National Audit of Resting Gated Heart Pool Study Analysis Using a Validated Dataset
Aims: Validation of quantitative results is an important process, especially when studies are performed in different locations. A voluntary national audit was undertaken to compare LVEF calculations from gated heart pool scans (GHPS) using software packages currently used throughout Australia and New Zealand.
Methods: Dataset supplied consisted of 12 GHPS studies from a 2005 UK audit(1), including worksheets and instructions for completion. Each site provided a brief description of version of software and details of analysis method for LVEF calculation. Total of 17 sites using 5 software packages with results from 93 technologists were assessed. Mean, median, standard deviation and average difference calculated for each study and software package.
Results: Over a range of 9 – 84.2 EF units, the national average LVEF was calculated as 47.4%, being 5.1 EF units higher than reference LVEF of 42.3%. Comparison of software packages shows a range of values and average difference from reference LVEF (see Table 1).
Conclusions: Differences in LVEF values between software packages has been demonstrated. Dataset can be used for validating LVEF, training operators, and prior to software upgrades to assess change in results.
References:
1. Skrypnuik JV, Bailey DL, Cosgriff PS, et al. UK Audit of left ventricular ejection fraction estimation from equilibrium ECG gated blood pool images. Nucl Med Commun 26(3):205-215, 2005.
Table 1: Comparison of 5 software packages used for calculation of LVEF
Package No Average LVEF (%) Difference from Reference
A 47.8 5.5
B 48.1 5.8
C 48.2 5.9
D 51.8 9.5
E 43.4 1.1