Editorial Policy

How a tool here is written, sourced, checked and dated

Last reviewed 21 August 2026. Applies to every tool at mediscuss.org/cdss.

The Short Version

76 clinical tools, every one of them naming the sources it was built from and the date it was last revised, and every correction to any of them published where you can read it.

  • Indian guidance is followed where it differs. NACO, NTEP, UIP, NVBDCP, ICMR, IAP and API are not adaptations of international protocols. They are different algorithms, and this system runs them.
  • Every threshold is traceable. Open the references on any tool and you can see what it rests on and judge it for yourself.
  • Nothing is fixed quietly. Corrections are published on What's new, including the ones that are embarrassing to publish.
  • Nobody pays for anything here. No advertising, no sponsorship, no industry funding, and no drug named for any reason other than that it is the drug you would reach for.
Who Writes It, and Who Reviews It

Each tool names its author and its reviewer on its own page, with the date of the last revision beside them. Attribution belongs on the tool, not on a policy page a clinician has no reason to open at the bedside.

Specialist review is being brought in tool by tool, and it is the part of this system that grows fastest when clinicians take part. A reviewer reads a finished tool against current guidance in their own field, says where it is wrong, out of date, unsafe or unclear, and is named on the tool they reviewed. Reviews are requested by tool, so nobody is ever sent something outside their field.

If you practise in a field these tools touch, an hour of your reading changes what every clinician who opens that tool afterwards sees at the bedside. Join MEDiscuss CDSS as a reviewer or a contributor.
Where the Content Comes From

Sources are taken in this order, and where they disagree the higher one governs unless the tool says otherwise and says why.

1 Indian national programme guidance NACO for HIV, NTEP for tuberculosis, UIP for immunisation, NVBDCP for vector-borne disease, the National Rabies Control Programme, Anemia Mukt Bharat, and the UDID framework. Where a national programme defines the algorithm, the algorithm is the programme's.
2 Indian professional society guidance ICMR, the Indian Academy of Pediatrics, the Association of Physicians of India, and the specialty societies, where they have published on the question.
3 International guidance and the primary literature WHO, and the major society guidelines, together with the trials and validation studies behind a score. Used in full where India has not published on the question, and named alongside the Indian position where the two diverge, so that you can see the divergence rather than only its result.
4 The product label, for every dose Doses, renal adjustments and the strengths a drug is actually dispensed in are taken from the label of the product available here, using generic names throughout.

Where international guidance changes something a national programme has not yet adopted, the tool follows the programme and says so in the output rather than in a footnote. That is how the HIV tools handle the places where WHO 2026 would choose a different drug from NACO: NACO governs, and the divergence is stated on the screen where the decision is being made.

How a Tool Is Built

A tool starts from a decision a clinician actually has to make, and usually it arrived as a request or as a report that something was missing.

  • The clinical logic comes first. What is being decided, what information decides it, what the bands are and where each boundary comes from.
  • Then the sources are fixed to the logic, threshold by threshold, and written into the tool's references rather than kept in a drawer.
  • Then the teaching text: what the result means, what it does not mean, and where an output can mislead in a particular patient.
  • Then it is built and exercised. Every tool is driven end to end in a real browser before it is published, and again on every subsequent change.
  • Then it is published with its version, its references and its revision date on the page, so a reader can judge how current it is without asking anybody.

Tools that return a management step rather than a number get more scrutiny than the arithmetic ones, for the obvious reason: a body surface area can be checked in your head, and an antiretroviral switch cannot.

How It Is Checked

Three things run continuously, and none of them is a substitute for the other two.

  • Automated checks on every change. Every tool is opened, filled and run in a headless browser, and the calculation logic behind each one is parsed, so that a tool cannot be published broken. These catch failures, not wrong medicine.
  • Critical audits of individual tools. A tool is taken apart against current sources, agent by agent or threshold by threshold, and what is found is corrected and recorded. The seven dosing errors found in the antimicrobial dosing tool in August 2026 came out of one of these, and they are published on What's new in the detail a clinician would need to re-check a patient.
  • Reports from clinicians using the tools. These have shaped this system more than anything else, and they go to the top of the queue.
Versions, Dates and Going Stale

Every tool carries a version number and the date it was last revised, both shown on the tool itself. Guidance moves, and a decision-support tool that does not say when it last looked is asking to be trusted on nothing.

A tool that has not been revised for eighteen months says so on its own page, in a notice asking you to verify its doses and thresholds against current guidance before use. That notice is produced by the system from the revision date, not added by hand, so it cannot be forgotten.

When a tool is withdrawn, its address keeps working and redirects to whatever now covers the question, so that a bookmark made at a bedside does not become a dead end.

Corrections

If a result looks wrong, it may well be. Reports of clinical errors are read first, ahead of everything else, and a correction to a dose or a threshold is made as soon as it is confirmed rather than held for a release.

Corrections are published, not folded away. Anything that changed a dose, a threshold or an interpretation appears on What's new, described plainly enough that you can tell whether it affects a patient you have already treated.

Report a clinical error

Independence and Funding
  • The tools are free and always will be. No subscription, no trial, no account required, and no feature held back for members.
  • No advertising and no sponsorship. Nothing on this site is paid for by anyone, and no company has any say in what is built or how it reads.
  • No industry funding, and no payment for placement. Drugs are named generically. Where a brand is unavoidable it is named because that is how the product is stocked, never because anyone asked.
  • Contributors and reviewers are not paid, and are not asked for anything either. They are named on the work they do, unless they ask not to be.

What the system does collect is set out in the privacy notice. Usage statistics, always. Your favourites and your own saved calculations, only if you choose to sign in. Nothing you type into a tool is sold, shared or used for research.

What This System Is Not

It is a set of decision-support tools, not a reference library. There are no monographs behind the calculators, and the coverage is restricted on purpose to areas where the content can be kept current. A tool nobody can keep up to date is worse than no tool. For a question outside that scope, the established references remain the right place to look.

Every output assumes a clinician who can judge whether it applies to the patient in front of them. The terms of use set out where responsibility for a clinical decision rests, and the answer is the one you would expect.