Syncope, a frequent diagnosis encountered in the emergency department (ED) and inpatient setting, creates risk assessment and disposition challenges. The San Francisco Syncope Rule (SFSR) was developed to help identify lower‐risk syncope patients less likely to have a serious outcome who might be considered for ED discharge. Large cohort validation data have been inconsistent and largely derived from academic center data. The presence of a prodrome, used in other validated risk stratification tools, portends improved outcomes in the setting of syncope.
The purpose of this pilot study was to identify low‐risk patients using a modified SFSR (MSFSR) that incorporates a prodrome and to evaluate resource utilization and costs for syncope patients evaluated in our ED.
We retrospectively evaluated all ED patients with a principle diagnosis of syncope from January through December 2009 at a 110‐bed community teaching hospital. Patients without documentation of loss of consciousness confirmed by a physician's history were excluded. Manual chart abstraction of patient data was performed. Patients were dichotomized as low (<1) or high (>1) modified SFSR, calculated by subtracting 1 from the SFSR when the patient had a documented prodrome. Basic statistics and simple chi‐square statistical analysis were used to compare patients with low and high MSFSR for specific outcomes. Three hundred patients with a principle diagnosis of syncope in 2009 were evaluated within our ED. One hundred and seventy‐four patients had confirmed syncope, of whom 136 (78%) had low MSFSR. Overall, the admission rate was high (33%), and expectedly MSFSR predicted patient admission (24% of low‐MSFSR patients admitted vs. 66% of high‐MSFSR patients, P < 0.0001). Use of neurologic imaging to evaluate syncope patients was high (79%) and not associated with MSFSR score (76% in low MSFSR vs. 89% in high MSFSR, P = 0.11). Among admitted patients, specialty consultation (e.g., cardiology or neurology) was high (78%), but MSFSR did not predict specialty consultation (80% of patients with low MSFSR vs. 76% of patients with high SFSR, P = 0.76). Whereas average cost of care was significantly different between MSFSR groups ($3066 for low MSFSR vs. $5682 for high MSFSR, P < 0.0001), this difference did not remain among admitted patients with low versus high MSFSR.
Although a high rate of hospital admission for syncope patients could be partially explained by risk score using an MSFSR, nearly a quarter of patients with low risk were still admitted. Neurologic imaging to evaluate syncope patients was extremely high and not mitigated in low‐risk patients. Significant opportunity exists to reduce unnecessary resource utilization and costs for syncope diagnosis in our small community hospital. Next steps include collaboration with ED personnel to prospectively measure facility and patient outcomes and institute tests of change in practice and consideration of a syncope clinical decision unit.
M. K. Moussa ‐ Section of Hospital Medicine, Division of General Medicine, Emory University School of Medicine, Atlanta, GA, author/no conflict of interest; F. Mufti ‐ Emory Johns Creek Hospital, Johns Creek, GA, coauthor/no conflict of interest; H. F. Shabbir ‐ Section of Hospital Medicine, Division of General Medicine, Emory University School of Medicine, Atlanta, GA, coauthor/no conflict of interest; D. D. Dressler ‐ Section of Hospital Medicine, Division of General Medicine, Emory University School of Medicine, Atlanta, GA, coauthor/no conflict of interest