Atrial fibrillation and flutter is a pervasive disease affecting 6.1 million people in the United States.Each year it is responsible for more than 750,000 hospitalizations and 130,000 deaths.In contrast to overall declining death rates for cardiovascular disease,4 AF as the “primary or contributing cause of death has been rising for more than two decades.”The annual economic burden of AF is six billion dollars; medical costs per AF patient are about $8,707 higher than for non-AF individuals.Thrombotic embolism of the cerebral circulation, or stroke, is the principal risk of AF and ranges from less than 2% to greater than 10% annually.AF is the cause of 100,000-125,000 embolic strokes each year, of which 20% are fatal.Anticoagulation to prevent these embolic events is standard of care unless contraindicated.However, it is not without risk, as even minor trauma can cause substantial and potentially life-threatening bleeding. Given that AF is the most common arrhythmia among the elderly,balancing these competing risks is challenging. Anticoagulation for AF is most commonly accomplished with a vitamin K antagonist, warfarin. However, its use requires patient education, medication compliance, dietary consistency, and close monitoring.CHA2 DS2 -VASc, ATRIA, HAS-BLED, ORBIT, and HEMORR2 HAGES are just some of the decision-support tools available to objectively weigh the risk of stroke and life-threatening bleeding from therpy.Newer, novel oral anticoagulant agents provide a benefit/risk profile that may surpass warfarin, especially when considering initiation in the emergency department. 16-18 In this issue of WestJEM, Smith and colleagues present a prospective observational evaluation of anticoagulation prescribing practices in non-valvular AF. Patients presenting to one of seven Northern California EDs with AF at high risk for stroke were eligible unless admitted, not part of Kaiser Permanente of Northern California ,cannabis grow indoor or already prescribed anticoagulation. During the 14-month study there were no departmental policies governing the initiation of anticoagulation in AF patients. University of Alabama School of Medicine, Department of Emergency Medicine, Birmingham, Alabama.
The authors report 27.2% of the 312 at high risk for stroke received a new anticoagulant at ED discharge, and only 40% were prescribed oral anticoagulation within 30 days of the index ED visit. Anticoagulation was more likely to be initiated in the ED if the patient was younger , had persistent AF at discharge, or when cardiology was consulted during the index visit. Furthermore, only 60.3% of patients were given patient education material on AF in their discharge instructions.Critics of Smith et al. will take issue with their inclusion criteria that required participation in KPNC. By definition, all members of KPNC are insured; they also have guaranteed access to timely primary care follow-up and are of higher socioeconomic means than the general population.Many of the factors that contribute to successful anticoagulation therapy – diet stability, monitoring of renal function, education and intervention of modifiable risk factors, smoking cessation, and fall risk – can all be assessed by a primary care physician and addressed with shared decision-making ensured in the KPNC system.While these limitations are acknowledged by the authors and narrow the generalizability of these findings, Smith and colleagues demonstrate the challenges of addressing ongoing chronic disease in the ED and highlight the complex decision making required. AF patients without insurance in the U.S. lack reliable access to primary care, and emergency physicians likely under-prescribe anticoagulation therapy due to an abundance of caution.Lacking the objective data to quantify these thromboembolic risk factors of AF, EPs are reluctant to initiate thromboprophylaxis, despite its known benefits, in light of the well-demonstrated risk for life-threatening bleeding.However, the risk is largely misperceived. Recent findings from the Spanish EMERG-AF trial demonstrate that initiating this therapy in the ED is at least as safe as in other settings and has clear mortality benefit at one year. Furthermore, that benefit does not come at the expense of reduced effectiveness over the course of one-year follow-up.In addition to highlighting the challenges of prescribing anticoagulation in the ED setting, Smith et al. also illustrate the opportunity for EPs to prevent future strokes in the setting of known AF. This opportunity is likely larger than reported considering the limitations of this investigation. Thankfully, there are clear guidelines to assist EPs based upon validated methods of risk-stratification.
Furthermore, of those patients receiving anticoagulation therapy in the first 30 days, more than half were initiated in the ED. While these subjects likely represent the least complex decision-making, these results also suggest some prescribing inertia; anticoagulation was continued by the primary care physician because it has already been initiated in the ED. Despite these limitations, Smith and colleagues demonstrate an immense target for EPs to improve stroke risk for at least 60% of AF patients discharged from the ED. Coupled with other evidence demonstrating that such practice is efficacious, safe, and cost effective, Smith makes a compelling case that thromboprophylaxis should be initiated in all but the most complex AF patients who will likely be admitted. EDs should develop policies to assure that AF patients can receive anticoagulation therapy on discharge. These local policies could include decision pathways that rely on guidelines, decision-support tools, and account for insurance status. As EPs, we should embrace the responsibility to provide thromboprophylaxis regardless of the likelihood of primary care follow-up. To defer that decision ignores the role emergency medicine plays in providing for the public health in the U.S., and frankly misses the mark. Arterial lines are important for monitoring and providing care to critically ill patients. Not only do they allow for rapid access to blood, but they also allow a provider continuous access to the patient’s blood pressure, which enables minute titration of vasoactive medications. Traditionally there are two locations for arterial line placement: femoral and radial arteries. The choice between sites is often made according to the provider’s preference with very little evidence guiding this decision.Although initial beliefs that arterial lines are immune to infection are certainly unfounded,vertical farming supplies there is evidence that the infection risk is proportionally similar to their central venous counterparts regarding location.It has also been shown repeatedly that central arterial monitoring provides different information from both peripheral and non invasive monitoring.Older studies have shown that the femoral artery is superior to the radial artery for blood pressure monitoring, but these results come from a different era of medicine when placement technique was different and the landscape of monitoring was not what it is today.
It is therefore important to reinvestigate femoral artery access in today’s environment. Line failure adds significant and unnecessary costs to the treatment of critically ill patients, including financial costs , time , and health. In the present study, we attempt to determine if one site is more prone to failure. We performed an ambispective, observational, cohort study to determine variance in failure rates between femoral and radial arterial lines. This study took place at a single center, a county teaching hospital with 12 adult ICU beds, and was approved by our institutional review board. Any patient with an arterial line placed anywhere in our hospital met our inclusion criteria. Providers at our site were not using ultrasound for arterial line placement routinely, so this metric was not evaluated. Although the specific indication for arterial line placement was not captured in our study, it is customary at our institution to place arterial lines for either ongoing titration of vasopressor agents or expected repeated evaluation of the management of patients with ventilatory support. Our institution uses the Arrow RA-04020 quick kit for radial arterial lines, which is a 20-gauge, 4.25 cm catheter. The Arrow select kit is used for femoral lines, which is also a 20-gauge catheter, though 12 cm in length. All patients in our study were admitted to an ICU bed and were therefore of high acuity. Exclusion criteria were patient age < 18 years old and line removal before 24 hours. We performed the retrospective arm of this study using the hospital billing database. Records from every patient who received and was successfully billed for an arterial line between January 2012 and June 2015 in our hospital were included. Research assistants , who were blinded to the study hypothesis , were provided a training presentation on how to extract relevant information from the electronic health record , including patient’s age, line insertion time, line removal time, and whether line removal was due to failure. We compiled their results into a database, and a pilot quality inprovement study was initially performed on every 20th patient in the study. The two principal investigators then reviewed the data to ensure that data acquisition was accurate between all RAs, demonstrating reliable inter-observer agreement regarding insertion and removal dates and classification of line failure. After confirming that our proposed method of data acquisition was precise, the RAs performed the complete review on the total cohort and the acquired data was kept in a spreadsheet without analysis until the prospective portion of the study was completed. The prospective arm of the study took place from June 2015 to March 2016. RAs obtained information on every adult patient in whom an arterial line was placed in our hospital during the enrollment period. To ensure capture of all patients, RAs would observe each ICU bed and ED resuscitation bay for new arterial lines three times daily. They compiled an ongoing list of known lines, noting the time of insertion, location of the line , patient age, and patient comorbidities. If the arterial line was found to have been removed, the RAs would document the time of removal and determine why the line had been removed , noting whether it was considered a failure and if it was replaced. The RAs obtained this information from nursing flow sheets or nursing interview at the time of their evaluation. Causes of failure included the following: 1) inaccuracy , 2) blockage , 3) site issue , and 4) accidental removal. We hypothesized a 2x greater failure rate of radial arterial lines compared to femoral amounting to a 50% reduction in failure rate by placing the line in the femoral artery. We postulated a 60% radial and 40% femoral distribution of line placement, based on observance of local practice. We calculated that 128 patients would provide sufficient power to detect the hypothesized failure rate if lines were split evenly between the two sites. We therefore planned to enroll 200 patients as the actual distribution was not known a priori. We chose an ambispective design as the EHR made retrospective data acquisition easy, allowing for greater power to the study. We subsequently used the prospective data to help validate our retrospective findings. In total, we evaluated 272 arterial lines over both the prospective and retrospective arms of our study, with 58 lines leading to failure for a combined total failure rate of 21.32%. Comorbidities between the two cohorts were similar, as shown in the Table. Our retrospective arm screened 304 arterial lines; however, only 196 met criteria for analysis over the three-and a-half years. The radial cohort had 43 failures and the femoral cohort had three failures , for an absolute risk reduction for failure of 25.4% if the femoral site was chosen. The prospective arm had 76 total lines, which included 39 radial and 37 femoral. The radial cohort had 10 failures and the femoral cohort had two failures. This similarly provided an absolute risk reduction of 20.2% in failure rate if a femoral line was placed instead of a radial arterial line. This outcome was consistent between the retrospective and prospective arms of the trial and led to a number needed to treat of 4.1 patients to prevent one line failure. Secondary outcomes evaluated include time to failure and cause of failure. Combined data showed the median time to failure for radial lines as two days compared to femoral lines having a median time to failure of four days. From the prospective data, the primary causes of failure for the radial lines were accidental removal , a line not drawing , and inaccurate readings. There were no radial lines removed due to “site issue” in our prospective arm; however, such issues were responsible for 15% of radial removals in our retrospective arm. Conversely, accidental removal accounted for only 5% of all removals in the retrospective cohort of radial lines but 40% of failures in the prospective arm.