0/2-Hour i-STAT hs-cTnI Pathway Validated in U.S. EDs

Key Takeaways

• In a prospective multisite observational cohort at three U.S. emergency departments, 578 patients had complete 0/2-hour assessments, index myocardial infarction occurred in 7.4% (43/578), and the original i-STAT hs-cTnI pathway classified 54.8% (317/578) to rule-out with a 99.4% negative predictive value for index myocardial infarction.
• The original pathway placed 36.3% (210/578) of patients in observation, where 4.3% (9/210) had adjudicated index myocardial infarction, and 8.8% (51/578) in rule-in, where the positive predictive value for index myocardial infarction was 62.7% (32/51; 95% CI 48.1%–75.9%).
• Exploratory optimization increased rule-out efficacy to 60.0% (347/578) with the same 99.4% negative predictive value (95% CI 97.9%–99.9%) and a higher positive predictive value of 74.4% (95% CI 58.8%–86.5%), while the algorithm was not used for disposition and the lower bound of the original rule-out negative predictive value confidence interval was 97.7%.

More than half of adults evaluated for possible acute coronary syndrome were placed in the rule-out zone by the original 0/2-hour i-STAT hs-cTnI pathway, with 54.8% (317/578) classified to rule-out and only 2 of 317 ruled-out patients having adjudicated index myocardial infarction. An exploratory recalibration widened rule-out classification and improved rule-in performance without changing the reported negative predictive value for index myocardial infarction. This Academic Emergency Medicine validation of a 0/2-hour i-STAT hs-cTnI algorithm represents a U.S. multisite assessment of bedside high-sensitivity troponin testing in emergency department chest pain evaluation.

In the ACUTE trial, investigators conducted a prospective multisite observational cohort study from February through September 2025 at three North Carolina emergency departments, accruing 602 adults and obtaining complete algorithm assessments in 578. Participants had symptoms concerning for possible acute coronary syndrome without STEMI; the cohort was 48% female, 40% non-White, and had a median age of 60 years, IQR 50–70. Point-of-care i-STAT 1 hs-cTnI testing was performed simultaneously with central laboratory Beckman Coulter hs-cTnI testing, and clinicians were blinded to the point-of-care results. Expert-adjudicated index visit myocardial infarction based on clinical hs-cTnI measures was the primary outcome, while 30-day cardiac death or myocardial infarction and 30-day major adverse cardiac events were secondary outcomes; 30-day major adverse cardiac events occurred in 9.3% (54/578).

For exploratory optimization, the authors tested a single-sample rule-out threshold below 7 ng/L, serial rule-out with a 0-hour value below 9 ng/L and a delta below 6 ng/L, and rule-in with a 0-hour value at or above 90 ng/L or a delta at or above 15 ng/L. With those cut points, rule-out efficacy increased to 60.0% (347/578) while the negative predictive value for index myocardial infarction remained 99.4% and the positive predictive value increased to 74.4%. The pattern suggested broader low-risk classification together with a more selective rule-in zone, with an optimized positive likelihood ratio of 36.2 (95% CI 19.7–66.7).

The authors noted that all study sites were in North Carolina, the sample size and event burden were modest, and the original rule-out negative predictive value had a lower confidence bound of 97.7%, leaving limited precision around the miss rate. They also emphasized that the pathway was observational and did not guide disposition because clinicians were blinded to point-of-care results, and that the primary analysis included some patients classifiable from very low or high 0-hour values without a completed 2-hour sample, which they said could have biased results. A sensitivity analysis restricted to 435 participants with completed 0-hour and 2-hour samples yielded similar results, and the cohort still showed high rule-out performance with exploratory gains in classification efficiency.

Clinician Questions

How were patients distributed across rule-out, observation, and rule-in zones with the original 0/2-hour i-STAT hs-cTnI pathway?

In hemodynamically stable adults without STEMI who were evaluated for possible acute coronary syndrome at three U.S. emergency departments, the original 0/2-hour i-STAT hs-cTnI pathway classified 54.8% (317/578) to rule-out, 36.3% (210/578) to observation, and 8.8% (51/578) to rule-in. Within the observation zone, 4.3% (9/210) had adjudicated index myocardial infarction, and the rule-in zone had a 62.7% positive predictive value for index myocardial infarction.

What cut points were used in the exploratory optimized 0/2-hour i-STAT hs-cTnI algorithm for suspected acute coronary syndrome?

In adults with symptoms concerning for possible acute coronary syndrome, the exploratory optimized 0/2-hour i-STAT hs-cTnI pathway used a single-sample rule-out threshold below 7 ng/L, serial rule-out with a 0-hour value below 9 ng/L and delta below 6 ng/L, and rule-in with a 0-hour value at or above 90 ng/L or delta at or above 15 ng/L. In an exploratory same-cohort optimization, these cut points classified 60.0% (347/578) to rule-out with an observed 99.4% negative predictive value for index myocardial infarction.

What were the 30-day cardiac death or myocardial infarction results for the original and optimized troponin pathways in this U.S. cohort?

In the North Carolina emergency department cohort, 30-day cardiac death or myocardial infarction occurred in 8.6% (50/578) of participants overall. For the original 0/2-hour i-STAT hs-cTnI pathway, the rule-out negative predictive value for 30-day cardiac death or myocardial infarction was 99.1% (95% CI 97.3%–99.8%) and the rule-in positive predictive value was 64.7% (95% CI 50.1%–77.6%). For the optimized pathway, the rule-out negative predictive value was 99.1% (95% CI 97.5%–99.8%) and the rule-in positive predictive value was 74.4% (95% CI 58.8%–86.5%).

What limits should frame interpretation of the U.S. validation of the 0/2-hour i-STAT hs-cTnI algorithm?

The U.S. validation of the 0/2-hour i-STAT hs-cTnI algorithm was observational, all three emergency department sites were in North Carolina, and clinicians were blinded to point-of-care hs-cTnI results so the pathway did not guide disposition. The modest sample size with a 7.4% index myocardial infarction rate left wide confidence intervals, including a 97.7% lower bound for the original rule-out negative predictive value. The primary analysis also included some patients classifiable from extreme 0-hour values without a completed 2-hour sample, which the authors said could have biased results.

Recommended Reading

• For more on troponin rule-out strategies: 0/1-Hour vs 0/3-Hour Troponin Pathways in Suspected ACS