Printer Friendly

Diagnostic Performance of High Sensitivity Compared with Contemporary Cardiac Troponin I for the Diagnosis of Acute Myocardial Infarction.

Cardiac troponin (cTn) [8] testing is central for ruling in and ruling out acute myocardial infarction (MI) (1). Serial cTn measurements are routinely obtained to evaluate for the presence or absence of myocardial injury as evidenced by concentrations >99th percentile (1-8). In the absence of myocardial injury and clinical features indicative of myocardial ischemia, patients are deemed ruled out for MI and considered for hospital discharge and/or further risk stratification. Conversely, if at least 1 cTn on serial measurements is >99th percentile, further evaluation and therapies are tailored according to whether the patient has MI or myocardial injury (non-MI cTn increase) due to another cardiac or noncardiac etiology.

Although numerous European and Asia-Pacific studies have demonstrated that ruling in and ruling out MI can be expedited using hs-cTn assays (7, 9), no large US study has compared the diagnostic performance of hs-cTnI vs contemporary cTnI assays following current guidelines (1) with the use of the 99th percentile to support the diagnosis of acute MI. It is important to understand the performance of contemporary and high-sensitivity assays as currently used globally that are based on (a) the use of the 99th percentile to dichotomize patients as having a negative ([less than or equal to] 99th percentile) or positive (>99th percentile) result and (b) the use of serial testing over time to determine if patients rule out when all results are negative or rule in when there is a rising and/or falling cTn pattern with at least 1 concentration >99th percentile.

Our goals were (a) to compare the diagnostic performance of an investigational hs-cTnI assay used in clinical practice globally outside the US to a contemporary cTnI assay used in US clinical practice for the diagnosis of type 1 MI (T1MI) and type 2 MI (T2MI) by use of cTn measurements alone and in combination with a normal 12-lead electrocardiogram (ECG), and (b) to determine the optimal deltas (serial cTnI concentration changes) for each assay.

Methods

STUDY DESIGN AND POPULATION

Following institutional review board approval, we prospectively included consecutive, unselected patients presenting from February 4, 2014, through May 9, 2014, through the emergency department (ED), in whom initial preset serial cTnI measurements (0, 3, 6, and 9 h) were ordered on clinical indication at Hennepin County Medical Center (Minneapolis, MN) to rule in or rule out acute MI (use of Abbott high-sensitivity troponin I assay in acute coronary syndromes [UTROPIA]; NCT02060760). For inclusion, patients needed a baseline cTnI measurement at presentation and at least 1 additional cTnI within 24 h of presentation before discharge, and at least one 12-lead ECG performed. All participants in our study had at least 2 cTn results on both assays, an additional inclusion criterion for the study. Most individuals had 3 or 4 measurements on both assays within the first 24 h. Exclusion criteria were: < 18 years old, ST-segment elevation MI, pregnancy, trauma, declined to participate, did not present through the ED, or were transferred from an outside hospital. For patients with more than 1 presentation during the study period, we included only the first.

CARDIAC TROPONIN I ASSAYS

Fresh EDTA plasma samples were simultaneously measured with both contemporary cTnI (clinically used) and hs-cTnI (investigational) assays (Abbott ARCHITECT i1000SR or i2000SR analyzers). The contemporary cTnI assay used by providers for clinical decision-making had the following assay characteristics: limit of detection (LoD) of 0.010 [micro]g/L and 99th percentile value of 0.030 [micro]g/L. The CV was 18.5% at 0.028 [micro]g/L. For hs-cTnI, sex-specific 99th percentiles were 16 ng/L for females and 34 ng/L for males. The CV was <20% at 1.9 ng/L LoD (10, 11).

EVENT ADJUDICATION

All cases with at least 1 cTnI >99th percentile were adjudicated according to the Third Universal Definition of MI consensus recommendations, including classification into MI subtypes, by 2 clinicians following review of all available medical records including 12-lead ECG, echocardiography, angiography, cTnI results, and clinical presentation (1). The adjudications for MI were based on all sample cTn measurements (2-plus) regardless of time post baseline over a 24-h period. Cases with an adjudication discrepancy were reviewed and adjudicated by a third senior clinician. All cases were adjudicated twice, using cTnI and hs-cTnI results separately, and blinded to the other adjudication (see the Methods in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/ content/vol63/issue10). In the adjudication of cases for each assay, different adjudicators were used who had no knowledge of the results of the other assay.

Cases having at least 1 cTnI >99th percentile were adjudicated on the basis of maximum cTnI concentrations from all patient samples measured following admission through 24 h (1). For other time frames (baseline; 0 and 3 h; 0, 3, and 6 h; 0, 3, 6, and 9 h), adjudication was based upon only the samples completed within the given timeframe or pattern (i.e., the test setup was not the same as the reference standard). For this reason, the sample size changed for each successive set of cutoffs. The reference standard never changed.

OUTCOMES

The diagnostic outcomes examined were (a) overall MI, (b) T1MI, and (c) T2MI during the index hospitalization. The safety outcome was a composite of MI or cardiac death at 30 days, including events occurring during the index hospitalization.

STATISTICAL ANALYSES

Statistics for both cTnI and hs-cTnI assays were based on the available preset serial measurements. The initial cTn result for all participants was set as the baseline concentration at presentation. All other test results were distributed and based on the time from the baseline sample as follows: 3-h sample (range, 0 to [less than or equal to] 4 h), 6-h sample (>4 to [less than or equal to] 7 h), and 9-h sample (>7 to [less than or equal to] 10 h). A positive test was defined as any result >99th percentile (single 99th percentile for contemporary cTnI and sex-specific 99th percentiles for hs-cTnI). Sensitivity, specificity, positive predictive value (PPV), and NPV were determined with 95% CI ascertained using binomial proportions. Analyses were examined for all MIs and T1MI and T2MI alone.

First, sensitivities, specificities, PPV, and NPV were run for each sampling time individually for both the contemporary (0/3/6/9h) and high-sensitivity assays (0/3/6h).

Second, composite sensitivities, specificities, PPV, and NPV were examined over time by use of serial testing. For the composite analyses, tests were considered positive if any single qualifying assay concentration was >99th percentile and considered negative only when all assay concentrations were negative ([less than or equal to] 99th percentile). For the composite analyses, the sample size became smaller over time compared with the individual draws over time post baseline, as we required that all cTn measurements be performed for both assays, within the order set timeframe, without missing a timeframe, to allow an appropriate comparison between assays. Composite analyses were performed at 0 and 3 h; 0, 3, and 6 h; and 0, 3, 6, and 9 h for both contemporary cTnI and hs-cTnI, along with 30-day risk stratification for MI and cardiac death. Baseline and composite times (0 and 3 h; 0, 3, and 6 h; and 0, 3, 6, 9 h) were also examined in combination with a normal ECG for both the diagnostic and safety outcomes. ECGs were categorized as normal following previously described criteria (12) (see Methods in the online Data Supplement). For early presenters, defined as individuals who had their first cTnI sample obtained [less than or equal to] 2 h after symptom onset, subgroup analyses were performed.

Analyses of serial changes in cTnI (deltas) were performed to maximize specificity and PPV for ruling in MI. These were subsequently graphed as ROC curves and used for area under the curve analysis. Both relative and absolute deltas were determined for both assays, by use of 0-h and 3-h samples for hs-cTnI and 0-h and 6-h samples for contemporary cTnI calculations. All analyses were performed by using SAS version 9.4.

Results

In total, 1631 patients presenting to the ED met the inclusion criteria. Using the contemporary cTnI assay, acute MI occurred in 210 patients (12.9%), including 71 T1MIs (4.4%) and 139 T2MIs (8.5%), with 279 patients (17.1%) classified as having myocardial injury. Using the hs-cTnI assay, acute MI occurred in 170 patients (10.4%), including 68 T1MIs (4.2%) and 102 (6.3%) T2MIs, with 245 patients (15.0%) having myocardial injury.

RULING OUT MI

Serial measurements. Diagnostic performance of contemporary cTnI and hs-cTnI assays for diagnosis of MI, including MI subtypes, by use of the composite of serial measurements at 0/3h; 0/3/6h; and 0/3/6/9h, is shown in Table 1.

Serial contemporary cTnI at 0/3h had a sensitivity of 93.5% (95% CI, 89.6-97.4) and a NPV of 98.7% (95% CI, 98.0-99.5) for MI, whereas 0/3/6h cTnI measurements had a sensitivity of 98.3% (95% CI, 95.8-100) and an NPV of 99.6% (95% CI, 99.0-100) (see Fig. 1A in the online Data Supplement). The addition of a normal ECG to 0/3h and 0/3/6h cTnI [less than or equal to] 99th percentile resulted in sensitivities of 98.7% (95% CI, 96.9-100) and 99.1% (95% CI, 97.4-100), respectively (Table 2, see Fig. 1B in the online Data Supplement). The 30-day risk stratification for MI or cardiac death for cTnI concentrations [less than or equal to] 99th percentile at 0/3h and at 0/3/6h is shown in Table 3. Using serial contemporary cTnI alone, 0/3/6h measurements resulted in a sensitivity of 98.3% (95% CI, 95.9-100) and an NPV of 99.6% (95% CI, 99.0-100) for 30-day acute MI or cardiac death. Adding a normal ECG to 0/3/6h cTnI measurements <99th percentile offered asensitivity of 99.1% (95% CI, 97.4-100) and an NPV of 99.5% (95% CI, 98.6-100) for 30-day acute MI or cardiac death (Table 3). Diagnostic performance and risk stratification for early presenters ([less than or equal to] 2 h) by use of cTnI measurements alone and in combination with a normal ECG are shown in Table 4 and Tables 1 and 2 in the online Data Supplement. Using contemporary cTnI, a sensitivity and an NPV of 100% for both the diagnostic and safety outcomes was achieved in early presenters using 0/3/6h cTnI [less than or equal to] 99th percentile and a normal ECG.

Serial hs-cTnI concentrations at 0/3h had a sensitivity of 95.4% (95% CI, 91.5-99.3) and a NPV of 99.4% (95% CI, 98.9 -99.9) for MI, whereas 0/3/6h hs-cTnI showed a sensitivity of 98.7% (95% CI, 96.1-100) and an NPV of 99.8% (95% CI, 99.4-100) (Table 1; see Fig. 1A in the online Data Supplement). Addition of a normal ECG to serial hs-cTnI measurements offered a sensitivity and an NPV of 100% (95% CI, 100-100) for both 0/3h and 0/3/6h measurements (Table 2; see Fig. 1B in the online Data Supplement). Serial hs-cTnI measurements alone at 0/3h and at 0/3/6h offered a sensitivity of 94.5% (95% CI, 90.3-98.8) and 97.4% (95% CI, 93.9-100), respectively, for the safety outcome of 30-day acute MI or cardiac death. Adding a normal ECG to serial hs-cTnI measurements offered both a sensitivity and a NPV of 100% (95% CI, 100-100) for both 0/3h and 0/3/6h measurements (Table 3). Diagnostic performance and risk stratification for early presenters ([less than or equal to] 2 h) by use of hs-cTnI alone and in combination with a normal ECG is shown in Table 4 and in Tables 1 and 2 in the online Data Supplement. Using hs-cTnI, the sensitivity and NPV were 100% (95% CI, 100-100) for early presenters using either 0/3h or 0/3/6h hs-cTnI alone or in combination with a normal ECG for the diagnostic outcome of acute MI. For the safety outcome, however, serial hs-cTnI concentration [less than or equal to] 99th percentile at 0/3h and 0/3/6h had asensitivity of 96.4% (95% CI, 89.6-100) and 95.7% (95% CI, 87.3-100), respectively (Table 4). Conversely, for early presenters ([less than or equal to] 2 h), the addition of a normal ECG to serial hs-cTnI measurements offered both a sensitivity and a NPV of 100% (95% CI, 100-100) with use of either 0/3h or 0/3/6h measurements (Table 4).

Individual concentrations at specific time points. Diagnostic performances of contemporary cTnI and hs-cTnI concentrations with use of the 99 th percentile for the diagnosis of MI, as well as T1MI and T2MI alone, by using individual concentrations at specific time points, are shown in Table 3 in the online Data Supplement. Using individual contemporary cTnI concentrations, the highest sensitivity was 82.5% (95% CI, 76.8-88.3) at the 6-h measurement, with NPVs ranging across time from 95.5% (95% CI, 94.3-96.6) at 0 h to 96.0% (95% CI, 94.6-97.4) at 9 h. Using individual hs-cTnI concentrations, the highest sensitivities for MI were 93.6% (95% CI, 89.0-98.2) and 93.7% (95% CI, 89.5-97.9) at 3 and 6 h, respectively, with NPVs ranging across time from 95.7% (95% CI, 94.6-96.8) at 0 h to 99.0% (95% CI, 98.4-99.7) at 6 h, including an NPV of 99.2% (95% CI, 98.5-99.8) at 3 h. At 3 h, the sensitivity and NPV for the contemporary cTnI assay were 82.4% (95% CI, 76.3-88.4) and 96.8% (95% CI, 95.6-98.0) for acute MI, whereas those for the hs-cTnI assay were 93.6% (95% CI, 89.0-98.2) and 99.2% (95% CI, 98.5-99.8), respectively.

Diagnostic performance for baseline cTnI and hs-cTnI measurements for MI, including subtypes, is shown in Table 1. For MI, T1MI, and T2MI, respectively, sensitivities for contemporary cTnI at baseline were 72.9% (95% CI, 66.8-78.9), 77.5% (95% CI, 67.8-87.2), and 70.5% (95% CI, 62.9-78.1), and those for hs-cTnI at baseline were 66.5% (95% CI, 59.4-73.6), 69.1% (95% CI, 58.1-80.1), and 64.7% (95% CI, 55.4-74.0), respectively. In combination with a normal ECG, sensitivities for contemporary cTnI at baseline were 94.3% (95% CI, 91.2-97.4), 92.9% (95% CI, 87.0-98.9), and 95.0% (95% CI, 91.3-98.6) for MI, T1MI, andT2MI, respectively, and those for hs-cTnI at baseline were 92.9% (95% CI, 89.1-96.8), 95.6% (95% CI, 90.7100), and 91.2% (95% CI, 85.7-96.7), respectively (Table 2).

