Machine learning compared with rule-in/rule-out algorithms and logistic regression to predict acute myocardial infarction based on troponin T concentrations.
Ontology highlight
ABSTRACT: Objective
Computerized decision-support tools may improve diagnosis of acute myocardial infarction (AMI) among patients presenting with chest pain at the emergency department (ED). The primary aim was to assess the predictive accuracy of machine learning algorithms based on paired high-sensitivity cardiac troponin T (hs-cTnT) concentrations with varying sampling times, age, and sex in order to rule in or out AMI.Methods
In this register-based, cross-sectional diagnostic study conducted retrospectively based on 5695 chest pain patients at 2 hospitals in Sweden 2013-2014 we used 5-fold cross-validation 200 times in order to compare the performance of an artificial neural network (ANN) with European guideline-recommended 0/1- and 0/3-hour algorithms for hs-cTnT and with logistic regression without interaction terms. Primary outcome was the size of the intermediate risk group where AMI could not be ruled in or out, while holding the sensitivity (rule-out) and specificity (rule-in) constant across models.Results
ANN and logistic regression had similar (95%) areas under the receiver operating characteristics curve. In patients (n = 4171) where the timing requirements (0/1 or 0/3 hour) for the sampling were met, using ANN led to a relative decrease of 9.2% (95% confidence interval 4.4% to 13.8%; from 24.5% to 22.2% of all tested patients) in the size of the intermediate group compared to the recommended algorithms. By contrast, using logistic regression did not substantially decrease the size of the intermediate group.Conclusion
Machine learning algorithms allow for flexibility in sampling and have the potential to improve risk assessment among chest pain patients at the ED.
SUBMITTER: Bjorkelund A
PROVIDER: S-EPMC7984484 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:AimsWhether a single cardiac troponin measurement can safely rule out myocardial infarction in patients presenting within a few hours of symptom onset is uncertain. The study aim was to assess the performance of troponin in early presenters.Methods and resultsIn patients with possible myocardial infarction, the diagnostic performance of a single measurement of high-sensitivity cardiac troponin I at presentation was evaluated and externally validated in those tested ≤3, 4-12, and >12 h from symptom onset. The limit-of-detection (2 ng/L), rule-out (5 ng/L), and sex-specific 99th centile (16 ng/L in women; 34 ng/L in men) thresholds were compared. In 41 103 consecutive patients [60 (17) years, 46% women], 12 595 (31%) presented within 3 h, and 3728 (9%) had myocardial infarction. In those presenting ≤3 h, a threshold of 2 ng/L had greater sensitivity and negative predictive value [99.4% (95% confidence interval 99.2%-99.5%) and 99.7% (99.6%-99.8%)] compared with 5 ng/L [96.5% (96.2%-96.8%) and 99.3% (99.1%-99.4%)]. In those presenting ≥3 h, the sensitivity and negative predictive value were similar for both thresholds. The sensitivity of the 99th centile was low in early and late presenters at 71.4% (70.6%-72.2%) and 92.5% (92.0%-93.0%), respectively. Findings were consistent in an external validation cohort of 7088 patients.ConclusionIn early presenters, a single measurement of high-sensitivity cardiac troponin I below the limit of detection may facilitate the safe rule out of myocardial infarction. The 99th centile should not be used to rule out myocardial infarction at presentation even in those presenting later following symptom onset.
| S-EPMC10406338 | biostudies-literature
Project description:AimsThe present acute myocardial infarction (AMI) rule-out strategies are challenged by the late temporal release of cardiac troponin. Copeptin is a non-specific biomarker of endogenous stress and rises early in AMI, covering the early period where troponin is still normal. An accelerated dual-marker rule-out strategy combining prehospital copeptin and in-hospital high-sensitivity troponin T could reduce length of hospital stay and thus the burden on the health care systems worldwide. The AROMI trial aimed to evaluate if the accelerated dual-marker rule-out strategy could safely reduce length of stay in patients discharged after early rule-out of AMI.Methods and resultsPatients with suspected AMI transported to hospital by ambulance were randomized 1:1 to either accelerated rule-out using copeptin measured in a prehospital blood sample and high-sensitivity troponin T measured at arrival to hospital or to standard rule-out using a 0 h/3 h rule-out strategy. The AROMI study included 4351 patients with suspected AMI. The accelerated dual-marker rule-out strategy reduced mean length of stay by 0.9 h (95% confidence interval 0.7-1.1 h) in patients discharged after rule-out of AMI and was non-inferior regarding 30-day major adverse cardiac events when compared to standard rule-out (absolute risk difference -0.4%, 95% confidence interval -2.5 to 1.7; P-value for non-inferiority = 0.013).ConclusionAccelerated dual marker rule-out of AMI, using a combination of prehospital copeptin and first in-hospital high-sensitivity troponin T, reduces length of hospital stay without increasing the rate of 30-day major adverse cardiac events as compared to using a 0 h/3 h rule-out strategy.
