Educational cardiometabolic lab interpreter
SAHC RiskLens turns your own lab values into clear, guideline-backed context: where each number sits against clinical thresholds and against real population data, with particular attention to South Asian heart risk. It explains — it never diagnoses.
The why
Standard reference ranges are built on general populations — and South Asians carry cardiometabolic risk at lower thresholds than those ranges assume. A number that looks "normal" on paper can still matter. RiskLens exists to close that gap: to give people honest, population-aware context about their own labs, and to help them walk into a doctor's visit already informed. It's the safety-engineered successor to the South Asian Heart Center's original SCORE tool.
Each value is classified against ACC/AHA guidelines and benchmarked against a real cohort — so you see where you stand, not just what you scored.
Ancestry as a risk-enhancer, lower BMI action points, and elevated Lp(a) — surfaced with guideline backing, benchmarked against a genuine South Asian cohort.
No diagnosis, no personal risk score, no treatment advice, no stored health data. These limits live in the code, not just the fine print.
The what & how
You enter your values; RiskLens returns a layered, descriptive read — always leading with a disclaimer and ending with an always-visible list of limitations.
Lipids, glucose, HbA1c, blood pressure, and BMI — plus optional advanced markers ApoB and Lp(a). Nothing is sent to a server for storage; results live only in your browser.
Every value gets an absolute, population-independent category — for example, "LDL 168 → High, ACC/AHA 2018." Clear, sourced, and consistent.
See where your value falls (p10–p90) against a selectable population: a public NHANES Asian proxy, or the South Asian Heart Center's own clinical cohort. Optional peer-matching narrows to your sex, age band, and medication use — with small groups suppressed for honesty.
Relevant, guideline-backed notes appear when they apply. Track dated draws over time in a file you own and control — never stored on a server.
A physician discussion guide, a copy-ready pre-visit brief for your clinician, and pointers to family screening and prevention programs — none of it prescriptive.
Who it helps
RiskLens is designed to make the patient–clinician conversation better on both sides — not to replace it.
Safety is the feature
RiskLens is deliberately not a medical device. Its limits aren't a matter of policy or wording that could be forgotten — they're enforced structurally, verified by nearly 300 automated tests and a dedicated AI-safety benchmark.
No condition is ever asserted or ruled out. No individual risk score. No treatment advice. A result cannot even be generated without a disclaimer attached.
A public proxy cohort is never dressed up as "South Asian." Small comparison groups are suppressed and disclosed rather than shown as if reliable.
The server stores nothing. Results live in your browser; your history is a file you export and keep. No language model ever touches the text you read.
Backed by CardioSafeBench — a reproducible benchmark that tests the tool against the failure modes of open-ended AI lab interpretation: overclaiming, hallucinated guidelines, unsafe advice, and missing South Asian context.
Enter a sample panel and watch RiskLens build a full, descriptive read — disclaimer, guideline classifications, cohort benchmarks, South Asian context, and a pre-visit brief.
Launch the live demo → Educational use only · No account · No data stored