Customize your JAMA Network experience by selecting one or more topics from the list below.
Verma AA, Masoom H, Rawal S, Guo Y, Razak F, for the GEMINI Investigators. Pulmonary Embolism and Deep Venous Thrombosis in Patients Hospitalized With Syncope: A Multicenter Cross-sectional Study in Toronto, Ontario, Canada. JAMA Intern Med. 2017;177(7):1046–1048. doi:10.1001/jamainternmed.2017.1246
The prevalence of pulmonary embolism (PE) among patients hospitalized with syncope is uncertain. The recently published Pulmonary Embolism in Syncope Italian Trial (PESIT)1 systematically evaluated patients hospitalized with a first episode of syncope and determined that 17.3% had a PE. It is not known how commonly patients hospitalized with syncope are investigated for PE or deep venous thrombosis (DVT) in routine practice.
We conducted a retrospective cross-sectional study at 4 hospital sites in Toronto, Ontario, Canada, that were participating in the General Medicine Inpatient (GEMINI) cohort study. The GEMINI study has linked electronic clinical data from hospital information systems with administrative data to study patients cared for by general medicine services in hospitals affiliated with the University of Toronto. We applied the same inclusion and exclusion criteria as PESIT in all general medicine patients hospitalized through the emergency department between April 1, 2011, and March 31, 2015. In Canada, hospitals use International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), codes to designate 1 primary diagnosis for the emergency department visit and 1 primary diagnosis at hospital discharge for mandatory reporting to the Canadian Institute for Health Information. The emergency department and hospital discharge diagnoses are not necessarily the same. We included patients if their primary diagnosis was syncope in either the emergency department or at hospital discharge, using the ICD-10 code R55. We excluded patients who were receiving anticoagulation at admission, had prior syncope, or were pregnant.
The main outcomes were the occurrence of investigation for venous thromboembolism (VTE) during hospitalization, defined as testing with a plasma D-dimer level, compression ultrasonography of the upper or lower limbs, computed tomographic pulmonary angiography (CTPA), or ventilation perfusion (V/Q) scan, and the diagnostic yield of these investigations. Because there is substantial overlap in the investigation and management of DVT and PE, we examined a composite of all investigations for VTE. The GEMINI study received research ethics board approval from all participating hospitals.
Of the 1650 patients hospitalized with syncope, 345 were excluded (179 with preadmission anticoagulation and 166 with previous syncope), and the final cohort included 1305 patients. At least 1 investigation for VTE was performed in 146 patients (11.2% [95% CI, 9.6%-13.0%]) (Table 1), including 120 who received imaging with CTPA, V/Q scan, and/or compression ultrasonography (9.2% [95% CI, 7.8%-10.9%]).
Pulmonary embolism was diagnosed in 11 of the 73 patients who received a CTPA or V/Q scan, and DVT was diagnosed in 10 of the 67 patients who received ultrasonography, representing a diagnostic yield of 15.1% (95% CI, 8.6%-25.0%) and 14.9% (95% CI, 8.3%-25.4%), respectively. The prevalence of VTE in the cohort was 1.4% (95% CI, 0.9%-2.2%) (Table 2).
In a large observational study, VTE was investigated in 11.2% of patients hospitalized with syncope. The diagnostic yield for VTE was 15.0% and the prevalence was 1.4%, not 41.9% and 17.1%, respectively, as found in the PESIT study.1 Our findings call into question the generalizability of the PESIT results. The 1.4% prevalence of VTE we found is consistent with previously published estimates, which range from 0.9% to 2.8%.2-4 Unlike the case in PESIT, all patients were not screened for VTE in our study or in these previous studies; thus, the prevalence may have been underestimated. A more illuminating comparison, however, is to examine the diagnostic yield of imaging for VTE. We expected that thrombosis would be more likely in the selected patients who received imaging for VTE based on clinical judgment in our study, not in the large proportion (40.9%) of patients who received imaging based on screening, as was the case in PESIT. The reverse was true, highlighting the differences in the study populations.
A limitation of our study is the use of administrative data to identify patients with syncope. Thus, some patients who were hospitalized for syncope may have been missed. The ICD-10 code for syncope is very specific, with a reported positive predictive value of 93% in the emergency department and 95% at hospital discharge.5 Although the ICD-10 code assigned at hospital discharge is only 63% sensitive for detecting patients hospitalized with syncope,5 we identified 590 patients (36% of participants in our study) with only an emergency department diagnosis of syncope. Thus, we included substantially more patients than if we had used hospital discharge diagnosis alone and improved the completeness of our cohort.
Assessing all patients hospitalized with syncope for VTE would represent a substantial departure from conventional practice. Based on the available data, there is little, if any, justification for routine testing for VTE in all patients hospitalized with a first episode of syncope.
Corresponding Author: Amol A. Verma, MD, MPhil, Li Ka Shing Knowledge Institute, St Michael’s Hospital, 30 Bond St, Room 345, Toronto, ON M5B1W8, Canada (email@example.com).
Accepted for Publication: March 9, 2017.
Published Online: May 8, 2017. doi:10.1001/jamainternmed.2017.1246
Author Contributions: Drs Verma and Masoom had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Verma and Masoom are co–first authors.
Study concept and design: Verma, Masoom, Razak.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Verma, Masoom, Razak.
Critical revision of the manuscript for important intellectual content: Verma, Masoom, Rawal, Guo.
Statistical analysis: Guo, Razak.
Obtained funding: Verma, Razak.
Administrative, technical, or material support: Razak.
Study supervision: Verma.
Conflict of Interest Disclosures: None reported.
Group Information: The GEMINI Investigators include Janice L. Kwan, MD, MPH, Department of Medicine, Mount Sinai Hospital, Department of Medicine, University of Toronto, Toronto, Ontario, Canada; Lauren Lapointe-Shaw, MD, Department of Medicine, University of Toronto; Terence Tang, MD, Trillium Health Partners, Department of Medicine, University of Toronto; Adina Weinerman, MD, MHSc, Sunnybrook Health Sciences Centre, Department of Medicine, University of Toronto.
Create a personal account or sign in to: