Differences between local investigator and core laboratory interpretation of the admission electrocardiogram in patients with unstable angina pectoris or non-Q-wave myocardial infarction (a Thrombin Inhibition in Myocardial Ischemia [TRIM] substudy)

Unstable Angina as a Component of Primary Composite Endpoints in Clinical Cardiovascular Trials: Pros and Cons

Background Unstable angina (UA) is a component of acute coronary syndrome that is only occasionally included in primary composite endpoints in clinical cardiovascular trials. The aim of this paper is to elucidate the potential benefits and disadvantages of including UA in such contexts. Summary UA comprises <10% of patients with acute coronary syndromes in contemporary settings. Based on the pathophysiological similarities, it is ideal as a part of a composite endpoint along with myocardial infarction (MI). By adding UA as a component of a primary composite endpoint, the number of events and feasibility of the trial should increase, thus decreasing size and costs of trials. Furthermore, UA has both economic and quality of life implications on a societal and an individual level. However, there are important challenges associated with the use of UA as an endpoint. With the introduction of high-sensitivity troponins, the number of individuals diagnosed with UA has decreased to rather low levels, with a reciprocal increase in the number of MI. In addition, UA is particularly challenging to define given the subjective assessment of the index symptoms, rendering a high risk of bias. To minimize bias, strict criteria are warranted, and events should be adjudicated by a blinded endpoint adjudication committee. Key messages UA should only be chosen as a component of a primary composite endpoint in cardiovascular trials after thoroughly evaluating the pros and cons. If it is chosen to include UA, appropriate precautions should be taken to minimize possible bias.

Golden Standard or Obsolete Method? Review of ECG Applications in Clinical and Experimental Context

Cardiovascular system and its functions under both physiological and pathophysiological conditions have been studied for centuries. One of the most important steps in the cardiovascular research was the possibility to record cardiac electrical activity. Since then, numerous modifications and improvements have been introduced; however, an electrocardiogram still represents a golden standard in this field. This paper overviews possibilities of ECG recordings in research and clinical practice, deals with advantages and disadvantages of various approaches, and summarizes possibilities of advanced data analysis. Special emphasis is given to state-of-the-art deep learning techniques intensely expanded in a wide range of clinical applications and offering promising prospects in experimental branches. Since, according to the World Health Organization, cardiovascular diseases are the main cause of death worldwide, studying electrical activity of the heart is still of high importance for both experimental and clinical cardiology.

Automatic Triage of 12-Lead ECGs Using Deep Convolutional Neural Networks

BACKGROUND

The correct interpretation of the ECG is pivotal for the accurate diagnosis of many cardiac abnormalities, and conventional computerized interpretation has not been able to reach physician‐level accuracy in detecting (acute) cardiac abnormalities. This study aims to develop and validate a deep neural network for comprehensive automated ECG triage in daily practice.

METHODS AND RESULTS

We developed a 37‐layer convolutional residual deep neural network on a data set of free‐text physician‐annotated 12‐lead ECGs. The deep neural network was trained on a data set with 336.835 recordings from 142.040 patients and validated on an independent validation data set (n=984), annotated by a panel of 5 cardiologists electrophysiologists. The 12‐lead ECGs were acquired in all noncardiology departments of the University Medical Center Utrecht. The algorithm learned to classify these ECGs into the following 4 triage categories: normal, abnormal not acute, subacute, and acute. Discriminative performance is presented with overall and category‐specific concordance statistics, polytomous discrimination indexes, sensitivities, specificities, and positive and negative predictive values. The patients in the validation data set had a mean age of 60.4 years and 54.3% were men. The deep neural network showed excellent overall discrimination with an overall concordance statistic of 0.93 (95% CI, 0.92–0.95) and a polytomous discriminatory index of 0.83 (95% CI, 0.79–0.87).

CONCLUSIONS

This study demonstrates that an end‐to‐end deep neural network can be accurately trained on unstructured free‐text physician annotations and used to consistently triage 12‐lead ECGs. When further fine‐tuned with other clinical outcomes and externally validated in clinical practice, the demonstrated deep learning–based ECG interpretation can potentially improve time to treatment and decrease healthcare burden.

Composite Primary End Points in Cardiovascular Outcomes Trials Involving Type 2 Diabetes Patients: Should Unstable Angina Be Included in the Primary End Point?

Reductions in cardiovascular (CV) outcomes in recently reported trials, along with the recent approval by the U.S. Food and Drug Administration of an additional indication for empagliflozin to reduce the risk of CV death in type 2 diabetes patients with evidence of CV disease, have renewed interest in CV outcome trials (CVOTs) of glucose-lowering drugs. Composite end points are a pragmatic necessity in CVOTs to ensure that sample size and duration of follow-up remain reasonable. Combining clinical outcomes into a composite end point increases the numbers of events ascertained and thus statistical power and precision. Historically, composite CV end points in diabetes trials have included a larger number of components, while more recent CVOTs almost exclusively use a composite of CV death, nonfatal myocardial infarction (MI), and nonfatal stroke-the so-called three-point major adverse CV event (3P-MACE) composite-or add hospitalization for unstable angina (HUA) to these three outcomes (4P-MACE). The inclusion of HUA increases the number of events for analysis, but noteworthy disadvantages include clinical subjectivity in ascertainment of HUA and its lower prognostic relevance compared with CV death, MI, or stroke. Furthermore, results from recent CVOTs indicate that glucose-lowering agents seem to have minimal impact on HUA. Its inclusion therefore potentially favors a shift of the hazard ratio (HR) toward the null, which is especially problematic in trials designed to demonstrate noninferiority. The primary outcome of 3P-MACE may offer a better balance than 4P-MACE between statistical efficiency, operational complexity, the likelihood of diagnostic precision (and therefore clinical relevance) for each of the component outcomes, clinical importance, and the aim to adequately capture any potential treatment effect of the intervention. Nevertheless, as individual medications may mechanistically differ in their impact on CV outcomes, no particular individual or composite end point can be seen as a "gold standard" for CVOTs of all glucose-lowering drugs.

