Brief Report

ED management of pediatric syncope: searching for a rationale

Monica Martin Goble MD?, Cathy Benitez MD, Max Baumgardner DO, Kathleen Fenske MD

Michigan State University College of Human Medicine, MI 48824, USA Department of Pediatrics and Human Development, MI 48824, USA

Received 28 March 2007; revised 19 June 2007; accepted 21 June 2007

Abstract

Objective: The aim of this study was to evaluate emergency department (ED) management of childhood syncope, focusing on diagnostic tests ordered, whether a reason for specific testing was recorded, and Hospital admission rates.

Methods: We reviewed ED records of patients aged 5 to 20 years who presented to a community hospital ED with syncope or near-syncope over a 1-year period (April 1, 2004, to March 31, 2005). Patient charts were nonelectronic (paper). We reviewed the elements of the recorded history and physical examination for each patient. The specific tests ordered in the ED were classified into 3 general testing categories for each patient as follows: (1) simple testing, with a hospital charge of $100 or less per test; (2) expanded testing, more than $100 per test, with a recorded explanation; and (3) expanded testing without a recorded explanation.

Results: The charts of 140 patients were reviewed. Of these, we excluded 27 based on exclusion criteria, including history of neurologic disorders. The mean age of the remaining 113 patients was 14.8 +- 3.3 years. Most (80%) presented with syncope; 20% had near-syncope. Ten percent were admitted to the hospital, over half for an electrocardiogram (ECG) interpreted as abnormal by an ECG machine and/or the ED staff. Overall, 17.5% of patients had simple testing, 32.5% had expanded testing with explanation, and 50% had expanded testing without explanation. Patients with syncope were more likely than patients with near-syncope to be in the expanded testing category (P b .008). The most commonly ordered tests in the ED in order of decreasing frequency were electrolytes (90%), ECG (85%), complete blood count (80%), urinalysis, urinary drug screen, or urinary human chorionic gonadotropin (76%), head computed tomography (CT, 58%), and chest x-ray (37%). The most expensive of these tests was the head CT; all head CT results were negative.

Conclusions: A relatively high number of our subjects were admitted (10%), most often because of questions raised by the ECG. Although an ECG is widely recommended for pediatric syncope presenting to the ED, this suggests that ECG interpretation by a pediatric cardiologist would be helpful before the decision to admit is made. In addition, 58% of our subjects had a head CT in the ED; all CT results were negative. This high percentage of head CTs for pediatric syncope has not been previously reported.

(C) 2008

* Corresponding author. Tel.: +1 517 355 2154.

E-mail address: [email protected] (M.M. Goble).

0735-6757/$ - see front matter (C) 2008 doi:10.1016/j.ajem.2007.06.012

Introduction

Syncope is common in children, especially those aged

10 years and older, and is usually benign. In a 2003 survey of medical students, approximately 20% of males and 50% of females reported a history of at least one syncopal episode [1]. Vasovagal (also called neurocardio- genic) syncope accounts for most Pediatric cases. It is characterized by systemic hypotension resulting in inade- quate cerebral blood flow from a reflex-mediated combi- nation of inappropriate vasodilatation and bradycardia. Vasovagal syncope must be differentiated from potentially life-threatening but treatable cardiac disorders. “Red flags” in the medical history, such as exercise-related syncope and/or family history of sudden death or arrhythmia, suggest cardiac testing or cardiac referral should be considered [2-5]. The mainstay of management includes reassurance, education, increased fluid intake, and behavior modification such as positional adjustments when symp- toms start [6].

Over the last 20 years, published studies have determined that a goal-directed approach to pediatric syncope is warranted and that a thorough history is critical for deciding if diagnostic testing is needed [7-9]. Rosen’s emergency medicine textbook emphasizes that the routine workup of pediatric and adolescent syncope should focus on the history and testing should be minimal [10]. The Agency for Healthcare Research and Quality guidelines for pediatric syncope is in agreement [11].

The objective of our study was to evaluate current management of childhood syncope in a community hospital emergency department (ED) with an emergency medicine residency program, focusing on diagnostic tests ordered, whether a reason for specific testing was recorded, and hospital admission rates.

Methods

With institutional review board approval, we reviewed the ED records of all patients aged 5 to 20 years who presented with syncope or near syncope from April 1, 2004, to March 31, 2005, to a community hospital ED. We excluded patients younger than 5 years to avoid patients with simple breath holding as a reason for syncope. Patient charts were nonelectronic (paper). The hospital has an accredited emergency medicine residency program enrolling 10 resi- dents per year. Each year, 60000 patients are seen in the ED; one third of these are children.

Exclusion criteria included patients who were uncon- scious, obtunded, or ventilated; patients with a known significant underlying medical illness, including cardiac or neurologic disorders; patients who were pregnant or under the influence of drugs or alcohol; patients with trauma; and patients with a history of seizures.

For the remaining patients we reviewed the elements of the recorded history listed in Table 1 and tabulated how many key elements appeared in the ED record for each patient.

The specific tests ordered in the ED were classified into 3 general testing categories for each patient as follows: (1) simple testing, with a hospital charge of $100 or less per test,

(2) expanded testing, more than $100 per test, with a recorded explanation, and (3) expanded testing without a recorded explanation. If a patient fell into the expanded testing category, we noted whether an explanation for the test that cost more than $100 was recorded in the history. For example, if a head computed tomography was ordered, we looked in the history and physical for notation of a possible head injury, headache, neurologic findings, or Seizure activity associated with the syncopal event.

For statistical analysis, we used ?2, maximal-likelihood models, and logistic regression to test for predictors of patient testing category (simple vs extended, and also simple vs extended, with and without explanation), predictors of hospital admission, and whether differences by sex were present. Statistical significance was defined as P b .05.

Results

The charts of 140 patients, aged 5 to 20 years, were reviewed. Of these, we excluded 27 based on exclusion criteria. The mean age of the remaining 113 patients was

14.8 +- 3.3 years. Thirty-five percent were male and 65% were female; race was not available. Eighty percent pre- sented with syncope and 20% with near-syncope.

Ninety-one patients (80%) had at least 3 of the 5 key elements listed in Table 1, documented in the medical history as recorded in the ED (paper) chart. Of note, 32 patients (28%) were noted to be ill, usually with a Viral illness, and 30 patients (27%) had a history of syncope recorded. The family history was documented for only 1 of the 113 patients. In that case there was a chart entry stating “family history negative.” The most commonly ordered tests in the ED in order of decreasing frequency were electrolytes (90%), electrocardio- gram (ECG, 85%), complete blood count (80%), urinalysis, urinary drug screen or urinary human chorionic gonado- tropin test (76%), head CT (58%), and chest x-ray (37%). When grouped by testing category, 19 patients (17.5%) had simple testing, 38 patients (32.5%) had expanded testing

Table 1 Key elements of recorded history

Position when symptoms occurred (standing, sitting, supine) Exercising when symptoms occurred (yes, no)

Fasting or dehydrated (yes, no) Intercurrent illness (yes, no) History of syncope (yes, no)

with recorded explanation, and 56 patients (50%) had expanded testing without recorded explanation.

Cardiology testing rates were low in the ED, excluding ECGs. ECGs were done in 85% of patients. One patient in our series was found to have a possible arrhythmogenic cause for syncope (Wolff-Parkinson-White syndrome on ECG). No causes for syncope other than vasovagal syncope were found in this series of patients.

As shown in Table 2, 11 patients (7 male and 4 female) were admitted. A majority (6) were admitted, according to the ED record, because the screening ECG results were read as abnormal by the ED provider. Of these Abnormal ECGs, only 1 was interpreted after admission by a pediatric cardiologist as abnormal.

The patient whose ECG finding was abnormal (patient 10) had Wolff-Parkinson-White syndrome (new diagnosis), which may have been related to the syncope. For the remaining 112 patients in our study the final diagnosis was not different from the presenting complaint.

Patients with syncope were more likely than patients with near-syncope to be in the expanded testing category (P b

.008). In addition, there was a trend for patients seen early in the day to undergo expanded testing compared with patients seen during the night (P = .08). The following variables did not predict testing category: age, sex, and elements of the history. However, our statistical power may not have been high enough to detect an effect of these variables.

We found a high rate for noncardiologic testing for pediatric syncope in the ED. Specifically, head CTs were done in 58% of all patients and all findings were normal.

Limitations

Our study has several limitations. It is a retrospective study with a small sample size, which limits statistical power. In addition, our findings are limited to one ED, without analysis

Table 2 Findings-patients admitted

by type of health care provider, and therefore may not be generalizable. Nonetheless, we believe that our findings will be of interest to providers in similar non-pediatric-based EDs.

Discussion

Prior studies have documented the limited yield of ancillary testing for both children and adults with syncope, emphasizing the need to be guided by the history [7-9]. A low yield for expensive ancillary testing is also supported by the current study. Yet the gap between recommendations and clinical practice persists.

There is general agreement that an ECG should be considered for patients with syncope. It has low yield but is highly sensitive for detecting heart disease and is relatively inexpensive [11]. Beyond the ECG, as underlying Cardiac abnormalities can be life threatening, the following informa- tion from the patient history should prompt additional cardio- logic evaluation: cardiac history, exercise-related syncope, or a family history of sudden death or Malignant arrhythmias.

