Article, Cardiology

Real-time cardiology overread of the electrocardiogram: where is the value added?

Unlabelled imageAmerican Journal of Emergency Medicine (2011) 29, 316-318

Editorial

Real-time cardiology overread of the electrocardiogram: where is the value added?

William J. Brady MDa,?, Amal Mattu MDb, Robert E. O’Connor MDa

aDepartment of Emergency Medicine, University of Virginia, Charlottesville, VA

bDepartment of Emergency Medicine, University of Maryland, Baltimore, MD

Received 16 September 2010; revised 23 September 2010; accepted 24 September 2010

Patients presenting to the emergency department (ED) with Suspected acute coronary syndrome (ACS) must be evaluated expeditiously. A primary goal of the early evaluation is the identification of significant ST-segment changes, with immediate identification of patients having ST-segment elevation myocardial infraction (STEMI). Rapid and accurate interpretation of the 12-lead electrocar- diogram (ECG) represents an extremely important skill during this early, time-sensitive evaluation. The emergency physician most often is responsible for ECG interpretation./a> at this crucial step in the process of care. The emergency physician must be an expert in the interpretation of the ECG in this and other urgent or emergent ED presentations.

A recent investigation published in this edition of the American Journal of Emergency Medicine, “[the] Effects of cardiology review of the ecg in patients with suspected acute coronary syndromes” by Feldman and colleagues [1], explores this important topic of ECG interpretation in the ED–with the addition of a “real-time” secondary review and interpretation of the ECG by a cardiologist. In this project, the authors at an academic medical center performed a prospective observational study investigating the impact of “real-time” cardiology review of the ECG in ED patients suspected of ACS. The review occurred on weekdays, from 8 AM to 11 PM. Of note, ECGs that were interpreted by the emergency physician as “normal” or “unchanged from baseline” (most ED ECGs) were excluded. Emergency

* Corresponding author.

E-mail address: [email protected] (W.J. Brady).

department reviewers of the ECG included both emergency medicine (EM) residents and/or EM attending physicians. The cardiology expert reviewer was a cardiology fellow with attending cardiology assistance as needed; study assistants faxed the ECG to the cardiology fellow for review and interpretation. The authors collected data including demo- graphics, clinical and ECG characteristics, and 3-month patient outcome; they compared EM physician, including EM resident, with cardiologist (ie, cardiology fellow) ECG interpretation. The primary outcome measure was any Change in management based on the cardiology reading, whereas the secondary outcome was the degree of concor- dance between cardiology and EM interpretations. Impor- tantly, the cardiologist interpretation of the ECG was considered the reference standard and final result of the 12- lead ECG. In other words, an independent review to determine the accuracy of each service’s (ie, cardiology and EM) interpretation was not performed; furthermore, discordant readings were not adjudicated.

One hundred forty-nine patients were initially entered into the study. Twenty-nine patients were excluded because they left the ED against medical advice (n = 11) or because cardiology was unavailable for ECG review (n = 18). Ultimately, 120 patients were included in the study and were used for data analysis.

Cardiology review resulted in a real-time callback to the ED in 6 cases with reported discordant EM-cardiology ECG interpretation (4% of the total sample and 5% of the analyzed group). In 3 cases (2.5%), disposition was changed to critical care unit, and no case of urgent/emergent reperfusion or

0735-6757/$ – see front matter (C) 2011 doi:10.1016/j.ajem.2010.09.029

Editorial 317

other time-sensitive escalation in therapies was added as a result of discordant review. The authors concluded that real- time ECG review by cardiology changed triage and Treatment decisions in a small number of ED patients evaluated for ACS; they further suggested that additional study of cardiology ECG review of ED ACS patients is needed.

This publication is of interest to the emergency physician for many reasons. First, the issue of emergency physician interpretation of the ECG warrants discussion. When one compares ECG interpretations for all patient presentation types between emergency physicians and cardiologists, several interesting themes are noted. There is an extremely high rate of interpretive agreement, extremely rare “miss” of significant abnormalities by the emergency physician, and minimal to no impact on patient care by cardiology “over- read” of the ED ECG. In general terms–that is, the interpretation of the ECG in all types of presentations– ECG interpretation by the emergency physician with a comparison with the cardiology post hoc review demonstrates a rate of discrepancy of only 4%; most of these discordant cases do not involve clinically significant discrepancies [2-5]. If one considers the literature discussing the emergency physician ECG interpretation in the patient suspected of ACS–as is the case in the article of Feldmam et al [1]– the emergency physician again demonstrates extremely high rates of agreement with the cardiologist. A number of previous publications report interpretive agreement rates approaching 95% [6-9]. These investigations support the contention that emergency physicians interpret the ECG in appropriate and correct fashion at the bedside and have the advantage of correlating the ECG findings with the clinical evaluation. Very high rates of agreement are found when these interpretations are compared with attending cardio- logists who review the ECG in the controlled environment of the electrocardiography laboratory the next day, often times with the benefit of past studies and other clinical information. Furthermore, when one considers discordant interpretations, little to no clinical significance is found. Thus, it appears that the medical literature overwhelmingly demonstrates that emergency physicians are, in fact, quite proficient in the interpretation of the ECG in both ACS and

other presentations [10].

