Tele‐Ultrasound Consult Implementation in a <scp>US</scp> Academic Emergency Department: A Pilot Feasibility Study

Point-of-care ultrasound (POCUS) is a time-saving and cost-effective diagnostic modality used in the delivery of care to patients in the emergency department (ED), yet its broader adoption is limited by insufficient training or supervision [1-4]. Remote ultrasound mentoring, called tele-ultrasound (tele-US), has the potential to bridge this gap in expertise and usage [5, 6]. Tele-US involves performance of POCUS by a clinician in one location and remote guidance by an expert in another location. In this study, we conducted a pilot feasibility study to preliminarily assess and refine a targeted approach to clinical utilization of tele-US. Using the implementation science framework, RE-AIM (Reach, Effectiveness, Adoption, Implementation and Maintenance) [7], we preliminarily evaluated feasibility and effectiveness of tele-US implementation in a tertiary care ED. We hypothesized that tele-US would be feasible in this environment and would lead to clinical effectiveness, specifically by reducing ED disposition time in patients presenting with cardiac complaints. This study was conducted from October 2023 to February 2024 at a US academic tertiary care ED with an annual patient census of ~120,000. We prospectively enrolled a convenience sample of emergency medicine (EM) clinicians who used the tele-US consult and patients (> 18 years) requiring a POCUS exam in the ED as determined by their treating EM clinician. Our institutional review board approved the study, and verbal informed consent was obtained from patients. The tele-US system consisted of four Butterfly iQ+ probes with Butterfly Tele-guidance software (Guilford, CT, USA) attached to iPads (Apple, Cupertino, CA) mounted on stands. An ultrasound faculty member with emergency ultrasound fellowship training was on call on weekdays from 12 to 3 pm or 3 to 6 pm, with these times chosen based upon needs assessment with EM clinicians prior to tele-US implementation (Data S1). Tele-US faculty received support through an institutional grant, and biomedical support was available through the ED. The tele-US faculty could remotely adjust study type, gain or depth, and annotate in real time or add color doppler. Scans were saved to the Butterfly iQ+ cloud and were available through Epic electronic medical record (EMR). Immediately after the tele-US consult, the tele-US faculty on-call conducted a post-call survey with the bedside clinician. In the RE-AIM implementation framework, we defined Reach as the proportion of clinicians that used tele-US. Effectiveness referred to the impact of tele-US on time to ED disposition. Adoption referred to the type of clinicians who used tele-US, the cases where tele-US was used, and the rate of tele-US use compared to POCUS and TTE. Implementation referred to clinician acceptability of tele-US images and external factors such as technical challenges that impact tele-US use. Maintenance was not studied as the pilot was for 5 months. To study effectiveness, we examined differences in time from triage to ED disposition (primary outcome) by imaging modality among patients who received a cardiac ultrasound. Tele-US was treated as the primary exposure, with POCUS and transthoracic echocardiography (TTE) as comparators. Data were queried from Epic EMR to identify patients who received a cardiac POCUS or TTE between October 2023 and February 2024 in the ED and combined with our tele-US database. Extracted variables included ED triage time, chief complaint, demographic information, relevant past medical history, and disposition time. We used the Charlson Comorbidity Score (based on diagnoses in patients' problem list) to adjust for comorbidity [8]. A descriptive analysis of this cohort was performed, evaluating patient demographics and visit characteristics including chief complaint (cardiac vs. non-cardiac). The association between exam type (categorized as tele-US, POCUS, or TTE) and time to ED disposition was evaluated using a Poisson Generalized Estimating Equations (GEE) model to account for the skewed distribution of the outcome and correlation introduced into the framework by authorizing provider. The following potential confounders were identified and adjusted for in the model: age (18–65, 65+), gender (male, female, other), combined race/ethnicity (Hispanic, non-Hispanic White, non-Hispanic Black, other), Charlson Comorbidity Score, exam order time (7a-3p, 3p-11p, 11p-7a), and whether or not the patient received an additional cardiac imaging exam after the initial exam. Coefficients from the regression model are presented as rate ratios with 95% confidence intervals (95% CI). A 13-item survey (Data S2) was developed and reviewed by the study team to assess Reach, Adoption, and feasibility. The first four questions addressed patient demographics and requesting clinician information. The remaining questions covered the presenting complaint, type of POCUS study requested, clinical question, US findings, impact on patient care and clinician confidence, adequacy of image quality, and technical challenges (free text). The survey was conducted via REDCap, a secure platform for survey management [9]. A descriptive analysis was performed for the survey responses. Age, sex, presenting complaint, and responses related to image acceptability and themes related to technical challenges were summarized using frequency counts and percentages. All statistical analyses were performed using