Limitations in the current screening practice of assessing left ventricular ejection fraction for a primary prophylactic implantable defibrillator in southern Ontario

Background Screening echocardiography (ECHO) is commonly performed to determine whether the patient's left ventricular ejection fraction (LVEF) is appropriate for primary prophylactic implantable cardiac defibrillator (ICD) referral. However, radionuclide ventriculography (RNA) is used by many implantation centres for decision making. Objective To determine whether current screening ECHO techniques are effective in identifying patients suitable for primary prophylactic ICD referral. Methods Correlation, sensitivity, specificity and likelihood ratios (LRs) of semiquantitative and numerical quantitative ECHO LVEFs were calculated for predicting RNA LVEFs that met implantation criteria (LVEF less than 30% and less than 35%). Results and Discussion Among 193 patients, the LRs for a semiquantitative ECHO predicting an RNA LVEF of less than 30% (negative LR was 0.21 to 0.69 and positive LR was 1.22 to 2.83) or RNA LVEF of less than 35% (negative LR was 0.24 to 0.73 and positive LR was 1.33 to 3.46) demonstrated that current screening ECHO techniques are ineffective. However, the positive predictive value of grade 4 ECHO was 93.0%, suggesting that these patients may not require further LVEF investigation before implantation. Among 102 patients, current quantitative ECHO techniques did not improve the screening characteristics. Conclusions Current screening ECHO techniques may not be adequate for screening patients for consideration of a primary prophylactic ICD, but a grade 4 ECHO finding has a high positive predictive value in meeting implantation LVEF criteria. Improved screening standards should increase the number of patients referred with appropriate LVEF for primary prophylactic ICD implantation.