies significantly around the globe, so the ability of this algorithm to decrease admissions will depend heavily on local practice patterns. Furthermore, as part of the development of this algorithm, an AFib clinic was developed to facilitate outpatient follow-up. The ability to manage AFib patients safely as outpatients is probably dependent on the reliability of outpatient follow-up available. There were 586 patients with a primary diagnosis of AFib in the preintervention year and 522 in the postintervention year. The admission rate (including observation status) decreased from 80.4% to 67.4%, an absolute reduction of 13% (p < 0.01). The rate of ED recidivism was low and unchanged between the two groups. At 3 days, 1% of patients had a return visit in both groups. At 30 days, return visits were 3.8% before and 3.6% after. No patients died, according to hospital records and a social security registry. Length of stay in the ED was also unchanged (4.7 hours vs. 4.6 hours). Some patients with AFib can certainly be managed safely as outpatients. This study suggests that the use of a standardized algorithm may safely decrease the admission rate for patients in AFib. The algorithm used here has not been externally validated, and it is not clear that a standardized algorithm provides any benefit over physician judgement. Chris Bond: Great episode and article to highlight. My big question is where does the Ottawa Aggressive Protocol come in for you? It is a highly safe and effective method and results in very high discharge rates (96.8%) and there were no thromboembolic events. Obviously only for < 48 hours onset subgroup. Dr. DeMeester's reply: Thanks for the response and question. Our practice does include electrical cardioversion but not chemical; this may be a regional or more US-based approach. In the recent article published by Stiel (Ann Emerg Med 2017), while the success rate was high, 40% patients were excluded because of unclear time of onset (common problem), and the study had a 30 day return rate of 27% with a 10% adverse event rate. My experience is that the US doesn't tolerate such high adverse and bounce back rates … We advocate oral rate control in the ED and then referral to Afib Clinic where patients are counseled regarding long-term [rate and rhythm control] treatment options. Paraphrased from Twitter: The algorithm looks at the CHADS2VASC score (<2) and the INR (>2) but doesn't mention the time from onset. Most of us are cardioverting based on time of onset. Why did you decide on these alternative criteria? Dr. DeMeester's reply: We did have two algorithms. One for < 48 hours and one for greater than. Kept it simple for the paper. Thanks for the question! Paper in a Pic from Kristy Challen: Evan Crowe (@Fearthecrowe): Love the spirit of reducing admissions for this! But … would like to know reasons for admission (eg fail rate control, intolerable sx, etc) so these can be addressed for validation and further study James French (@grade1view): Yes … I'd also add that fast AF is often because of other driving pathophysiology. Is preload too high or too low? Is after load too high or too low? Also what is the atrial internal diameter? Atria bigger than 6cm probably just need rate control … This before-and-after observational trial demonstrated that fewer patients with AFib were admitted to hospital after the introduction of a new management algorithm, without any increase in ED return visits. Managing more patients with primary AFib in the outpatient setting is a reasonable goal.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.