Thread by @dr_benoy_n_shah

Time for my first Twitter Poll!

I believe we should abandon the terms 'sensitivity' and 'specificity' of non-invasive imaging tests. 20th century language, no longer fit for purpose! I'll explain why after poll closes.

Do you agree?

@BSEcho @ASE360 @BritishCardioSo @escardio

@TheBJCA @BJCA_Women_LTFT @WessexSpRs #echofirst #whyCMR #cardiotwitter #Fellowsfirst

@BSCImaging @BSCMR @vass_vassiliou @MayooranShan @navtejc @BCIS_uk

Oh no...I 'lost' my first poll! 😫🙈

Anyway, so 1/3 agreed & 2/3 disagreed.

My thoughts: fundamental problem I have with sensitivity & specificity is the inherent inference that the test is 'right' or 'wrong', as defined by an angiogram. We know now that just isn't true...(1)

We've always known some patients with angina-like CP, that improves with anti-anginal drugs, have normal or near normal coronaries...cardiac syndrome X as some called it. In the functional imaging era, we also see this not infrequently. I had the privilege of recruiting for...(2)

the #ISCHEMIA trial years ago as a Research Fellow in Northwick Park. We lost count of the number of SYMPTOMATIC patients with barn door abnormal stress echocardiograms (approved as well by Core lab remember) that subsequently had normal coronaries. I think in the end, >10%...(3)

of ISCHEMIA patients had no significant CAD. So we know this happens.

Of course one possible explanation is someone over-interpreting the stress test...a genuine "false +ve". But the other option is there IS a problem, it's just not due to epicardial coronary disease! (4)

So yes, I do get a bit animated about false +ve and false -ve as concepts as the "accuracy" of the test goes far beyond simply what the coronaries show.

Indeed, a very under-recognised study (IMHO) was this one from @pattypellikka and colleagues, showing that whether...(5)

a patient had CAD or not, the outcomes were not significantly different if the stress test (in this case stress echo) was abnormal. Thus, some data suggesting that there is something going on in these patients with abnormal stress tests. Maybe we haven't worked it out yet (6)

So, in summary, I'm not a fan of sensitivity or specificity as they're based on angiography as the arbiter of whether your test (SE, SPECT or CMR) was "right" or "wrong"... In a patient with really abnormal stress test but normal coronaries, may not be a false positive stress (7)

test, but rather a false negative (i.e. falsely reassuring) angiogram!!

Not sure my interventional cardiology colleagues will like that suggestion though! 🤣

Thoughts welcome!