| S-EPMC10568000 | biostudies-literature
Project description:BackgroundThe objective of this study was to quantify the sensitivity of very low concentrations of high-sensitivity cardiac troponin T (hsTnT) at ED arrival for acute myocardial infarction (AMI) in a large cohort of chest pain patients evaluated in real-world clinical practice.MethodsThis retrospective study included consecutive ED patients with suspected cardiac chest pain evaluated in four urban EDs, excluding those with ST-elevation AMI, cardiac arrest or abnormal kidney function. The primary outcomes were AMI at 7, 30, and 90 days. Secondary outcomes included major adverse cardiac events (MACE; all-cause mortality, AMI, and revascularization) and the individual MACE components. Test characteristics were calculated for hsTnT values from 3 to 200 ng/L .ResultsA total of 7,130 patients met inclusion criteria. AMI incidences at 7, 30, and 90 days were 5.8, 6.0, and 6.2%. When the hsTnT assay was performed at ED arrival, the limit of blank of the assay (3 ng/L) ruled out 7-day AMI in 15.5% of patients with 100% sensitivity and negative predictive value (NPV). The limit of detection of the assay (5 ng/L) ruled out AMI in 33.6% of patients with 99.8% sensitivity and 99.95% NPV for 7-day AMI. The limit of quantification (the Food and Drug Administration [FDA]-approved cutoff for lower the reportable limit) of 6 ng/L ruled out AMI in 42.2% of patients with 99.8% sensitivity and 99.95% NPV. The sensitivities of the cutoffs of <3, <5, and <6 ng/L for 7-day MACE were 99.6, 97.4, and 96.6%, respectively. The NPVs of the cutoffs of <3, <5, and <6 ng/L for 7-day MACE were 99.8, 99.5, and 99.4%, respectively. A secondary analysis was performed in a subgroup of 3,549 higher-risk patients who underwent serial troponin testing. In this subgroup, a cutoff of 3 ng/L ruled out 7-day AMI in 9.6% of patients with 100% sensitivity and NPV, a cutoff of 5 ng/L ruled out 7-day AMI in 23.3% of patients with 99.7% sensitivity and 99.9% NPV, and a cutoff of 6 ng/L ruled out 7-day AMI in 29.8% of patients with 99.7 and 99.9% NPV. In the higher-risk subgroup, the sensitivities of cutoffs of <3, <5, and <6 ng/L for 7-day MACE were 99.8, 97.4, and 96.6%, respectively. In this higher-risk subgroup, the NPV of cutoffs of <3, <5, and <6 ng/L for 7-day MACE were 99.7, 98.5, and 98.4%, respectively.ConclusionsWhen used in real-world clinical practice conditions, hsTnT concentrations < 6 ng/L (below the lower reportable limit for an FDA-approved assay) at the time of ED arrival can rule out AMI with very high sensitivity and NPV. The sensitivity for MACE is unacceptably low, and thus a single-troponin rule-out strategy should only be used in the context of a structured risk evaluation.
| S-EPMC5656889 | biostudies-literature
Project description:BackgroundHigh-sensitivity cardiac troponin assays enable myocardial infarction to be ruled out earlier, but the safety and efficacy of this approach is uncertain. We investigated whether an early rule-out pathway is safe and effective for patients with suspected acute coronary syndrome.MethodsWe performed a stepped-wedge cluster randomized controlled trial in the emergency departments of 7 acute care hospitals in Scotland. Consecutive patients presenting with suspected acute coronary syndrome between December 2014 and December 2016 were included. Sites were randomized to implement an early rule-out pathway where myocardial infarction was excluded if high-sensitivity cardiac troponin I concentrations were <5 ng/L at presentation. During a previous validation phase, myocardial infarction was ruled out when troponin concentrations were <99th percentile at 6 to 12 hours after symptom onset. The coprimary outcome was length of stay (efficacy) and myocardial infarction or cardiac death after discharge at 30 days (safety). Patients were followed for 1 year to evaluate safety and other secondary outcomes.ResultsWe enrolled 31 492 patients (59±17 years of age [mean±SD]; 45% women) with troponin concentrations <99th percentile at presentation. Length of stay was reduced from 10.1±4.1 to 6.8±3.9 hours (adjusted geometric mean ratio, 0.78 [95% CI, 0.73-0.83]; P<0.001) after implementation and the proportion of patients discharged increased from 50% to 71% (adjusted odds ratio, 1.59 [95% CI, 1.45-1.75]). Noninferiority was not demonstrated for the 30-day safety outcome (upper limit of 1-sided 95% CI for adjusted risk difference, 0.70% [noninferiority margin 0.50%]; P=0.068), but the observed differences favored the early rule-out pathway (0.4% [57/14 700] versus 0.3% [56/16 792]). At 1 year, the safety outcome occurred in 2.7% (396/14 700) and 1.8% (307/16 792) of patients before and after implementation (adjusted odds ratio, 1.02 [95% CI, 0.74-1.40]; P=0.894), and there were no differences in hospital reattendance or all-cause mortality.ConclusionsImplementation of an early rule-out pathway for myocardial infarction reduced length of stay and hospital admission. Although noninferiority for the safety outcome was not demonstrated at 30 days, there was no increase in cardiac events at 1 year. Adoption of this pathway would have major benefits for patients and health care providers. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03005158.
| S-EPMC8177493 | biostudies-literature
Project description:ImportanceEmergency department (ED) investigations of patients with suspected acute myocardial infarction (AMI) are time consuming, partly because of the turnaround time of laboratory tests. Current point-of-care troponin assays shorten test turnaround times but lack precision at lower concentrations. Development of point-of-care troponin assays with greater analytical precision could reduce the decision-making time in EDs for ruling out AMI.ObjectiveTo determine the clinical accuracy for AMI of a single troponin concentration measured on arrival to ED with a new-generation, higher precision point-of-care assay with a 15-minute turnaround time.Design, setting, and participantsThis observational study occurred at a single urban regional ED. Adults presenting acutely from the community to the ED with symptoms suggestive of AMI were included. Troponin concentrations were measured on ED arrival with both a novel point-of-care assay (i-STAT TnI-Nx; Abbott Point of Care) and a high-sensitivity troponin I assay (Architect hs-cTnI; Abbott Diagnostics).Main outcomes and measuresThe primary outcome was type 1 AMI during index presentation. We compared the discrimination ability of the TnI-Nx assay with the hs-cTnI assay using the area under receiver operator characteristic curve (AUC) and sensitivity, negative predictive value, and the proportion of negative test results at thresholds with 100% sensitivity.ResultsOf 354 patients (255 [72.0%] men; mean [SD] age, 62 [12] years), 57 (16.1%) experienced an AMI. Eighty-five patients (24.0%) presented to the ED less than 3 hours after symptom onset. No difference was found between the AUC of the TnI-Nx assay (0.975 [95% CI, 0.958-0.993]) and the hs-cTnI assay (0.970 [95% CI, 0.949 to 0.990]; P = .46). A TnI-Nx assay result of less than 11 ng/L identified 201 patients (56.7%) as low risk, with a sensitivity of 100% (95% CI, 93.7%-100%) and a negative predictive value of 100% (95% CI, 98.2%-100%). In comparison, an hs-cTnI assay result of less than 3 ng/L identified 154 patients (43.5%) as low risk, with a sensitivity of 100% (95% CI, 93.7%-100%) and a negative predictive value of 100% (95% CI, 97.6%-100%).Conclusions and relevanceA novel point-of-care troponin assay that can produce a result 15 minutes after blood sampling had comparable discrimination ability to an hs-cTnI assay for ruling out AMI after a single blood test. Use in the ED may facilitate earlier decision making and could expedite the safe discharge of a large proportion of low-risk patients.
