Improved Utilization of Serial Testing Without Increased Admissions after Implementation of High-Sensitivity Troponin I: a Controlled Retrospective Cohort Study

BACKGROUND: Guidelines recommend high-sensitiv- ity cardiac troponin (hs-cTn) for diagnosis of myocardial infarction. Use of hs-cTn is increasing across the U.S., but questions remain regarding clinical and operational impact. Prior studies have had methodologic limitations and yielded conflicting results.

To evaluate the impact of transitioning from conventional cardiac troponin (cTn) to hs-cTn on test and resource utilization, operational efficiency, and patient safety. DESIGN: Retrospective cohort study in two New York City hospitals during the months before and after tran- sition from conventional cTn to hs-cTn at Hospital 1. Hospital 2 served as a control. PARTICIPANTS: Consecutive emergency department (ED) patients with at least one cTn test resulted. INTERVENTION: Multifaceted hs-cTn intervention bun- dle, including a 0/2-h diagnostic algorithm for non-ST- elevation myocardial infarction, an educational bundle, enhancements to the electronic medical record, and nurs- ing interventions to facilitate timed sample collection. MAIN MEASURES: Primary outcomes included serial cTn test utilization, probability of hospital admis - sion, ED length of stay (LOS), and among discharged patients, probability of ED revisit within 72 h result - ing in hospital admission. Multivariable regression models adjusted for age, sex, temporal trends, and interhospital differences. KEY RESULTS: The intervention was associated with increased use of serial cTn testing (adjusted risk differ - ence: 48 percentage points, 95% CI: 45–50, P < 0.001) and ED LOS (adjusted geometric mean difference: 50 min, 95% CI: 50–51, P < 0.001). There was no significant asso- ciation between the intervention and probability of admis- sion (adjusted relative risk [aRR]: 0.99, 95% CI: 0.89–1.1, P = 0.81) or probability of ED revisit within 72 h resulting in admission (aRR: 1.1, 95% CI: 0.44–2.9, P = 0.81).

Implementation of a hs-cTn inter - vention bundle was associated with an improvement in serial cTn testing, a neutral effect on probability of hospital admission, and a modest increase in ED LOS. KEY WORDS: high sensitivity cardiac troponin; test utilization; emergency department; patient outcomes; length of stay; admission; myocardial infarction J Gen Intern Med DOI: 10.1007/s11606-023-08535-3 © The Author(s), under exclusive licence to Society of General Internal Medicine 2023

The 5th generation cardiac troponin T (hs-cTnT) and high- sensitivity cardiac troponin I (hs-cTnI) assays have been approved for clinical use by the U.S. Food and Drug Admin- istration since 2017 and 2018, respectively. High-sensitivity cardiac troponin (hs-cTn) assays are more precise at low concentrations than conventional assays and allow shortened time interval to repeat assessment. Recent guidelines and rec- ommendations endorse using hs-cTn with rapid diagnostic algorithms to evaluate patients undergoing workup for non- ST-elevation myocardial infarction (NSTEMI) in the emer- gency department (ED).1–4 Rapid diagnostics may allow for improved risk stratification and faster disposition decisions, potentially leading to improved quality and efficiency of care. Hospitals across the U.S. are transitioning from con - ventional cardiac troponin (cTn) to hs-cTn, 5 but questions remain concerning the impact of such a change on clinical and operational outcomes. While there is hope that switch - ing to hs-cTn will shorten time to ED disposition decision, there also has been concern that implementing hs-cTn will increase unnecessary admissions.5–7 Although European and Australian studies have shown reduced ED length of stay (LOS) and a variable impact on admissions,7–15 these studies may not be applicable in the U.S. given the vastly different healthcare systems. Furthermore, results from U.S. studies examining these outcomes have been less consistent. 5, 16–20 Laura Warren and Brett G. Fischer contributed equally to this work. Received May 2, 2023 Accepted November 9, 2023 739 39(5):739–46 Published online , 2023November 22 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort Study JGIM Another potential advantage of a shortened time interval required for repeat testing is increased adherence to use of serial cTn testing, which is the standard of care for acute coronary syndrome (ACS) evaluation. 21 The use of serial cTn testing varies by practitioner and clinical situation, but several U.S. studies have shown an increase in use of serial cTn testing after implementation of hs-cTn.16, 17, 19, 20 Impor- tantly, to our knowledge, all prior studies have used before- and-after study design, which lacks a comparison (control) group, limiting the ability to account for temporal trends unrelated to the intervention. In February 2021, NewYork-Presbyterian/Weill Cornell Medical Center (Hospital 1) implemented a hs-cTnI assay, together with several clinical, educational, and electronic medical record (EMR) cointerventions. The aim of this study is to understand the impact of this multifaceted interven - tion on test and resource utilization, operational efficiency, and patient safety in the ED of an urban U.S. hospital with a diverse patient population. To address the limitations of single-arm before-and-after study design, we used a sister hospital as the control—NewYork-Presbyterian/Lower Man- hattan Hospital (Hospital 2), which shares ED clinicians but did not implement hs-cTn. Additional hospital characteris - tics are described in the Supplement.

Patient Population and Data Collection We performed a retrospective cohort study at two urban medical centers that share ED clinicians. Consecutive patients with at least one cTn test resulted were eligible for inclusion. Eligible patients were excluded if they walked out, were missing a disposition order to either admit or discharge, or had an ED LOS > 72 h, which was thought to represent inaccuracies in the dataset. Patients with multiple unique ED visits with cTn results were counted multiple times. Patients who were ineligible for inclusion (due to not having any cTn test results) were separately analyzed in a post-hoc analysis. Demographic, laboratory, and outcomes data were extracted from the institution’s data warehouse using Structured Query Language queries. This study was reviewed and determined to meet exemp - tion requirements at HHS 45 CFR 46.104(d) by the Weill Cornell Medicine institutional review board. Intervention At Hospital 1, Siemens Centaur hs-cTnI was introduced on February 22, 2021, to replace Siemens Centaur conventional cardiac troponin I (cTnI) (Siemens Healthineers, Erlangen, Germany). At Hospital 2, Abbott Architect conventional cTnI (Abbott Laboratories, IL, U.S.) was used throughout the study. Period 1 was defined as November 1, 2020, to February 21, 2021, and Period 2 was defined as February 22, 2021, to May 31, 2021 (Fig.  1). Prior to the transition at Hospital 1, and throughout the study at Hospital 2, a clinical pathway for the evaluation of chest pain and ACS recom - mended cTn testing at 0 and 4 h for patients presenting with suspicion for NSTEMI. For about 6 months prior to the transition, a multidis - ciplinary team that included stakeholders from emergency medicine, hospital medicine, cardiology, pathology and laboratory medicine, nursing, and information technology groups met at least monthly. These meetings led to the fol - lowing interventions: (i) development of a 0/2-h diagnostic algorithm for NSTEMI (Figure S1); 22 (ii) development and launch of an educational bundle to those who order, collect, or interpret cTn in multiple practice settings; (iii) enhance - ments to the EMR, including creation of a serial cTn order set, automatic calculation of the 2-h delta cTn value, and hyperlink to the diagnostic algorithm included in cTn result reports; and (iv) additional nursing interventions to facilitate appropriate timing of 2-h sample collection, including mes- saging alerts and restrictions on early label printing. These will be referred to as the hs-cTn intervention bundle in this study. hs-cTn intervention Bundle*i mplemented on Feb2 2, 2021 Hospital 1 SiemensC entaur conventional cTnI Period 1 Nov1 ,2 020, to Feb2 1, 2021 Period 2 Feb2 2, 2021,t oM ay 31,2 021 SiemensC entaur hs-cTnI No interventionHospital 2 Abbott Architect conventional cTnI Abbott Architect conventional cTnI Figure 1 Study design. cTnI = cardiac troponin I; hs-cTn = high-sensitivity cardiac troponin; hs-cTnI = high-sensitivity cardiac troponin I. *The hs-cTn intervention bundle included a 0/2-h diagnostic algorithm for non-ST-elevation myocardial infarction (shown in Figure S1), an educational bundle, enhancements to the electronic medical record, and nursing interventions to facilitate timed sample collection. 740 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort StudyJGIM Outcome Measures There were four primary outcomes: probability of serial (≥ 2) cTn testing, defined as having a second cTn resulted prior to disposition decision (either admit or discharge) being entered into the EMR by the ED clinician; probability of hospital admission for any reason; ED LOS, defined as time from ED arrival to time of disposition decision; and among discharged patients, probability of ED revisit within 72 h resulting in admission for any reason (primary safety outcome), which was measured using a separate dataset that enabled capture of ED revisits to any of the 8 hospitals in the NewYork-Presbyterian health system. For probability of serial cTn testing, a sensitivity analysis redefined the out - come as having a second cTn resulted within 8 h of the first, regardless of the time of disposition