Editorial, Evidence, and Release Review

Child Growth Lab is independently operated. One individual developer has final responsibility for source selection, implementation review, public wording, test acceptance, corrections, and the decision to publish a release. The site does not present automated checks or AI-assisted work as clinical judgment.

The current status is an internal editorial, evidence, implementation, and release review—not an external medical opinion. The public date above changes only after a substantive review of this page and its stated process.

8 / 8 calculator-specific internal algorithm reviews have been completed. For each calculator, review covers the source and version, intended population, age and unit rules, date handling, formulas, thresholds, interpolation, rounding, boundary cases, displayed outputs, and expressly excluded uses.

That count does not mean every disease, special population, or possible result has been assessed. It does not mean each output has been confirmed by a doctor, that the site is a diagnostic device, or that it has undergone clinical trials or regulatory approval.

Sources are reviewed in this order when they are available and relevant:

1. Official government and data-publisher standards, datasets, and instructions.
2. Professional-organization standards and practice guidance.
3. Primary model publications and technical documentation.
4. Official updates, corrections, validation material, and worked examples.
5. High-quality systematic reviews or supporting literature used for context.

Provenance checks record the source identity, version, retrieval information, permission where required, checksums where practical, and the scope in which a source may be used. Commercial calculators, blogs, and search snippets do not override an original source.

Citing CDC, WHO, AAP, PES, RCPCH, or another organization does not imply its endorsement of this independent site. Source pages can change between scheduled reviews; maintenance dates record completed work rather than a guarantee of real-time monitoring. Calculator-specific evidence is listed in Methodology.

The review translates each accepted source into an explicit implementation contract: required inputs, unit conversion, age or date mapping, formulas, source data tables, interpolation, branch conditions, boundary inclusion, error states, rounding, and visible outputs. Changes are inspected as diffs so a page explanation, test, and production implementation remain aligned without accidentally changing scope.

Current formula versions are read from the central site configuration:

• cdc-lms-v2
• who-igrowup-v2
• corrected-age-v3
• tanner-mid-parental-v3
• cdc-kids-bmi-v3
• growth-velocity-v3

Production calculations and independent validation code are kept separate where the source permits a second implementation. Code running successfully is not, by itself, evidence of medical correctness.

Official software outputs, publisher examples, or reference vectors are used when available. Otherwise, validation may independently reconstruct expected values from original data or use separate formula-based scripts rather than calling the production calculation function.

Regression coverage includes, as applicable:

• Fixed numerical vectors and source-table anchors.
• Lower and upper endpoints, threshold equality, and just-outside boundaries.
• Metric and US-unit equivalence, date arithmetic, birthdays, and leap years.
• Invalid, missing, non-finite, and out-of-scope inputs.
• Rounding, submitted-input snapshots, result clearing, and stale-result prevention.

Tests establish the explicit software contract for reviewed cases. They do not prove clinical validity for every child, condition, measurement method, or care setting.

Public wording is checked to distinguish percentiles, z-scores, screening references, diagnoses, trends, and predictions. Calculator outputs remain reference comparisons or arithmetic summaries unless an authoritative source and the implemented method support a narrower statement.

Review removes unsupported treatment, referral, feeding, or calorie advice; guarantees; absolute normal-or-abnormal labels; and stigmatizing language. Disclaimers and limitations must remain visible, and explanations should remain understandable to both caregivers and professional users. Wording review cannot replace clinical judgment or make incomplete measurements clinically sufficient.

Release acceptance can include unit and content-contract tests, independent validators, lint, TypeScript, a production build, metadata and canonical checks, sitemap and robots checks, internal-link validation, Playwright browser flows, Preview deployment review, responsive layouts, keyboard operation, axe checks, and console or page-error review. Production smoke checks tie the accepted route and marker set to the intended commit and public domain.

Privacy review checks whether child health inputs are incorrectly collected or sent, where measurements are processed, the difference between browser-local calculation and ordinary infrastructure logs, and whether the public statement remains accurate. See the Privacy Policy. Passing release gates cannot cover every device, assistive-technology combination, or future browser change.

AI tools may help retrieve and organize research, compare source passages, draft wording or code, propose tests, and inspect diffs. AI is not a medical source. Its output is checked against primary sources, repository contracts, and executable tests before the individual developer decides whether to keep and publish it.

AI does not publish automatically, accept clinical responsibility, or supply an independent expert opinion. Using multiple tools or automated checks does not mean that multiple clinicians reviewed the work.

Advertising may help support the free site, but it does not determine evidence sources, formulas, thresholds, result categories, safety wording, limitations, or whether a correction is made. Advertising must remain visually distinct from editorial content and must not imply that an advertiser or source organization endorses Child Growth Lab.

Ownership, funding context, audience, and site purpose are described on the About page.

Reports may concern a source, algorithm, displayed explanation, privacy statement, accessibility barrier, broken link, or other technical behavior. Contact support@childgrowthlab.com or use the Contact page.

1. Compare the report with the original source and current implementation.
2. Determine the affected algorithm, data, page, explanation, privacy statement, or user flow.
3. Add or update a regression check when practical.
4. Make the scoped code, content, provenance, or accessibility correction.
5. Run the full release acceptance process for the proposed correction.
6. Update the affected public review date after substantive re-review.
7. Verify the released fix on the Production domain.

Support is not a medical or emergency service. Do not email medical records, names, dates of birth, or other identifying child information. Not every report requires a change, and an unresolved report may remain under investigation until reliable evidence is available.