The next consideration involves the cardiologist as the expert ECG interpreter–in essence, the “gold standard” against which all other ECG interpretations are measured. The methodology of the study of Feldman et al [1] assumes that the cardiologist is, in fact, the gold standard with regards to interpretation–at this time, there is no such evidence in the literature supporting this assumption.

It is reasonable to assume that in every specialty, there is variation in specific skills. For example, it is reasonable to assume that the invasive cardiologists are more skilled than other types of cardiologists at detecting subtleties of ischemia, that electrophysiologists are more skilled than other types of cardiologists at diagnosing dysrhythmias, and so on. Furthermore, those cardiologists who have dedicated

their careers to teaching and writing about electrocardiogra- phy, such as the late Henry Marriott, Galen Wagner, Edward Chung, and others, would certainly be considered more “expert” than other cardiologists. It stands to reason, therefore, that not every cardiologist can be considered a “gold standard” upon which to compare the skills of other clinicians, including the emergency physician.

A recent publication [11] noted that cardiologists also experience difficulty in the interpretation of the ECG in patients suspected of ACS who present with STE. Fifteen cardiologists from North America, Europe, and the Middle East were asked to interpret 116 ECGs demonstrating STE; this STE met the criteria for Fibrinolytic therapy if STEMI was suspected clinically. The interpreting cardiologists had no direct contact with the patients but were told that the patients had symptoms consistent with an acute cardiac presentation. The primary question asked of the cardiologist involved candidacy for Primary percutaneous coronary intervention for suspected STEMI; if electrocardio- graphic STEMI was not suspected, the cardiologist was asked to determine the cause of the STE. In this study population, only 7% had confirmed STEMI; the remainder had noninfarction causes of the STE. Considering the pooled results of the cardiologist interpretation of these ECGs, large variation was found not only regarding suitability for Primary PCI, ranging from 7.8% to 33%, but also in correct interpretation of the STEMI and non-STEMI patterns. In fact, on average, the cardiologists missed 25% of the true STEMIs. In their commentary in the article, the authors noted their surprise and discomfort regarding these imperfect results, “[w]e hypothesized that these electrocardiographers would be both highly accurate and show consistency among individual readers. We found neither to be true. If experienced readers, using the current criteria and guidelines, cannot accurately and consistently distinguish between STEMI and [non-STEMI], less-experienced readers cannot be expected to do so.”

This data set confirms that not all cardiologists should be considered “gold standards,” and in fact, it is also reasonable to question whether we should consider all cardiologists “experts” at electrocardiography when “hand-picked” cardiologists ended up missing 25% of STEMIs. This study also demonstrates that the cardiologist experiences similar issues that the emergency physician faces on a daily basis–the challenge of STE differentiation of STEMI from non-STEMI electrocardiographic patterns.

Now, if we consider the ECG training experience for EM residents, internal medicine residents, and cardiology fellows, we do note interesting comparisons regarding the ECG interpretations early in their respective professional careers. This comparison is based on the interpretation of both electrocardiographic rhythm strips as well as 12-lead ECGs; it also uses the lowest possible number of ECGs interpreted by the clinicians in training. The EM resident at the University of Maryland or the University of Virginia will likely interpret 4500 ECGs during a standard 3-year

318 Editorial

residency training program, whereas the internal medicine resident is likely to interpret 3000 ECGs during a similar training period. The cardiology fellow, in his/her training program, will likely interpret an additional 4500 ECGs. Thus, based strictly on the number of ECGs viewed and interpreted, the recent EM residency graduate likely has more interpretation experience than the “junior” cardiology fellow. This comparison is of value during and soon after completion of training; its value, however, is markedly reduced in mid and later career in that the practicing emergency physician and cardiologist will amass varying levels of ECG interpretation experience depending on their practice setting and patient volume, among other variables. The practical issues regarding the system of consultation for ECG interpretation described in the study merits discussion. The emergency physician’s use of the cardiolo- gist for a “real-time” overread would place yet another step in the overall process of Patient evaluation and management in the ED. The actual contact with the cardiologist, confirma- tion of ECG transmission and receipt, secondary interpreta- tion, and review of interpretation agreement represent very significant challenges for an ED–all performed and secured within an appropriate time period. Also, an extrapolation of this approach could involve the cardiologist “over-reading” the ECG suspected of STEMI before the initiation of Emergent reperfusion therapy. Such a mandatory ECG review would most likely delay access to fibrinolysis or PCI and take the ED approach back 3 decades with cardiology review of all ECGs before initiation of fibrino- lysis. Lastly, the cardiologist would need to be immediately available 24 hours a day, 7 days a week, or 365 days a year with an extremely short “turn-around” time on all ECG reviews. Also, do we not only forward suspected ACS ECGs but other ECG scenarios? Cardiac arrest, acute poisoning, metabolic disarray, and so on, constitute other examples. This universal, near-immediate availability would present problems for any group of clinicians. In fact, in the study of Feldman et al [1], 12% of the initial study population did not have their ECG reviewed by the cardiologist because of “cardiology unavailable,” illustrating that complete coverage of all ECG interpretation is not possible–not at academic centers employing cardiology fellows and certainly not in community hospitals employing attending cardiologists. In addition, from the existing literature, such a mandatory overread interpretation does not appear to add significantly