RULING IN MI

Delta troponin. ROC curve analysis using relative (%) and absolute concentration deltas ([micro]g/L) are shown in Fig. 1, Panels A and B for the contemporary cTnI assay. Absolute deltas were superior to relative deltas (AUC 0.557 vs 0.8312, P = 0.01). Absolute cTnI values, however, were too low (0.010-0.005 M-g/L) to be clinically useful. In contrast to the specificity of 84.4% (95% CI, 82.5-86.3) achieved by the individual baseline contemporary cTnI concentration at presentation and the composite specificity of 78.7% (95% CI, 75.4, 82.0) for the diagnosis of MI obtained by serial measurements at 0, 3, and 6 h (Table 1), delta cTnI values using 0/6h percent relative changes offered an improved specificity for MI at concentration changes >150% (see Table 4 in the online Data Supplement and Fig. 2A in the online Data Supplement), with a lower proportion of patients qualifying at higher delta values (see Fig. 2B in the online Data Supplement).

Delta values for the diagnosis of MI by use of absolute concentration changes (ng/L) and percent relative change values for hs-cTnI are shown in Table 5 in the online Data Supplement and Fig. 1, Panels C and D. ROC curve analysis showed that absolute deltas were superior to relative deltas (AUC 0.8979 vs 0.6317, P = 0.01). In contrast to the specificity of 86.9% (95% CI, 85.1, 88.6) achieved by a baseline hs-cTnI concentration at presentation, and specificity of 85.7% (95% CI, 83.5, 87.9) by serial measurements at 0/3h (Table 1), delta hs-cTnI values offered an improved specificity for MI at concentration changes > 5 ng/L. Although an increase in absolute delta hscTnI concentration offered improved specificity with rising values, a lower proportion of patients qualified at higher values (see Fig. 3B in the online Data Supplement).

Delta analyses based on whether the baseline hscTnI was negative or increased demonstrated that delta hs-cTnI values obtained to maximize specificity differed according to the baseline hs-cTnI concentrations (see Table 5 in the online Data Supplement). For patients with negative hs-cTnI at presentation (n = 873), an absolute delta cTnI concentration > 5 ng/L provided a specificity of 94.2% (95% CI, 92.6-95.8). Conversely, for patients with increased hs-cTnI concentration at presentation (n = 195), an absolute delta cTnI concentration >5 ng/L provided a specificity of 54.2% (95% CI, 45.3-63.2), with absolute delta cTnI concentration >100 ng/L required to obtain a specificity of 94.1% (95% CI, 89.8-98.3).

Discussion

Our study has several important and unique findings. First, we demonstrate that both contemporary cTnI (clinically used in the US) and hs-cTnI (investigational in the US, yet used worldwide) assays are excellent in safely ruling out MI following consensus recommendations predicated on the use of serial testing and the 99th percentile, when used in combination with a normal ECG. In many countries, including the US, contemporary cTn assays remain the standard of care. Supporting current US practice, our study demonstrates that serial cTnI measurements performed every 3 h for a maximum of 3 measurements (0, 3, and 6 h) are excellent in safely ruling out MI when combined with a normal ECG, as demonstrated by the NPV and sensitivity [greater than or equal to] 99% for both diagnostic and safety outcomes, including 30-day risk of MI or cardiac death. This approach, consistent with current standard of care, allows to safely rule out MI in 29% of patients if serial 0/3/6h cTnI measurements are [less than or equal to] 99th percentile and the ECG is normal.

hs-cTn assays have been endorsed and adopted worldwide because they allow for a more rapid rule in and rule out of MI. In our current study, we demonstrate that by using an hs-cTnI assay, MI may be safely ruled out within 3 h by measuring hs-cTnI concentration at 0 and 3 h, if concentrations at both time points are below sex-specific 99 th percentiles in combination with a normal ECG. This strategy provides both a sensitivity and an NPV of 100% for acute MI within 3 h, including excellent performance among early ([less than or equal to] 2 h) presenters for which most studies do not provide early diagnosis data. This approach allows to safely rule out MI in a third of patients as demonstrated by a sensitivity and an NPV of 100% for acute MI or cardiac death at 30 days. Practically, we demonstrate that in the setting of normal ECG, rather than taking >6 h to safely rule out MI by using 3 serial contemporary cTnI measurements (0, 3, and 6 h), the same performance can be obtained within 3 h using only 2 hs-cTnI measurements (0 and 3 h); the latter approach may lead to more rapid discharge in selected patients, expedite triage in the ED, reduce overcrowding, and improve cost-effective resource utilization.

Second, we highlight the importance of using cTn testing in combination with ECG findings, with the goal of [less than or equal to] 1% MI miss rate. Using serial cTnI measurements alone at 0/3/6h provided a sensitivity of 98.3% for MI and 98.3% for the safety outcome of MI or cardiac death at 30 days, a sensitivity that some have argued is not acceptable owing to a desired miss rate of <1% and/or sensitivity of [greater than or equal to] 99%, and yet it represents the current standard of care (13). Similarly, using serial hs-cTnI measurements alone at 0/3h and at 0/3/6h, sensitivities were 95.4% and 98.7%, respectively, with corresponding sensitivities of 94.5% and 97.4%, respectively, for safety outcomes. Although our CIs do not eliminate the possibility of the diagnostic sensitivity being as high as 99%, we lack the statistical power to confirm this point as the lower bound is in the mid-90%. This highlights the need for a consensus for establishing what is an acceptable miss rate when defining acceptable rule-out strategies.

In contrast to our findings, Pickering et al., by use of hs-cTnI in a pooled, cohort analysis (1061 non-US patients in cohorts ranging from 64 to 409 patients), with disease prevalence ranging from 6% to 19%, showed an NPV of 99.0% and a sensitivity of 93.2% for MI for a 0- and 3-h hs-cTnI concentration <26 ng/L (overall 99th percentile), suggesting the 99th percentile was suboptimal for clinical use (14). Our study, however, suggests that with a normal ECG, if hs-cTnI concentrations are [less than or equal to] 99th percentile (sex-specific cutoffs) at 0 and 3 h, then both the NPV and sensitivity are 100% for the diagnostic and safety outcome.

Third, we demonstrate that cTnI testing by using the 99th percentile should be exclusively used with serial measurements when evaluating patients for suspected MI. None of the individual concentrations obtained at any of the specific time points examined offered an acceptable sensitivity with either assay. Using the contemporary cTnI assay, the highest sensitivity was 82.5% at 6 h, whereas using the hs-cTnI assay, the highest sensitivity was 93.7% at 6 h. These findings support the use of preset serial cTn measurements (serially every 3 h as part of a standard order set available to physicians in our study) rather than isolated individual cTn measurements that provide a suboptimal sensitivity for acute MI.

Fourth, we provide novel insights regarding the performance on MI subtypes. If the main goal is solely to provide a safe exclusion of T1MI, rather than all MIs, then a strategy using serial hs-cTnI at 0/3h with a normal ECG that provides a sensitivity of 100% for both T1MI and overall MI or cardiac death at 30 days appears preferable over using serial contemporary cTnI at 0/3/6h with a normal ECG; the latter provides a sensitivity of 97.7% for T1MI with a sensitivity of 97.8% for MI or cardiac death at 30 days. At present, there is no consensus as to whether diagnostic studies should focus their analyses on all MIs vs T1MI alone, with heterogeneous designs seen in this regard across the literature (15).

