LABORATORIO DE URGENCIAS

 

 

 

Improvement of Emergency Department Chest Pain Evaluation Using Hs-cTnT and a Risk Stratification Pathway

 

                                                     Zhengqiu Zhou, MD,1 Kevin S. Hsu, MD :1 Joshua Eason, DO; Brian Kauh, MD;                                                                       Joshua Duchesne, MD ; Mikiyas Desta, MD; William Cranford, MS; Alison Woodworth, PHD;
                                                     James D. Moore, MD; Seth T. Stearley, MD and Vedant A. Gupta, MD

                                                                                     Journal of Emergency Medicine, Vol. 66, No. 6, pp. e660–e669, 2024

 

 

Chest pain is among the most common reasons for pre- sentation to the emergency department (ED) worldwide, accounting for approximately 8–10 million visits a year, and comprises 5–8% of ED visits in the United States (1).
The extensive differential diagnosis of chest pain, ranging from benign conditions to life-threatening emer- gencies such as acute myocardial infarction (AMI), makes the evaluation a challenge, particularly because the inten- sity of symptoms do not always correlate with severity of the etiology or prognosis (2). Standard workup approach for AMI includes history, physical, measurement of serial cardiac biomarkers, risk stratification scores, and electro- cardiogram (ECG) findings. Ultimately, 75–85% of these patients do not have an AMI (3 , 4). However, the compre- hensive testing to work up these patients is estimated to account for 10 to 13 billion dollars in health care costs annually in the United States (5 , 6). Multiple rapid rule-in and rule-out diagnostic strate- gies have been trialed in EDs to quickly risk-stratify patients toward likelihood of major cardiac involvement (7 ,8). These strategies have largely leveraged clinical risk prediction tools and higher sensitivity assays. The use of clinical risk prediction tools, such as the HEART score (History, Electrocardiogram, Age, Risk factors, Troponin), have become widely prevalent. The HEART score is a validated chest pain risk stratification tool that utilizes a combination of pretest probability, serial car- diac troponin concentrations, and ECG characteristics (9). However, no specific chest pain evaluation algo- rithm has been universally adopted, and significant health care resources continue to be dedicated toward compre- hensive evaluations of patients with chest pain to avoid future major adverse cardiac events (MACE), which are most frequently AMIs, hospitalizations for heart failure, and cardiac-related deaths ( 3–6 ,10). The recent Ameri- can Heart Association/American College of Cardiology guidelines of chest pain management specifically high- light the importance of developing and testing risk strat- ification pathways that balance efficiency of health care resource utilization with patient safety (9). The Roche (F. Hoffman-LaRoche AG, Basel, Switzer- land) high-sensitivity cardiac troponin (hs-cTnT) is a 5th generation troponin assay that was approved by the U.S. Food and Drug Administration in 2017 after multiple large, multicenter studies and meta-analyses validated its use (11–13). It has been identified to have sensitivity of 91% and specificity of 74% for AMI in the ED setting (14).
The hs-cTnT has higher sensitivity but lower specificity than the older 4th generation Roche Elecsys cTnT assay, which has been found to have sensitivities and specificities of 89% and 80%, respectively (14). This has been the natural progression of the troponin literature re- sulting in what used to be very specific but not sensitive, to an assay that is now very sensitive, but at the expense of specificity (15). High-sensitivity troponin assays, such as the Roche assay, has become the gold standard test to eval- uate for myocardial damage, and many institutions have transitioned or are in the process of transitioning to these assays in the United States (9 ,16).
However, few studies have examined the variables that impact the observed ED throughput advantages of the hs-cTnT, especially in con- junction with other risk stratification pathways. Our institution, a large academic quaternary care cen- ter, transitioned from the Roche Elecsys 4th generation cTnT (measured at 0,3 and 6 h) to Roche Elecsys 5th generation hs-cTnT (measured at 0 and 2 h) in Decem- ber 2018.
In this study, we aimed to evaluate the use of a novel risk stratification pathway that we created, named the “Acute Chest Pain Optimal Care Pathway”(Figure 1), which utilizes a patient-specific HEART score in combi- nation with hs-cTnT level, with regards to efficiency of ED resource utilization in the evaluation of chest pain pa- tients.

