Reference range values of troponin measured by sensitive assays in elderly patients without any cardiac signs/symptoms

A Comparison of Five Algorithmic Methods and Machine Learning Pattern Recognition for Artifact Detection in Electronic Records of Five Different Vital Signs: A Retrospective Analysis

BACKGROUND

Research on electronic health record physiologic data is common, invariably including artifacts. Traditionally, these artifacts have been handled using simple filter techniques. The authors hypothesized that different artifact detection algorithms, including machine learning, may be necessary to provide optimal performance for various vital signs and clinical contexts.

METHODS

In a retrospective single-center study, intraoperative operating room and intensive care unit (ICU) electronic health record datasets including heart rate, oxygen saturation, blood pressure, temperature, and capnometry were included. All records were screened for artifacts by at least two human experts. Classical artifact detection methods (cutoff, multiples of SD [z-value], interquartile range, and local outlier factor) and a supervised learning model implementing long short-term memory neural networks were tested for each vital sign against the human expert reference dataset. For each artifact detection algorithm, sensitivity and specificity were calculated.

RESULTS

A total of 106 (53 operating room and 53 ICU) patients were randomly selected, resulting in 392,808 data points. Human experts annotated 5,167 (1.3%) data points as artifacts. The artifact detection algorithms demonstrated large variations in performance. The specificity was above 90% for all detection methods and all vital signs. The neural network showed significantly higher sensitivities than the classic methods for heart rate (ICU, 33.6%; 95% CI, 33.1 to 44.6), systolic invasive blood pressure (in both the operating room [62.2%; 95% CI, 57.5 to 71.9] and the ICU [60.7%; 95% CI, 57.3 to 71.8]), and temperature in the operating room (76.1%; 95% CI, 63.6 to 89.7). The CI for specificity overlapped for all methods. Generally, sensitivity was low, with only the z-value for oxygen saturation in the operating room reaching 88.9%. All other sensitivities were less than 80%.

CONCLUSIONS

No single artifact detection method consistently performed well across different vital signs and clinical settings. Neural networks may be a promising artifact detection method for specific vital signs.

EDITOR’S PERSPECTIVE

New reference limits for cardiac troponin T and N-terminal b-type natriuretic propeptide in elders

BACKGROUND AND AIMS

Our aim was to define reference limits for cardiac troponin T (cTnT) and N-terminal pro B-type natriuretic peptide (proBNP) that would better reflect their concentrations in older people. In addition, the incidence of acute myocardial infarctions (AMIs) was studied using these reference limits in an older population with and without previous heart diseases.

MATERIALS AND METHODS

A population-based study with a ten-year follow-up. The reference population was formed by 763 individuals aged over 64 years, with no diagnoses of heart or kidney diseases.

RESULTS

There was a significant increase in cTnT and proBNP concentrations with age. The 99 % reference limits for cTnT were 25 ng/L, 28 ng/l, 38 ng/l, and 71 ng/l for men in five-year-intervals starting from 64 to 69 years to 80 years and older, and 18 ng/L, 22 ng/l, 26 ng/l, and 52 ng/L for women, respectively. The 97.5 % reference limits for proBNP were 272 ng/L, 287 ng/l, 373 ng/l and 686 ng/L for men, and 341 ng/L, 377 ng/l, 471 ng/l, and 794 ng/L for women, respectively. Elevated proBNP was statistically significantly associated with future AMIs in subjects with and without a previous heart disease.

CONCLUSIONS

Age-specific reference limits for cTnT and proBNP are needed to better evaluate cardiac symptoms.

Determination of age- and sex-specific 99th percentiles for high-sensitive troponin T from patients: an analytical imprecision- and partitioning-based approach

BACKGROUND

Detection of acute myocardial infarction (AMI) is mainly based on a rise of cardiac troponin with at least one value above the 99th percentile upper reference limit (99th URL). However, circulating high-sensitive cardiac troponin T (hs-cTnT) concentrations depend on age, sex and renal function. Using an analytical imprecision-based approach, we aimed to determine age- and sex-specific hs-cTnT 99th URLs for patients without chronic kidney disease (CKD).

METHODS

A 3.8-year retrospective analysis of a hospital laboratory database allowed the selection of adult patients with concomitant plasma hs-cTnT (<300 ng/L) and creatinine concentrations, both assayed twice within 72 h with at least 3 h between measurements. Absence of AMI was assumed when the variation between serial hs-cTnT values was below the adjusted-analytical change limit calculated according to the inverse polynomial regression of analytical imprecision. Specific URLs were determined using Clinical and Laboratory Standards Institute (CLSI) methods, and partitioning was tested using the proportion method, after adjustment for unequal prevalences.

RESULTS

After outlier removal (men: 8.7%; women: 6.6%), 1414 men and 1082 women with estimated glomerular filtration rate (eGFR) ≥60 mL/min/1.73 m2 were assumed as non-AMI. Partitioning into age groups of 18-50, 51-70 and 71-98 years, the hs-cTnT 99th URLs adjusted on French prevalence were 18, 33, 66 and 16, 30, 84 ng/L for men and women, respectively. Age-partitioning was clearly required. However, sex-partitioning was not justified for subjects aged 18-50 and 51-70 years for whom a common hs-cTnT 99th URLs of about 17 and 31 ng/L could be used.

CONCLUSIONS

Based on a laboratory approach, this study supports the need for age-specific hs-cTnT 99th URLs.

High-sensitivity cardiac troponin T for diagnosis of NSTEMI in the elderly emergency department patient: a clinical cohort study

PURPOSE

The aim of this study is to evaluate the impact of age on the diagnostic performance of high-sensitivity troponin T (hsTnT) under routine conditions.

MATERIALS AND METHODS

Data of 4118 consecutive emergency department (ED) patients who underwent a routine TnT measurement between 11 October 2012 and 30 November 2013 were analysed. Diagnostic accuracy of hsTnT was compared in four age categories (<50, 50-64, 65-74, ≥75 years of age) for different cut-off values. Primary endpoint was a main hospital diagnosis of NSTEMI.

RESULTS

The median age of the study population (n = 4118) was 61 years (IQR: 45-75 years). NSTEMI was diagnosed in 3.3% (n = 136) of all patients. There were significant differences in hsTnT concentrations between age-groups (p < 0.001) in all patients, but not in NSTEMI patients (p = 0.297). 72.2% of all patients ≥75 years of age (583/808) without NSTEMI had hsTnT concentrations above the 99th percentile of a healthy reference population. Specificity at 14 ng/L was 93.6% (95% CI: 92.12-94.87) in patients below 50 years of age and 27.9% (95% CI: 24.78-31.08) in patients 75 years of age and older.

CONCLUSIONS

Patients' age needs to be considered at least one influencing factor on hsTnT concentrations at admission and should be included in the clinical interpretation of hsTnT concentrations for further clinical workup beneath other influencing factors like comorbidities and symptom onset time. The implementation of age-specific cut-off values could be considered for single troponin testing at admission but is associated with an increased risk of underdiagnosis of NSTEMI.

Clinical Interpretation of Elevated Concentrations of Cardiac Troponin T, but Not Troponin I, in Nursing Home Residents

OBJECTIVE

Cardiac troponins T (cTnT) and I (cTnI) are the preferred biomarkers to detect myocardial damage. The present study explores the value of measuring cardiac troponins (cTn) in nursing home residents, by investigating its relation to heart failure and 1-year mortality using 1 cTnT and 2 cTnI assays that are widely used in clinical practice.

DESIGN

All participants underwent extensive clinical examinations and echocardiographic assessment for the diagnosis of heart failure. cTn was measured using high-sensitive (hs)- cTnT (Roche), hs-cTnI (Abbott), and sensitive cTnI (Beckman) assays. The glomerular filtration rate was estimated (eGFR) using serum creatinine and cystatin C concentrations. Data on all-cause mortality were collected at 1-year follow-up.

PARTICIPANTS AND SETTING

Participants were 495 long-term nursing home residents, older than 65 years, of 5 Dutch nursing home organizations.

RESULTS

Median (IQR) concentrations were 20.6 (17.8-30.6), 6.8 (4.1-12.5), and 4.0 (2.0-8.0) ng/L for hs-cTnT, hs-cTnI, and cTnI, respectively. In total, 79% had elevated hs-cTnT concentrations, whereas only 9% and 5% of hs-cTnI and cTnI concentrations were elevated. Most important and independent determinants for higher hs-cTnT and hs-cTnI concentrations were heart failure and renal dysfunction. Whereas both heart failure (odds ratio [OR] 3.4) and eGFR lower than 60 mL/min/1.73 m(2) (OR 3.6) were equal contributors to higher hs-cTnT concentrations (all P < .001), hs-cTnI and cTnI were less associated with renal dysfunction (OR of, respectively, 1.9 and 2.1; P < .01) in comparison with heart failure (OR 4.3 and 4.7, respectively, P < .001). Furthermore, residents with higher hs-cTnT or hs-cTnI concentrations (fourth quartile) had respectively 4 versus 2 times more risk of 1-year mortality compared with lower concentrations.

CONCLUSION

Regardless of their cardiac health, hs-cTnT but not hs-cTnI concentrations were elevated in almost all aged nursing home residents, questioning the use of the current diagnostic cutoff in elderly with high comorbidity. Nonetheless, measuring cardiac troponins, especially hs-cTnT, had a promising role in assessing future risk of mortality.

