Acoustic-based rule-out of stable coronary artery disease: the FILTER-SCAD trial

Overtesting of low-risk patients with suspect chronic coronary syndrome (CCS) is widespread. The acoustic-based coronary artery disease (CAD)-score has superior rule-out capabilities when added to pre-test probability (PTP). FILTER-SCAD tested whether providing a CAD-score and PTP to cardiologists was superior to PTP alone in limiting testing.

Overtesting of low-risk patients with suspect chronic coronary syndrome (CCS) is widespread. The acoustic-based coronary artery disease (CAD)-score has superior rule-out capabilities when added to pre-test probability (PTP). FILTER-SCAD tested whether providing a CAD-score and PTP to cardiologists was superior to PTP alone in limiting testing.

The implementation strategy of providing cardiologists with a CAD-score alongside SDE did not reduce testing overall but indicated a possible role in patients with low CCS likelihood. Further strategies are warranted to address resistance to modifying diagnostic pathways in this patient population.

At six Danish and Swedish outpatient clinics, patients with suspected new-onset CCS were randomized to either standard diagnostic examination (SDE) with PTP, or SDE plus CAD-score, and cardiologists provided with corresponding recommended diagnostic flowcharts. The primary endpoint was cumulative number of diagnostic tests at one year and key safety endpoint major adverse cardiac events (MACE).

In total, 2008 patients (46% male, median age 63 years) were randomized from October 2019 to September 2022. When randomized to CAD-score (n = 1002), it was successfully measured in 94.5%. Overall, 13.5% had PTP ≤ 5%, and 39.5% had CAD-score ≤ 20. Testing was deferred in 22% with no differences in diagnostic tests between groups (P for superiority = .56). In the PTP ≤ 5% subgroup, the proportion with deferred testing increased from 28% to 52% (P < .001). Overall MACE was 2.4 per 100 person-years. Non-inferiority regarding safety was established, absolute risk difference 0.49% (95% confidence interval -1.96-0.97) (P for non-inferiority = .003). No differences were seen in angina-related health status or quality of life. Conclusions: The implementation strategy of providing cardiologists with a CAD-score alongside SDE did not reduce testing overall but indicated a possible role in patients with low CCS likelihood. Further strategies are warranted to address resistance to modifying diagnostic pathways in this patient population.