Do cardiac risk scores only muddy the waters?

Chest pain risk scores have become almost ubiquitous in their use in EDs internationally. Having internally and externally validated numerous risk scores, together with their use with high-sensitivity troponin (hs-cTn) assays (both T and I hs-cTn assays are available from a variety of manufacturers) over the past decade, I have turned against their use in the clinical environment, for now. Is this because I am disgruntled that not one centre, I am aware of, uses the modified Goldman risk score I spent 3 years of my life evaluating?1 Perhaps. My work evaluating the modified Goldman risk score was published over 5 years ago and at that time we found that this risk score when combined with a single hs-cTnT below the 99th percentile cut-off value taken at presentation to ED had a very high diagnostic accuracy for the rule-out of 30-day major adverse cardiac events (MACE), with a sensitivity of 98.8%. So why did it not take off? There are likely to be several reasons. The score itself is cumbersome, without a catchy acronym and these findings were from a single-centre study. However, one key finding from this earlier work was that an undetectable (below the limit of detection; LoD) hs-cTnT result in isolation, without the need for a cumbersome risk score, had an equally high diagnostic accuracy for the rule-out of 30-day MACE and allowed over 30% of patients with chest pain to be potentially discharged after a single hs-cTnT test. The premise that single hs-cTnT below the LoD, in combination with a non-ischaemic ECG, has since been evaluated in …