Chapter 20
Syncope

Lucas Oliveira Junqueira e Silva^1,2 and Fernanda Bellolio¹

¹ Department of Emergency Medicine, Mayo Clinic, Rochester, MN, USA

² Department of Emergency Medicine, Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil

Background

Syncope is a transient loss of consciousness associated with a return to preexisting neurological function. It accounts for up to 2% of all emergency department (ED) visits. Syncope is a symptom and has a wide variety of causes, ranging from benign to life threatening. The evaluation of patients with “unstable” syncope where there is a clear etiology (e.g., ongoing chest pain, gastrointestinal bleeding, or cardiac rhythm disturbances) can typically be focused on correcting or treating the underlying cause. The evaluation of syncope that is “stable” poses a greater diagnostic conundrum to the emergency physician because in approximately half of these cases, the cause for syncope is unclear, even after a thorough ED evaluation.^1–3 Even in stable patients with syncope, there are several potentially lethal causes including cardiac arrhythmias, myocardial infarction, ruptured ectopic pregnancy, subarachnoid hemorrhage, and pulmonary embolism. The clinical presentation can sometimes be confused with other conditions where there is a loss of consciousness like seizures (see Chapter 49), vertigo, dizziness, coma, or shock, and can be the result of trauma, alcohol intoxication, or other toxic substances.

As a result of the diagnostic uncertainty and the multiple potentially serious etiologies of syncope, patients are frequently admitted to the hospital for further evaluation, monitoring, and additional testing. As an inpatient, patients may receive further diagnostic testing such as echocardiogram, electroencephalogram, cardiac monitoring, and cardiac stress testing.⁴ The yield of hospitalizations in this population with diagnostic uncertainty, however, has been questioned.⁵ Specific treatments, such as pacemakers or defibrillators, can be used if a cardiac arrhythmia is determined as the cause for syncope (see Chapter 22), or changes in medication that impact blood pressure, heart rate, or underlying rhythm disturbances may be made to reduce the risk of syncope in the future. Over the last decade, numerous studies have been performed to identify lower‐risk syncope patients who may be safe for discharge after ED evaluation, as well as to identify those at higher risk for short‐term major adverse events.

Clinical question

Can elements of the history, physical exam, and basic workup risk stratify adult patients who present with syncope to the ED?

A review⁶ in the JAMA Rational Exam Series has identified key elements from the ED evaluation that both increase and decrease the probability of cardiac syncope (Table 20.1). Gibson et al.⁷ conducted a systematic review on predictors of short‐term outcomes (30 days) after syncope, and among 17 studies, the following predictors were found to have the highest yield, (i) clinical findings: elevated blood urea nitrogen (BUN) (positive likelihood ratio [LR+] 2.86), history of congestive heart failure (CHF) (LR+ 2.65), initial low blood pressure (LR+ 2.62), history of arrhythmia (LR+ 2.32), abnormal troponin (LR+ 2.49); (ii) symptoms: dyspnea (LR+ 2.29); and (iii) physical exam findings: hypotension (LR+ 2.62), altered respiratory rate (LR+ 2.26).

Toarta et al.⁸ investigated the emergency physician’s syncope diagnosis and its association with 30‐day serious outcome after ED disposition. In a large prospective cohort of 5010 patients of whom 177 (3.5%) had serious short‐term outcomes, the ED diagnostic category assigned by the emergency physician strongly correlated with the probability of 30‐day serious outcome. There were no reported deaths in the vasovagal syncope group, and patients in this group had the lowest serious outcome rates. Patients with a clear history of vasovagal syncope as identified by an emergency physician had a very good prognosis in this study.

Table 20.1 Elements of history, physical exam, and basic workup and their ability to increase or decrease the probability of cardiac syncope

Source: Data from [6].

Increased probability of cardiac syncope	Decreased probability of cardiac syncope
Age at first episode ≥35 (LR+ 3.3)	Age < 35 (LR+ 0.13)
History of atrial fibrillation or flutter (LR+ 7.3)	Warm place (LR+ 0.17)
Severe structural heart disease (LR+ 3.3–4.8)	Pain or medical procedure (LR+ 0.12)
Dyspnea (LR+ 3.5)	After using the toilet (LR+ 0.05)
Chest pain prior to syncope (LR+ 3.4–3.8)	Headache prior to syncope (LR+ 0.17)
Witnessed cyanosis during syncope (LR+ 6.2)	Feeling cold prior to syncope (LR+ 0.16)
Elevated cardiac troponin (LR+ 2.3–5.4)	Mood change or prodromal preoccupation with details prior to syncope (LR+ 0.09)
Elevated BNP (LR+ 1.8–47)	Normal cardiac troponin (LR+ 0.15–0.39)
Elevated BNP (LR+ 1.8–47)	Normal BNP (LR+ 0.16–0.20)

A 2010 systematic review⁹ identified nine different clinical decision rules (CDRs) for adults presenting with syncope to the ED. Among all of these instruments, only two rules were externally validated at that time (The San Francisco Syncope Rule [SFSR] and the Osservatorio Epidemiologico sulla Sincope nel Lazio [OESIL] risk score). The SFSR focused on predicting 7 and 30‐day outcomes, while the OESIL looked at 1‐year endpoints. An expert consortium developed standardized reporting guidelines for ED syncope stratification research in 2012,¹⁰ and it was determined that ED‐based risk stratification tools should identify serious outcomes occurring at the ED visit or within 7–30 days after discharge. For this reason, in this chapter, we will consider only those CDRs that have their outcomes defined in accordance with this timeframe. This includes the following CDRs in various stages of derivation and validation: the SFSR, the Boston syncope criteria, the Risk stratification of syncope in the ED (ROSE) rule, the Short‐Term Prognosis of Syncope (STePS), the Syncope Risk Score, the Canadian Syncope Risk Score (CSRS), and the FAINT score.

San Francisco Syncope Rule (SFSR)

In 2004, Quinn et al. derived¹¹ and, 2 years later, validated, the SFSR. The authors used outcomes at 7 days as the standard by which to assess whether a patient with syncope requires hospital admission. The outcomes included mortality, myocardial infarction, arrhythmia, pulmonary embolism, stroke, subarachnoid hemorrhage, significant hemorrhage, or return to the ED. In the derivation study, the authors followed 684 patients with syncope or near syncope who were evaluated in the ED. Of the 684, there were 79 (11.5%) serious outcomes. They performed a kappa analysis (test of interrater agreement) and only used variables with good agreement (0.5–1.0) for the decision rule. The rule, which required the absence of all the risk factors listed in Table 20.2, was 96% (confidence interval [CI] 92–100%) sensitive and 62% (CI 58–66%) specific for identifying serious outcomes at 7 days (Table 20.3). If the rule had been applied to the derivation cohort, it could have safely decreased admission rates for syncope by 10%.

The initial validation of the SFSR conducted by the same group included 791 consecutive ED visits for syncope, with 53 (6.7%) serious reported outcomes within 30 days. The authors found that the rule was 98% sensitive (CI 89–100%) and 56% specific (CI 52–60%). Some limitations of the study included that it was conducted at only one hospital. Because they used a composite outcome that included multiple serious outcomes, the study was not powered to detect any outcome (such as pulmonary embolism) individually. The authors advocated that the rule should be used as a risk stratification instrument rather than as an admission guideline, citing the fact that there are many reasons for admission to the hospital.

Saccilotto et al.¹³ conducted a systematic review and found 11 validation studies from five different countries (Australia, Canada, Italy, the United Kingdom, and the United States) looking at the performance of the SFSR. For nine studies including all patients presenting with syncope to the ED, the pooled sensitivity and specificity were 85% (CI 76–92%) and 51% (CI 39–64%), respectively. When looking at the five studies that included a subgroup analysis on patients without a cause for syncope identified in the ED, the pooled sensitivity and specificity were 88% (CI 70–96%) and 54% (CI 44–63%), respectively. The authors reported substantial between‐study heterogeneity that resulted in a 95% prediction interval for sensitivity of 55–98%. In patients with all high‐risk factors absent, the probability of a serious outcome was 5% or lower, and it was 2% or lower when the rule was used for patients where no cause of syncope was identified after initial ED evaluation.

Table 20.2 San Francisco Syncope Rule

Abnormal electrocardiogram (ECG)	Interpretation: the presence of at least one of these variables identified the patient as high risk (SFSR positive)
Complaint of shortness of breath
Hematocrit less than 30%
Triage systolic blood pressure of less than 90 mmHg
History of congestive heart failure

Table 20.3 Performance of San Francisco Syncope Rule (SFSR) in derivation and validation studies

Rule	N	Sensitivity	Specificity	NPV	PPV	LR−	LR+
*Quinn et al. (the United States)¹¹ – 7‐day serious outcomes^
SFSR	684	76/79, 96.2%	375/605, 61.9%	99.2%	24.8%	0.06	2.53
*Quinn et al. (the United States)*¹² – 30‐day serious outcomes
SFSR	713	52/53, 98%	370/660, 56%	99.7%	15%	0.03	2.23
*Serrano et al. (meta‐analysis, 10 studies)*⁹
SFSR	5468	86%	49%	NR	NR	0.28	1.74
*Saccilotto et al. (meta‐analysis, 12 studies)*¹³
SFSR	5316	87%	53%	97%	19%	NR	NR
*Tan et al. (Singapore)*¹⁴ – 7‐day serious outcomes
SFSR	1194	130/138, 94.2%	536/1056, 50.8%	98.5%	20%	0.1	1.9
*Costantino et al. (individual data meta‐analysis, three studies)*¹⁵ – 10‐day serious outcome
SFSR	2348	76%	53%	NR	NR	0.45	1.61
*Costantino et al. (individual data meta‐analysis, three studies)*¹⁵ – 30‐day serious outcome
SFSR	2348	74%	61%	NR	NR	0.43	1.89

^* Serious outcomes were defined as death, myocardial infarction, arrhythmia, pulmonary embolism, stroke, subarachnoid hemorrhage, significant hemorrhage, or any condition causing or likely to cause a return ED visit, and hospitalization for a related event.

NR = not reported by the systematic review and numbers unavailable for calculations; NPV = negative predictive value; PPV = positive predictive value; LR = likelihood ratio.

When looking individually across the different validation studies of SFSR,¹³ sensitivity levels ranged from as low as 52% to as high as 100%. However, it is important to note that none of the validation studies applied the SFSR as originally outlined in the derivation study, which may explain its variable performance across different settings.

In 2013, the SFSR was prospectively validated in two Asian EDs.¹⁴ A total of 1194 patients (age ≥ 12) presenting with syncope or near syncope to the ED were included for analysis, of whom 138 (11.6%) had a serious outcome at 7 days. Sensitivity was 94.2% (CI 89–97%) and specificity was 50.8% (CI 47.7–53.8%). However, when excluding events that happened in the ED, sensitivity went down to 93.6% (CI 82.8–97.8%). In this cohort, there were nine adverse events in eight patients who were not identified by the rule as high‐risk. Emergency physician’s judgment, independent of the rule, correctly identified all 138 patients and admitted them to the hospital.

In a 2014 individual data meta‐analysis,¹⁵ Costantino et al. compared the SFSR with clinical judgment to predict 10‐ and 30‐day serious outcomes. SFSR had a sensitivity of 76% for 10‐day outcomes and 74% for 30‐day outcomes, with a specificity of 53% for 10‐day outcomes and 61% for 30‐day outcomes. Independent clinical judgment of the emergency physician, however, had a much better performance (sensitivity of 95% and specificity of 55% for 10‐day outcomes; sensitivity of 94% and specificity of 50% for 30‐day outcomes).

In 2018, a large prospective cohort study¹⁶ included 1490 patients (age ≥ 40) presenting with syncope to 13 different EDs in eight countries (Switzerland, Spain, Germany, Italy, Poland, New Zealand, Australia, and the United States), and compared nine risk stratification scores with early clinical judgment in regards to the ability of predicting death, major cardiac adverse events (MACE), and the diagnosis of cardiac syncope. When looking at the accuracy of analyzed scores for the prediction of MACE at 30 days, the SFSR had the lowest area under the curve (AUC) when compared to other tools. Its ability to predict death and MACE at 2‐year follow‐up included sensitivities of 84% (death) and 85.7% (MACE). Again, emergency physician judgment outperformed all risk scores when predicting cardiac syncope (AUC 0.87). The authors published an analysis from the same cohort¹⁷ evaluating the prognostic accuracy of cardiac biomarkers (B‐type natriuretic peptide [BNP] and troponin) when compared to different risk scores. The SFSR had significantly inferior prognostic accuracy for MACE (AUC 0.64) when compared to biomarkers such as BNP, N‐terminal proB‐type natriuretic peptide (NT‐proBNP), high‐sensitivity troponin I (hs‐cTnI), and high‐sensitivity troponin T (hs‐cTnT) (AUCs 0.75–0.79).