decision, to account for patients for whom a second cTn was ordered by the inpatient team. Secondary outcomes included test turnaround time (time from when a sample was received by the laboratory to when the result was reported in the EMR) and time inter- val between collection of the first and second tests (among those who had serial testing). Post-hoc outcome measures including temporal trends in SARS-CoV-2 test results and overall ED volume were added to evaluate the potential for residual confounding related to the coronavirus disease 2019 (COVID-19) pandemic. Troponin Assays The limit of detection (LOD), limit of quantification (LOQ), and the 99th percentile upper reference limit (URL) of the Siemens Centaur conventional cTnI assay are 0.006 µg/L (6 ng/L), 0.03 µg/L (30 ng/L) and 0.04 µg/L (40 ng/L), respectively. Values < 0.03 µg/L were reported as < 0.03 µg/L in the EMR. The LOQ of the Siemens Centaur hs-cTnI assay is 2.5 ng/L, and the 99th percentile URLs are 58 ng/L for males and 39 ng/L for females. For Abbott Architect con - ventional cTnI, Abbott did not perform the Limit of Detec - tion (LoD) and Limit of Quantitation (LoQ) 99% confi - dence interval; instead, sensitivity testing was performed to describe the values at the low end of the curve. The analyti - cal sensitivity was calculated to the 95% level of confidence as < 0.01 ng/mL (< 10 ng/L). Statistical Analysis All analyses were performed using R version 4.2.1. Descrip- tive statistics were used to summarize baseline characteris - tics and raw outcomes data. Median and IQR were reported for continuous variables, and percentages were reported for categorical variables. For probability of serial cTn testing, probability of admission, and probability of ED revisit within 72 h result - ing in admission, multiple Poisson regression modeling was employed. The regression models contained ED site (Hospital 1 vs. Hospital 2), time (Period 2 vs. Period 1), site-time interaction, age, and sex as covariates. The interac- tion term coefficient for Hospital 1, Period 2, was interpreted as the independent effect of implementation of the hs-cTn intervention bundle when adjusting for age, sex, temporal trends, and interhospital differences. Results are reported as adjusted relative risk (aRR) and 95% CI. If the result for the variable of interest was statistically significant, an additional step was added to the models to produce the adjusted risk difference (aRD) and 95% CI. For ED LOS, multiple linear regression modeling was performed using the same five covariates as the Poisson models. The LOS outcome was log-transformed to reduce right skewness and better satisfy the modeling assumption of linearity. Coefficients were then exponentiated to back- transform the results onto an untransformed scale. Adjusted geometric mean ratio (aGMR) with 95% CI is reported.16 If the result for the variable of interest was statistically signifi- cant, an additional step was added to the model to produce the adjusted geometric mean difference (aGMD) and 95% CI. Given concern that the association between interven - tion and ED LOS may be modified by number of cTn tests and ED disposition, two exploratory subgroup analyses were performed based on the number of cTn tests resulted prior to disposition decision (1 vs. ≥ 2) and disposition (admit vs. discharge). To test for interaction, linear regression models were constructed containing a three-way interaction between ED site, time, and the subgroup variable. Additional post-hoc analytical methods, including assess- ment of probability of admission at Hospital 1 only (without a control group), comparison of temporal trends in SARS- CoV-2 test results and overall ED volume between hospitals, and assessment of probability of admission and ED LOS among patients without cTn testing, are described in the Supplement. For the main analysis of primary outcomes and subgroup tests for interaction only, P values are reported and rep - resent the probability of finding an effect at least as large as that observed if there were truly no difference between groups. By convention, P < 0.05 was considered statistically significant.

Patient Characteristics At Hospital 1, out of 38,956 encounters, 9,757 (25%) were included, and at Hospital 2, out of 18,106 encounters, 4,321 (24%) were included (Fig.  2). Baseline characteristics for included patients are described in Table  1 and Table S1. Overall, patients at Hospital 1 were slightly older, more likely to be female, more likely to identify their primary language as English, and more likely to have commercial insurance or Medicare than at Hospital 2. At Hospital 2, 741 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort Study JGIM patients were more likely to identify their primary language as Cantonese or Mandarin and more likely to have Medic - aid than at Hospital 1. Within each site, these demographic variables were similar between time periods. Data on race and ethnicity were deemed unreliable due to a high degree of missingness. Primary Outcomes: Probability of Serial cTn Testing, Probability of Admission, ED LOS, and Probability of 72‑h ED Revisit Resulting in Admission (Safety Outcome)

for primary outcomes are summarized in Table  2, and regression model results for all included variables are reported in Table S2. The percent of encounters with ≥ 2 cTn tests resulted prior to ED disposition decision at Hospital 1 was 18% dur- ing Period 1 and 65% during Period 2. In the multivariable model, the intervention was associated with an absolute increase of 48 percentage points (95% CI: 45–50) in serial cTn utilization (P < 0.001). In the sensitivity analysis allow- ing for a second cTn test to have resulted up to 8 h after the first, serial cTn testing generally occurred more frequently, but the overall effect was similar (aRD: 49 percentage points, 95% CI: 46–51). The probability of admission at Hospital 1 was 55% dur - ing Period 1 and 51% during Period 2. However, in the mul- tivariable model, the intervention had no significant effect on probability of admission (aRR: 0.99, 95% CI: 0.89–1.1, P = 0.81). The median ED LOS at Hospital 1 was 293 min (IQR: 201, 409) during Period 1 and 316 min (IQR: 241, 424) during Period 2. In the multivariable model, the intervention was associated with an increase of 50 min (95% CI: 50–51) in geometric mean ED LOS ( P < 0.001). However, in the exploratory subgroup analysis by number of cTn tests, ED LOS decreased after the intervention in both patients with and without serial cTn testing, and there was no statistically significant interaction between subgroups (Pinteraction = 0.59). In the exploratory subgroup analysis by ED disposition, ED LOS increased after the intervention in both patients who were admitted and discharged, and there was no statistically significant interaction between subgroups (Pinteraction = 0.40). (Table S3). Among those encounters that resulted in discharge from the ED at Hospital 1, the probability of ED revisit within 72 h resulting in admission was 0.8% during Period 1 and 38,956 ED encounters were screened Hospital 1 10,043 Had ≥1 cTnt est resulted Hospital 2 286 Were excluded 200 Walked out aftere valuation 78 Had missing disposition 7 Had ED LOS> 72 hours 1 Walked out beforee valuation 9,757 Were included in final analysis 28,913 Were ineligible fori nclusion 28,913 Did not have a cTnt est resulted 18,106 ED encounters were screened 4,460 Had ≥1 cTnt est resulted 139 Were excluded 116W alkedo ut aftere valuation 15 Had missing disposition 6 Had ED LOS> 72 hours 2 Walked outb eforee valuation 4,321W erei ncludedi n final analysis 13,646 Were ineligible fori nclusion 13,646 Did noth ave a cTnt est resulted Figure 2 Flow diagram of the study cohort at each site. ED = emergency department; cTn = cardiac troponin; LOS = length of stay. Table 1 Baseline Characteristics * Conventional cardiac troponin I was used in these groups † High-sensitivity cardiac troponin I was used in this group Hospital 1 (Intervention) Hospital 2 (Control) Characteristic Overall N = 9,757 Period 1 (Pre)* N = 5,163 Period 2 (Post)† N = 4,594 Overall N = 4,321 Period 1* N = 2,246 Period 2* N = 2,075 Age, Median (IQR) 64 (49, 77) 64 (49, 77) 64 (49, 77) 61 (47, 76) 62 (48, 77) 60 (47, 76) Female sex, N (%) 5,200 (53) 2,709 (52) 2,491 (54) 2,043 (47) 1,047 (47) 996 (48) 742 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort StudyJGIM 1.0% during Period 2. In the multivariable model, the intervention had no significant effect on probability of ED revisit within 72 h resulting in admission (aRR: 1.1, 95% CI: 0.44–2.9, P = 0.81). Secondary Outcomes: Test Turnaround Time, Time Between Serial cTn Test Collection, and Post‑Hoc Outcomes Test turnaround time was similar at both hospitals during both time periods (Table S4). Among those with serial cTn testing, the median time between collection of the first and second sample at Hospital 1 was 3.6 h (IQR: 3.0, 4.2) dur - ing Period 1 and 2.0 h (IQR 1.8, 2.2) during Period 2. The median time between collection of the first and second sam- ple at Hospital 2 during both time periods was similar to that of Hospital 1 during Period 1 (Figure S2). Results of post-hoc analyses are described in the Supplement.

In a retrospective cohort study with one intervention site and one control site, implementation of a hs-cTn intervention bundle was associated with a large increase in serial cTn testing, neutral effect on probability of hospital admission, and modest increase in ED LOS. For discharged patients, the intervention had a neutral effect on probability of ED revisit within 72 h resulting in admission, although the effect estimate was imprecise due to the small number of events. A major strength of this study is the comparison to a con- trol hospital where the intervention was not undertaken. This has significant advantages over single-arm before-and-after studies, most notably in controlling for potential tempo - ral trends unrelated to the intervention. To the best of our knowledge, this is the first study to employ this design to evaluate the effect of implementation of hs-cTn on these outcomes. Recent cardiology and emergency medicine society guidelines recommend use of serial cTn testing in patients presenting with symptoms suspicious for NSTEMI. 2–4, 21 Thus, implementation of the hs-cTn intervention bundle was associated with an improvement in care consistent with national guidelines. Consistent with our findings, previous U.S. before-and-after studies have also demonstrated large increases in serial cTn testing after switching from conven - tional cTn to hs-cTn.16, 17, 19, 20 Within the hs-cTn intervention bundle, there were a number of potential contributors to the observed change in ordering patterns, including employment of a diagnos - tic algorithm that prescribes serial cTn testing in most patients in whom the ED clinician is considering the diagnosis of NSTEMI, implementation of a serial cTn Table 2 Primary Outcomes cTn = cardiac troponin; ED = emergency department; LOS = length of stay * Conventional cardiac troponin I was used in these groups † High-sensitivity cardiac troponin I was used in this group ‡ Independent effect of implementation of the high sensitivity cardiac troponin intervention bundle at Hospital 1 after adjusting for age, sex, tempo - ral trends, and interhospital differences § Adjusted risk difference expressed as percentage points, derived from Poisson regression model ‖ Adjusted relative risk, derived from Poisson regression model ¶ ED LOS was defined as the time from ED arrival to time of decision entered into electronic medical record to either admit or discharge. ED LOS median and IQR is expressed in minutes # Adjusted geometric mean difference, derived from linear regression model on log-transformed data ** Only patients discharged from the ED during the index visit were eligible for this outcome: for Hospital 1, Period 1, N = 2,346; for Hospital 1, Period 2, N = 2,299; for Hospital 2, Period 1, N = 1,107; for Hospital 2, Period 2, N = 1,101. Within each hospital and period combination, the sum of discharged and admitted patients does not exactly equal the total N because the data used for this outcome were derived from a separate dataset that enabled capture of ED revisits to any of the 8 hospitals in the NewYork-Presbyterian network, but to which exclusion criteria were not applied Hospital 1 (Intervention) Hospital 2 (Control) Outcome Period 1 (Pre)* N = 5,163 Period 2 (Post)† N = 4,594 Period 1* N = 2,246 Period 2* N = 2,075 Adjusted Effect‡ (95% CI) P value Serial (≥ 2) cTn tests, N (%) 916 (18) 2,999 (65) 302 (13) 403 (19) 48 (45–50)§ < 0.001 Hospital admission, N (%) 2,833 (55) 2,344 (51) 1,122 (50) 949 (46) 0.99 (0.89–1.1)‖ 0.81 ED LOS¶, Median (IQR) 293 (201, 409) 316 (241, 424) 263 (187, 376) 269 (189, 369) 50 (50–51)# < 0.001 72-h ED revisit resulting in hospital admission**, N (%) 41 (0.8) 47 (1.0) 11 (0.5) 11 (0.5) 1.1 (0.44–2.9)‖ 0.81 743 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort Study JGIM order set in the EMR, the shorter time interval needed to check a second cTn test, the educational bundle, and general increased awareness of national guidelines over time. Among these hypotheses, Hospital 2 may have also been somewhat affected by the educational bundle and clinical practice influ- ence from the 0/2-h diagnostic algorithm for NSTEMI (given that ED clinicians rotated between both sites), as well as increased guideline awareness over time. However, the fact that the change at Hospital 1 remained large and statisti - cally significant when controlling for Hospital 2 suggests that the majority of the effect was isolated to features specific to actual implementation of the hs-cTn intervention bundle, though it is not possible to further isolate the effect to any of the bundle’s individual components. Findings from previous U.S. studies assessing the impact of hs-cTn implementation on probability of admission and ED LOS have been less consistent.5, 16–20 For example, Suh, et al., showed no change in median provider-to-disposition decision time (defined similarly to our ED LOS outcome) and no increase in hospital admissions, 17 while Ford, et al., reported increased ED LOS and decreased admissions. 