Fifth, we demonstrate that the use of delta cTnI (contemporary and high-sensitivity assays) best optimizes diagnostic specificity for acute MI; with delta cTn used to differentiate acute myocardial injury, including MI, from those with chronic increases (2, 3). In comparison to the specificity achieved by individual cTnI concentrations at presentation and by serial testing using both the contemporary and high-sensitivity assays, overall delta cTnI values using relative changes > 150% and delta hs-cTnI values using absolute changes > 5 ng/L provided a superior specificity. For contemporary cTnI assays, relative (%) changes have been historically recommended, whereas for high-sensitivity assays, absolute concentration changes (ng/L) have been shown to be superior (16-18). Our study supports the superiority of absolute changes over relative changes.

Sixth, we show that in comparison to patients with baseline cTnI concentrations [less than or equal to] 99th percentile, for patients with a baseline concentration >99th percentile, delta values required to achieve an improved specificity differed. In contrast to the absolute delta > 5 ng/L to obtain a specificity of 89% in the entire cohort or 94% in those with an initial negative concentration [less than or equal to] 99th per centile, an absolute delta of 50-100 ng/L was necessary to achieve specificities of 89% and 94%, respectively, in those with baseline hs-cTnI concentrations >99th percentile. For patients with myocardial injury at presentation, a different metric may be needed in contrast to patients with an initial negative sample.

Seventh, our data confirm previous observations that the rate of MI does not increase by using the Abbott hs-cTnI assay compared with the contemporary cTnI assay (10). The use of the Abbott hs-cTnI assay provides clinicians with a greater number of reportable numeric cTnI concentrations, via a shift of results from < LoD to those between the LoD and the 99th percentile, and does not increase the cTnI concentrations above the 99th percentile (11), thereby decreasing the analytical noise that can lead to more cTnI increases based on the contemporary assay (10). Our findings using a hs-cTnI assay contrast with recent observations showing an increased rate of MI when transitioning from a contemporary cTnT assay (fourth generation) to the hs-cTnT assay (fifth generation) (19).

Finally, we have recently reported that the head-to-head concordances between the contemporary and hscTnI assays in our UTROPIA cohort were the highest in the diagnosis ofT1MI (K = 0.72, 98% concordance), followed by T2MI (K = 0.53, 93% concordance), and myocardial injury (K = 0.58, 89% concordance) (20). Discordant diagnosis were mainly due to differences in adjudication between T2MI and myocardial injury (20), with myocardial injury defined as any cTnI increase >99th percentile that was not adjudicated as an MI; the most common etiologies for myocardial injury being renal failure, congestive heart failure, and neurological conditions.

Possible limitations of our current study are as follows. First, our findings are based on 1 manufacturer's hs-cTnI assay, and may not be transferable to other hscTnI or contemporary cTnI or cTnT assays. Second, we recognize adjudicating MI by using the 99th percentile threshold for cTnI and hs-cTnI and prespecified delta criteria may be considered a limitation. This approach is, however, intrinsic to all studies that investigate cTn and deltas as key criteria to make the diagnosis of MI according to the Third Universal Definition of MI consensus recommendations, which require the presence of a rise and/or fall of cTn with at least 1 value above the 99th percentile. Compared with many studies that do not provide clear details as to how a rise and/or fall was determined to support the diagnosis of acute MI, the strengths of our study are that it describes how (a) hs-cTnI changes over time and (b) on the basis of prior data, using the hs-cTnI assay's biological variation offers adjudicators a guideline on what cTn concentration changes lie within biological variation. Third, our uniform event-adjudication protocol, based on the maximum cTnI concentrations from all patient samples measured following admission through 24 h and independently adjudicating diagnostic accuracy on the basis of 2 different cTnI assays in one single population, with the adjudicators blinded to the results from the other assay, was used to simulate real practice conditions for each assay. While our process is unique, we acknowledge it differs from the common practice of a single assay gold standard. Fourth, we recognize the potential for incorporation bias due to the gradual drop off in numbers up to the 9 h sample in the order sets, for example, dropping from 1631 patients who had the baseline sample to 711 who had the 0-, 3-, 6-, and 9-h set. The subjects who were dropped for the 0-, 3-, 6-, and 9-h time set were not dropped owing to the lack ofa second sample (as our inclusion criteria required subjects to have at least 2 cTn measurements within 24 h), but because the second, third, or fourth samples were drawn outside the acceptable time window from the baseline sample, or because there was lack of a sufficient sample to run both assays. Real-world practice, however, as we move toward hs-cTn assays, will require a full and appropriately timed sample set to obtain accurate diagnostic information.

In conclusion, both contemporary cTnI and hs-cTnI assays were excellent in ruling out MI following consensus recommendations predicated on the use of serial testing and the 99th percentile. In comparison to current US approaches using contemporary assays, hscTnI allows for a safe and more rapid rule out by use of serial testing at 0 and 3 h along with sex-specific 99th percentiles. The use of delta cTnI values improves the diagnostic specificity for MI, with different deltas needed according to whether the initial sample is normal or increased.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contribution to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.

Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:

Employment or Leadership: F.S. Apple, HyTest Ltd. and Clinical Chemistry, AACC.

Consultant or Advisory Role: Y. Sandoval, Roche Diagnostics; S.W. Smith, Alere, Roche Clinical Diagnostics, Siemens; F.S. Apple, Philips Healthcare Incubator, Metanomics Healthcare.

Stock Ownership: None declared.

Honoraria: F.S. Apple, Instrumentation Laboratory, Abbott POC.

Research Funding: UTROPIA study (NCT02060760) partially funded through a grant from Abbott Diagnostics and the Minneapolis Medical Research Foundation. S.A. Love, Research PI through Minneapolis Medical Research Foundation (MMRF), not salaried, and Biokit, Hytest Ltd., Instrumentation Laboratory; F.S. Apple, Research PI through Minneapolis Medical Research Foundation (MMRF), not salaried: Abbott Diagnostics, Roche Diagnostics, Siemens Healthcare, Alere, Ortho-Clinical Diagnostics, Nanomix, Becton Dickinson, Singulex.

Expert Testimony: None declared.

Patents: None declared.

Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, and final approval of manuscript.

References

(1.) Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol 2012; 60:1581-98.

(2.) Apple FS, Sandoval Y, Jaffe AS, Ordonez-Llanos, IFCC Task Force on Clinical Applications of Cardiac BioMarkers. Cardiac troponin assays: guideto understanding analytical characteristics and their impact on clinical care. Clin Chem 2017; 63:73-81.

(3.) Apple FS, Jaffe AS, Collinson P, Mockel M, Ordonez-Llanos J, Lindahl B, et al. IFCC educational materials on selected analytical and clinical applications of high-sensitivity cardiac troponin assays. Clin Biochem 2015; 48:201-3.

(4.) Apple FS, Collinson PO, IFCC Task Force on Clinical Applications of Cardiac Biomarkers. Analytical characteristics of high-sensitivity cardiac troponin assays. Clin Chem 2012; 58:54-61.

(5.) Amsterdam EA, Wenger NK, Brindis RG, Casey DE Jr., Ganiats TG, Holmes DR, Jr., et al. 2014 AHA/ACC guideline for the management of patients with non ST-elevation acute coronary syndromes: executive summary: a report of the America College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014; 130:2354-94.

(6.) Sandoval Y, Apple FS. The global need to define normality: the 99th percentile value of cardiac troponin. Clin Chem 2014; 60:455-62.

(7.) Sandoval Y, Smith SW, Apple FS. Present and future of cardiac troponin in clinical practice: a paradigm shiftto high-sensitivityassays.AmJ Med2016; 129:354-65.

(8.) Roffis M, Patrono C, Collet JP, Mueller C, Valgimigli M, Andreotti F, et al. 2015 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: Task Force for the Management of Acute Coronary Syndromes in Patients Presenting without Persistent ST-Segment Elevation of the European Society of Cardiology (ESC). Eur Heart J 2016; 37:267-315.

(9.) Morrow DA. Clinicians guide to early rule-out strategies with high-sensitivity cardiac troponin. Circulation 2017; 135:1612-6.

(10.) Sandoval Y, Smith SW, Schulz KM, Murakami MM, Love SA, Nicholson J, Apple FS. Diagnosis of type 1 and type 2 myocardial infarction using a high-sensitivity cardiac troponin I assay with sex-specific 99th percentiles based on the third universal definition of myocardial infarction classification system. Clin Chem 2015; 61: 657-63.

(11.) Love SA, Sandoval Y, Smith SW, Nicholson J, Cao J, Ler R, et al. Incidence of undetectable, measurable, and increased cardiac troponin I concentrations above the 99th percentile using a high-sensitivity vs. a contemporary assay in patients presenting to the emergency department. Clin Chem 2016; 62:1115-9.

(12.) Sandoval Y, Smith SW, Shah AS, Anand A, Chapman AR, Love SA, et al. Rapid rule-out of acute myocardial injury using a single high-sensitivity cardiac troponin I measurement. Clin Chem 2017; 63:369-76.

(13.) Than M, Herbert M, Flaws D, Cullen L, Hess E, Hollander JE, et al. What is an acceptable risk of major adverse cardiac event in chest pain patients soon after discharge from the emergency department? a clinical survey. Int J Cardiol 2013; 166:752-4.

(14.) Pickering JW, Greenslade JH, Cullen L, Flaws D, Parsonage W, George P, et al. Validation of presentation and 3 h high-sensitivity troponin to rule-in and rule-out acute myocardial infarction. Heart 2016; 102:1270-8.

(15.) Sandoval Y, Thygesen K. Myocardial infarction type 2 and myocardial injury. Clin Chem 2017; 63:101-7.

(16.) Reichlin T, Irfan A, Twerenbold R, Reiter M, Hochholzer W, Burkhalter H, et al. Utility of absolute and relative changes in cardiac troponin concentrations in the early diagnosis of acute myocardial infarction. Circulation 2011; 124:136-45.

(17.) Mueller M, Biener M, Vafaie M, Doerr S, Keller T, Blankenberg S, et al. Absolute and relative kinetic changes of high-sensitivity cardiac troponin T in acute coronary syndrome in patients with increased troponin in the absence of acute coronary syndrome. Clin Chem 2012; 58:209-18.

(18.) Apple FS, Smith SW, Pearce LA, Murakami MM. Delta changes for optimizing clinical specificity and 60-day risk of adverse events in patients presenting with symptoms suggestive of acute coronary syndrome utilizing the ADVIA Centaur TnI-Ultraassay. Clin Biochem 2012; 45:711-3.

(19.) Reichlin T, Twerenbold R, Reiter M, Steuer S, Bassetti S, Balmelli C, et al. Introduction of high-sensitivity troponin assays: impact on myocardial infarction incidence and prognosis. Am J Med 2012; 125:1205-13.

(20.) Sandoval Y, Smith SW, Thordsen SE, Bruen CA, Carlson MD, Dodd KW, et al. Type 1 and 2 myocardial infarction and myocardial injury: clinical transition to high-sensitivity cardiac troponin I. [Epub ahead of print] Am J Med July 21, 2017 as doi: http://dx.doi.org/ 10.1016/j.amjmed.2017.05.049.

Yader Sandoval, [1] Stephen W. Smith, [2,3] Sarah E. Thordsen, [1] Charles A. Bruen, [2,4] Michelle D. Carlson, [1] Kenneth W. Dodd, [2,4] Brian E. Driver, [2] Katherine Jacoby, [2,4] Benjamin K. Johnson, [1] Sara A. Love, [5,6] Johanna C. Moore, [2] Anne Sexter, [7] Karen Schulz, [7] Nathaniel L. Scott, [2,4] Jennifer Nicholson, [5] and Fred S. Apple [5,6] *

[1] Division of Cardiology, Hennepin County Medical Center and Minneapolis Heart Institute, Abbott Northwestern Hospital, Minneapolis, MN; [2] Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, MN; [3] Department of Emergency Medicine, University of Minnesota, Minneapolis, MN; [4] Department of Medicine, Hennepin County Medical Center, Minneapolis, MN; [5] Department of Laboratory Medicine and Pathology, Hennepin County Medical, Minneapolis, MN; [6] Department of Labora tory Medicine and Pathology, University of Minnesota, Minneapolis, MN; [7] Minneapolis Medical Research Foundation, Minneapolis, MN.

* Address correspondence to this author at: Hennepin County Medical Center, Clinical Laboratories P4, 701 Park Ave., Minneapolis, MN 55415. Fax 612 904 4229; e-mail apple004@umn.edu.

Received February 12,2017; accepted June 9,2017.

Previously published online at DOI: 10.1373/clinchem.2017.272930

[8] Nonstandard abbreviations: cTn, cardiac troponin; hs-cTn, high-sensitivity cardiac troponin; MI, myocardial infarction; ECG, electrocardiogram; T1MI, type 1 myocardial infarction; T2MI, type 2 myocardial infarction; ED, emergency department; LoD, limit of detection; PPV, positive predictive value; NPV, negative predictive value.

Caption: Fig. 1. ROC curves for contemporary (0/6h) (A and B) and high-sensitivity (0/3h) (C and D) cTnI assays showing percent relative (%) and absolute concentration ([micro]g/L or ng/L) deltas for the diagnosis of acute MI.
Table 1. Diagnostic accuracy for acute MI, including T1MI and
T2MI, by use of baseline and serial contemporary cTnI and hs-cTnI
measurements at 0/3h; 0/3/6h; and 0/3/6/9h.

                                 Baseline measurement [less than
                                 or equal to] 99th URL
                                       cTnI               hs-cTnI

Acute MI
Number, n (%) of patients        1256/1631 (77.0)    1326/1631 (81.3)
  qualifying
NPV (95% CI)                     95.5 (94.3, 96.6)   95.7 (94.6, 96.8)
Sensitivity (95% CI)             72.9 (66.8, 78.9)   66.5 (59.4, 73.6)
PPV (95% CI)                     40.8 (35.9, 45.8)   37.1 (31.6, 42.5)
Specificity (95% CI)             84.4 (82.5, 86.3)   86.9 (85.1, 88.6)
T1MI
Number, n (%) of patients        1215/1492 (81.4)    1290/1,529 (84.4)
  qualifying
NPV (95% CI)                     98.7 (98.0, 99.3)   98.4 (97.7, 99.1)
Sensitivity (95% CI)             77.5 (67.8, 87.2)   69.1 (58.1, 80.1)
PPV (95% CI)                     19.9 (15.2, 24.6)   19.7 (14.6, 24.7)
Specificity (95% CI)             84.4 (82.5, 86.3)   86.9 (85.1, 88.6)
T2MI
Number, n (%) of patients        1240/1560 (79.5)    1305/1563 (83.5)
  qualifying
NPV (95% CI)                     96.7 (95.7, 97.7)   97.2 (96.4, 98.1)
Sensitivity (95% CI)             70.5 (62.9, 78.1)   64.7 (55.4, 74.0)
PPV (95% CI)                     30.6 (25.6, 35.7)   25.6 (20.3, 30.9)
Specificity (95% CI)             84.4 (82.5, 86.3)   86.9 (85.1, 88.6)

                                     0/3h measurements [less than
                                       or equal to] 99th URL
                                       cTnI               hs-cTnI

Acute MI
Number, n (%) of patients         795/1103 (72.1)     827/1068 (77.4)
  qualifying
NPV (95% CI)                     98.7 (98.0, 99.5)   99.4 (98.9, 99.9)
Sensitivity (95% CI)             93.5 (89.6, 97.4)   95.4 (91.5, 99.3)
PPV (95% CI)                     46.4 (40.9, 52.0)   43.2 (36.9, 49.4)
Specificity (95% CI)             82.6 (80.2, 85.0)   85.7 (83.5, 87.9)
T1MI
Number, n (%) of patients         790/1003 (78.8)     824/1003 (82.2)
  qualifying
NPV (95% CI)                     99.4 (98.8, 99.9)   99.8 (99.4, 100)
Sensitivity (95% CI)             90.6 (82.7, 98.4)   95.5 (89.3, 100)
PPV (95% CI)                     22.5 (16.9, 28.2)   23.5 (17.3, 29.7)
Specificity (95% CI)             82.6 (80.2, 85.0)   85.7 (83.5, 87.9)
T2MI
Number, n (%) of patients         790/1050 (75.2)     825/1024 (80.6)
  qualifying
NPV (95% CI)                     99.4 (98.8, 99.9)   99.6 (99.2, 100)
Sensitivity (95% CI)             95.0 (90.7, 99.3)   95.4 (90.3, 100)
PPV (95% CI)                     36.5 (30.7, 42.4)   31.2 (24.7, 37.6)
Specificity (95% CI)             82.6 (80.2, 85.0)   85.7 (83.5, 87.9)

                                        0/3/6h measurements
                                 [less than or equal to] 99th URL
                                       cTnI               hs-cTnI

Acute MI
Number, n (%) of patients         472/711 (66.4)      462/627 (73.7)
  qualifying
NPV (95% CI)                     99.6 (99.0, 100)    99.8 (99.4, 100)
Sensitivity (95% CI)             98.3 (95.8, 100)    98.7 (96.1, 100)
PPV (95% CI)                     46.9 (40.5, 53.2)   45.5 (37.9, 53.1)
Specificity (95% CI)             78.7 (75.4, 82.0)   83.7 (80.6, 86.8)
T1MI
Number, n (%) of patients         471/641 (73.5)      461/585 (78.8)
  qualifying
NPV (95% CI)                     99.8 (99.4, 100)     100 (100, 100)
Sensitivity (95% CI)             97.7 (93.3, 100)     100 (100, 100)
PPV (95% CI)                     25.3 (18.8, 31.8)   27.4 (19.6, 35.3)
Specificity (95% CI)             78.7 (75.4, 82.0)   83.7 (80.6, 86.8)
T2MI
Number, n (%) of patients         471/667 (70.6)      462/593 (77.9)
  qualifying
NPV (95% CI)                     99.8 (99.4, 100)    99.8 (99.4, 100)
Sensitivity (95% CI)             98.6 (95.8, 100)    97.6 (93.0, 100)
PPV (95% CI)                     35.2 (28.5, 41.9)   31.3 (23.4, 39.2)
Specificity (95% CI)             78.7 (75.4, 82.0)   83.7 (80.6, 86.8)

                                   0/3/6/9 h measurements [less than
                                      or equal to] 99th URL
                                       cTnI               hs-cTnI

Acute MI
Number, n (%) of patients         338/520 (65.0)     319/434 (73.5)
  qualifying
NPV (95% CI)                     99.7 (99.1, 100)    100 (100, 100)
Sensitivity (95% CI)             98.9 (96.7, 100)    100 (100, 100)
PPV (95% CI)                     47.8 (40.6, 55.1)   46.1 (36.9, 55.2)
Specificity (95% CI)             78.0 (74.1, 81.9)   83.7 (80.0, 87.4)
T1MI
Number, n (%) of patients         338/469 (72.1)     319/409 (78.0)
  qualifying
NPV (95% CI)                     99.7 (99.1, 100)    100 (100, 100)
Sensitivity (95% CI)             97.3 (92.1, 100)    100 (100, 100)
PPV (95% CI)                     27.5 (19.8, 35.1)   31.1 (21.6, 40.7)
Specificity (95% CI)             78.0 (74.1, 81.9)   83.7 (80.0, 87.4)
T2MI
Number, n (%) of patients         337/483 (69.8)     319/406 (78.6)
  qualifying
NPV (95% CI)                      100 (100, 100)     100 (100, 100)
Sensitivity (95% CI)              100 (100, 100)     100 (100, 100)
PPV (95% CI)                     34.9 (27.2, 42.7)   28.7 (19.2, 38.2)
Specificity (95% CI)             78.0 (74.1, 81.9)   83.7 (80.0, 87.4)

Table 2. Diagnostic accuracy for acute Ml, including T1MI and T2MI, by
use of baseline and serial contemporary cTnl and hs-cTnl measurements
at 0-3h; 0-3-6h; and 0-3-6-9h, in addition to a normal 12-lead ECG.

                                         Baseline measurement
                                      [less than or equal] 99th
                                         URL and normal ECG

                                       cTnl               hs-cTnl
Acute Ml
Number, n (%) of patients         528/1631 (32.4)     548/1631 (33.6)
  qualifying
NPV (95% Cl)                     97.7(96.5, 99.0)    97.8(96.6, 99.0)
Sensitivity (95% Cl)             94.3(91.2, 97.4)    92.9(89.1, 96.8)
PPV (95% Cl)                     18.0(15.7, 20.2)    14.6(12.5, 16.7)
Specificity (95% Cl)             36.3(33.8, 38.8)    36.7 (34.2, 39.2)
T1 MI
Number, n (%) of patients         521/1492(34.9)      539/1529(35.3)
  qualifying
NPV (95% Cl)                     99.0(98.2, 99.9)     99.4(98.8, 100)
Sensitivity (95% Cl)             92.9(87.0, 98.9)     95.6(90.7, 100)
PPV (95% Cl)                      6.8 (5.2, 8.4)      6.6 (5.0, 8.1)
Specificity (95% Cl)             36.3(33.8, 38.8)    36.7 (34.2, 39.2)
T2MI
Number, n (%) of patients         523/1560(33.5)      545/1563(34.9)
  qualifying
NPV (95% Cl)                     98.7(97.7, 99.7)    98.4(97.3, 99.4)
Sensitivity (95% Cl)             95.0(91.3, 98.6)    91.2(85.7, 96.7)
PPV (95% Cl)                     12.7(10.7, 14.8)     9.1 (7.4, 10.9)
Specificity (95% Cl)             36.3(33.8, 38.8)    36.7 (34.2, 39.2)

