Breadth of ambition and strength of evidence are different axes, and Cardiovolt and Anumana sit at different points on each, which is what makes this comparison useful rather than a simple ranking. Cardiovolt's public proposition spans mortality, heart failure, valve disease, arrhythmia and non-cardiac disease from a single ECG, a wide net still substantially in development and translation. Anumana's low-ejection-fraction algorithm is narrower and further along the evidence ladder's hardest rung: a pragmatic randomised trial showing the intervention changed real-world clinical detection, which is a materially stronger claim than discrimination on retrospective data, however large.
The comparison, dimension by dimension
Breadth of disease targets: Cardiovolt wins decisively, its ambition spans a genuinely wider set of conditions than any single-purpose competitor. Regulatory authorisations: Anumana holds named clearance for its low-ejection-fraction algorithm, a concrete evidence-maturity marker Cardiovolt's broader platform has not yet matched. Development population: both draw on large retrospective cohorts, with Anumana's low-EF work reporting multi-site retrospective performance around an AUROC of 0.93 for that specific indication. Prospective and randomised evidence: this is where the comparison sharpens most, because Anumana has something Cardiovolt does not yet have for any single indication, a completed pragmatic randomised trial. Clinical action following a positive result and negative-result limitations: both require the same discipline this whole cluster insists on, a positive flag triggering defined follow-up, a negative flag never treated as ruling disease out. Health-system integration: Anumana's narrower, trial-tested indication is arguably easier to integrate cleanly than Cardiovolt's still-expanding target list.
The EAGLE trial, read in full
The result worth understanding precisely rather than headlining loosely: the EAGLE pragmatic randomised trial involved 22,641 adults across 120 primary-care teams, and found that providing clinicians with the AI-ECG result increased diagnosis of low ejection fraction, without increasing overall echocardiography use. That combination, more diagnoses without more testing volume, is the finding that matters clinically, because it suggests the tool redirected existing echocardiography toward patients who actually had the condition rather than simply generating more tests. The reported relative increase in diagnosis was approximately 31%, and the distinction this article insists on stating plainly is that a 31% relative increase is not a 31-percentage-point absolute improvement, a conflation that inflates apparent effect size dramatically and is exactly the kind of number that needs its baseline stated before it means anything. A randomised design also does something retrospective discrimination studies cannot: it controls for the confounders that plague observational "AI improved outcomes" claims elsewhere in this cluster, which is why EAGLE deserves to be treated as a stronger category of evidence, not merely a bigger study.
The editorial verdict
Best breadth of predictive research: Cardiovolt, without much competition, its target list is simply wider than any single comparator's. Best current example of prospective clinical utility: Anumana's low-ejection-fraction programme, on the strength of EAGLE specifically, a randomised trial showing the intervention changed detection is a different and stronger claim than any of Cardiovolt's currently published evidence offers. Most important unresolved question for both: whether earlier detection, of low ejection fraction for Anumana, of the wider target list for Cardiovolt, actually produces better long-term outcomes rather than simply more diagnoses earlier in their natural history, the gap between detection and benefit that this entire diagnostic-AI cluster keeps returning to, because it is the gap that determines whether a discrimination result becomes a genuine clinical advance.
What this comparison teaches about reading the category
The generalisable lesson: breadth of ambition and strength of evidence are not substitutes for each other, and a company with a narrower but trial-tested claim has, for that specific claim, stronger evidence than a company with a wider but earlier-stage platform, regardless of which company's overall proposition sounds more impressive. Applying the evidence ladder per indication rather than per company is the discipline this comparison demonstrates, and it is the discipline worth carrying into every other comparison in this cluster: ask which specific claim is being evaluated, then ask which rung of validation that specific claim has actually reached.
Frequently asked questions
Does EAGLE mean Anumana's tool improves patient outcomes?
EAGLE demonstrated improved diagnosis without increased testing volume, a strong intermediate result; it did not itself measure downstream outcomes such as mortality or heart-failure hospitalisation, which would require further, longer-horizon study.
Could Cardiovolt eventually run a trial like EAGLE for its own targets?
In principle yes, and doing so for its highest-priority indications would be the natural next evidence-ladder step; until then, its broader claims should be read as earlier-stage than Anumana's low-EF-specific claim.
Should a health system choose one platform over the other?
The choice depends on which specific clinical problem is being solved: a defined, trial-evidenced intervention for low ejection fraction points toward Anumana's demonstrated use case, while broader exploratory risk-stratification points toward Cardiovolt's wider but earlier-stage ambition.