Every calculator on RECATOOLS produces a number, and a number is only as good as the rule that made it. This page describes where those rules come from, who checks them, how often, and — importantly — what this process is not.
1. Where a formula comes from
A RECATOOLS calculator implements a published, named formula, not a rule of our own devising. Wherever a tool computes a result, the intent is that the result is traceable to one of:
- the original publication that defined the formula — the paper, with authors, journal, year and where possible a DOI or PubMed ID;
- the issuing body that maintains it — a standards organisation, a regulator, a statutory authority, or a clinical guideline committee; or
- an arithmetic identity that needs no source, in which case the tool should say so rather than invent a citation.
Where a tool has a recorded source, it is stored against that tool and shown on the tool page. Where a tool has none, that absence is a gap in our records, not a claim that no source exists.
2. What a review actually consists of
A review is not a glance at the page. It is a documented pass in which we:
- read the formula from the shipped code, not from the tool's own description of itself — the two can disagree, and when they do it is usually the description that is right and the code that is wrong;
- check every coefficient, threshold and band boundary against the original publication or the issuing body;
- ask specifically whether the formula has been superseded. This is the check that finds the most, and it is the one a purely arithmetic review misses entirely: a formula can be implemented perfectly and still be the wrong formula for today;
- check that guideline citations are current, including the year — a correct rule credited to a superseded document is still a defect;
- trace quoted reference ranges and event rates to a named source, and check that pooled or averaged figures are labelled as such; and
- verify that the tool's stated limitations are accurate — that it fails where it should fail, and says so.
When a review finds something wrong, the outcome is a fix, not a caveat added beside a wrong answer. A review that finds nothing wrong is recorded too, with a note saying what was checked.
3. Who does the review
RECATOOLS is operated by RECASYS, a sole proprietorship registered in Singapore. Tool reviews are performed by the site operator against the primary sources named above, with AI assistance in reading code and locating source material.
We want to be exact about what that means, because it is the part that is easiest to imply falsely. Our clinical calculators have not been reviewed or endorsed by a physician, pharmacist, nurse or any other licensed healthcare professional. RECATOOLS employs no clinicians and retains none. What we can tell you is narrower and, we hope, still worth something: that a named person has read the code, compared it line by line against the paper or guideline it claims to implement, written down what they checked, and dated it. That is a documentary standard, not a clinical one. Anyone telling you the difference does not matter is selling something.
The same applies in the other regulated areas we cover: our finance and tax tools are not reviewed by a licensed financial adviser, accountant or tax agent, and our legal-adjacent tools are not reviewed by a lawyer.
4. How often a tool is re-checked
Each reviewed tool carries its own review cycle, and the cycle is set from how fast that particular formula's owner actually moves — not from a single house default applied to everything.
- A score whose issuing bodies revised its application twice inside two years is on a 12-month cycle.
- A statutory calculation that changes with an annual budget is re-checked against that budget's calendar.
- An arithmetic identity published in 1916, or a clinical scale unchanged since 1974, sits on a 36-month cycle — and that cycle exists to re-verify our implementation, since the formula itself cannot be revised.
A cycle is a commitment about when we will look again. It is not a promise that the tool is correct in the interval, and a tool that is within its cycle has not thereby been checked today.
5. What we do not claim
Read this section as the operative one.
- No calculator here replaces professional judgement. A clinical score is a structured way of organising information a clinician already has; it does not diagnose, does not prescribe, and does not know anything about the person using it. A finance tool does not know your circumstances.
- These are not medical devices and are not registered, certified or cleared as such by any regulator. They are educational reference tools.
- A "last reviewed" date describes an act, not a guarantee. It means a person read the formula on that date and recorded what they found. It does not mean the tool is error-free, and it does not mean the underlying guidance has not changed since.
- Where a tool shows no review date, none has been recorded. We would rather show you nothing than show you a date we cannot stand behind. Most tools on this site are in that position; we are working through them, and we will not backfill dates to close the gap cosmetically.
- Absence of a caveat is not a warranty. We may simply not have found the problem yet.
6. Telling us we are wrong
If you believe a tool produces an incorrect result, or cites a formula or guideline that has been superseded, please write to [email protected]. A specific report — the tool, the inputs you used, the number you got, and the number you expected with its source — is worth far more to us than a general one, and we will act on it faster.
Corrections to tool logic are treated the same way as corrections to articles: see our Corrections Policy. For how we source and review written content rather than tool formulas, see our Editorial Policy.