to the patient’s early management.

The cardiologist completing this review would also need to be a practicing cardiologist, not a cardiologist in training; the cardiologist must also be engaged in acute and critical care cardiology, including regular ECG interpretation. Again, in the report of Feldman et al [1], a cardiology fellow–that is, a cardiologist in training–was the most frequent “expert” interpreter. The abilities of a board- certified, experienced emergency physician are at least equal, if not superior to, a cardiology fellow with respect to ECG interpretation.

The bottom line of this article is its message, and the reality of ED is 2-fold: (1) the emergency physician is the appropriate interpreter of all ECGs performed in the ED, and

(2) in selected cases, the emergency physician should feel comfortable with and have a mechanism to consult the cardiologist for a collaborative review of the ECG in certain problematic and/or challenging situations, especially in cases needing urgent intervention.

Emergency physicians by Training and practice must commit themselves to continually improve their ECG interpretation skills so that they do not need to rely on any other specialists to provide the interpretations. The EM resident must strive to go far beyond “proficiency” and aim for “expertise” in the interpretation of the ED ECG before completion of training. Furthermore, we should not return to the 1980s with a mandatory cardiology overview of all ED ECGs. At this time, cardiology overread, secondary review, or additional interpretation–whatever we would call this system–is not needed for all ACS patients or all ECGs in the ED setting. Rather, a selected approach should be followed in which we work together caring for our patients. We should continue the partnership that has developed and continues to develop between EM and cardiology as we further improve emergency cardiac care… and our ability to manage our patients.

References

  1. Feldman J, Bernard S, Mitchell P, et al. Effects of cardiology review of the ECG in patients with suspected acute coronary syndromes. Am J Emerg Med [in press], doi:10.1016/j.ajem.2010.09.023.
  2. Schaffer JA, Valenzuela TD, Wright AL, et al. Emergency physician interpretation of prehospital, paramedic-acquired electrocardiograms. Prehosp Disast Med 1992;7:251-60.
  3. Kuhn M, Morgan MT, Hoffman JR. Quality assurance in the emergency department: Evaluation of the ECG Review process. Ann Emerg Med 1992;21:10-5.
  4. Westdorp EJ, Gratton MC, Watson WA. Emergency department interpretation of electrocardiograms. Ann Emerg Med 1992;21:541-4.
  5. Todd KH, Hoffman JR, Morgan MT. Effect of cardiologist ECG review on emergency department practice. Ann Emerg Med 1996;27:16-21.
  6. Aufderheide TP, Keelan MH, Hendley GE, et al. Milwaukee prehospital chest pain project: Phase I. Feasibility and accuracy of prehospital thrombolytic candidate selection. Am J Cardiol 1992;69:991-9.
  7. Brady WJ, Perron A, Chan T. Electrocardiographic ST segment elevation: correct identification of AMI and Non-AMI syndromes by emergency physicians. Acad Emerg Med 2001;8:349-60.
  8. Brady WJ, Perron AD, Martin ML, Beagle C, Aufderheide TP. Electrocardiographic ST segment elevation in emergency department chest pain center patients: etiology responsible for the ST segment abnormality. Am J Emerg Med 2001;19:25-8.
  9. Brady WJ, Perron A, Ullman E. Errors in emergency physician interpretation of ST segment elevation in ED chest pain patients. Acad Emerg Med 2000;7:1256-60.
  10. Michelson EA, Brady WJ. Emergency physician interpretation of the electrocardiogram. Acad Emerg Med 2002;9:317-9.
  11. Jayroe JB, Spodick DH, Nikus K, et al. Differentiating ST elevation myocardial infarction and nonischemic causes of ST elevation by analyzing the presenting electrocardiogram. Am J Cardiol 2009;103: 301-6.

Leave a Reply

Your email address will not be published. Required fields are marked *