This is an algorithm for deciding whether a patient suffers from a specific disease or not:

A patient has rheumatoid arthritis if at least four out of the following seven symptoms are present:

  • Symmetric swelling (arthritis)
  • Rheumatoid nodules
  • The presence of rheumatoid factor (in serum)
  • ...

There are similar classifications for many conditions. I remember seeing a specific term used for this kind of definition. It was fairly technical and could be unique to the fields of medicine or medical informatics.

What is this definition type called?


The word was polythetic, as opposed to monothetic.

(of a class of things) having many, but not all properties in common (wiktionary)

A polythetic class is defined in terms of a broad set of criteria that are neither necessary nor sufficient. Each member of the category must possess a certain minimal number of defining characteristics, but none of the features has to be found in each member of the category. (iva.dk)

| improve this answer | |
  • Wow, "fairly technical" is right! – Gnawme Oct 27 '11 at 21:45

They're probably normally called diagnostic criteria, individually and collectively.

Any one symptom associated with a medical condition may also be called an indicator, but that has undesirable resonance with contra-indications (reasons why you might not adopt some particular treatment in some particular case).

I think most healthcare professionals wouldn't particularly want to encourage the use of a specialist term for 'Tick box' medicine, since they see this as something fundamentally undesirable.

| improve this answer | |
  • The word I'm looking for specifies that the diagnostic criteria require at least a specific fraction of the symptoms to be present, as opposed to all of them being present. – user4727 Oct 27 '11 at 17:06
  • I'll add that this is quite a common way for diagnosing certain "fuzzy" conditions. The DSM-IV lists three major symptoms of depression, two of which should be present for the diagnosis to be given. Of course there'd be an overall judgement, too -- I'll edit my question if it seems a bit too crass. – user4727 Oct 27 '11 at 17:16
  • Yes, I'm familiar with the methodology. But say there's some particular disease with a well-known list of 10 "most common" indicators, and someone tries to promote the idea that if you have 6 of them you have the disease, but with 5 you don't. I think most doctors would poo-poo the idea of using such a "tick box/cutoff point" style of diagnosis - they'd rather trust their own judgement on a case-by-case approach. It's more relevant to governments - when deciding if you qualify for disability payments, for example. – FumbleFingers Oct 27 '11 at 17:20
  • And there are always people who take one of these "5 out of 10" lists, pick the 5 most general or vague criteria, and then claim that everyone or everything qualifies. Like hypocondriacs who conclude they have some rare disease because they met "#2 headaches, #5 trouble sleeping, and #7 excessive tiredness". Or conversely someone will say that diagnostic criteria were obviously invented by people with some ulterior motive because almost everyone in the world could meet them. – Jay Oct 27 '11 at 18:41

These are called classification criteria, and are used in performing a differential diagnosis. For example, a differential diagnosis is performed to determine whether someone has myasthenia gravis.

For your specific example of rheumatoid arthritis (RA), I found this statement:

The standard and accepted means of defining RA is by use of classification criteria. Classification criteria enable the stratification of groups of individuals into those with and those without RA... [italics mine]

(The classification criteria for RA were revised in 2010; the methodology makes fascinating reading.)

| improve this answer | |
  • 1
    The example presented is of a simple diagnosis (patient has RA vs. patient does not have RA), not a differential diagnosis (patient has RA vs. patient has some other condition). – Pops Oct 27 '11 at 18:43
  • @LordTorgamus Are you referring to the OP's example? Perhaps I misread it, but I thought the OP was asking what the term is when using some set of symptoms to differentiate conditions. – Gnawme Oct 27 '11 at 18:52
  • 1
    @Gnawme: The term might apply in both cases, but where I saw it, it was used to differentiate between having RA and not having RA. Classification criteria might be what I'm looking for, although it doesn't immediately ring a bell. – user4727 Oct 27 '11 at 18:56
  • FYI, read the paper at the link referenced in my answer. The determination of whether RA is present is much more fine-grained than in your example. (The decision tree shown in the paper is fairly involved.) – Gnawme Oct 27 '11 at 19:05
  • I can't seem to find a definition for classification criteria. Are you certain that it doesn't apply to a diagnostic rule like "you have scarlet fever if and only if you have a red tongue"? – user4727 Oct 27 '11 at 19:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy