Please read this announcement about upgrades to JournalFeed (CME, Amal Mattu, new website)! We also need to do a price increase due to rising costs. Price changes started Dec. 1, 2023 for new subscribers and on Jan 1, 2024 for current subscriber auto-renewals. If cost is a barrier for you, please read the announcement. We want to help! ~Clay Smith
Written by Jason Lesnick
Spoon Feed
This secondary analysis of the STOP CP cohort study found that, in isolation, using a 0/2 hour hs-cTnT rule-out protocol did not achieve a NPV ≥ 99% for 30-day cardiac death or MI (CDMI). When combined with a low risk History, ECG, Age, and Risk factor (HEAR) score, it did – but at the cost of significantly decreased efficacy (ruling out only 30.7% of patients).
Did you HEAR how we can make the 0/2 hour hs-cTnT safe in the US?
This study was a pre-planned subgroup analysis of the multi-center, cohort STOP CP study that prospectively enrolled 1,307 possible ACS patients ≥ 21 years old across 8 US academic EDs from January 2017 to September 2018. The 0/2 hour hs-cTnT algorithm has been developed in Europe and Australia as well as validated in Canada. This study used the Gen 5 STAT assay from Roche. Treating clinicians were blinded to hs-cTnT results; thus, care was not influenced by these results but was determined by local practice patterns and contemporary troponin results. CDMI was defined according to the ACCORD trial definition, with the exception of death due to stroke (which was in this study classified as a non-cardiac death).
The algorithm would rule out ACS according to the below figure.
The primary objective of this study was to evaluate if a 0/2 hour hs-cTnT algorithm in a U.S. cohort of patients could safely rule out 30-day CDMI. A secondary objective was to evaluate the diagnostic performance for index visit CDMI, CDMI at 90 days, and major adverse cardiovascular events (MACE, includes cardiac death, MI, or coronary revascularization) at index, 30, and 90 days. The final aim of the study was to determine if combining the hs-cTnT 0/2 protocol with the HEAR score improved diagnostic performance. Guidelines aim for a NPV ≥ 99% for 30-day MACE as a safe number for evaluating ACS in the ED.
This study found the 0/2 hour algorithm ruled out 61.4% of all patients but that CDMI at 30 days occurred in this group of ‘ruled out’ patients with a NPV of 98.1% (95%CI 96.9-98.9%). Meanwhile, for the 12.4% of patients who were in the rule-in zone, the PPV for 30-day CDMI was 61.7% (95%CI 53.8-69.2%). When the 0/2 hour hs-cTnT algorithm was combined with a low risk (0-3) HEAR score, the NPV for CDMI at 30 days improved to 99.3% (95%CI 97.8-99.8%) but only ruled out 30.7% of patients. In the combined approach, 14.8% of patients were ruled in, and the PPV for 30-day CDMI was 63.6%.
The results of this study differed from European, Australian, and Canadian populations when applying the 0/2 hour hs-cTnT algorithm in that generally those studies showed a near 100% NPV for 30-day CDMI. The baseline patient characteristics of this study had significantly higher rates of HTN, diabetes, known CAD, prior MI, and peripheral vascular disease when compared to international studies. This study also had a higher rate of 30-day CDMI and MACE compared to prior U.S. cohorts as well, and this increased prevalence could have impacted the NPV.
How will this change my practice?
This study and its parent STOP CP have, in my opinion, slowed enthusiasm for copying the rapid rule-out hs-cTnT algorithms that work incredibly well in Europe. I am looking forward to new research that will hopefully implement creative combinations of ECG, labs, +/- parts of decision rules to help us safely, rapidly rule out ACS in U.S. populations.
Editor’s Note: These studies blow my mind in that they make no sense to me. Please listen to this week’s JournalFeed podcast for my rant questioning the entire premise of the study. ~Nick Zelt
Source
Performance of the 0/2-Hour hs-cTnT Accelerated Diagnostic Protocol in a Multisite United States Cohort. Acad Emerg Med. 2023 Nov 4. doi: 10.1111/acem.14827. Epub ahead of print.