Methods

* Study Design and Setting
The Acute Chest Pain Optimal Care Pathway was a novel chest pain evaluation path- way Roche 5th generation hs-cTnT developed by the Cardiology and Emergency Departments at our large quaternary academic center. A retrospective chart review was conducted on patients who met the appropriate inclusion criteria and exclusion criteria 6 months prior to and 6 months after the implementation date of December 11, 2018.
* Selection of Participants
Our study population included all patients 18 years of age and older that presented to our ED with a chief com- plaint of “non-traumatic chest pain”between June 1, 2018 and June 1, 2019, approximately 6 months prior to and 6 months after implementation date. While recognizing that potential acute coronary syndromes manifest with many different presenting symptoms, only chest pain patients were included in the analysis to be consistent with pre- vious studies. Exclusion criteria included traumatic chest pain, patients who suffered from cardiac arrest, and pa- tients who were diverted to emergent invasive coronary angiography prior to additional workup. The group of pa- tients with ED encounters prior to December 11, 2018 were designated as the “pre-implementation”group, and those seen in the ED on or after December 11, 2018 were the “post-implementation”group. Each chest pain evaluation for patients seen on multiple occasions were considered independent encounters.
* Interventions
The Acute Chest Pain Optimal Care Pathway was implemented on December 11, 2018. According to Path- way, patients whose presentation warranted evaluation for acute coronary syndrome (ACS) without definitive is- chemic changes on serial ECGs underwent the Roche 5th generation hs-cTnT. With the hs-cTnT assay, measure- ments were made at 0 h and 2 h. Patients were triaged into one of three groups based on their initial troponin concentration: troponin < 6 ng/L regardless of gender (undetectable), troponin ≥6 ng/L but less than the gender- specific 99th percentile cut-off (13 ng/L for females,18 ng/L for males), or troponin greater than the gender- specific 99th percentile cut-off (15).
Patients with initial troponin < 6 ng/L were deemed very low risk ( < 1% MACE at 1 year), and deemed appropriate for discharge from the ED with outpatient Cardiology referral at pa- tient/provider preference. For the other two groups, based on presence or absence of a significant delta troponin, defined as > 10 ng/L, and the HEART score, patients then were either discharged with an outpatient Cardiol- ogy referral, further evaluated with coronary computed tomography angiography, or referred for inpatient Cardi- ology evaluation. Leading up to implementation of the Acute Chest Pain Optimal Care Pathway, provider education was pro- vided via departmental e-mails and a series of in-person
educational sessions for faculty and trainees across the Emergency Medicine Department, Division of Hospital Medicine, and Cardiovascular Medicine Department.
The principal investigator (VAG) of this study was readily available to answer questions throughout the early imple- mentation period. Prior to pathway implementation, there was not a stan- dardized chest pain workflow strategy.
Troponins were measured by Roche Elecsys 4th generation cTnT and per- formed up to three times at 0 h, 3 h, and potentially, 6 h, based on provider discretion. The Institutional Review Board approved the study and waived need for informed consent, given that no patient contact was made by study personnel and patient data were both aggregated and sufficiently de-identified prior to statistical analysis.
* Outcomes
The primary outcome was overall ED length of stay (LOS). ED LOS was defined as arrival time (initial regis- tration time) until time of admission, discharge, or elope- ment. Boarding time was not included in the overall ED LOS. Secondary outcomes included Cardiology consult rates, admission rates, number of ED boarders, and num- ber of eloped patients. Subsequently, a time analysis was performed to better understand where changes in the over- all ED LOS were observed, comparing arrival time with consult time, and consult time with time to disposition (for those discharged) and time to admission (for those admit- ted).
Comprehensive MACE evaluations from the same data set are being independently reported in a separate manuscript,but are included here as well for a more complete understanding of performance of the pathway. MACE was identified by comprehensive review of our electronic medical record, which encompasses ED, inpatient, and outpatient data. We also reviewed the Kentucky Health Information Exchange, which is a consortium encompassing majority hospitals throughout the Common- wealth of Kentucky to allow for data sharing.
* Analysis
Chi-squared test was used for categorical data. Mann- Whitney U test was used for time analysis due to assumed unequal variances. Binary logistic or multivariate linear regression models were used where indicated. Significance was defined in this study as p < 0.05. As a sensitivity analysis, a mixed-effect model was used to account for multiple encounters per patient. Data were analyzed using SPSS (version 27; IBM, Armonk, NY) and the R programming language, version 4.1.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

 * Characteristics of Study Subjects
Patient data from June 1, 2018 to June 1, 2019 were as- sessed. Two cohorts (pre- and post implementation) were identified using December 11, 2018 as the implementa- tion date. Baseline population characteristics are shown in Table 1 . There was a total of 3977 patients that met initial screening inclusion criteria during our 1-year study period. After reviewing for completeness of data and ex- clusion criteria, 3236 patients were included in the final analysis: 1707 of which were prior to the Acute Chest Pain Optimal Care Pathway implementation and 1529 after; 304 patients had more than one ED visit during the study period. There were 217 patients seen in both preimplementation and postimplementation periods. Each hospital encounter was treated as independent ED evaluations for chest pain.
* Main Results
Primary outcome analysis revealed a significant re- duction in the mean overall LOS in the ED from 317 min to 286 min, an absolute reduction of 31 min (95% confidence interval [CI] 22–41 min), in the postimple- mentation group when compared with the control group ( p < 0.001). The median LOS was 309 min preimplementa- tion (interquartile range 228–384 min) compared with 266 min postimplementation (interquartile range 215–335; p < 0.001).
Using a multivariate linear regression model, after adjusting for age, gender, and cardiology consulta- tion, a statistically significant association between LOS in the two test groups remained ( p < 0.001). The calculated coefficient estimate was−34 (95% CI−24 to−44), suggesting a decrease in ED LOS of 34 min per patient af- ter implementation of the Acute Chest Pain Optimal Care Pathway independent of the above variables.
For our secondary analyses, there was a significant reduction of cardiology consults from 26.9% of (n = 459) in the preimplementation cohort to 16.0% (n = 244) in the postimplementation cohort (odds ratio [OR] 0.516, 95% CI .434–.614, p < 0.001; Table 2 ). Using a binary logistic regression model, after adjusting for age, gender, and LOS, a statistically significant association between pre - and postimplementation continues to exist with the rate of cardiology consults. The odds of a cardiology consult were 54% lower in the postimplementation cohort ( p < .001; OR 0.46, 95% CI 0.38–0.56). Furthermore, the overall rates of admitted patients significantly decreased from 30.1% (n = 513) preim- plementation to 22.7% (n = 347) postimplementation (OR .683, 95% CI .583–.801, p < 0.001). This absolute reduction in admission rate was 7.4%, reducing one admission for every 13.5 chest pain patients seen.
When examining the subgroup of admitted patients, there was no statistically significant difference in the percentage of patients admitted to a primary cardiology service com- pared with any other hospital service ( p = 0.144).
The number of ED boarders as a proportion of all nontrau- matic chest pain patients significantly decreased, from 25.1% (n = 429) preimplementation to 18.6% (n = 285) postimplementation (OR 0.683, 95% CI 0.576–0.808, p < 0.001).
However, when analysis was repeated for boarders as a portion of only admitted patients, there was no significant reduction in the number of ED boarders (OR 0.872, 95% CI 0.606–1.248, p < 0. 453), signifying that the reduction in boarders is mostly attributed to reduction of total admissions.
There was no significant change in the number of eloped patients (OR 0.907,95% CI 0.435–1.890,p = 0.793).
For the time analysis, there was no significant difference noted in the time from patient arrival to Cardiology consultation ( p = 0.214), time from consult to patient dis- position ( p = 0.226), or time from consult to admission ( p = 0.436). These findings are summarized in Figure 2.
Supplementary there was a low rate of MACE, defined as a composite of cardiovascular mortality, acute myocardial infarction, and unplanned revascularization at 30 days in both co- horts.
There were 3 MACE events in the preintervention cohort, and 5 MACE events in the postintervention cohort (0.2% vs. 0.3%, p = ns). The MACE rates pre- and postin- tervention are well below the < 1–2% acceptable missed ACS rate proposed by the American College of Emer- gency Physicians 2018 Clinical Policy Document (17).

Discussion

In this retrospective analysis conducted at our quaternary care academic medical center, we investigated the efficacy of utilizing the Roche 5th generation hs-cTnT in conjunction with a novel cardiac risk stratification pathway for patients presenting with nontraumatic chest pain.
The key findings from this study can be summarized as follows:
1) Combining hs-cTnT with our innovative risk stratifi- cation pathway not only enhances ED efficiency, but also reduces the need for hospital admissions and additional troponin testing.
2) the improved ED throughput primarily results from the early discharge of low-risk patients without the involvement of consultants.
3) importantly, our pathway achieves these outcomes safely, without a significant impact on downstream MACE, as will be detailed in a separate manuscript. This investigation helps elucidate discrepancies observed in existing literature and provides valuable guidance for the development of clini- cal pathways in other ED settings. Prior to implementing the Acute Chest Pain Opti- mal Care Pathway, the evaluation of chest pain involved troponin measurements at 0, 3, and 6 h, determined at the discretion of the provider. However, our analysis revealed that over 90% of patients (1587 of the 1707 patients) did not undergo a third troponin assessment. After the transition to a protocol incorporating troponin measurement at 0 and 2 h, we observed a median decrease in LOS of 31 min. Additionally, this intervention led to a reduction in cardiology consultations and patient admission rates. These changes reflect a positive trend toward high-value health care,resulting in significant cost savings for both the overall health care system and patients, all while maintaining consistent levels of downstream MACE, hospital admissions, and cardiology consultations. Although several prior studies have evaluated the safety and efficacy of hs-cTnT in combination with a risk stratification pathway, very few of these have focused on the downstream effects on ED throughput or the drivers of any reduction observed. Of the other studies that did evaluate the effects of hs-cTnT implementation on ED throughput, there were mixed results. Ola et al. found that, at their single academic institution, LOS and stress testing rates decreased, but diagnosis rate of AMI and an- giography utilization increased after the transition from cTnT to hs-cTnT (18).
Ford et al., at a community site, noted increased median ED LOS in the initial 6 months after hs-cTnT implementation, though overall admissions decreased and no significant difference in cardiology con- sultation trends were found (19).
Crowder et al., in a study from three tertiary care centers, noted a decreased mean ED LOS after hs-cTnT implementation, but no significant change in cardiology consultation or admission trends after implementation (20). There has also been a parallel line of investigation utilizing HEART score and other clinical risk scores. The results of these studies are also mixed, but have consistently shown a reduction in admission rates and the correlation with downstream cardiovascular events (21–24). However, the impact on ED throughput is less clear. The largest multicenter study showed no differ- ence in rates of early discharge or median length of stay in the ED (25). The study used a stepped-wise cluster randomization method, which involved randomization of different sites at various times due to differences in when centers transitioned over to utilizing the HEART score.
Our study highlighted the potential for synergistic ben- efit of combining high sensitivity troponin and a risk stratification pathway on ED throughput efficiency—specifically decreasing ED LOS, utilization of cardiology consultation services, admission rates, and boarding rates.
The results of this study (and pathway) are likely due to the step-wise approach guiding when to integrate clinical risk scores. The high-sensitivity assays not only carry di- agnostic utility, but strong prognostic performance, espe- cially undetectable values. Our protocol allows for rapid triage, quick assessment, and decreased cognitive load for a significant cohort of patients. Similarly, there is a limited role for clinical risk scores among those that would meet the 4th Universal Definition for acute myocardial in- farction (26). Reserving more nuanced assessment for a smaller, truly intermediate risk cohort allows for application of the risk scores to a cohort in whom it is likely to more meaningfully impact decision-making.
The time series assessment further supports the idea that decreasing cognitive load in a lower-risk cohort drives most of the benefit seen from the pathway. There was no difference noted in the times to consult or time to disposition in anyone for whom a consult was ordered. It is truly in guiding those low-risk patients to a rapid disposition without involving a consultant in whom a reduction in ED resources and expedited throughput was observed. This is an important component of the study that can be translated to other similar investigations. Pathways that are designed to quickly triage low-risk individuals with minimal involvement of additional resources are most likely to increase ED throughput.
This decrease in median length of stay in the ED, and decreased admission and consult rates aids in decompressing a very busy system. Utilizing absolute reduction and total number of chest pain patients seen in the ED over a year, the impact can be significant. The reduction in median length of stay translates to over 2319 h saved annually in the assessment of chest pain patients. The reduction in admission rates projects out to 240 admissions avoided annually, and projects to 353 fewer consults annually. This is a dramatic reduction in resource utilization for what is consistently one of the most common indications for ED evaluation.
These benefits were seen while maintaining a low MACE rate at 30 days after discharge.

