I’ve Syncopized and Got Up, but Can I Go Home? Evaluating the Validation of the Canadian Syncope Risk Score

Author: Jonathan McMahon, DO, MA

Edited by: Robert Allen, MD

Case

A 46-year-old female with a past medical history of hyperlipidemia presents to the ED with a chief complaint of syncope that happened about an hour ago. She states that she was walking to the kitchen when she suddenly lost consciousness. Her four friends were with her and were able to safely lower her to the ground to avoid traumatic injuries (as seen above). They tell you that she regained consciousness very quickly after she was laid down. They did not notice any shaking or incontinence. She had been feeling well and did not notice any prodromal symptoms, including headache, chest pain, shortness of breath, weakness, change in vision or hearing, or slurred speech.

Physical Exam:

Vitals: BP 110/76 mm Hg, HR 82/min, RR 16/min, T 98.9 F, SpO2 100% on RA

Gen: comfortable, speaking full sentences, AAOx3

HEENT: no lateral tongue lacerations or signs of biting, PERRL

Neck: No JVD. Supple.

CV: +S1/S2. No murmurs.

Resp: CTA BL. No chest wall tenderness.

Abd: soft, non-tender, non-distended.

Ext: No peripheral edema.

Neuro: CN II-XII grossly intact, no focal neurological deficits, 5/5 strength in extremities, normal finger-to-nose, negative pronator drift, negative Romberg.

So now your question is should she stay or should she go home?

Introduction

Half of those who present with syncope are hospitalized, and only 10% of those who had a syncopal event has a serious underlying condition.[2] Last month, our Food and Journal Club met and discussed the 2020 validation paper for the Canadian Syncope Risk Score (CSRS). The paper’s goal: find a way to risk-stratify this common complaint. Here are some of the highlights of the discussion.

Study design

Patients

This study was a prospective, cohort validation study involving nine Canadian emergency departments between March 2014 and April 2018.[2] Almost 4,000 subjects aged 16 or greater who presented to the ED within 24 hours of a syncopal event were eligible for the study. Exclusion criteria included those who had another cause of loss of consciousness, including head trauma, seizure, and prolonged loss of consciousness. Other exclusion criteria included if the syncope led to traumatic injuries and situations that prevented obtaining an accurate history, i.e. language barrier and substance use.

Outcomes

Primary outcome: serious conditions identified after 30 days

Results

Out of the 3,819 patients in the study, 139 had serious outcomes at 30 days (3.6%): 2.8% arrhythmic and 0.8% non-arrhythmic. 89.9% of the medium-, high-, and very-high-risk groups had serious arrhythmia. Of those, arrhythmia was usually identified within 15 days. There were a large number of patients who did not have troponin testing (41%) performed. There were statistically significant differences in the rates of serious outcomes associated with the different CSRS risk levels (p<0.001).

Pros

This was a large study that tried to validate a clinical decision tool for evaluating a patient with a prevalent presenting ED complaint. Two physicians, blinded to the CSRS scores and factors, independently determined if a serious outcome had occurred. The CSRS allowed for physician clinical judgment to be factored into the risk stratification, which previously had been shown to be correlated with serious outcomes in syncope.[3] For example, suspicion of vasovagal syncope decreased the score, and suspicion of cardiac syncope increased the score. Finally, the sources of funding had no impact on the study and its reporting.

Cons/Limitations

Twenty percent of the possible enrollees were not included in this study. Additionally, the Canadian Syncope Risk Score derivation paper was released in September 2016, during the time data was being collected for the validation paper.[4] While the authors did not include the score on their data collection papers, there is no way they could ensure blinding to the score. A “Hawthorne-like” effect could have impacted evaluation and disposition.[5] There was also a large amount of data missing: 4.9% of patients did not have an ECG reported, and 41% of patients did not have a troponin measured despite being needed to calculate the score. The authors claim they may have been missing in younger, healthier patients, and clinicians did not believe troponin or ECG were necessary for workup and disposition. This cannot be proven and may represent a partial verification bias.[6] This was especially important as an ECG alone is needed for four points on the score. The authors addressed one large limitation, missing troponins, by conducting a sensitivity analysis of non-troponin tested patients. This analysis showed that there were similar outcomes with random troponin results factored into the score.

The average age for the study was 53.9 years old, ranging from 16 to 101 years old. The score may have different outcomes based on age. It would be interesting to see a subgroup analysis of the data based on age. One more limitation was that the 95% confidence interval range of the sensitivity in the lowest-risk group had a lower boundary of 97.4%; therefore, 2.6% of patients may have a serious outcome despite having the lowest CSRS score. This may not be considered by some to be an acceptable false-negative rate.

Conclusions

It seems the authors have successfully validated the CSRS. Depending on the score, and thereby the risk, clinicians can use this tool to predict 30-day outcomes, especially in the very-low- and low-risk groups which had the largest number of patients. Given patients with very-low- and low-risk scores had such a small probability of a serious outcome (0.2% and 0.7%, respectively), the score can be used to justify discharge with good return precautions. The medium-risk subgroup had a higher risk of 30-day serious outcomes with 8%, so clinicians should strongly consider an observation period for these patients. However, there should be shared decision making with all patients, regardless of the score. As mentioned, the lowest score possible may carry an unacceptable false-negative rate of > 2%.

Follow up studies should include using the CSRS and observing the impact on patient care. Does it change patient-centered outcomes such as hospitalization rates at a long-term interval?

So What Happened to Our Patient?

During your workup, troponin level is negative and other labs are unremarkable. The patient’s BP comes back to normal after one liter of lactated ringers and time. After applying the CSRS, you find that this patient is at low risk for a serious outcome. After a discussion with the patient, she chooses to go agreeing to close follow-up with her primary care physician.

References

1. Jenkins, D. & Gerred, D. Normal adult 12-lead ECG. https://ecglibrary.com/norm.php. Published 2017. Accessed September 27, 2020.

2. Thiruganasambandamoorthy, T., et al. Multicenter emergency department validation of the Canadian Syncope Risk Score. JAMA Intern Med. 2020;180(5), 737-744. DOI: 10.1001/jamainternmed.2020.0288

3. Tora, C., et al. Syncope prognosis based on emergency department diagnosis: a prospective cohort study. Academic Emergency Medicine. 2017; 25(4). DOI: 10.1111/acem.13346

4. Thiruganasambandamoorthy, T., et al. Development of the Canadian Syncope Risk Score to predict serious adverse events after emergency department assessment of syncope. CMAJ. 2016;188(12), E289-E298. DOI: 10.1503/cmaj/151469

5. Spencer, E.A. & Mahtani, K. Hawthorne bias. Catalogue of Bias Collaboration. https://catalogofbias.org/biases/hawthorne-effect/. Published 2017. Accessed September 29, 2020.

6. Pluddermann, A., McCall, M., O’Sullivan, J., & Banerjee, A. Verification bias. Catalogue of Bias Collaboration. https://catalogofbias.org/biases/verification-bias/. Published 2019. Accessed October 2, 2020.