Author: Craig Davidson, Dip Datta / Codes: CC5, HAP8 / Published: 28/01/2015
Study Objective: Authors aimed to derive and validate a Clinical Decision Rule (CDR) for suspected cardiac chest pain to risk stratify patients in the ED. Aiming to identity those who could be discharged from the ED and those at high risk of ‘Major adverse cardiac events’ (MACE).
Study design: CDR derivation and validation using two sequential prospective observational cohort studies.
Setting: Cohort 1: Manchester Royal Infirmary from 2006 to 2007. Cohort 2 Stepping Hill Hospital Stockport 2010
Summary: This study developed a decision rule containing 8 variables which when put into an computerised calculator can generate a risk of MACE:
High sensitivity Troponin T
Heart-type fatty acid binding protein (H-FABP)
SBP < 100 mmHg
Pain radiating to Rt arm or shoulder
When applied to the second validation cohort the risk stratification accurately reflected the patients actual outcomes. Of the patients identified as ‘very low risk’ by the CDR none had an AMI and 1.6% had MACE (which were coronary stenosis which required no intervention). Sensitivity for MACE was 98%. Of the patients identified as ‘high risk’ 91% had an AMI and 96% had MACE. The CDR appeared to perform just as well with hs-trop T.
In the validation cohort this would have indicated safe D/C in 27% of patients which may therefore reduced admissions for serial troponins.
In the validation cohort the ‘very low risk’ group had a slightly higher % MACE than the ‘low risk’ group. This suggests the CDR design may be so specific to the derivation cohort that it loses power on other groups. This is known as ‘overfitting’. The authors perspective is that the overall AUC was still good in the validation study, and point out further external validation is needed anyway.
The derivation study was slightly under powered according to the authors calculations.
Included patients presenting with ‘Chest pain suspected to be cardiac in origin’. How appropriately this is applied is likely to vary between clinicians and departments and could therefore affects results.
Included patients which were obviously having an AMI which may dilute potential power as a rule out rule for MACE.
Rules using biochemical tests may not work as well using different assays in different labs.
H-FABP is not widely available – although there is now an automated assay.
The risk stratification of those not D/Ced is dependant on the healthcare system.
The MACS rule shows a lot of promise in identifying those at very low risk of MACE, and therefore facilitating appropriate D/C after a single blood test. Even in the era of HS-Trop this represents an opportunity to streamline care. It also helps risk stratify those not able to be discharged (within a UK system).
However limited generalisability, alternatives (such as the HEART Score) and the development of HS-trop mean MACS, in its current form isn’t ready for implementation in our ED practice.
Locally we use the combination of low GRACE score and low 6 hour hs-Trop to facilitate early discharge. We are not convinced the implementation of this rule, with the resource use it entails, would make as big an impact on our discharges as it could have done in the validation cohort.
The authors are already working on an impact assessment. Further studies will hopefully examine its power compared to the alternatives and in external populations. We are looking forward to the results.
– By the authors
– Critique and the Authors reply.