Limitations

The current study has several limitations:
- First, as part of the study design we simultaneously implemented two interventions—the hs-cTnT assay as well as the Acute Chest Pain Optimal Care Pathway. Though the two are directly linked, as the Acute Chest Pain Optimal Care Pathway utilizes the hs-cTnT results as an element to aid in risk stratification, our data do not allow us to differentiate the independent effects of the two interventions on our outcomes.
- Second, during our chart review we noted poor consistency in HEART score documentation in provider notes; therefore, we were unable to verify how strictly the Acute Chest Pain Optimal Care Pathway was followed.
- Third, a degree of patient mixing between the pre- and postintervention groups occurred due to each ED evaluation of chest pain being treated as independent events.
- Lastly, the determination of cardiology consultation burden was based on the presence of the order being placed in the electronic medical record system by an ED provider. There may have been patients that were consulted, but the order was not placed in the medical chart and was therefore missed in our study. Although there was also a review of the chart for a consult note, it was not included in the time analysis due to unclear time of request.
- Lastly, our study only included patients that presented with “non-traumatic chest pain”as their chief complaint. There are numerous other presentations that may rep- resent ACS, such as “shortness of breath,”“dyspnea,”“epigastric pain,”and “syncope,”in which troponins were obtained; however, these patients were not included in our data. Only nontraumatic chest pain patients were included in our study, to extrapolate patients with the highest likeli- hood of MACE and to mirror other studies done on acute chest pain.

Conclusions

Our study adds to the growing body of evidence that the hs-cTnT assay, particularly when combined with utilization of risk stratification pathways, has potential to improve efficiency of chest pain evaluation in the ED by decreasing LOS as well as cardiology consults and ad- missions. This benefit is seen largely by leveraging the negative predictive value of hs-cTnT, allowing for early discharge of low-risk patients without involvement of consultants, and providing guidance for intermediate-risk patients.
Further studies are needed to evaluate whether other more accelerated pathways may be possible to further improve ED throughput.

NOTE: Figures and tables can be consulted in the original publication in the aforementioned journal.

References

 1.Hsia RY, Hale Z, Tabas JA. A national study of the prevalence of life-threatening diagnoses in patients with chest pain. JAMA Intern Med 2016;176:1029–32 .

2.Boie ET. Initial evaluation of chest pain. Emerg Med Clin North Am 2005;23:937–57 .

3.Chase M, Robey JL, Zogby KE, Sease KL, Shofer FS, Hollander JE. Prospective validation of the Thrombolysis in Myocardial Infarction Risk Score in the emergency department chest pain population. Ann Emerg Med 2006;48:252–9 .

4.Hollander JE. The continuing search to identify the very-low-risk chest pain patient. Acad Emerg Med 1999;6:979–81 .

5.Roberts RR, Zalenski RJ, Mensah EK, et al. Costs of an emergency department-based accelerated diagnostic protocol vs hospitalization in patients with chest pain: a randomized controlled trial. JAMA 1997;278:1670–6 .

6.Storrow AB, Gibler WB. Chest pain centers: diagnosis of acute coronary syndromes. Ann Emerg Med 2000;35:449–61 .

7.Cullen L, Greenslade JH, Than M, et al. The new Vancouver Chest Pain Rule using troponin as the only biomarker: an external valida-tion study. Am J Emerg Med 2014;32:129–34 .