| S-EPMC6583693 | biostudies-literature
Project description:AimsThis study tested the hypothesis that combining stress-induced biomarkers (copeptin or glucose) with high-sensitivity cardiac troponin (hs-cTn) increases diagnostic accuracy for non-ST-elevation myocardial infarction (NSTEMI) in patients presenting to the emergency department.Methods and resultsThe ability to rule-out NSTEMI for combinations of baseline hs-cTnT or hs-cTnI with copeptin or glucose was compared with the European Society of Cardiology (ESC) hs-cTnT/I-only rule-out algorithms in two independent (one Norwegian and one international multicentre) diagnostic studies. Among 959 patients (median age 64 years, 60.5% male) with suspected NSTEMI in the Norwegian cohort, 13% had NSTEMI. Adding copeptin or glucose to hs-cTnT/I as a continuous variable did not improve discrimination as quantified by the area under the curve {e.g. hs-cTnT/copeptin 0.91 [95% confidence interval (CI) 0.89-0.93] vs. hs-cTnT alone 0.91 (95% CI 0.89-0.93); hs-cTnI/copeptin 0.85 (95% CI 0.82-0.87) vs. hs-cTnI alone 0.93 (95% CI 0.91-0.95)}, nor did adding copeptin <9 mmol/L or glucose <5.6 mmol/L increase the sensitivity of the rule-out provided by hs-cTnT <5 ng/L or hs-cTnI <4 ng/L in patients presenting more than 3 h after chest pain onset (target population in the ESC-0 h-algorithm). The combination decreased rule-out efficacy significantly (both P < 0.01). These findings were confirmed among 1272 patients (median age 62 years, 69.3% male) with suspected NSTEMI in the international validation cohort, of which 20.7% had NSTEMI. A trend towards increased sensitivity for the hs-cTnT/I/copeptin combinations (97-100% vs. 91-97% for the ESC-0 h-rule-out cut-offs) was observed in the Norwegian cohort.ConclusionAdding copeptin or glucose to hs-cTnT/I did not increase diagnostic performance when compared with current ESC guideline hs-cTnT/I-only 0 h-algorithms.
| S-EPMC8929978 | biostudies-literature
Project description:BackgroundThe instant, single-sampling rule-out of acute myocardial infarction (AMI) is still an unmet clinical need. We aimed at testing and comparing diagnostic performance and prognostic value of two different single-sampling biomarker strategies for the instant rule-out of AMI.MethodsFrom the Biomarkers in Acute Cardiac Care (BACC) cohort, we recruited consecutive patients with acute chest pain and suspected AMI presenting to the Emergency Department of the University Medical Center Hamburg-Eppendorf, Hamburg, Germany. We compared safety, effectiveness and 12-month incidence of the composite endpoint of all-cause death and myocardial infarction between (i) a single-sampling, dual-marker pathway combining high-sensitivity cardiac troponin I (hs-cTnI) and ultra-sensitive copeptin (us-Cop) at presentation (hs-cTnI ≤ 27 ng/L, us-Cop < 10 pmol/L and low-risk ECG) and (ii) a single-sampling pathway based on one-off hs-cTnI determination at presentation (hs-cTnI < 5 ng/L and low-risk ECG). As a comparator, we used the European Society of Cardiology (ESC) 0/1-h dual-sampling algorithm.ResultsWe enrolled 1,136 patients (male gender 65%) with median age of 64 years (interquartile range, 51-75). Overall, 228 (20%) patients received a final diagnosis of AMI. The two single-sampling instant rule-out pathways yielded similar negative predictive value (NPV): 97.4% (95%CI: 95.4-98.7) and 98.7% (95%CI: 96.9-99.6) for dual-marker and single hs-cTnI algorithms, respectively (P = 0.11). Both strategies were comparably safe as the ESC 0/1-h dual-sampling algorithm and this was consistent across subgroups of early-comers, low-intermediate risk (GRACE-score < 140) and renal dysfunction. Despite a numerically higher rate of false-negative results, the dual-marker strategy ruled-out a slightly but significantly higher percentage of patients compared with single hs-cTnI determination (37.4% versus 32.9%; P < 0.001). There were no significant between-group differences in 12-month composite outcome.ConclusionsInstant rule-out pathways based on one-off determination of hs-cTnI alone or in combination with us-Cop are comparably safe as the ESC 0/1 h algorithm for the instant rule-out of AMI, yielding similar prognostic information. Instant rule-out strategies are safe alternatives to the ESC 0/1 h algorithm and allow the rapid and effective triage of suspected AMI in patients with low-risk ECG. However, adding copeptin to hs-cTn does not improve the safety of instant rule-out compared with the single rule-out hs-cTn at very low cut-off concentrations.
