Talk:Interpreting test results

From Ganfyd

I have a lot more content that I was proposing to include, but it's not accessible to me right now. I shall continue this later... --Penglish 10:50, 18 February 2006 (CET)

Done now. --Penglish 13:08, 18 February 2006 (CET)

I wonder if some of the topics would be better hived off into sub-pages as the page grows, e.g. interpreting binary tests, Baye's theory etc.Mark ong

I'd agree that the Bayesian analysis stuff should be hived off to Bayesian analaysis, as it is worthy of a page itself.--A.l.brown 19:33, 18 February 2006 (CET)

PS: Done the bayesian analysis calculator at http://calculate.ganfyd.org/?form=bayesian+analysis --A.l.brown

Requesting tests

How about a separate page for Requesting tests? Mark ong

Good idea, Mark! --Penglish 13:08, 18 February 2006 (CET)

SpIn and SpOut...

I saw the pages midgley added a link to. I'm afraid I can't make out exactly what SpIn and SpOut are from the pages. Likelihood ratios? Predictive values? Does a link to them add anything, when they are not (to me at least) very meaningful? --Penglish 12:54, 18 February 2006 (CET)

The EBM crowd eg on evidence-based-health list make a lot of them - I find them complicated, whcih is why there isn't a screed of mathematics and examples of working them out on those pages, but there will be I think eventually. It seems to be regarded as very fundamental to the value of a test, so I'd leave them in. Midgley 19:44, 18 February 2006 (CET)

I suspect they are also called these ... Positive Predictive Value ; Negative Predictive Value as in that calculator which looks v clever. Midgley 19:48, 18 February 2006 (CET)

Oddly the abbreviation is hard to find. I don't think it was me fantasising - I do, but not about maths - but I can't find it in the list archive. Shall we assume it is a shorthand way of writing those expressions above, whcih is basically how I remember it - if not, what abbr, would one use for them? Midgley 20:48, 18 February 2006 (CET)

Are you referring to SpPins and SnNouts? I think you probably are, in which case see [1] --Jb 20:54, 18 February 2006 (CET)

Aha! Thanks,

It still doesn't clearly define what SpPins and SnNouts are. OK - SpPin has something to do with ruling a diagnosis in, or reliably confirming it; and SnNout has something to do with ruling out, or reliably refuting it. But is it an largely subjective concept, or can it be quantified? The reference Jb gave isn't clear about this. --Penglish 01:28, 19 February 2006 (CET)

Has anybody managed to find out what SpPin and SnNout stand for, yet? --Penglish 10:56, 19 February 2006 (CET)

Clearly it would help if I had remembered it straight, and/or properly understood it to start with. I retain the clear impression that this idea was rather important, and I think that jb's reference does actually define and explain them quite well. When one of the EBH list people gets sucked into our orbit they will sort it out in a falsh no doubt, meanwhile I'll try to name the pages correctly. Midgley 14:27, 19 February 2006 (CET)

OK, got it. It was important, and the explanations are now better, with an example. They deserve adding back in, alongside predictive value. Possibly they should simply be merged in, but the article is beginning to get long.Midgley 14:54, 19 February 2006 (CET)

EBH stuff

Trawling the Ebh list though throws up lots of stuff as ever, eg

http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0508&L=EVIDENCE-BASED-HEALTH&P=R29870&I=-3

"calculatation of the probability of a disease is only useful if it will allow us to make decisions i.e. to act (that is, to convert the probabilities expressed on the continuum scale into categorical "yes/no" decisions). How high probability is "high" or, how low is "low", as you are well aware, is not only function of LR of a diagnotic test, but also of benefits and harms of treatment under consideration. As a classic teaching goes, "don't order a diagnostic test if it will not affect your decision" (or, as we showed, "don't order a diagnostic test if harms of treatments outweigh its benefits"). This is, of course, where classic Pauker& Kassirer's "threshold" model comes handy. Some years ago, we modified Fagan's nomogram (based on Paul Glasziou's work) to allow easy calculation of "rule-in" and "rule-out" thresholds (see Djulbegovic B and Desoky A. Equation and nomogram for calculation of testing and treatment thresholds. Med Decis Making 1996;16:198-199) (attached)

Note that the nomogram requires use of NET benefits and harms. As many different summary measures may be used to express benefits and harms, one can adopt the threshold model accordingly (see for example, Djulbegovic B, Hozo I and Lyman G. Linking Evidence-based Medicine Therapeutic Summary Measures to Clinical Decision Analysis. MedGenMed, January 13, 2000 http://www.medscape.com/Medscape/GeneralMedicine/journal/2000/v02.n01/mgm0113.djul/mgm0113.djul-01.html) "

The Medscape article looks excellent - but is almost entirely about treatment decisions, with a very small section on testing. Is it referred to somewhere else on ganfyd? --Penglish 11:02, 19 February 2006 (CET)

Not yet. I dropped it here because I was here... Arguably we could do with a "heap" to drop things that look as thoguh they deserve to be somewhere, rather than attaching them to a particular article. Midgley 13:48, 19 February 2006 (CET)

Yes, I like the "heap" idea! --Penglish 12:20, 20 February 2006 (CET)