8.Reichlin T, Schindler C, Drexler B, et al. One-hour rule-out and rule-in of acute myocardial infarction using high-sensitivity cardiac troponin T. Arch Intern Med 2012;172:1211–18 .

9.Gulati M, Levy PD, Mukherjee D, et al. 2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR guideline for the evaluation and diagnosis of chest pain: executive summary: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021;144:e368–454 .

10.Bernstein SL, Aronsky D, Duseja R, et al. The effect of emergency department crowding on clinically oriented outcomes. Acad Emerg Med 2009;16:1–10 .

11.Reichlin T, Hochholzer W, Bassetti S, et al. Early diagnosis of my-ocardial infarction with sensitive cardiac troponin assays. N Engl J Med 2009;361:858–67 .

12.Sandoval Y, Jaffe AS. Using high-sensitivity cardiac troponin t for acute cardiac care. Am J Med 2017;130:1358–65 12e1 .

13.Al-Saleh A, Alazzoni A, Al Shalash S, et al. Performance of the high-sensitivity troponin assay in diagnosing acute myocar-dial infarction: systematic review and meta-analysis. CMAJ Open 2014;2:E199–207 .

14.Hammerer-Lercher A, Ploner T, Neururer S, et al. High-sensitivity cardiac troponin T compared with standard troponin T testing on emergency department admission: how much does it add in every-day clinical practice? J Am Heart Assoc 2013;23 .

15.Balk EM, Ioannidis JP, Salem D, Chew PW, Lau J. Accuracy of biomarkers to diagnose acute cardiac ischemia in the emergency de-partment: a meta-analysis. Ann Emerg Med 2001;37:478–94 .

16.Apple FS, Ler R, Murakami MM. Determination of 19 car-diac troponin I and T assay 99th percentile values from a com-mon presumably healthy population. Clin Chem 2012;58:1574– 1581 .

17.Tomaszewski CA, Nestler D, Shah KH, Sudhir A, Brown MDAmer-ican College of Emergency Physicians Clinical Policies Subcom-mittee (Writing Committee) on Suspected Non-ST-Elevation Acute Coronary Syndromes. Clinical policy: critical issues in the eval-uation and management of emergency department patients with suspected non-ST-elevation acute coronary syndromes. Ann Emerg Med 2018;72:e65–106 .

18 Ola O, Akula A, De Michieli L, et al. Clinical impact of high-sen-sitivity cardiac troponin I implementation in the community. J Am Coll Cardiol 2021;77:3160–70

19. 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 .

20Crowder KR, Jones TD, Lang ES, et al. The impact of high-sensitiv-ity troponin implementation on hospital operations and patient out-comes in 3 tertiary care centers. Am J Emerg Med 2015;33:1790–4 .

21.Smulowitz PB, Dizitzer Y, Tadiri S, Thibodeau L, Jagminas L, No-vack V. Impact of implementation of the HEART pathway using an electronic clinical decision support tool in a community hospital setting. Am J Emerg Med 2018;36:408–13 .

22.Bylund WE, Cole PM, Lloyd ML, et al. Effect of implementation of HEART chest pain protocol on emergency department disposition, testing and cost. West J Emerg Med 2021;22:308–18

23.Backus BE, Six AJ, Kelder JC, et al. A prospective validation of the HEART score for chest pain patients at the emergency department. Int J Cardiol 2013;168:2153–8 .

24.Mahler SA, Riley RF, Hiestand BC, et al. The HEART Pathway randomized trial: identifying emergency department patients with acute chest pain for early discharge. Circ Cardiovasc Qual Outcomes 2015;8:195–203 . 25. Poldervaart JM, Reitsma JB, Backus BE, et al. Effect of using the HEART Score in patients with chest pain in the eergency depart-ment: a stepped-wedge, cluster randomized trial. Ann Intern Med 2017;166:689–97

25.Thygesen K, Alpert JS, Jaffe AS, et al. Fourth Universal Definition of Myocardial Infarction. 2018. Circulation 2018;138:e618–51 .

.