| S-EPMC9395923 | biostudies-literature
Project description:BackgroundWe aimed to prospectively validate a novel 1-hour algorithm using high-sensitivity cardiac troponin T measurement for early rule-out and rule-in of acute myocardial infarction (MI).MethodsIn a multicentre study, we enrolled 1320 patients presenting to the emergency department with suspected acute MI. The high-sensitivity cardiac troponin T 1-hour algorithm, incorporating baseline values as well as absolute changes within the first hour, was validated against the final diagnosis. The final diagnosis was then adjudicated by 2 independent cardiologists using all available information, including coronary angiography, echocardiography, follow-up data and serial measurements of high-sensitivity cardiac troponin T levels.ResultsAcute MI was the final diagnosis in 17.3% of patients. With application of the high-sensitivity cardiac troponin T 1-hour algorithm, 786 (59.5%) patients were classified as "rule-out," 216 (16.4%) were classified as "rule-in" and 318 (24.1%) were classified to the "observational zone." The sensitivity and the negative predictive value for acute MI in the rule-out zone were 99.6% (95% confidence interval [CI] 97.6%-99.9%) and 99.9% (95% CI 99.3%-100%), respectively. The specificity and the positive predictive value for acute MI in the rule-in zone were 95.7% (95% CI 94.3%-96.8%) and 78.2% (95% CI 72.1%-83.6%), respectively. The 1-hour algorithm provided higher negative and positive predictive values than the standard interpretation of highsensitivity cardiac troponin T using a single cut-off level (both p < 0.05). Cumulative 30-day mortality was 0.0%, 1.6% and 1.9% in patients classified in the rule-out, observational and rule-in groups, respectively (p = 0.001).InterpretationThis rapid strategy incorporating high-sensitivity cardiac troponin T baseline values and absolute changes within the first hour substantially accelerated the management of suspected acute MI by allowing safe rule-out as well as accurate rule-in of acute MI in 3 out of 4 patients.Trial registrationClinicalTrials.gov, NCT00470587.
| S-EPMC4435896 | biostudies-literature
Project description:Although guidelines recommend fixed cardiac troponin thresholds for the diagnosis of myocardial infarction, troponin concentrations are influenced by age, sex, comorbidities and time from symptom onset. To improve diagnosis, we developed machine learning models that integrate cardiac troponin concentrations at presentation or on serial testing with clinical features and compute the Collaboration for the Diagnosis and Evaluation of Acute Coronary Syndrome (CoDE-ACS) score (0-100) that corresponds to an individual's probability of myocardial infarction. The models were trained on data from 10,038 patients (48% women), and their performance was externally validated using data from 10,286 patients (35% women) from seven cohorts. CoDE-ACS had excellent discrimination for myocardial infarction (area under curve, 0.953; 95% confidence interval, 0.947-0.958), performed well across subgroups and identified more patients at presentation as low probability of having myocardial infarction than fixed cardiac troponin thresholds (61 versus 27%) with a similar negative predictive value and fewer as high probability of having myocardial infarction (10 versus 16%) with a greater positive predictive value. Patients identified as having a low probability of myocardial infarction had a lower rate of cardiac death than those with intermediate or high probability 30 days (0.1 versus 0.5 and 1.8%) and 1 year (0.3 versus 2.8 and 4.2%; P < 0.001 for both) from patient presentation. CoDE-ACS used as a clinical decision support system has the potential to reduce hospital admissions and have major benefits for patients and health care providers.
| S-EPMC10202804 | biostudies-literature
Project description:BackgroundThe 1-hour (h) algorithm triages patients presenting with suspected acute myocardial infarction (AMI) to the emergency department (ED) towards "rule-out," "rule-in," or "observation," depending on baseline and 1-h levels of high-sensitivity cardiac troponin (hs-cTn). The economic consequences of applying the accelerated 1-h algorithm are unknown.Methods and findingsWe performed a post-hoc economic analysis in a large, diagnostic, multicenter study of hs-cTnT using central adjudication of the final diagnosis by two independent cardiologists. Length of stay (LoS), resource utilization (RU), and predicted diagnostic accuracy of the 1-h algorithm compared to standard of care (SoC) in the ED were estimated. The ED LoS, RU, and accuracy of the 1-h algorithm was compared to that achieved by the SoC at ED discharge. Expert opinion was sought to characterize clinical implementation of the 1-h algorithm, which required blood draws at ED presentation and 1h, after which "rule-in" patients were transferred for coronary angiography, "rule-out" patients underwent outpatient stress testing, and "observation" patients received SoC. Unit costs were for the United Kingdom, Switzerland, and Germany. The sensitivity and specificity for the 1-h algorithm were 87% and 96%, respectively, compared to 69% and 98% for SoC. The mean ED LoS for the 1-h algorithm was 4.3h-it was 6.5h for SoC, which is a reduction of 33%. The 1-h algorithm was associated with reductions in RU, driven largely by the shorter LoS in the ED for patients with a diagnosis other than AMI. The estimated total costs per patient were £2,480 for the 1-h algorithm compared to £4,561 for SoC, a reduction of up to 46%.ConclusionsThe analysis shows that the use of 1-h algorithm is associated with reduction in overall AMI diagnostic costs, provided it is carefully implemented in clinical practice. These results need to be prospectively validated in the future.
| S-EPMC5679593 | biostudies-literature