Trends in Utilization and Yield of Inpatient Transthoracic Echocardiography for Syncope: Impact of Cardiac Biomarkers

Transthoracic echocardiography (TTE) is frequently obtained in the inpatient evaluation of syncope, despite 2017 American College of Cardiology/American Heart Association/Heart Rhythm Society syncope guidelines (2017 Guidelines) advising TTE only if cardiac etiology is suspected after initial “history, physical examination, or electrocardiogram (ECG).”¹ Since the 2017 Guidelines, cardiac biomarkers have been shown to predict prognosis in syncope patients and are now commonly obtained.^2,3 We sought to assess the impact of cardiac biomarkers, including N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin T (hscTnT), on TTE utilization and yield among hospitalized patients with syncope.

Inpatients with TTEs for syncope (International Classification of Diseases-10 = R55) at the University of Chicago Medicine after 2017 Guidelines publication (August 2017-January 2020) were assessed for suspicion of cardiac etiology prior to TTE. Cardiac suspicion after “standard evaluation” (SE), as advised by the 2017 Guidelines, was defined as the presence of known cardiovascular disease, an abnormal cardiovascular physical examination, or an abnormal ECG. “Biomarker-only” cardiac suspicion was defined as abnormal cardiac biomarkers, with normal SE. The Institutional Review Board approved this study.

Abnormal cardiovascular examination (murmur, jugular venous distension, rales, edema), ECG (heart rate < 50 or >110, QRS ≥120 msec, any fascicular block, Mobitz type II or third-degree atrioventricular block, more than one premature ventricular contraction, corrected QT >480 msec, Q-waves or ST-T abnormalities, or any impression of “infarction” or “ischemia,” Wolff-Parkinson-White, or Brugada pattern), and any known cardiovascular disease (coronary artery disease, congestive heart failure, arrhythmia, valve disease, cardiomyopathy, congenital heart disease, prior cardiac procedure or device) were recorded.

Abnormal biomarkers were defined according to the reference standard in our electronic medical record system, as NT-proBNP ≥ 125 pg/mL, cardiac troponin T (cTnT) > 0.1 ng/mL, or hs-cTnT ≥ 14 ng/L. On August 27, 2018, Hs-cTnTreplaced cTnT for all troponin testing at our institution.

A “clinically important abnormal TTE” (CI-ABN TTE) was defined as left ventricular ejection fraction ≤ 40%, moderate or higher RV dysfunction, moderate or higher valvular stenosis or regurgitation, moderately sized or greater pericardial effusion, pulmonary artery systolic pressure ≥50 mm Hg, regional wall motion abnormality, any vegetation/thrombus/mass, severe left ventricular hypertrophy, or intracavitary gradient > 20 mm Hg.

Of 23,467 inpatient TTEs during the study period, 758 had a first TTE for syncope and formed the study group. Of these, 24% had a CI-ABN TTE.

A cardiac suspicion based solely on SE was present in 50% of patients, and 24% had biomarker-only suspicion, leaving 26% with no identifiable reason for suspicion of cardiac syncope (Figure 1). CI-ABN TTEs were found in 36%, 18%, and 6% of SE suspicion, biomarker-only suspicion, and no suspicion patients, respectively (Figure 1).

Figure 1.

Suspicion of a cardiac etiology of syncope and frequency of CI-ABN TTEs after the 2017 Guidelines.

Cardiac biomarkers were routinely obtained in the initial syncope evaluation (NT-proBNP, 53%; troponins, 93%). NT-proBNP testing was abnormal in 72% of those tested. After hs-cTNT replaced cTnT at our institution, abnormal troponin results increased six-fold (59% of those tested with hs-cTnT vs 10% with cTnT [P < .001]). The change from cTnT to hs-cTnT testing also resulted in increased biomarker-only suspicion TTEs (14% vs 32%, P < .001) and decreased no cardiac suspicion TTEs (37% vs 19%, P < .001), with no change in SE suspicion TTEs. After hs-cTnT was introduced, only 4% of no cardiac suspicion patients had a CI-ABN TTE.

We recently reported that hospitalized patients with syncope and normal initial evaluation have a low likelihood of having an abnormal TTE, supporting the 2017 Guidelines recommendation for more selective TTE use.^1,4 Yet interestingly, TTEs ordered for syncope increased after the 2017 Guidelines.⁴ In this study, we identify routine ordering of cardiac biomarkers, and the adoption of hs-cTNT testing, as likely causes for this increase. We also find that cardiac biomarkers effectively stratify the likelihood of CI-ABN TTEs among inpatients with syncope.

The 2011 Appropriate Use Criteria for TTE established syncope as an “appropriate” indication in all patients.⁵ The 2017 Guidelines advise more selective TTE use, only if cardiac concerns are raised on initial SE, and discourage routine biomarker testing.¹ Since the 2017 Guidelines, biomarkers have been found to have prognostic value after syncope.^2,3 We found biomarker testing is now routine among syncope patients, and nearly a third appear to have positive biomarkers as the only abnormal cardiac finding prompting TTE.

Our results may inform inpatient TTE use for syncope. CI-ABN TTEs are common after an abnormal SE, supporting the 2017 Guidelines recommendation for TTE independent of biomarkers in these patients. If initial SE is normal, normal biomarkers identify patients in whom TTE is very low yield. Patients with positive biomarkers and normal SE largely account for increased TTE use and have an intermediate likelihood of a CI-ABN TTE. Further investigation of higher abnormal biomarker cutoffs might be a strategy to optimize TTE utilization.

We chose to report the group with SE suspicion only because the 2017 Guidelines advise using SE to guide the decision to obtain a TTE, while discouraging routine use of biomarkers.¹ Using this standard, half of inpatient TTEs for syncope in our institution do not have a current guideline-supported indication and thus potentially risk denial of reimbursement. We show that biomarker testing in syncope patients is now routine, likely due to new evidence supporting their prognostic significance since the 2017 Guidelines.^2,3 Thus, a majority of TTEs with a normal SE have abnormal biomarkers as the likely indication. Conversely, there is little debate that patients admitted for syncope with an abnormal SE warrant TTE, as the 2017 Guidelines already advise. While further study to define the optimal role of biomarker testing in the evaluation of syncope is needed, our findings suggest current guidelines need to be updated to reflect current practice.

Limitations include our retrospective study design, which may not capture all clinical factors that lead to TTE ordering. Additionally, our definition of CI-ABN TTE is necessarily subjective, as it is difficult to clearly attribute the cause of syncope to any one TTE finding.

In summary, these findings suggest that inpatient TTE utilization for syncope beyond the current guideline recommendations in response to cardiac biomarkers may be useful. However, optimal strategies incorporating cardiac biomarkers to avoid unnecessary TTE are needed. This is particularly important among inpatients, where TTEs may receive limited reimbursement in the era of value-based healthcare.