Article Text

Download PDFPDF

14 Efficacy of handheld echocardiography at grading left ventricular and left-sided valvular dysfunction compared to standard transthoracic echocardiography before and after educational intervention at a central london teaching hospital
  1. Holli Evans1,
  2. Benjamin Low1,
  3. Balrik Kailey2,
  4. Punam Pabari1
  1. 1Imperial College Healthcare NHS Trust
  2. 2NIHR ACF Cardiology, Imperial College London


Introduction The use of handheld echocardiography (hTTE) is increasing, due to its diagnostic value in a number of clinical scenarios coupled with its availability, portability and relatively low cost1. The well-documented limitations of hTTE compared to standard TTE (sTTE) include operator experience in both image acquisition and interpretation2. Our first aim was to quantify the discrepancy when assessing left-sided cardiac pathology. Secondly, we aimed to identify whether this was amenable to improvement following educational intervention.

Methods We retrospectively identified 119 patients who underwent both hTTE and sTTE. Cardiology specialist trainees (STs) performed hTTE and this was compared to sTTE performed by blinded BSE accredited sonographers (gold standard). The parameters assessed and the grading system used is detailed below (table 1). Concordance between hTTE and sTTE was evaluated, both pre- and post-education, by the weighted Kappa statistic. Educational intervention included information given to all cardiology trainees highlighting both the overall cohort performance and confidential individualised feedback. Following this a further 29 patients who underwent both hTTE and sTTE were identified.

Abstract 14 Table 1

Severity grading system used to assess left ventricular systolic function (LVSF), mitral regurgitation (MR), aortic regurgitation (AR) and aortic stenosis (AS)

Results The average error for all parameters was 0.34 with fair agreement (k = 0.38). Figure 1 and Table 2 highlight the average error for each parameter. Where discrepancy existed, hTTE tended to underestimate severity of each parameter, with the exception of LVSF which was equivocal. The greatest discrepancy was seen when grading MR, and of 80 discordant scans, 57 were due to an underestimation (figure 2). The majority of error was in severe MR. As predicted, performance improved through years of training; ST3s consistently had the highest error compared to other grades, particularly when assessing MR. This improved significantly after education (0.66 to 0.33). Preliminary analysis post-education has shown an increase in overall agreement and a reduction in error when grading valvular pathology, but an increase in error when grading LVSD.

Abstract 14 Figure 1

Graph demonstrating average error in each parameter before and after intervention

Abstract 14 Table 2

Table showing average error for each parameter pre- and post-intervention with respective weighted kappa statistics (κ)

Abstract 14 Figure 2

Graph demonstrating the proportion of underestimation and overestimation in discordant scans across each parameter

Conclusion Our results are concordant with previous studies demonstrating that diagnostic accuracy of hTTE is heavily influenced by operator experience. It is less known how much training should be given to operators before hTTE assessment is reliable enough to base clinical decisions upon. However with simple education, we hope to demonstrate that discrepancy between hTTE and sTTE can be reduced. Moving forwards, we plan to introduce a dedicated training day for new ST3 cardiology trainees and observe how this influences performance. The discrepancy and underestimation with hTTE raises the question of clinical implications, particularly of underestimating MR. It may be the case that cardiology trainees should be more prudent when commenting on MR, spend slightly longer obtaining images if MR is present and liaise more closely with medical colleagues informing them about the limitations of HHE.

Conflict of Interest None

  • Bedside Echocardiography
  • Diagnostic Imaging
  • Valvular Assessment

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.