R (version 4.3.1), a widely used statistical language, with the “comorbidity” package for the Charlson Comorbidity Index and “gee” for generalized estimating equations regression analyses. At the time of the study, 210 emergency medicine clinicians worked in the ED: 60 residents, 80 advance practice providers (APPs), and 70 attendings. During the study period, 76 tele-US consults were performed by 45 emergency medicine clinicians (Data S3). Reach was 4% (3/70), 71% (57/80), and 26% (16/60) among the attendings, APPs, and residents, respectively. In evaluating the effectiveness outcome, 54 cardiac tele-US exams were performed during the tele-US consult pilot. Fifty cardiac exams were included in the final analysis. There were 2293 cardiac POCUS and 1193 TTE exams performed for patients presenting with cardiac complaints during the study period, which includes times with and without tele-US availability (Data S4). The median times to disposition (IQR) were 310 (296) minutes for tele-US, 299 (342) minutes for POCUS, and 414 (1160) minutes for TTE (Data S4). From the adjusted Poisson GEE model, time to disposition was estimated to be 1.95 times longer for patients who received a TTE compared to patients who received a tele-US, 95% CI [1.47, 2.59] (Table 1). ED disposition times for tele-US and POCUS were similar. For adoption, 75% of the tele-US studies (57/76) were by APPs, 21% (16/76) were by residents and 4% (3/46) were attendings. The most common presenting complaint was shortness of breath (30.3%; 23/76) and chest pain (25.0%; 19/76) (Data S3). The most frequent tele-US study was cardiac (69.7%; 53/76), followed by 17.1% (13/76) for lung. For the cardiac studies, the rate of tele-US use compared to POCUS and TTE exams ordered during tele-US availability was 7.5% (53/702) and 15.8% (53/336), respectively. In assessing tele-US acceptability, for 90.8% (69/76) of consults, clinicians stated the image quality was sufficient to make interpretation. In 53.0% (40/76) of consults, clinicians reported technical challenges. The most common reported technical challenges were lagging or “glitchy” videos (32.5%, 13/40) and poor image quality (27.5%, 11/40) (Data S5). To our knowledge, this is the first study to study tele-US implementation in a tertiary care ED in the United States. Our study demonstrated that tele-US, using the Butterfly Tele-guidance software, was feasible in this high-volume ED. Although not powered for effectiveness testing, we observed a decrease in ED disposition time compared to comprehensive TTE. While prior studies in community and academic EDs have shown that POCUS decreases ED length of stay [1, 10], this is the first study to show that tele-US may decrease ED disposition time. This finding might be explained by tele-US decreasing the need for emergent advanced imaging leading to expedited diagnosis, faster treatment times, and decreasing time to disposition. Such advanced or comprehensive imaging studies may be canceled, deferred, or deemed less urgent because the bedside clinician felt confident with POCUS findings guided by tele-US support. This study has some limitations. First, the number of tele-US cases was relatively small compared to POCUS cases (76 vs. 2293) during the study period, and no formal sample size calculation was performed given the feasibility design and restricted timeframe. However, the analytic results reflect the entirety of available data. Despite adjusting for provider and comorbidities, selection bias may remain as clinicians who were more comfortable with POCUS or tele-US may have been more likely to use it and less acutely ill patients may have been more likely to receive tele-US. Secondly, the study was conducted at a single tertiary care ED with a robust POCUS infrastructure, limiting generalizability to other settings. Third, feasibility assessment relied heavily on clinician surveys, which may be subject to response bias, as those with positive experiences were more likely to respond. We also did not capture the severity of reported technical issues, the number of unsuccessful tele-US attempts or how clinician POCUS competency influenced use. Finally, we did not evaluate cost effectiveness of tele-US implementation and could not assess Maintenance in the RE-AIM framework given the limited timeframe. In summary, this study found that tele-US is feasible in a high-volume US tertiary care ED, with potential impact on reducing ED disposition time compared to comprehensive TTE. Future research should look at its impact on patient outcomes in a powered trial and whether it can be scaled to other ED settings. O.F.E.: study concept, design, writing and revising the manuscript, analysis and interpretation of the data, acquisition of funding. M.M.: data analysis and interpretation, critical revision of the manuscript. K.U.: revision of the manuscript. Y.H.: revision of the manuscript. H.S.: critical revision of the manuscript. A.L.: critical revision of the manuscript. J.E.H.: study concept, design, critical revision of the manuscript, acquisition of funding. O.F.E. takes responsibility for the paper as a whole. This work was supported by Clinician–Teacher Development Award from the Massachusetts General Hospital, Center for Diversity and Inclusion, and Mass General Physicians Organization (239054). O.F.E. consults for Caption Health/GE Health and Sonio. The other authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. Appendix S1: acem70209-sup-0001-Supinfo1.docx. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.