Abstract
Background
Early identification of cardiac syncope is critical because it is associated with adverse outcomes in the emergency department. However, the optimal observation duration remains unclear. In this study, we evaluated factors associated with ED observation time in patients with syncope, determined the timing of cardiac findings, and identified the clinical features that distinguish cardiac from neurally mediated or orthostatic syncope.
Methods
A cohort study was conducted in the emergency department of an urban hospital, including 400 adult patients with syncope. Patients with transient loss of consciousness due to syncopal causes were excluded. The primary outcome was the presence and timing of cardiac findings (rhythm, structural cardiogenic, or both) during the observation period in the emergency department. Data were extracted from the electronic health records. Logistic regression was used to identify factors associated with cardiac syncope and time spent in the emergency department. Cumulative and incidence rate analyses of the time until a cardiac finding within 24 h were estimated using a time-to-event function.
Results
Cardiac syncope was associated with older age (OR: 1.04 per year; 95% CI: 1.02–1.07), prior pacemaker or ICD implantation (OR: 4.17; 95% CI: 1.50–11.6), absence of autonomic symptoms (OR: 0.16; 95% CI: 0.08–0.31), and abnormal electrocardiogram findings (OR: 6.84; 95% CI: 3.00–15.6), and higher rates of hospital admission (OR: 31.7; 95% CI: 14.1–71.3) and pacemaker implantation (OR: 63.0; 95% CI: 18.5–245). Cardiac findings were confirmed in 70.5% of patients initially classified as having cardiac syncope (OR: 33.0; 95% CI: 15.3–71.2). Most cardiac findings (70.9%) occurred within the first 12 h (incidence density: 1.04 vs. 0.54 per 100 patient-hours). The cumulative incidence increased from 12.5% at 12 h to 23.3% at 24 h, then slowed.
Conclusion
Most cardiac findings were identified within the first 12 h of observation. These results support a selective strategy in which prolonged monitoring (≥ 12 h) may be reserved for patients with predefined clinical conditions, electrocardiographic markers, or concerns regarding underlying heart disease at initial assessment.
Clinical trial number
Not applicable.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12873-026-01533-8.
Keywords: Syncope, Emergency service, Hospital, Observation, Risk assessment, Cohort studies
Background
Syncope is a transient loss of consciousness caused by a decrease in blood pressure, resulting in cerebral hypoperfusion and loss of postural tone [1]. It typically manifests with rapid onset, short duration, and spontaneous complete recovery [2]. Approximately 40% of the population experiences at least one syncopal episode during their lifetime [3]. Syncope is estimated to account for approximately 1%–5% of all hospital Emergency Department (ED) visits [4], with 27%–35% of the cases requiring hospitalization, with an average length of stay ranging from 3 to 4 days [5].
Syncope can be initially reflex (neurally mediated), orthostatic hypotension, or cardiac in origin. Reflex syncope includes vasovagal, situational, carotid sinus syndrome, and non-classical forms without clear triggers. Orthostatic syncope may result from the use of certain medications, volume depletion, or autonomic dysfunction, whereas cardiac syncope results from arrhythmias, structural heart disease, or cardiopulmonary conditions [2]. Syncope diagnosis begins with a detailed history of current and past episodes, including clinical features that preceded or accompanied the event. This is followed by a physical examination, including assessment of blood pressure in both the supine and standing positions, and an electrocardiogram to identify potential cardiac causes. These initial steps are crucial because the underlying cause of syncope is the primary determinant of adverse short-term outcomes [6]. Current clinical guidelines recommend a structured set of criteria for classifying syncope and assessing the risk of adverse events. These criteria include the characteristics of the syncopal episode, the patient’s medical record, physical examination findings, and electrocardiogram results [2].
Syncope of cardiac origin is a high–risk form of syncope based on its underlying etiology and the likelihood of adverse clinical outcomes [2, 7]. Syncope caused by neurally mediated or orthostatic hypotension can be safely discharged without additional diagnostic testing or prolonged observation [7]. In contrast, approximately 5% of syncopal episodes are considered high risk because they suggest a serious condition [8, 9]. These presentations are commonly associated with underlying cardiovascular conditions, such as cardiac arrhythmias, cardiomyopathies, valvular heart disease, and acute thromboembolic events. These conditions often require a comprehensive assessment, including extended monitoring and observation, to identify potentially associated abnormalities, such as arrhythmias. Despite the availability of multiple risk assessment tools and scoring systems, current evidence suggests that risk stratification scores do not offer superior prognostic value compared with clinical judgment in predicting short–term serious outcomes following a syncopal episode [2].
The duration of clinical observation in the emergency department should balance diagnostic accuracy and the time required to identify potential cardiac abnormalities in patients with syncope. However, the optimal observation duration has not been established, although durations of 6–24 h have been suggested as suitable. This study aimed to evaluate factors associated with the duration of emergency department observation in patients with syncope. Additionally, we sought to determine the timing of significant cardiac findings that may arise during the observation period and identify the clinical characteristics that differentiate neurally mediated or orthostatic syncope from cardiac syncope.
Methods
Study design and setting
A cohort study was conducted using the electronic health records of patients with syncope episodes registered in the syncope episode registry who visited the emergency department (ED) of an adult tertiary-care hospital in Barcelona, Spain, between January and December 2018. The hospital does not provide pediatric emergency services and serves a population of approximately 540,000 inhabitants [10]. Patients who arrive with an episode of transient loss of consciousness undergo triage and are transferred by default to the ED floor, where patients with Spanish Triage System levels 1–3, corresponding to emergencies requiring immediate, very urgent, or urgent attention, are usually attended. As soon as personnel and a monitoring station are available, patients are placed in the monitoring station for continuous electrocardiography (ECG), blood pressure (BP), and pulse oximetry (SpO₂) monitoring.
Eligibility and selection criteria
During the study period, 1,652 ED visits were coded with syncope (ICD–10: R55) in the electronic health record (EHR) registry.
Patients were eligible if they had a confirmed diagnosis of syncope (ICD–10: R55) recorded in the ED and were aged ≥ 18 years. The study was conducted in an adult tertiary-care hospital that does not provide pediatric emergency services; therefore, all registry entries corresponded to adult patients, and no additional age-based exclusions were required.
Records were excluded if the final ED diagnosis indicated a non-syncopal cause of transient loss of consciousness, including traumatic, neurological, endocrine/metabolic, psychiatric, or intoxication-related etiologies, as determined by ICD–10 codes documented in the EHR.
Variables
We collected data on sociodemographic characteristics, syncope classification at ED arrival (neurally mediated, orthostatic, cardiac, or unclassified), pre-existing cardiac device status (pacemaker or implantable cardioverter–defibrillator), syncope triggers, history of prior syncopal episodes, and associated symptoms from electronic health records. Additional variables included autonomic symptoms (e.g., dizziness, warmth, sweating, nausea, or vomiting), hypotension, electrocardiographic findings, laboratory test results, and cardiac murmurs. Other tests performed during ED observation, including chest radiography, computed tomography, cardiac magnetic resonance imaging (MRI), coronary angiography, echocardiography, and head-up tilt testing, were also recorded. Cardiac findings, including structural heart disease (severe aortic stenosis, severe valvular disease, hypertrophic cardiomyopathy, ischemic heart disease, and dilated cardiomyopathy) and cardiac rhythm disorders (bradycardia, tachycardia, or tachycardia–bradycardia syndrome), were documented. The duration of clinical observation, pacemaker or defibrillator implantation, and mortality were also recorded.
Data source, extraction, and inclusion processes
Data were extracted from the hospital’s electronic health record (EHR) system. From the 1,652 ED visits coded with syncope (ICD–10: R55) during the study period, a simple random sample was selected for detailed manual review and eligibility assessment. Random selection was performed using a computer-generated randomization procedure applied to the complete registry list, ensuring that each record had an equal probability of inclusion in the audited cohort.
Within the audited sample, records were reviewed in random order to avoid systematic patterns in data assessment and to reduce potential reviewer-related bias. After application of the predefined inclusion and exclusion criteria, all eligible cases were retained in the final analysis. To preserve real-world representativeness within the audited cohort and minimize selection bias, no eligible cases were excluded solely because of minor missingness or incomplete documentation.
Sample Size
Based on 1,652 health record registries with syncope in the study year and assuming a 5% incidence of arrhythmic events and structural heart disease, a minimum sample size of 358 health records was required to achieve a 2% margin of error in the estimated proportion [9].
Syncope classification
The research team members independently reviewed the data to classify the type of syncope as neurally mediated, orthostatic, or cardiac, based on the clinical features noted during the initial evaluation upon the patient’s arrival in the ED, as recorded in the health record. To ensure that the reviewers acted under the same informational conditions as the treating clinicians and to minimize potential information bias, only information documented in the ED electronic health record was used for this classification, without incorporating any additional data. This classification was performed retrospectively according to the guidelines using all clinical information available from the ED encounter, including history, physical examination, vital signs, ECG, and laboratory results [2]. The research team comprised four clinicians with experience in emergency and cardiovascular care: two family physicians and two cardiologists. Each pair independently reviewed separate subsets of records, and discrepancies were resolved by consensus under the supervision of a senior supervisor. The attending physician’s initial classification was analyzed only to assess concordance with the guideline-based classification and was not used to define study outcomes. The attending physician’s initial classification was analyzed solely to assess concordance with the guideline-based classification and was not used to define study outcomes. The applied criteria are listed in Table A0 in the appendix. The reviewers were blinded to any pre-existing syncope classification documented in the records. Subsequent analyses were based on the final consensus classification.
Statistical analysis
We used the kappa test to assess the agreement between the classifications. Patient characteristics and factors associated with syncope types, categorized as neurally mediated, orthostatic, or cardiac, were analyzed using logistic or linear regression models, with syncope classification as the outcome and clinical variables as the explanatory factors. A similar approach was used to examine differences in observation duration (< 12 h vs. ≥12 h), with observation time as the outcome variable. All results are expressed as odds ratios (ORs) with 95% confidence intervals (CIs), and statistical significance was set at p-value < 0.05.
The duration of clinical observation in the ED associated with cardiac findings during the 24–hour observation period was studied using a cumulative incidence function. Time–to–event was defined as the number of hours from ED admission to the occurrence of cardiac findings or censoring. The results are graphically reported. All analyses were conducted using R software [11].
Ethics
All procedures were carried out in accordance with the principles of the Declaration of Helsinki. The ethics committee approved the study protocol.
Results
During the study period, 1,652 ED records were coded with syncope (ICD–10: R55). A random sample of 768 records was audited to assess eligibility. After exclusion of 368 cases due to non-syncopal causes of transient loss of consciousness (traumatic, neurological, endocrine, or psychiatric etiologies), 400 adult patients with confirmed syncope were included in the final analysis. The detailed selection process is presented in Figure A1 (Appendix).
The concordance between the syncope classification on initial evaluation, as recorded in the health records, and that based on the guidelines was high, with an overall agreement rate of 90% (κ = 0.78; 95% CI: 0.70–0.85). We found the most significant discrepancies between neurally mediated and orthostatic syncope, with a 47% disagreement rate, suggesting a tendency to classify neurally mediated cases as orthostatic. Four discrepancies (9.09%) involved cardiac syncope: three cases lacked a recorded classification, and one was misclassified as orthostatic.
The mean age of the patients was 63 years (SD = 22.4), with a similar sex distribution (males: 53.25%; females: 46.75%). Hypertension was the most common comorbidity (50.8%), followed by dyslipidemia (34.8%) and a previous record of syncope (26.3%). Structural heart disease or coronary artery disease was present in 21.8% of patients, and 19.3% were current smokers. Only 4.8% had a pacemaker or defibrillator, and 1.0% had a family history of sudden cardiac death. Postural changes (18.3%) and prolonged standing or heat exposure (14.8%) were the most frequently reported triggers. Autonomic symptoms were common, occurring in 81.0% of patients. Less frequent symptoms included hypotension (6.5%), chest pain (4.5%), and dyspnea (2.5%). All the characteristics are listed in Table 1.
Table 1.
Clinical characteristics, risk factors, and clinical presentation of patients with syncope
| Characteristics |
N = 400 n (%) |
95% CI |
|---|---|---|
| Age, years. Mean (SD) | 63.01 (22.37) | 60.81–65.21 |
| Sex (females) | 213 (53.25%) | 48.23%–58.22% |
| Medical history and risk factors | ||
| Hypertension | 203 (50.75%) | 45.74%–55.75% |
| Dyslipidemia | 139 (34.75%) | 30.09%–39.64% |
| Previous history of syncope | 105 (26.25%) | 22.00%–30.85% |
| Structural heart disease or coronary artery disease (CAD) | 87 (21.75%) | 17.80%–26.12% |
| Current smoking status | 77 (19.25%) | 15.50%–23.46% |
| Type 1 or 2 diabetes mellitus | 65 (16.25%) | 12.77%–20.24% |
| Toxic habits | 33 (8.25%) | 5.75%–11.39% |
| Cardiac device carrier (Pacemaker or ICD) | 19 (4.75%) | 2.88%–7.32% |
| Family history of sudden cardiac death | 4 (1.00%) | 0.27%–2.54% |
| Syncope Triggers | ||
| Postural changes from the supine or seated position | 73 (18.25%) | 14.59%–22.39% |
| Prolonged standing, heat exposure, or crowded spaces | 59 (14.75%) | 11.42%–18.61% |
| During or after food intake (postprandial) | 48 (12.00%) | 8.98%–15.59% |
| Before, during, or after physical exertion | 43 (10.75%) | 7.89%–14.21% |
| Following a sudden and unexpected stimulus of pain | 33 (8.25%) | 5.75%–11.39% |
| During coughing, defecation, or urination | 21 (5.25%) | 3.28%–7.91% |
| Head rotation or pressure of the carotid sinus | 3 (0.75%) | 0.15%–2.18% |
| Signs and Symptoms | ||
| Autonomic Symptomsa | 324 (81.00%) | 76.81%–84.73% |
| Hypotension | 26 (6.50%) | 4.29%–9.38% |
| Lower extremity edema | 9 (2.25%) | 1.03%–4.23% |
| Dyspnea | 10 (2.50%) | 1.21%–4.55% |
| Chest Pain | 18 (4.50%) | 2.69%–7.02% |
| Abdominal discomfort | 24 (6.00%) | 3.88%–8.80% |
a: dizziness, warm sensation, sweating, nausea, or vomiting
Factors associated with syncope type
Patients with cardiac syncope on initial evaluation were older than those with neurally mediated or orthostatic syncope (mean difference: 15.7 years; OR: 1.04, 95% CI: 11.0–20.4). A prior pacemaker or ICD implant was associated with the cardiac group (13.6% vs. 3.7%, OR: 4.17, 95% CI: 1.50–11.6). A history of syncope was also associated with cardiac syncope (47.7% vs. 23.6%; OR, 2.96; 95% CI, 1.56–5.61). In contrast, autonomic symptoms were less frequent (47.7% vs. 85.1%; OR: 0.16, 95% CI: 0.08–0.31). Cardiac murmurs and conduction abnormalities, including a prolonged PR interval (OR: 6.84, 95% CI: 3.00–15.6) and a prolonged QRS duration (OR: 5.19, 95% CI: 2.34–11.5), were more prevalent among patients with cardiac syncope. Patients classified as having cardiac syncope had a cardiac cause confirmed during ED observation in 70.5% of cases, compared with 6.7% of those initially classified as having neurally mediated or orthostatic syncope (OR: 33.0, 95% CI: 15.3–71.2). Attending physicians more frequently requested additional diagnostic tests, such as chest radiography (OR: 8.45; 95% CI: 2.57–27.8) and echocardiography (OR: 20.0; 95% CI: 9.57–41.7), for patients with cardiac syncope. Patients with cardiac syncope were also more likely to require extended ED observation (≥ 12 h; OR: 8.79, 95% CI: 4.38–17.6), post–syncope pacemaker implantation (OR: 63.0, 95% CI: 18.5–245), and hospitalization (OR: 31.7, 95% CI: 14.1–71.3). Table 2 compares the clinical and diagnostic characteristics of patients with neurally mediated or orthostatic syncope and those with cardiac syncope. Table 3 shows the cardiac findings according to syncope classification at admission. Table A1 in the Appendix shows no significant associations between the characteristics of neurally mediated and orthostatic syncope. Although a slight difference in hemoglobin levels was observed (OR: 0.98; 95% CI: 0.97–0.998), this represents an average variation of approximately 6 g/L, which is clinically negligible and does not justify the diagnostic or therapeutic changes.
Table 2.
Association between syncope risk classification on initial evaluation and clinical and diagnostic characteristics
| Characteristics and Results of ED Observation and Findings | All N = 400 |
Neurally mediated and orthostatic syncope | Cardiac syncope | OR (95% CI) | P–value |
|---|---|---|---|---|---|
| n = 356 (89%) | n = 44 (11%) | ||||
| Sex (females) | 213 (53.25%) | 190 (53.37%) | 23 (52.27%) | 0.96 (0.51–1.81) | 0.890 |
| Age, years; mean (SD) | 63.01 (22.37) | 61.28 (22.62) | 77.02 (13.92) | 1.04 (1.02–1.07) | < 0.001 |
| Previous pacemaker or ICD | 19 (4.75%) | 13 (3.65%) | 6 (13.64%) | 4.17 (1.50–11.6) | 0.014 |
| Previous history of syncope | 105 (26.25%) | 84 (23.60%) | 21 (47.73%) | 2.96 (1.56–5.61) | 0.001 |
| Autonomic Symptoms a | 324 (81.00%) | 303 (85.11%) | 21 (47.73%) | 0.16 (0.08–0.31) | < 0.001 |
| Hypotension | 26 (6.50%) | 24 (6.74%) | 2 (4.55%) | 0.66 (0.15–2.89) | 0.631 |
| Cardiac Murmurs | 35 (8.75%) | 21 (5.9%) | 14 (31.82%) | 7.39 (3.35–16.1) | < 0.001 |
| Prolonged PR interval (> 200 ms) | 32 (9.36%) | 20 (6.56%) | 12 (32.43%) | 6.84 (3.00–15.6) | < 0.001 |
| Prolonged QRS duration (> 100 ms) | 34 (9.44%) | 22 (6.94%) | 12 (27.91%) | 5.19 (2.34–11.5) | < 0.001 |
| Prolonged QT Interval (> 450 ms) | 16 (4.98%) | 12 (4.29%) | 4 (9.76%) | 2.41 (0.74–7.88) | 0.173 |
| Hemoglobin, g/L Mean (SD), n = 343 | 129.60 (17.29) | 129.12 (17.35) | 132.86 (16.70) | 1.01 (0.99–1.03) | 0.180 |
| Creatinine, mg/dL. Mean (SD), n = 340 | 1.07 (0.60) | 1.06 (0.61) | 1.11 (0.55) | 1.12 (0.69–1.81) | 0.648 |
| Cardiac findings at follow-up in the emergency department | 55 (13.75%) | 24 (6.74%) | 31 (70.45%) | 33.0 (15.3–71.2) | < 0.001 |
| Other tests conducted during the observation | |||||
| Chest radiography | 261 (65.25%) | 220 (61.80%) | 41 (93.18%) | 8.45 (2.57–27.8) | < 0.001 |
| CT scan | 50 (12.50%) | 40 (11.24%) | 10 (22.73%) | 2.32 (1.07–5.06) | 0.045 |
| Cardiac MRI | 2 (0.50%) | 2 (0.56%) | 0 (0.00%) | –* | 0.792 |
| Coronary angiography | 5 (1.25%) | 0 (0.00%) | 5 (11.36%) | –* | < 0.001 |
| Echocardiography | 47 (11.75%) | 22 (6.18%) | 25 (56.82%) | 20.0 (9.57–41.7) | < 0.001 |
| Tilt Test | 3 (0.75%) | 3 (0.84%) | 0 (0.00%) | –* | 0.704 |
| ED Stay ≥ 12 h | 107 (26.75%) | 76 (21.35%) | 31 (70.45%) | 8.79 (4.38–17.6) | < 0.001 |
| ED stay, mean (SD) n = 363b. | 7.69 (7.58) | 13.63 (7.58) | 1.07 (1.03–1.12) | 0.002 | |
| Pacemaker implantation following syncope | 19 (4.75%) | 3 (0.84%) | 16 (36.36%) | 63.0 (18.5–245) | < 0.001 |
| Hospital Admission | 37 (9.25%) | 13 (3.65%) | 24 (54.55%) | 31.7 (14.1–71.3) | < 0.001 |
a: Dizziness, warm sensation, sweating, nausea or vomiting
b: Calculated only for patients referred to the home, excluding hospitalized patients
*: 95% confidence interval not computed due to complete or quasi-complete separation (zero cell count)
The percentages represent the proportions within each group. ICD: Implantable Cardioverter–Defibrillator. CT: Computed tomography. MRI: magnetic resonance imaging
Table 3.
Distribution of cardiac findings according to the syncope type
| Cardiac findings | All | Neurally mediated and orthostatic syncope | Cardiac syncope | P-value |
|---|---|---|---|---|
| N = 400 | n = 356 | n = 44 | ||
| Cardiac Rhythm Disorder | < 0.001 | |||
| No cardiac rhythm disorder | 353 (88.25%) | 335 (94.10%) | 18 (40.91%) | |
| Bradycardia | 25 (6.25%) | 7 (1.97%) | 18 (40.91%) | |
| Tachycardia | 20 (5.00%) | 14 (3.93%) | 6 (13.64%) | |
| Tachycardia-Bradycardia Syndrome | 2 (0.50%) | - | 2 (4.55%) | |
| Structural cardiogenic disorder | < 0.001 | |||
| No structural cardiogenic disorder | 389 (97.25%) | 353 (99.16%) | 36 (81.82%) | |
| Severe aortic stenosis | 4 (1.00%) | 0 (0.00%) | 4 (9.09%) | |
| Severe Valvopathy | 1 (0.25%) | 1 (0.28%) | - | |
| Hypertrophic Cardiomyopathy | 1 (0.25%) | 1 (0.28%) | - | |
| Ischemic heart disease (IHD) | 4 (1.00%) | 1 (0.28%) | 3 (6.82%) | |
| Dilated Cardiomyopathy | 1 (0.25%) | - | 1 (2.27%) | |
| Structural Cardiogenic Disorder with Bradycardia | 2 (0.50%) | - | 2 (4.55%) | 0.012 |
| Structural cardiogenic disorder with tachycardia | 2 (0.50%) | - | 2 (4.55%) | 0.012 |
| Structural Cardiogenic Disorder with Tachycardia-Bradycardia Syndrome | 1 (0.25%) | - | 1 (2.27%) | 0.110 |
Values are presented as n (%). The P-values indicate comparisons between the Neurally mediated/orthostatic and cardiac syncope groups. Percentages represent proportions within each group. “–” indicates zero cases
Observation and monitoring time in the emergency department
Patients with ED stay ≥ 12 h were older than those discharged earlier (mean age: 75.0 vs. 58.6 years; OR for age: 1.04 per year, 95% CI: 1.03–1.06). Prior pacemaker or ICD use was more prevalent among patients with prolonged stays (10.3% vs. 2.7%; odds ratio [OR]: 4.08, 95% CI: 1.60–10.4), as was a previous record of syncope (36.5% vs. 22.5%; OR: 1.97, 95% CI: 1.22–3.19). Autonomic symptoms were associated with the observation < 12 h group (85.3% vs. 69.2%), with an OR of 0.39 (95% CI: 0.23–0.65). The initial classification of cardiac syncope was strongly associated with prolonged ED observation (28.9% vs. 4.4%; OR, 8.79; 95% CI, 4.38–17.6). Clinical findings, such as cardiac murmurs (OR: 17.0, 95% CI: 4.84–59.6), prolonged PR and QRS intervals, and QT prolongation, were more prevalent in the observation group with ≥ 12 h of observation. Cardiac findings were significantly more frequent among patients with prolonged ED stays (32.7% vs. 6.8%; OR: 6.64, 95% CI: 3.61–12.2), pacemaker implantations (14.9% vs. 1.0%; OR: 17.0, 95% CI: 4.84–59.6), and hospital admissions (32.7% vs. 0.7%; OR: 70.7, 95% CI: 16.6–301). Table 4 shows the analysis results for the observation duration.
Table 4.
Association between time spent in the ED and clinical, diagnostic characteristics, and results of the observations
| Characteristics and Results of ED Observation and Findings | ED stay<12 h | ED stay≥12 h | OR (95% CI) | P–value |
|---|---|---|---|---|
| n = 293 (73.25%) | n = 107 (26.75%) | |||
| Male sex | 154 (52.56%) | 59 (55.14%) | 1.11 (0.71–1.73) | 0.65 |
| Age, years; mean (SD) | 58.63 (23.02) | 75.00 (15.00) | 1.04 (1.03–1.06) | < 0.001 |
| Previous use of a pacemaker or ICS | 8 (2.73%) | 11 (10.28%) | 4.08 (1.60–10.4) | 0.004 |
| Previous history of syncope | 66 (22.53%) | 39 (36.45%) | 1.97 (1.22–3.19) | 0.006 |
| Autonomic Symptoms | 250 (85.32%) | 74 (69.16%) | 0.39 (0.23–0.65) | < 0.001 |
| Hypotension | 13 (4.44%) | 13 (12.15%) | 2.98 (1.33–6.65) | 0.010 |
| Cardiac classification of syncope on initial evaluation | 13 (4.44%) | 31 (28.97%) | 8.79 (4.38–17.6) | < 0.001 |
| Cardiac Murmurs | 3 (1.02%) | 16 (14.95%) | 17.0 (4.84–59.6) | < 0.001 |
| Prolonged PR interval (> 200 ms) | 19 (6.48%) | 16 (14.95%) | 2.54 (1.25–5.14) | 0.012 |
| Prolonged QRS duration (> 100 ms) | 13 (5.20%) | 19 (20.65%) | 4.74 (2.24–10.1) | < 0.001 |
| Prolonged QT Interval (> 450 ms) | 18 (6.87%) | 16 (16.33%) | 2.64 (1.29–5.43) | 0.01 |
| Hemoglobin, g/L Mean (SD) | 130.97 (17.13) | 126.5 (17.34) | 4.53 (1.60–12.9) | 0.131 |
| Creatinine, mg/dL. Mean (SD) | 1.03 (0.62) | 1.14 (0.54) | 1.33 (0.92–1.93) | 0.028 |
| Cardiac Finding | 20 (6.83%) | 35 (32.71%) | 6.64 (3.61–12.2) | < 0.001 |
| Pacemaker implantation following syncope | 3 (1.02%) | 16 (14.95%) | 17.0 (4.84–59.6) | < 0.001 |
| Hospital Admission | 2 (0.68%) | 35 (32.71%) | 70.7 (16.6–301) | < 0.001 |
ICD: implantable cardioverter–defibrillator. Percentages represent proportions within each group. a: dizziness, warm sensation, sweating, nausea, or vomiting
Clinicians detected cardiac findings in 26 patients (7.45%) who were monitored in the ED over 24 h. Of these, 40.9% (n = 18) were identified within the first 12 h of observation, corresponding to a cumulative incidence of 12.5%. By two hours, the cumulative incidence was 2.3% (95% CI: 0.9%–3.9%); this increased to 12.5% by 12 h (95% CI: 7.49%–18.9%) and reached 23.3% by 24 h (95% CI: 12.9%–35.3%). Figure 1 shows the cumulative and incidence rates of cardiac findings over a 24 h observation period. The stacked bars display the hourly incidence rates of cardiac syncope by type, with red and green representing cardiac syncope and orthostatic or neurally mediated syncope, respectively. The black line represents the cumulative incidence of cardiac findings during the 24-h observation period. The incidence rate rose steadily over the first 12 h, with minor peaks at approximately 2, 7, and 10 h, after which event detection declined. A slight apparent increase between hours 19 and 20 was due to a small number of late events, rather than an actual rise in incidence. By 12 h, more than 70% of all cardiac findings were identified, most of which occurred in patients with cardiac syncope.
Fig. 1.
Incidence of cardiac findings within 24 h of ED observation in patients with syncope. Stacked bars show the hourly incidence rate of newly detected cardiac findings during observation, stratified by the initial ED classification of syncope (red = cardiac syncope; green = orthostatic or neurally mediated syncope). The black line represents the cumulative incidence function, indicating the progressive detection of cardiac findings over time. Bars reflect the number of new cardiac findings within each time interval; patients were not reclassified according to the finding itself
Discussion
This study showed that most cases of syncope presenting to the ED were classified as neurally mediated or orthostatic syncope. These cases were commonly associated with autonomic symptoms and typically required 12 h of clinical observation. Patients with neurally mediated or orthostatic syncope underwent the same initial assessments. However, additional cardiac tests, such as echocardiography or chest radiography, were performed only when clinical or electrocardiographic abnormalities were identified. Some cases initially classified as neurally mediated or orthostatic syncope later showed cardiac findings.
Concordance between syncope classifications
The high concordance between the recorded classification in health records and European guidelines should not overshadow the notable disagreement between neurally mediated and orthostatic syncope. This pattern of misclassification was consistent with previous reports, suggesting that these two forms of syncope are often perceived as clinically similar, with overlapping features such as autonomic symptoms and rapid recovery without sequelae [12]. Although the observed discrepancies align with the low incidence of adverse outcomes, misclassification may result in suboptimal management, as patients with orthostatic hypotension may benefit from volume expansion and physical countermeasures, as well as from identification of the underlying cause [13]. In contrast, the management of neurally mediated primarily involves patient education, avoidance of known triggers, and reassurance [2].
Clinical characteristics associated with syncope type
Most patients were initially classified as having orthostatic or neurally mediated syncope, in contrast to a smaller proportion diagnosed with cardiac syncope, despite the latter group showing a markedly higher risk of hospital admission (OR 31.7). Syncope risk stratification tools estimate that 11%–23% of patients presenting to the ED are high-risk [13]. However, our findings indicate that prolonged observation in the ED was more commonly applied to patients initially classified as having cardiac syncope, reflecting the greater diagnostic complexity and need for targeted testing in this group. This group typically includes patients without autonomic prodromes or identifiable triggering events, with abnormal findings on physical examination or electrocardiography, or with a history of structural heart disease. Therefore, we propose that advanced diagnostic investigations should be performed selectively based on clinical indications rather than applied systematically to all patients.
Observation time in the emergency department
Extended ED stays exceeding 12 h were significantly associated with older age, absence of autonomic prodromes, and cardiac findings during the initial assessment and subsequent hospital admission. Although cardiac findings continued to accumulate beyond 12 h, approximately 70% were detected within the first half of the observation period, and the rate of new diagnoses declined thereafter. Extending the observation period from 12 to 24 h increased the cumulative incidence but with a lower incremental relative yield. These findings suggest that a 12-hour observation strategy captures most clinically relevant events; however, prolonged monitoring is recommended for selected patients, particularly those with abnormal electrocardiograms, structural heart disease, or unclear initial classification [14]. Extended observation in this group enables comprehensive diagnostic assessment, supports the timely initiation of targeted treatment, and ultimately contributes to patient safety [15]. Conversely, a shorter observation period remains a reasonable and efficient approach for low-risk patients with neurally mediated or orthostatic syncope.
A small proportion of patients classified as having non-cardiac syncope subsequently required pacemaker implantation. These cases likely represent diagnostic reclassification following additional evaluation rather than true misclassification at presentation, as conduction abnormalities may not be evident on the initial ED assessment. These findings highlight the inherent diagnostic uncertainty of syncope at presentation and reinforce the importance of continued observation in selected patients until cardiac causes are excluded.
Strengths and limitations
This study is among the few that have assessed the precise timing of cardiac findings (identified arrhythmias or structural abnormalities) in patients with syncope during ED observation. However, some limitations of this study should be acknowledged. Because the exclusion criteria were based on ICD–10 codes and final diagnoses documented by the attending physicians, misclassification of the underlying cause of loss of consciousness cannot be entirely excluded. Similarly, although the initial syncope classification at ED arrival was analyzed, it was not used to determine inclusion or outcome. All patients coded with syncope (ICD–10: R55) were included, regardless of the presumed etiology, to ensure consistency with real-world diagnostic practice and minimize subjective interpretation by the research team. We assessed only ED-based cardiac findings, without post-discharge follow-up, and did not assess patients for late-presenting arrhythmic events. Although clinicians detected most cardiac findings within the first 12 h of observation, only one-quarter of the findings were detected by 24 h. Therefore, discharging patients within 12 h could theoretically miss a proportion of late-occurring events. We did not conduct any post-discharge follow-up. These considerations prevent any inference regarding the safety of discharge after 12 h observation period. Some associations showed substantial ORs with wide CIs, particularly for outcomes such as pacemaker implantation and hospital admission. These values likely reflect a sparse data bias owing the limited number of events in specific subgroups. Therefore, the magnitudes of these estimates should be interpreted with caution, emphasizing their direction rather than their absolute size. Reliance on the research team’s consensus rather than the attending physician’s judgment for the final syncope classification may limit the reflection of the whole clinical context, potentially affecting real-world applicability. Additionally, the absence of a dedicated syncope unit may have constrained prolonged monitoring and post-ED follow-up [2].
Conclusion
Most patients presenting to the emergency department with syncope were classified as having neurally mediated syncope, with high concordance between the established classification criteria and medical record data. Clinicians detected the most relevant cardiac findings within the first 12 h of observation; however, a meaningful proportion occurred beyond this period. These findings support a selective approach in which a 12 h observation period is appropriate for patients without predefined clinical conditions, electrocardiographic abnormalities, or suspected structural heart disease on initial assessment, while prolonged monitoring (≥ 12 h) may be considered for others.
Supplementary Information
Below is the link to the electronic supplementary material.
Author contributions
ACT and TLS conceived and designed the study. DCS supervised this study. ACT, TLS, BAB, LS, and LCG collected data. LGP analyzed the data and performed statistical analyses. PVP, IAL, and LGP drafted the manuscript, and all authors contributed substantially to revising it. All authors have read and approved the final version of this manuscript. LGP takes responsibility for the entire paper.
Funding
None.
Data availability
The dataset used and analyzed in this study is available from the corresponding author upon reasonable requests.
Declarations
Ethical approval
This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Hospital Clínic de Barcelona (reference number: HCB/2020/0692). The Ethics Committee waived the requirement for informed consent to participate, as permitted under Spanish and EU regulations, specifically, Law 14/2007 of July 3 on Biomedical Research, and Organic Law 3/2018 of December 5 on the Protection of Personal Data and Guarantee of Digital Rights. The waiver was granted because the study involved the analysis of anonymized data collected during routine care, with no impact on the participants’ rights or welfare.
Consent for publication
Not applicable.
Competing interests
LGDP reports receiving funding from Chiesi for an independent study unrelated to the present work. All other authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Möckel M, Catherine Janssens KA, Pudasaini S, Garcia-Castrillo Riesgo L, Moya Torrecilla F, Golea A, et al. The syncope core management process in the emergency department: a consensus statement of the EUSEM syncope group. Eur J Emerg Med. 2024;31:250–9. 10.1097/MEJ.0000000000001146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brignole M, Moya A, De Lange FJ, Deharo JC, Elliott PM, Fanciulli A, et al. 2018 ESC Guidelines for the diagnosis and management of syncope. Eur Heart J. 2018;39:1883–948. 10.1093/EURHEARTJ/EHY037. [DOI] [PubMed] [Google Scholar]
- 3.Heeger CH, Rillig A, Ouyang F, Kuck KH, Tilz RR. [Syncope: epidemiology, definition, classification, pathophysiology and prognosis]. Herz. 2014;39:423–8. 10.1007/S00059-014-4087-0. [DOI] [PubMed] [Google Scholar]
- 4.Reed MJ, Karuranga S, Kearns D, Alawiye S, Clarke B, Möckel M, et al. Management of syncope in the Emergency Department: a European prospective cohort study (SEED). Eur J Emerg Med. 2024;31:136–46. 10.1097/MEJ.0000000000001101. [DOI] [PubMed] [Google Scholar]
- 5.Probst MA, Kanzaria HK, Gbedemah M, Richardson LD, Sun BC. National Trends in Resource Utilization Associated with Emergency Department Visits for Syncope. Am J Emerg Med. 2015;33:998. 10.1016/J.AJEM.2015.04.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Numeroso F, Mossini G, Giovanelli M, Lippi G, Cervellin G. Short-term Prognosis and Current Management of Syncopal Patients at Intermediate Risk: Results from the IRiS (Intermediate-Risk Syncope) Study. Acad Emerg Med. 2016;23:941–8. 10.1111/ACEM.13013. [DOI] [PubMed] [Google Scholar]
- 7.Bayard M, Gerayli F, Holt J. Syncope: Evaluation and Differential Diagnosis. Am Fam Physician. 2023;108:454–63. [PubMed] [Google Scholar]
- 8.Sandhu RK, Sheldon RS. Syncope in the Emergency Department. Front Cardiovasc Med. 2019;6:497130. 10.3389/FCVM.2019.00180/BIBTEX. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Furlan L, Trombetta L, Casazza G, Dipaola F, Furlan R, Marta C, et al. Syncope time frames for adverse events after emergency department presentation: An individual patient data meta-analysis. Med (Lithuania). 2021;57:1235. 10.3390/MEDICINA57111235/S1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hospital Clínic de Barcelona. https://www.clinicbarcelona.org/en. Accessed 28 Feb 2022.
- 11.R Core Team. R Foundation for Statistical Computing. R: A Language and Environment for Statistical Computing. 2021.
- 12.Nwazue VC, Raj SR. Confounders of Vasovagal Syncope: Orthostatic Hypotension. Cardiol Clin. 2013;31:89. 10.1016/J.CCL.2012.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Thiruganasambandamoorthy V, Sivilotti MLA, Le Sage N, Yan JW, Huang P, Hegdekar M, et al. Multicenter Emergency Department Validation of the Canadian Syncope Risk Score. JAMA Intern Med. 2020;180:737–44. 10.1001/JAMAINTERNMED.2020.0288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Numeroso F, Casagranda I. Improving the approach to non-cardiac syncope in the emergency department observation unit. World J Emerg Med. 2023;14:128–32. 10.5847/WJEM.J.1920-8642.2023.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Canzoniero JVL, Afshar E, Hedian H, Koch C, Morgan DJ. Unnecessary hospitalization and related harm for patients with low-risk syncope. JAMA Intern Med. 2015;175:1065–7. 10.1001/JAMAINTERNMED.2015.0945. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset used and analyzed in this study is available from the corresponding author upon reasonable requests.