Where do my condition risk scores come from?

Cardiogram’s Report Card uses a machine learning algorithm to estimate your risk levels for certain health conditions based on your wearable data. These scores reflect how similar your data patterns are to those of users who have self-reported medical diagnoses.

🧠 How the algorithm works

Our AI model was developed using:

• Heart rate data
• Step counts and activity patterns
• Sleep trends (if available)
• Demographic info like age, sex, height, and weight
• Self-reported diagnoses from Cardiogram users

Using this data, the algorithm identifies patterns and correlations between certain behaviors or trends and known medical conditions.

📊 What does your score represent?

Your condition risk score shows how closely your health data resembles that of users who have reported a specific diagnosis — such as hypertension or sleep apnea.

For example, the algorithm may learn that users with hypertension often show elevated resting heart rate, fewer daily steps, or specific recovery patterns after exercise.

If your data aligns with those trends, your estimated risk score may be higher — but that doesn’t mean you have the condition.

⚠️ Not a diagnosis

These results reflect statistical correlations, not clinical determinations. The Report Card is a tool to help you stay informed and proactive about your health — not a replacement for a doctor.

If you're concerned about your results, we recommend discussing them with your healthcare provider.

💬 Have questions?

We’d be happy to explain your score in more detail. Reach out to us at [email protected]