16 In the largest of these studies, McCarthy, et al. demonstrated a statistically significant decrease in ED LOS, but the clinical and operational importance of this finding was unclear, with a difference in median ED LOS of < 30 min. 5 Many factors may have contributed to these discrepant observations, such as differences in clinical and EMR pathways employed, turn- around time of serial cTn results, patient population, the edu- cation and training provided to clinical teams surrounding the transition, and study design and statistical methods, including the challenges in correcting for multifactorial operational effi- ciencies contributing to variations in ED LOS. As the first two-arm cohort study with a control group to evaluate the effect of implementation of hs-cTn on the prob- ability of admission and ED LOS among patients with cTn testing, our findings greatly enhance the quality of the over- all body of evidence. The post-hoc analysis using before- and-after design without a control group was performed to demonstrate this point: had the present study followed the same design as prior studies, an incorrect conclusion— decreased probability of admission after the intervention— could have been drawn. While our controlled analysis did not demonstrate a reduction in probability of admission after the intervention, given the present study’s strengths, the lack of effect should mitigate concerns that implementing hs-cTn might increase hospital admissions. 5–7 On the other hand, ED LOS increased after the implemen- tation of the hs-cTn intervention bundle, despite a reduction of 1.6 h in median time between serial cTn test collection at Hospital 1 and a stable test turnaround time. Post-hoc analysis demonstrated that ED LOS was stable in patients who did not have cTn testing. The clinical and operational importance of the increased ED LOS remains unclear, with a difference in geometric mean ED LOS of < 1 h. The most plausible explanation for the observed increase in ED LOS is the increase in serial cTn testing instead of single cTn testing, suggesting that the observed increase in ED LOS may have been an unintended but worthwhile cost. Our hospital implemented a 0/2-h algorithm, but the finding of increased ED LOS may warrant reconsideration of using a 0/1-h algorithm. Interestingly, when analyzed by subgroup based on number of cTn tests, ED LOS appeared to decrease in both groups with and without serial testing. However, this exploratory subgroup analysis should be interpreted with caution due to concern for confounding by other patient fac- tors, given the drastic shift in the proportion of patients with serial cTn testing between pre- and post-intervention periods at Hospital 1. Furthermore, number of cTn tests is likely a mediator rather than a confounder in the relationship between the intervention and ED LOS, and stratification by mediators may introduce collider bias.23 As with many similar studies, our study has several addi- tional limitations. First, we conducted this study in a single hospital system, which may limit generalizability. Further - more, interpretation of our data on race and ethnicity is lim- ited by a large degree of missingness, a previously described problem that is not unique to our dataset.24 Second, as with any observational study, we cannot rule out residual confounding. While Hospital 2 shares the same ED clinicians and is within the same geographic location as Hospital 1 (Manhattan, New York), the patient popula - tion may vary in terms of demographics, diagnoses, and acuity. For example, more patients at Hospital 2 identified Cantonese or Mandarin as their primary language (which we included as a complement to race and ethnicity data). While these differences may represent potential causes of residual confounding, we did attempt to control for inter - hospital differences in our statistical models. In addition, while inclusion of Hospital 2 was intended to control for potential temporal trends unrelated to the intervention, there could have been variation in temporal trends between sites. However, we are reassured that, at least with ED vol - ume and COVID-19 infections, this did not seem to be the case. Third, we were not able to include patient-level clinical data, limiting our ability to restrict inclusion to patients with suspicion for ACS, assess for cardiovascular risk factors, evaluate the cause of 72-h ED revisits resulting in admission, and to assess clinical outcomes. Fourth, we did not evaluate utilization of additional down- stream testing, and we did not assess financial impact for the hospital or patients. Fifth, the use of ED revisit within 72 h resulting in acute hospital admission as a surrogate quality measurement, while well recognized in ED literature, is limited by the dataset not including patients who returned to other institu - tions, which existing research suggests occurs in 11–24% of all ED return visits. 25–27 Our study utilized data to capture 744 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort StudyJGIM return visits across 8 New York City EDs, but this may have provided incomplete mitigation. These limitations will serve as areas of further research.