                                            0/3h measurements
                                    [less than or equal] 99th URL
                                            and normal ECG

                                       cTnl               hs-cTnl
Acute Ml
Number, n (%) of patients         351/1103(31.8)      350/1068(32.8)
  qualifying
NPV (95% Cl)                      99.4(98.6, 100)      100(100, 100)
Sensitivity (95% Cl)             98.7 (96.9, 100)      100(100, 100)
PPV (95% Cl)                     20.1 (17.2, 22.9)   15.2(12.6, 17.8)
Specificity (95% Cl)             36.7 (33.7, 39.8)   36.5(33.5, 39.5)
T1 MI
Number, n (%) of patients         351/1003(35.0)      305/1003(34.9)
  qualifying
NPV (95% Cl)                      99.4(98.6, 100)      100(100, 100)
Sensitivity (95% Cl)              96.2(91.1, 100)      100(100, 100)
PPV (95% Cl)                      7.8 (5.8, 9.9)      6.7 (4.8, 8.7)
Specificity (95% Cl)             36.7 (33.7, 39.8)   36.5(33.5, 39.5)
T2MI
Number, n (%) of patients         349/1050(33.2)      305/1024(34.2)
  qualifying
NPV (95% Cl)                       100(100, 100)       100(100, 100)
Sensitivity (95% Cl)               100(100, 100)       100(100, 100)
PPV (95% Cl)                     14.3(11.7, 16.9)     9.6 (7.4, 11.9)
Specificity (95% Cl)             36.7 (33.7, 39.8)   36.5 (33.5, 39.5)

                                         0/3/6h measurements
                                    [less than or equal] 99th URL
                                            and normal ECG

                                       cTnl               hs-cTnl
Acute MI
Number, n (%) of patients         208/71 1 (29.3)     194/627 (30.9)
  qualifying
NPV (95% Cl)                      99.5(98.6, 100)      100(100, 100)
Sensitivity (95% Cl)             99.1 (97.4, 100)      100(100, 100)
PPV (95% Cl)                     22.5(18.8, 26.1)    17.6(14.0, 21.1)
Specificity (95% Cl)             34.7 (30.9, 38.5)   35.2(31.2, 39.2)
T1 MI
Number, n (%) of patients         208/641 (32.5)       194/585(33.2)
  qualifying
NPV (95% Cl)                      99.5(98.6, 100)      100(100, 100)
Sensitivity (95% Cl)             97.7 (93.3, 100)      100(100, 100)
PPV (95% Cl)                      9.9 (7.1, 12.8)     8.7 (5.9, 11.5)
Specificity (95% Cl)             34.7 (30.9, 38.5)   35.2(31.2, 39.2)
T2MI
Number, n (%) of patients          207/667(31.0)       194/593(32.7)
  qualifying
NPV (95% Cl)                       100(100, 100)       100(100, 100)
Sensitivity (95% Cl)               100(100, 100)       100(100, 100)
PPV (95% Cl)                     15.2(11.9, 18.5)     10.5(7.5, 13.5)
Specificity (95% Cl)             34.7 (30.9, 38.5)   35.2(31.2, 39.2)

                                    0/3/6/9h measurements [less than
                                          or equal] 99th URL
                                            and normal ECG

                                       cTnl               hs-cTnl
Acute MI
Number, n (%) of patients          150/520(28.8)       132/434(30.4)
  qualifying
NPV (95% Cl)                      99.3(98.0, 100)      100(100, 100)
Sensitivity (95% Cl)              98.9(96.7, 100)      100(100, 100)
PPV (95% Cl)                     23.5(19.2, 27.8)    17.6(13.3, 21.8)
Specificity (95% Cl)             34.5(30.0, 39.0)    34.7 (29.9, 39.4)
T1 MI
Number, n (%) of patients          150/469(32.0)       132/409(32.3)
  qualifying
NPV (95% Cl)                      99.3(98.0, 100)      100(100, 100)
Sensitivity (95% Cl)              97.3(92.1, 100)      100(100, 100)
PPV (95% Cl)                      11.3(7.8, 14.8)    10.1 (6.6, 13.7)
Specificity (95% Cl)             34.5(30.0, 39.0)    34.7 (29.9, 39.4)
T2MI
Number, n (%) of patients          149/483(30.9)       132/406(32.5)
  qualifying
NPV (95% Cl)                       100(100, 100)       100(100, 100)
Sensitivity (95% Cl)               100(100, 100)       100(100, 100)
PPV (95% Cl)                     15.3(11.4, 19.1)     9.1 (5.7, 12.5)
Specificity (95% Cl)             34.5(30.0, 39.0)    34.7 (29.9, 39.4)

URL, upper reference limit.

Table 3. Safety outcomes: 30-day risk stratification for acute MI
or cardiac death (including event occurring during the index
hospitalization) by use of baseline and serial contemporary cTnI
and hs-cTnI measurements 0-3h and at 0-3-6h alone, in addition to
normal 12-lead ECG.

                                          Baseline measurement
                                   [less than or equal to] 99th URL
                                       cTnl               hs-cTnl

                            cTnl and hs-cTnl measurements alone
Acute Ml
Number, n (%) of patients        1256/1631 (77.0)    1326/1631 (81.3)
  qualifying
NPV (95% CI)                     95.5 (94.3,96.6)    95.6 (94.5,96.7)
Sensitivity (95% CI)             73.0 (67.0, 79.0)   66.1 (59.0,73.2)
T1MI
Number, n (%) of patients        1215/1493 (81.4)    1290/1530 (84.3)
  qualifying
NPV (95% CI)                     98.7 (98.0, 99.3)   98.3 (97.6,99.0)
Sensitivity (95% CI)             78.1 (68.6, 87.6)   68.6 (57.7,79.5)
T2MI
Number, n (%) of patients        1241/1565 (79.3)    1307/1568 (83.4)
qualifying
NPV (95% CI)                     96.6 (95.6, 97.6)   97.0 (96.1,97.9)
Sensitivity (95% CI)             71.0 (63.7,78.4)    63.9 (54.8,73.0)
      cTnl and hs-cTnl measurement in combination with a normal ECG
Acute MI
Number, n (%) of patients         528/1631 (32.4)     548/1631 (33.6)
  qualifying
NPV (95% CI)                     97.7 (96.5, 99.0)   97.8 (96.6,99.0)
Sensitivity (95% CI)             94.3 (91.2, 97.4)   93.0 (89.2,96.8)
T1MI
Number, n (%) of patients         521/1493 (34.9)     539/1530 (35.2)
  qualifying
NPV (95% CI)                      99.0 (98.2,100)     99.4 (98.8,100)
Sensitivity (95% CI)             93.2 (87.4, 99.0)    95.7 (91.0,100)
T2MI
Number, n (%) of patients         523/1565 (33.4)     545/1568 (34.8)
  qualifying
NPV (95% CI)                     98.7 (97.7, 99.7)   98.4 (97.3,99.4)
Sensitivity (95% CI)             95.2 (91.7, 98.7)   91.7 (86.5,96.9)

                                          0/3h measurements
                                   [less than or equal to] 99th URL
                                       cTnl               hs-cTnl

                            cTnl and hs-cTnl measurements alone
Acute MI
Number, n (%) of patients         795/1103 (72.1)     827/1068 (77.4)
  qualifying
NPV (95% CI)                     98.7 (98.0,99.5)    99.3 (98.7,99.9)
Sensitivity (95% CI)             93.5 (89.6,97.4)    94.5 (90.3,98.8)
T1MI
Number, n (%) of patients         790/1003 (78.8)     824/1003 (82.2)
  qualifying
NPV (95% CI)                     99.4 (98.8,99.9)     99.6 (99.2,100)
Sensitivity (95% CI)             90.7 (83.0,98.5)     93.3 (86.1,100)
T2MI
Number, n (%) of patients         790/1053 (75.0)     825/1027 (80.3)
qualifying
NPV (95% CI)                     99.4 (98.8,99.9)     99.5 (99.0,100)
Sensitivity (95% CI)             95.2 (91.1,99.3)    94.2 (88.7,99.7)
      cTnl and hs-cTnl measurement in combination with a normal ECG
Acute MI
Number, n (%) of patients         351/1103 (31.8)     350/1068 (32.8)
  qualifying
NPV (95% CI)                      99.4 (98.6,100)      100 (100,100)
Sensitivity (95% CI)              98.7 (96.9,100)      100 (100,100)
T1MI
Number, n (%) of patients         351/1003 (35.0)     350/1003 (34.9)
  qualifying
NPV (95% CI)                      99.4 (98.6,100)      100 (100,100)
Sensitivity (95% CI)              96.3 (91.3,100)      100 (100,100)
T2MI
Number, n (%) of patients         349/1053 (33.1)     350/1027 (34.1)
  qualifying
NPV (95% CI)                       100 (100,100)       100 (100,100)
Sensitivity (95% CI)               100 (100,100)       100 (100,100)

                                      0/3/6h measurements [less
                                     than or equal to] 99th URL
                                       cTnl               hs-cTnl

                            cTnl and hs-cTnl measurements alone
Acute MI
Number, n (%) of patients         472/711 (66.4)      462/627 (73.7)
  qualifying
NPV (95% CI)                      99.6 (99.0,100)     99.6 (99.0,100)
Sensitivity (95% CI)              98.3 (95.9,100)     97.4 (93.9,100)
T1MI
Number, n (%) of patients         471/641 (73.5)      461/585 (78.8)
  qualifying
NPV (95% CI)                      99.8 (99.4,100)     99.8 (99.4,100)
Sensitivity (95% CI)              97.8 (93.5,100)     97.1 (91.6,100)
T2MI
Number, n (%) of patients         471/669 (70.4)      462/595 (77.6)
qualifying
NPV (95% CI)                      99.8 (99.4,100)     99.6 (99.0,100)
Sensitivity (95% CI)              98.6 (96.0,100)     95.6 (89.5,100)
      cTnl and hs-cTnl measurement in combination with a normal ECG
Acute MI
Number, n (%) of patients         208/711 (29.3)      194/627 (30.9)
  qualifying
NPV (95% CI)                      99.5 (98.6,100)      100 (100,100)
Sensitivity (95% CI)              99.1 (97.4,100)      100 (100,100)
T1MI
Number, n (%) of patients         208/641 (32.4)      194/585 (33.2)
  qualifying
NPV (95% CI)                      99.5 (98.6,100)      100 (100,100)
Sensitivity (95% CI)              97.8 (93.5,100)      100 (100,100)
T2MI
Number, n (%) of patients         207/669 (30.9)      194/595 (32.6)
  qualifying
NPV (95% CI)                       100 (100,100)       100 (100,100)
Sensitivity (95% CI)               100 (100,100)       100 (100,100)

URL, upper reference limit.

Table 4. Safety outcomes: 30/day risk stratification for acute MI or
cardiac death (including event occurring during the index
hospitalization) by use of baseline and serial contemporary cTnI and
hs/cTnI measurements at 0/3h and at 0/3/6h alone, in addition to
normal ECG in early presenters.

                                       Baseline measurement
                                 [less than or equal to] 99th URL

Early presenters n = 262               cTnI               hs-cTnI
                          cTnI measurements alone
Acute MI
Number, n (%) of patients         208/262 (79.4)      222/262 (84.7)
  qualifying
NPV (95% CI)                     92.3 (88.7, 95.9)   91.0 (87.2, 94.8)
Sensitivity (95% CI)             61.0 (46.0, 75.9)   50.0 (34.5, 65.5)
                          cTnI measurement and a normal ECG
Acute MI
Number, n (%) of patients          81/262 (30.9)       85/262 (32.4)
  qualifying
NPV (95% CI)                     95.1 (90.3, 99.8)   94.1 (89.1, 99.1)
Sensitivity (95% CI)             90.2 (81.2, 99.3)   87.5 (77.3, 97.8)

                                          0/3h measurements
                                 [less than or equal to] 99th URL

Early presenters n = 262               cTnI               hs-cTnI
                          cTnI measurements alone
Acute MI
Number, n (%) of patients         123/182 (67.6)      131/180 (72.8)
  qualifying
NPV (95% CI)                      99.2 (97.6,100)     99.2 (97.8,100)
Sensitivity (95% CI)              96.8 (90.6,100)     96.4 (89.6,100)
                          cTnI measurement and a normal ECG
Acute MI
Number, n (%) of patients          53/182 (29.1)       52/180 (28.9)
  qualifying
NPV (95% CI)                      98.1 (94.5,100)      100 (100,100)
Sensitivity (95% CI)              96.8 (90.1,100)      100 (100,100)

                                      0/3/6h measurements [less
                                 than or equal to]  99th URL

Early presenters n = 262               cTnI               hs-cTnI
                          cTnI measurements alone
Acute MI
Number, n (%) of patients          76/119 (63.9)       76/108 (70.4)
  qualifying
NPV (95% CI)                       100 (100,100)      98.7 (96.1,100)
Sensitivity (95% CI)               100 (100,100)      95.7 (87.3,100)
                          cTnI measurement and a normal ECG
Acute MI
Number, n (%) of patients          32/119 (26.9)       30/108 (27.8)
  qualifying
NPV (95% CI)                       100 (100,100)       100 (100,100)
Sensitivity (95% CI)               100 (100,100)       100 (100,100)

URL, upper reference limit.
COPYRIGHT 2017 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Proteomics and Protein Markers
Author:Sandoval, Yader; Smith, Stephen W.; Thordsen, Sarah E.; Bruen, Charles A.; Carlson, Michelle D.; Dod
Publication:Clinical Chemistry
Article Type:Report
Date:Oct 1, 2017
Words:9094
Previous Article:Gene Expression Signatures in Circulating Tumor Cells Correlate with Response to Therapy in Metastatic Breast Cancer.
Next Article:Molecular Diagnostics and Genetics Multiplexed Elimination of Wild-Type DNA and High-Resolution Melting Prior to Targeted Resequencing of Liquid...
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters