Washington University Emergency Medicine Journal Club– October 19th, 2022
Vignette
Ms. Jones is a70-year-old female with a history of HTN and type II DM who presents to the ED after a syncopal event which occurred 30 minutes prior to arrival. The event was witnessed by the patient’s husband who states she had collapsed soon after getting up from her recliner to go to the restroom. He notes that she fell against the wall and landed on her buttocks. She did not hit her head and had no seizure activity. The patient regained consciousness within seconds and returned to a normal mental status. She does not remember the event and denies any precipitating symptoms, specifically no chest pain, shortness of breath, palpitations, headache, dizziness or vision changes. She denies any recent illness and her ROS is negative. She denies any hx of CAD, CHF or CVA/TIA. She is on Toprol XL and glyburide.
On exam the patient is awake, alert and oriented x 3. She is afebrile with a BP of 145/89, HR 68, RR 14 and a SaO2 of 98% on room air. Her d-stick is 122. The remainder of her physical exam is normal. Her ECG and labs are unremarkable including a Hct of 38%, a normal BMP, and negative initial troponin. Given her age, you decide to admit the patient for cardiac monitoring and serial cardiac enzymes. The Medicine Triage Attending calls you back ten minutes later and asks if you considered applying the Canadian Syncope Risk Score for this patient, which would indicate a low risk of 30-day serious adverse events. Based on this, she asks if you would be willing to send the patient home to follow-up in the medicine clinic within the next week. You ask your attending, who counters with the following: “Have any studies validated that this rule is safe and effective to use in our population?” Knowing that this is a valid question, you turn your attention to the medical literature…
PICO Question
Population: Adult patients presenting to the ED with syncope
Intervention: Use of the Canadian Syncope Risk Score to identify high-risk syncope patients
Comparison: Unaided clinical intuition (physician gestalt) or other syncope risk scores
Outcome: Serious adverse outcome within 30 days (death, myocardial infarction,
arrhythmia, structural heart disease, pulmonary embolism, serious hemorrhage and
procedural interventions)
Search Strategy
PubMed was searched using the terms “Canadian Syncope Risk Score,” resulting in
58 citations (https://tinyurl.com/4km5ejvz). Along with the original derivation
study, two validation studies were chosen for inclusion. A 4th article not related to
the Canadian Syncope Risk Score, but instead assessing the beneKit of admission in
older patients, was selected as well.
Article 1: Thiruganasambandamoorthy V, Kwong K, Wells GA, Sivilotti MLA, Mukarram M, Rowe BH, Lang E, Perry JJ, Sheldon R, Stiell IG, Taljaard M. Development of the canadian syncope risk score to predict serious adverse events after emergency department assessment of syncope. CMAJ. 2016;188(12):E289-98. Answer Key.
Article 2: Thiruganasambandamoorthy V, Sivilotti MLA, Le Sage N, Yan JW, Huang P, Hegdekar M, Mercier E, Mukarram M, Nemnom M-, McRae AD, Rowe BH, Stiell IG, Wells GA, Krahn AD, Taljaard M. Multicenter emergency department validation of the canadian syncope risk score. JAMA Intern Med . 2020;180(5):737-44. Answer Key.
Article 3: Zimmermann T, du Fay de Lavallaz J, Nestelberger T, Gualandro DM, Lopez-Ayala P, Badertscher P, Widmer V, Shrestha S, Strebel I, Glarner N, Diebold M, Miró Ò, Christ M, Cullen L, Than M, Javier Martin-Sanchez F, Di Somma S, Frank Peacock W, Keller DI, Bilici M, Costabel JP, Kühne M, Breidthardt T, Thiruganasambandamoorthy V, Mueller C, Belkin M, Leu K, Lohrmann J, Boeddinghaus J, Twerenbold R, Koechlin L, Walter JE, Amrein M, Wussler D, Freese M, Puelacher C, Kawecki D, Morawiec B, Salgado E, Martinez-Nadal G, Inostroza CIF, Mandrión JB, Poepping I, Rentsch K, von Eckardstein A, Buser A, Greenslade J, Reichlin T, Bürgler F, BASEL IX Investigators. International validation of the canadian syncope risk score A cohort study. Ann Intern Med . 2022;175(6):783-94. Answer Key.
Article 4: Probst MA, Su E, Weiss RE, Yagapen AN, Malveau SE, Adler DH, Bastani A, Baugh CW, Caterino JM, Clark CL, Diercks DB, Hollander JE, Nicks BA, Nishijima DK, Shah MN, Stiffler KA, Storrow AB, Wilber ST, Sun BC. Clinical benefit of hospitalization for older adults with unexplained syncope: A propensity-matched analysis. Ann Emerg Med . 2019;74(2):260-9. Answer Key.
Bottom Line
Syncope is a common presenting complaint, representing 0.6% to 1.0% of ED visits
in North America (Thiruganasambandamoorthy 2008). In the US, nearly half of all
patients presenting with syncope end up admitted to the hospital (Reed 2010). Over
the last 2 decades, multiple clinical decision rules have been derived in an attempt to
risk stratify these patients and increase discharge rates without worsening
outcomes. One of the most notable of these is the San Francisco Syncope Rule, which
appeared quite sensitive on initial derivation (Quinn 2004) and initial validation
(Quinn 2006). Unfortunately, subsequent attempts at validation in a separate ED
population failed to reproduce the results of these initial studies (Birnbaum 2008).
Over the subsequent years, several other clinical decision rules have been evaluated
(OESIL, the Boston Syncope Rule, the Rose Score, STEPs, the EGSYS score), none of
which have been widely validated.
More recently, the Canadian Syncope Risk Score (CSRS) was derived in 6 large
Canadian EDs (Thiruganasambandamoorthy 2016). Among over 4000 patients
presenting with syncope with no serious adverse outcome identiKied in the ED, the
rule had a 99.2% sensitivity at a threshold of -2 or higher, and a sensitivity of 97.7%
at a threshold of -1 or higher, to predict the composite outcome of serious conditions
related to syncope within 30 days of disposition from the ED. This composite
included death, arrhythmia, myocardial infarction, serious structural heart disease,
aortic dissection, pulmonary embolism, severe pulmonary hypertension, severe
hemorrhage, subarachnoid hemorrhage, any other serious condition causing
syncope, and procedural interventions for the treatment of syncope.
!3
The derived rule is fairly simple, including cardiac history, low blood pressure, ECG
Kindings, troponin values, and clinician gestalt (see Kigure).
Subsequent prospective validation of the CSRS was undertaken in 9 large Canadian
EDs , i n c l u d i n g t h e 6 EDs i nvo lve d i n t h e de r iva t i o n s t u dy
(Thiruganasambandamoorthy 2016). Among 3800 patients in the Kinal analysis, the
rule had a 99.2% sensitivity at a threshold of -2 or higher and a sensitivity of 97.8%
at a threshold of -1 or higher to predict 30-day advert events. Among those with a
“very low” risk score (-3 or -2), the risk of a 30-day adverse event was 0.3%, while
the risk among those with a “low” risk score of 0 or -1 was 0.7%.
Further validation of the CSRS has now been undertaken in a large international
cohort conducted at EDs in Switzerland, Spain, Germany, Italy, Poland, New Zealand,
Australia, and one center in the US (Baylor College of Medicine) with 2283 patient
encounters included (Zimmerman 2022). For patients with a very low risk score
(CSRS ≤ 0), the risk of an adverse 30-day outcome was 1.1%, giving a negative
predictive value of 99% (95% CI 99% to 100%). Of note, the risk of an adverse 30-
day event among patients discharged was 1.6%, higher than among those with a
very low CSRS. The authors report that use of the CSRS score would have
theoretically led to a 14.9% absolute increase in the discharge rate compared to
standard practice, while still reducing the risk of an adverse 30-day event.
While the evidence thus far suggests that the CSRS is highly sensitive at predicting
30-day adverse events, with potential reductions in both admission rates and event
rates, there is still a paucity of evidence among US patients. Only one US medical
center was included in the international validation study, and only Canadian
hospitals were involved in the derivation and initial validation studies. Canadian and
European health systems are quite different from healthcare in the US, where many
patients are underinsured or lack quick access to primary care follow-up. Lack of
easy follow-up often leads to higher admission rates in the US compared to other
countries, and this disparity will need to be accounted for when considering any
syncope clinical decision rule. Further research in the US will need to conKirm the
!4
theoretical Kindings from the international validation before a large number of US
emergency physicians are willing to adopt the CSRS.
We reviewed one Kinal study attempting to evaluate the potential beneKits of hospital
admission following syncope (Probst 2019). This secondary analysis of data
collected prospectively in 11 academic US EDs compared the rates of 30-day adverse
outcomes among older patients (60 or older) who were admitted or discharged
following a presentation for syncope. There were 2500 patients enrolled, of whom
75% were admitted. Following propensity score matching for predictive
confounders, the risk of a serious adverse event at 30 days was found to be slightly
higher among hospitalized patients, although this difference was not statistically
signiKicantly: risk difference 2.07%, 95% CI -0.24% to 4.38%. While interesting, and
despite the use of propensity matching to attempt to balance the groups, this study
is at high risk for selection bias, as clinical gestalt concerning whether patients were
at higher or lower risk of adverse outcomes could not be accounted for with this
statistical technique. It is unlikely that this study alone would convince emergency
physicians not to admit most older patients with syncope.