Dr Smith of Dr Smith’s ECG Blog has pointed out that an absolute measurement of 5mm as a criteria reduces sensitivity and specificity and argues that this criteria should be based on discordant elevation greater than 0.25 of the S wave amplitude (Figure \ref{fig:smith_mods}; \cite{Smith_2012}). Using this instead of the 5mm criterion as the “Smith-modified Sgarbossa criteria” improved diagnostic accuracy and sensitivity in their study, reporting a positive likelihood ratio of 9 and negative likelihood ratio of 0.1. This is annotated, a little confusingly to me, as ST/S ≤ -0.25. That is, the positive deflection of the ST segment is more than a quarter of the negative deflection of the S. It is argued that any excessively discordant ST segment based on proportion (ST elevation or depression) could be interpreted as diagnostic of STEMI.