This rate of ECGs obtained in our study (85%) is in agreement with published guidelines. A new finding is that more than half of the 11 hospital admissions in our series were arranged because of equivocal ECGs; all but one of these ECGs (which showed Wolff-Parkinson-White syn- drome) was later read as normal by pediatric cardiology. Thus, interpretation of ECGs by a pediatric cardiologist before planning admission, for example, via a Web-based or good-quality faxed transmission, may have averted up to 5 of the 11 admissions.

Others have reported discrepancy rates from 13% to 24% for ECG interpretation in pediatric EDs [12 13]. Given the wide availability of fax machines and electronic transmis- sions, when either is available we propose attempting to have the equivocal ECG officially interpreted before arranging for a hospital admission on the basis of the ECG. (The ECG

Patient	Age	Sex	ED shift	Upright (Y/N)	Fasting (Y/N)	Sick (Y/N)	Exercising (Y/N)	History of syncope (Y/N)	ED testing category	Syncope (S) or near-syncope (NS)	Reason for admission
1	17	F	M	Y	-	N	N	-	2	S	R/O seizure
2	18	M	A	Y	-	N	Y	-	2	NS	DIB/chest pain
3	18	M	M	Y	-	Y	N	Y	3	S	LVH by ECG
4	15	F	A	N	-	N	N	-	3	S	QTc, 0.45
5	16	M	M	N	-	N	N	Y	3	S	Recurrent syncope
6	15	M	N	Y	Y	Y	N	N	2	S	QTc, 0.44
7	18	M	A	Y	-	N	N	-	3	S	Flat ST-T segment
8	13	F	M	-	-	Y	N	Y	3	S	QTc, 0.457
9	14	M	A	Y	-	N	Y	Y	3	S	Syncope with exertion
10	17	M	M	Y	-	N	N	Y	3	S	WPW
11	18	F	M	-	-	N	N	Y	3	S	Abdominal pain
Dashes indicate not specified; M, morning; A, afternoon; N, night; N, no; Y, yes; R/O, rule out; DIB, difficulty in breathing; QTc, QT interval on ECG corrected for heart rate; WPW, Wolf-Parkinson-White pattern on ECG; LVH, left ventricular hypertrophy.

tracing needs to be of acceptable quality for off-site interpretation.) Because many children admitted for observa- tion for syncope are admitted to wards with telemetry, the decision to admit can be a costly one both in terms of hospital charges and inconvenience to the family.

Regarding other testing procedures, we found a high rate of head CTs (58% of patients) in this series. This rate is twice that reported in other series [2,3]. This is an important finding because head CTs have a hospital charge of $1100 to $1500. In our study we included patients with near-syncope (20% of the total) and excluded children with a neurologic history; this makes the high rate of head CTs more remarkable.

Two recent reports of pediatric syncope are similar in design to ours except that both included patients with a history of seizures. They also caution against Routine testing other than routine ECGs. The first report, a 2004 Belgian study by Massin et al [14], reviewed 226 ED pediatric patients and found that in no case was disease-related syncope diagnosed by ancillary studies. In that study, 22% of patients had a CT scan (1 positive test result, findings not specified). Regarding a neurologic workup, they recommend that patients with Prolonged LOSs of consciousness, seizure activity, and a postictal phase be referred for neurologic evaluation and an electroencephalogram, but they suggest a head CT is not indicated unless focal neurologic deficits are found. The second report, a 2005 study by Steinberg and Knilans [15], reviewed 169 ED and clinic patients with syncope and reported that only 4% of all tests were diagnostic. The average cost of testing per patient was

$1055. Neuroimaging studies were done in 27% of patients (2 positive test results, findings not specified). Both studies conclude that the evaluation of pediatric syncope remains expensive and testing has a low diagnostic yield. Both recommend an approach that focuses on the use of testing to verify findings from the history and physical examination to exclude Life-threatening causes.

Table 3 shows an algorithm for the initial evaluation of children with syncope, highlighting the unusual instances when more than a careful history and ECG may be indicated. Although we cannot generalize our findings to other locations, our study highlights a need for better availability and use of pediatric guidelines in the ED, particularly when pediatric patients are seen by providers who are not pediatric based. As the medical field becomes more guideline-driven and moves to fully electronic patient records, an increase in standardized management of pediatric ED patients may follow. The Agency for Healthcare Research and Quality, a branch of the US Department of Health and Human Services, posts the following recommendation for the management of pediatric syncope on its Web site: “The diagnosis and differentiation of benign from more serious causes is made

primarily by the history and Standard ECG” [11].

The medical literature has many examples of large gaps existing between best evidence and practice. Clinical practice guidelines have been adopted by many as one means of bridging this gap. Barriers to guideline adherence include

Table 3 Testing recommendations based on history and

physical findings

History and physical findings

Known heart disease or Palpitations, exertional

chest pain or syncope, or Family history of

sudden cardiac death

->

Recommended initial approach

ECG

Consider cardiology referral or admission for observation

Prolonged loss of consciousness or

Seizure or

Post-syncopal lethargy or Lack of a common syncope

precipitant or Neurological deficit

ECG, EEG

Basic metabolic panel

-> CT Scan or MRI Consider neurology referral or

Hospital admission

Fasting, ill or ECG

Possible drug ingestion or Basic metabolic panel Coma or -> Toxicologic analysis Suspected diabetes/metabolic Hospital admission/

problem monitoring

Prolonged Upright posture or Other common syncope

precipitant

->

ECG

lack of guideline awareness and a reluctance to change practice [16]. One article emphasizes the need for closer collaboration between ED providers and other health care providers [17].

Published studies of pediatric syncope have relatively small sample sizes but the findings are consistent. It seems clear from our study and others’ that (1) ED interpretation of ECGs suspected of being abnormal should be verified if possible before planning admission based on that ECG, (2) routine admission and/or extensive testing in the ED for simple pediatric syncope is not warranted except for a very small subset of patients, including those with exercise-related syncope or a family history of sudden death or malignant arrhythmias, and (3) because syncope may recur during adolescence, anticipatory guidance from the ED provider to the patient and family including specific ways to minimize recurrence, although not evaluated as part of this study, is likely to be cost-effective.

References

1. Ganzeboom KS, Colman N, Reitsma JB, Shen WK, Wieling W. Prevalence and triggers for syncope in medical students. Am J Cardiol 2003;91:1006-8.
2. Chronister TE. Pediatric neurologist as consultant in evaluation of syncope in infants and children: when to refer. Prog Pediatric Cardiology August 2001;13(2):133-8.
3. Tanel RE, Walsh EP. Syncope in the pediatric patient. Cardiology Clinics May 1997;15(2):277-94.
4. Driscoll DJ, Jacobsen SJ, Porter CJ, Wollan PC. Syncope in children and adolescents. J Am Coll Cardiol 1997;29:1039-45.
5. Massin MM, Malekzadeh-Milani S, Benatar A. Cardiac syncope in pediatric patients. Clin Cardiol 2007;30:81-5.
6. Wieling W, Ganzeboom KS, Saul JP. Reflex syncope in children and adolescents. Heart 2004;90:1094-100.
7. Gordon TA, Moodie DX, Passalacqua M, Sterba R, Rothner AD, Erenberg G, et al. A retrospective analysis of the cost-effective workup of syncope in children. Clev Clin J Med 1987;54:391-4.
8. Lerman-Sagie T, Lerman P, Mukamel M, Blieden L, Mimouni M. A prospective evaluation of pediatric patients with syncope. Clin Pediatr 1994;33:67-70.
9. Ritter S, Tani LY, Etheridge SP, Williams RV, Craig JE, Minich L. What is the yield of screening echocardiography in pediatric syncope? Pediatrics 2000;105(5):e58-e60.
10. Marx: Rosen’s emergency medicine: concepts and clinical practice. 5th ed. Mosby, Inc; 2002.
11. Guidelines on management (diagnosis and treatment) of syncope- update 2004. Available at: http://www.guideline.gov/. Bibliographic

source: Brignole M, Alboni P, Benditt DG, et al. Europace 2004;6(6): 467-537

1. Horton LA, Mosee S, Brenner J. Use of the electrocardiogram in a pediatric emergency department. Arch Peidatr Adolesc Med 1994;148: 184-8.
2. Wathen JE, Rewers AB, Yetman AT, Schaffer MS. Accuracy of ECG interpretation in the pediatric emergency department. Ann Emerg Med 2005;46:507-11.
3. Massin M, Bourguignont A, Coremans C, Comte L, Lepage P, Gerard

P. Syncope in pediatric patients presenting to an emergency department. Journal of Pediatrics August 2004;145(2):223-8.

1. Steinberg L, Knilans TK. Syncope in children: diagnostic tests have a high cost and low yield. Journal of Pediatrics March 2005:355-8.
2. Sinuff T, Kahnamoui K, Cook D, et al. Practice guidelines as multipurpose tools: a qualitative study of noninvasive ventilation. Crit Care Med 2007;35:776-82.
3. Mehta RH. The impact of emergency department structure and care processes in delivering care for Non-ST-segment elevation acute coronary syndromes. Am Heart J 2006;152(4):648-60.