Consequences of high-sensitivity troponin T testing applied in a primary care population with chest pain compared with a commercially available point-of-care troponin T analysis: an observational prospective study

Background

There is a demand for a highly sensitive and specific point-of care test to detect acute myocardial infarction (AMI). It is unclear if a high-sensitivity troponin assay will have enough discriminative power to become a decision support in primary care. The aim of this study was to evaluate a high-sensitivity troponin T assay performed in three primary health care centres in southeast Sweden and to compare the outcome with a point-of-care troponin T test.

Methods

This study included 115 patients who consulted their general practitioner for chest pain, dyspnoea on exertion, unexplained weakness and/or fatigue in the last 7 days. Troponin T was analysed by a point-of-care test and a high-sensitivity method together with N-terminal pro-B-type natriuretic peptide (NT-proBNP) and creatinine. All patients were checked for AMI or unstable angina (UA) within 30 days of study enrolment. Univariate and multivariate logistic regression was carried out to examine possible connections between troponin T ≥ 15 ng/L, clinical variables and laboratory findings at baseline. In addition, 21 patients with troponin T ≥ 15 ng/L and no signs of AMI or UA were followed up for 2–3 years.

Results

Three patients were diagnosed with AMI and three with UA. At the ≥ 15 ng/L cut-off, the troponin T method had 100% sensitivity, 75% specificity for AMI and a positive predictive value of 10%. The troponin T point-of-care test missed one case of AMI and the detection limit was 50 ng/L. Troponin T ≥ 15 ng/L was correlated to age ≥65 years (odds ratio (OR), 10.9 95% CI 2.28–51.8) and NT-proBNP in accordance with heart failure (OR 8.62 95% CI 1.61–46.1). Fourteen of the 21 patients, without signs of AMI or UA at baseline, still had increased troponin T at follow-up after 2–3 years.

Conclusions

A high-sensitivity troponin T assay could become useful in primary care as a point-of-care test for patients <65 years. For patients older than 65–70 years, a higher decision limit than ≥15 ng/L should be considered and used in conjunction with clinical parameters and possibly with NT-proBNP.

High-sensitivity cardiac troponin assays: answers to frequently asked questions

Cardiac troponin (cTn) assays have quickly gained in analytical sensitivity to become what are termed 'high-sensitivity cardiac troponin' (hs-cTn) assays, bringing a flurry of dense yet incomplete literature data. The net result is that cTn assays are not yet standardized and there are still no consensus-built data on how to use and interpret cTn assay results. To address these issues, the authors take cues and clues from multiple disciplines to bring responses to frequently asked questions. In brief, the effective use of hs-cTn hinges on knowing: specific assay characteristics, particularly precision at the 99th percentile of a reference population; factors of variation at the 99th percentile value; and the high-individuality of hs-cTn assays, for which the notion of individual kinetics is more informative than straight reference to 'normal' values. The significance of patterns of change between two assay measurements has not yet been documented for every hs-cTn assay. Clinicians need to work hand-in-hand with medical biologists to better understand how to use hs-cTn assays in routine practice.

The calculation of the cardiac troponin T 99th percentile of the reference population is affected by age, gender, and population selection: a multicenter study in Italy

BACKGROUND

The aim of this study is to determine the 99th upper-reference limit (URL) for cardiac troponin T (cTnT) in Italian apparently healthy subjects.

METHODS

The reference population was selected from 5 cities: Bolzano (n=290), Milano (CAMELIA-Study, n=287), Montignoso (MEHLP-Study, n=306), Pisa (n=182), and Reggio Calabria (MAREA-Study, n=535). Subjects having cardiac/systemic acute/chronic diseases were excluded. Participants to MEHLP project underwent cardiac imaging investigation. High-sensitive cTnT was measured with Cobas-e411 (Roche Diagnostics).

RESULTS

We enrolled 1600 healthy subjects [54.6% males; age range 10-90years; mean (SD): 36.4 (21.2) years], including 34.6% aged <20years, 54.5% between 20 and 64years, and 10.9% over 65years. In the youngest the 99th URL was 10.9ng/L in males and 6.8ng/L in females; in adults 23.2ng/L and 10.2ng/L; and in elderly 36.8ng/L and 28.6ng/L. After the exclusion of outliers the 99th URL values were significantly decreased (P<0.05) in particular those of the oldest (13.8ng/L and 14ng/L). MEHLP participants were divided in healthy and asymptomatic, according to known cardiovascular risk factors (HDL, LDL, glucose, C-reactive protein): the 99th URL of cTnT values of these subgroups was significantly different (19.5 vs. 22.7, P<0.05).

CONCLUSIONS

99th URL of cTnT values was strongly affected by age, gender, selection of subjects and the statistical evaluation of outliers.