In this retrospective observational cohort study with one intervention site and one control site, implementation of a hs-cTn intervention bundle when switching from conven - tional cTnI to hs-cTnI was associated with increased adher- ence to guideline-recommended serial cTn testing without significant effect on probability of admission or 72-h ED revisit leading to admission, and a modest increase in ED LOS. Taken together with the evidence from prior U.S. studies in real-world ED settings, transitions to hs-cTn are consistently associated with improvements in adherence to guideline-recommended serial cTn testing, and while the impact on hospital admissions and ED LOS remains unclear, it seems most likely to be small. Supplementary Information The online version contains supplementary

available at https:// doi. org/ 10. 1007/ s11606- 023- 08535-3.

We thank Arthur T. Evans, MD, MPH (Weill Cornell Medicine) for his invaluable guidance in developing the sta - tistical analysis plan and Patrick Rumble, BS (Weill Cornell Medi - cine), for his instrumental assistance in data acquisition. Corresponding Author: He S. Yang, PhDPeter A. D. Steel, MDZ - hen Zhao, PhD; Department of Pathology and Laboratory Medicine Department of Emergency Medicine, Weill Cornell Medicine Weill Cornell Medicine, New York, NY, USA, New York, NY, USA (e-mail: [email protected] [email protected] zhz9010@ med.cornell.edu). Data Availability The datasets generated and analyzed during the current study are available from the corresponding authors on rea- sonable request. Declarations: Conflict of Interest: LW is pending a patent for “Methods for detect- ing and treating endometriosis” Publication number: 20210096137. JS had a registration fee waived as a speaker at XGM Conference May 2023 and is the Co-Chair of the didactics Sub-committee for the Society of Academic Emergency Medicine 2022–2023 (unpaid). RJK has stock options in Cleerly Health, Inc. AC has received consulting fees from Lei- ca Biosystems (Immunohistochemistry) and Boehringer-Ingelheim (he- matopoietic neoplasms) and lecture fees at McGill University, Medical College of Wisconsin, and Northshore University Hospital (Evanston, IL). HSY received a speaker fee from Siemens Healthineers. P ADS received consulting fees for ET health, speaker fees for the American College of Emergency Physicians Directors Academy, and is on the David Lynch Foundation National Advisory Board. ZZ has sponsored research sup- ported by Novartis, Waters, Siemens Healthineers, Polymedco, Waters, Roche and ET Healthcare and has received consulting/speaker fee from Siemens Healthineers, Roche and ET Healthcare.

1. Kontos MC, de Lemos JA, Deitelzweig SB, et al. 2022 ACC Expert Con- sensus Decision Pathway on the Evaluation and Disposition of Acute Chest Pain in the Emergency Department: A Report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol. 2022;80(20):1925-1960. 2. Collet JP, Thiele H, Barbato E, et a l. 2020 ESC Guidelines for the management of acute coronary syndromes in patients pre - senting without persistent ST-segment elevation . Eur Heart J. 2021;42:1289-1367. 3. Sandoval Y, Apple FS, Mahler SA, Body R, Collinson PO, Jaffe AS . High-Sensitivity Cardiac Troponin and the 2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guidelines for the Evaluation and Diag - nosis of Acute Chest Pain. Circulation. 2022;146:569-581. 4. Gulati M, Levy PD, Mukherjee D, et al. 2021 AHA/ACC/ASE/ CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diag - nosis of Chest Pain: A Report of the American College of Cardiology/ American Heart Association Joint Committee on Clinical Practice Guidelines . Circulation. 2021;144:e368-e454. 5. McCarthy C, Li S, Wang TY, et al. Implementation of High-Sensitiv - ity Cardiac Troponin Assays in the United States . J Am Coll Cardiol. 2023;81(3):207-219. 6. Yang HS, Shemesh A, Li J, et al. No increase in the incidence of car - diac troponin I concentration above the 99th percentile by Siemens Centaur high-sensitivity compared to the contemporary assay . Clin Biochem. 2021;89:77-80. 7. Khan E, Lambrakis K, Nazir SA, et al. Implementation of more sen - sitive cardiac troponin T assay in a state-wide health service . Int J Cardiol. 2022;347:66-72. 8. Peck D, Knott J, Lefkovits J . Clinical impact of a high-sensitivity troponin assay introduction on patients presenting to the emergency department . Emerg Med Australas. 2016;28:273-8. 9. Greenslade JH, Parsonage W, Foran L, et al. Widespread Introduc - tion of a High-Sensitivity Troponin Assay: Assessing the Impact on Patients and Health Services . J Clin Med. 2020;9(6):1883. 10. Jülicher P, Greenslade JH, Parsonage WA, Cullen L . The organi - sational value of diagnostic strategies using high-sensitivity tro - ponin for patients with possible acute coronary syndromes: a trial- based cost-effectiveness analysis . BMJ Open. 2017;7:e013653. 11. Roos A, Holzmann M J. Healthcare and Resource Use in Patients With Stable High-Sensitivity Cardiac Troponin T Levels. Am J Cardiol. 2020;128:67-74. 12. Anand A, Lee KK, Chapman AR, et al. High-Sensitivity Cardiac Troponin on Presentation to Rule Out Myocardial Infarction: A Stepped-Wedge Cluster Randomized Controlled Trial . Circulation. 2021;143:2214-2224. 13. Ambavane A, Lindahl B, Giannitsis E, et al. Economic evaluation of the one-hour rule-out and rule-in algorithm for acute myocardial infarction using the high-sensitivity cardiac troponin T assay in the emergency department . PLoS One. 2017;12:e0187662. 14. Conway R, Byrne D, Cournane S, O’Riordan D, Coveney S, Silke B. Is there excessive troponin testing in clinical practice? Evidence from emergency medical admissions. Eur J Intern Med. 2021;86:48-53. 15. Johannessen TR, Halvorsen S, Atar D, et al. Cost-effectiveness of a rule-out algorithm of acute myocardial infarction in low-risk patients: emergency primary care versus hospital setting . BMC Health Serv Res. 2022;22:1274. 16. Ford JS, Chaco E, Tancredi DJ, Mumma BE . Impact of high-sen - sitivity cardiac troponin implementation on emergency department length of stay, testing, admissions, and diagnoses . Am J Emerg Med. 2021;45:54-60. 17. Suh EH, Tichter AM, Ranard LS, et al. Impact of a rapid high- sensitivity troponin pathway on patient flow in an urban emergency department . J Am Coll Emerg Physicians Open. 2022;3:e12739. 18. Bevins NJ, Chae H, Hubbard JA, et al . Emergency Depart - ment Management of Chest Pain With a High-Sensitivity Tro - ponin-Enabled 0/1-Hour Rule-Out Algorithm . Am J Clin Pathol. 2022;157:774-780. 19. Ola O, Akula A, Michieli LD, et al. Clinical Impact of High-Sensitivity Cardiac Troponin T Implementation in the Community . J Am Coll Cardiol. 2021;77:3160-3170. 20. Younis A, Farooq S, Bisognano JD, et al. Outcomes Associated with

of the 5(th) Generation High-Sensitivity Cardiac Tro - ponin in Patients Presenting with Cardiovascular Disorders . J Emerg Med. 2022;62:657-667. 21. Tomaszewski CA, Nestler D, Shah KH, Sudhir A, Brown MD . Clini- cal Policy: Critical Issues in the Evaluation and Management of Emer- gency Department Patients With Suspected Non-ST-Elevation Acute Coronary Syndromes . Ann Emerg Med. 2018;72:e65-e106. 745 Warren et al.: High-Sensitivity Troponin Implementation: Controlled Cohort Study JGIM 22. Fischer BG, Evans AT . High-Sensitivity Cardiac Troponin Algo - rithms and the Value of Likelihood Ratios. J Gen Intern Med. 2023;38(9):2189-2193. 23. Rohrer JM. Thinking clearly about correlations and causation: Graphi- cal causal models for observational data . Advances in Methods and Practices in Psychological Science. 2018;1:27-42. 24. Polubriaginof FCG, Ryan P, Salmasian H, et al. Challenges with qual- ity of race and ethnicity data in observational databases. J Am Med Inform Assoc. 2019;26:730-736. 25. Chartier LB, Ovens H, Hayes E, et al. Improving Quality of Care Through a Mandatory Provincial Audit Program: Ontario’s Emer - gency Department Return Visit Quality Program . Ann Emerg Med. 2021;77:193-202. 26. Rising KL, Karp DN, Powell RE, Victor TW, Carr BG . Geography, Not Health System Affiliations, Determines Patients’ Revisits to the Emergency Department . Health Serv Res. 2018;53:1092-1109. 27. Shy BD, Loo GT, Lowry T, et al. Bouncing Back Elsewhere: Multilevel Analysis of Return Visits to the Same or a Different Hospital After Initial Emergency Department Presentation . Ann Emerg Med. 2018;71:555- 563.e1. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law. 746