ClinVar stars describe review status, not your disease probability

Draft • Research checked 4 October 2026 • AI-assisted DomDNA editorial content. No independent clinical review or publication is claimed. General education, not individual medical advice.

Gold stars beside a genomic database entry can look like a familiar review rating. In ClinVar, they represent a specific kind of metadata: the review status supporting a submitted or aggregated classification. They are not customer ratings, and they are not a percentage probability that a person has a condition.

The words beside the stars matter. The same number of stars can correspond to different situations, and classification types have their own rules.

Read the text label before the symbol

The NCBI review-status documentation explains the system. Review status can reflect whether criteria were provided, agreement among submissions or an expert-panel or practice-guideline submission.

This is useful information about the record. It does not replace checking the variant, the condition and the kind of classification being displayed. A germline classification, an oncogenicity classification and a somatic clinical-impact classification should not be treated as a single generic verdict.

A hypothetical reading example shows the trap. A person sees two stars and writes “two out of four chance of disease”. No such calculation follows from the display. The stars do not form a probability scale.

Another person sees zero stars and assumes that the variant is harmless. That conclusion also does not follow. The text can indicate missing criteria or a lack of classification rather than a benign result.

Who supplied the interpretation?

ClinVar collects submissions. The submission instructions describe the information submitters provide and how review status is assigned. The presence of a record on an official website is not a statement that NCBI independently reproduced every underlying experiment.

For a useful reading note, record the submitting organisation and the assertion criteria where available. Preserve the condition named in the submission. A variant can be discussed in different clinical contexts, and a summary stripped of those contexts can mislead.

If submissions disagree, read the nature of the disagreement rather than counting votes as if every submission had the same scope and evidence. A database summary is a starting point for understanding the record, not a self-service clinical decision.

This does not make the database unhelpful. Its value includes making the provenance and differences visible so that specialists and researchers can examine them.

Expert review is a process description

The NCBI expert-panel and guideline application guidance sets out requirements for those designations. The designation identifies a recognised review process; it does not supply an individual prognosis.

Do not paraphrase an expert-panel classification as “clinically reviewed for me”. The panel did not examine the reader, their specimen or their medical history. It reviewed a variant classification within its stated domain.

Likewise, a classification does not confirm that a row in a consumer file is a correctly measured result. Establishing the identity and technical validity of a person's finding is a separate matter.

This is a useful point to preserve when making a screenshot. The database's authority concerns the record it presents, not every possible use of that record elsewhere.

A hypothetical screenshot repair

Suppose someone has saved a cropped image containing a variant name, a classification word and stars. It is missing the condition, review-status text, submission details and access date.

A better private note would retain those missing fields or link to the full record. It would also say whether the source is the laboratory report or a separate database lookup. Those two sources should not silently merge into one claim.

If the laboratory interpretation differs from the database view, save both with their dates and ask the reporting service about the difference. Do not overwrite a medical report with a current web label on your own.

What you can do with this

Choose one ClinVar entry for a reading exercise and copy the exact review-status wording into a private note. Add the classification type and condition. Do not convert the stars into a risk percentage or a treatment decision.

The practical gain is a better question: “What does this review status tell us about the submitted evidence, and how does it relate to the laboratory's interpretation?” That keeps metadata in its proper role.

For general lifestyle learning, DomDNA’s educational quiz does not search ClinVar or classify your variants.