Lack of Concordance Between Local Investigators, Angiographic Core Laboratory, and Clinical Event Committee in the Assessment of Stent Thrombosis: Results From the TRACER Angiographic Substudy

BACKGROUND

Stent thrombosis (ST) is an important end point in cardiovascular clinical trials. Adjudication is traditionally based on clinical event committee (CEC) review of case report forms and source documentation rather than angiograms. However, the degree to which this method of adjudication is concordant with the review of independent angiographic core laboratories (ACLs) has not been studied. This report represents the first assessment of variability between local investigators (LIs), a CEC, and an ACL.

METHODS AND RESULTS

Serial angiograms of 329 patients with acute coronary syndrome without ST-segment-elevation who underwent percutaneous coronary intervention at entry in the Trial to Assess the Effects of Vorapaxar in Preventing Heart Attack and Stroke in Particpants With Acute Coronary Syndrome (TRACER) and who met criteria for possible ST subsequent to the index event were reviewed by an ACL. The ACL was blinded to the assessment by both LIs and the CEC regarding the presence or absence of ST. CEC adjudication was based on Academic Research Consortium definitions of ST, using case report form data and source documents, including catheterization laboratory reports. The ACL, CEC, and LIs agreed on the presence or absence of ST in 52.9% events (κ=0.32; 95% confidence interval, 0.26-0.39). The ACL and CEC agreed on 82.7% of events (κ=0.57; 95% confidence interval, 0.47-0.67); the ACL and LIs agreed on 61.1% of events (κ=0.25; 95% confidence interval, 0.16-0.34); and the CEC and LIs agreed on 62% of events (κ=0.28; 95% confidence interval, 0.21-0.36).

CONCLUSIONS

ST reporting by an ACL, a CEC, and LIs is discordant. The assessment of ST is more often detected by direct review of angiograms by an ACL.

CLINICAL TRIAL REGISTRATION

URL: http://www.clinicaltrials.gov. Unique identifier: NCT00527943.

The value of core lab stress echocardiography interpretations: observations from the ISCHEMIA Trial

BACKGROUND

Stress echocardiography (SE) is dependent on subjective interpretations. As a prelude to the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) Trial, potential sites were required to submit two SE, one with moderate or severe left ventricular (LV) myocardial ischemia and one with mild ischemia. We evaluated the concordance of site and core lab interpretations.

METHODS

Eighty-one SE were submitted from 41 international sites. Ischemia was classified by the number of new or worsening segmental LV wall motion abnormalities (WMA): none, mild (1 or 2) or moderate or severe (3 or more) by the sites and the core lab.

RESULTS

Core lab classified 6 SE as no ischemia, 35 mild and 40 moderate or greater. There was agreement between the site and core in 66 of 81 total cases (81%, weighted kappa coefficient [K] =0.635). Agreement was similar for SE type - 24 of 30 exercise (80%, K = 0.571) vs. 41 of 49 pharmacologic (84%, K = 0.685). The agreement between poor or fair image quality (27 of 36 cases, 75%, K = 0.492) was not as good as for the good or excellent image quality cases (39 of 45 cases, 87%, K = 0.755). Differences in concordance were noted for degree of ischemia with the majority of discordant interpretations (87%) occurring in patients with no or mild LV myocardial ischemia.

CONCLUSIONS

While site SE interpretations are largely concordant with core lab interpretations, this appears dependent on image quality and the extent of WMA. Thus core lab interpretations remain important in clinical trials where consistency of interpretation across a range of cases is critical.

TRIAL REGISTRATION

ClinicalTrials.gov NCT01471522.

Accuracy of cardiogoniometry compared with electrocardiography in the diagnosis of coronary artery disease

Background:

Cardiogoniometry (CGM) is a novel spatiotemporal electrocardiographic method utilizing computer-assisted three-dimensional data on cardiac potentials.

Objectives:

This study compares the accuracy of CGM and electrocardiography (ECG) by detecting coronary artery disease (CAD) with reference to angiography as a well-known gold standard.

Patients and Methods:

A total of 390 patients undergoing coronary angiography with CAD were enrolled. CGM was performed a few hours prior to coronary angiography. A standard 12-lead ECG was recorded after the CGM. The CGM and ECG results were recorded and analyzed by an independent investigator blinded to all patient data and the results of the coronary angiography.

Results:

The coronary angiography showed a normal coronary artery in 263 patients (67.4%). A median of CGM score was 1 (0–2), the minimum score was 0 and maximum score was 8. A total of 90 patients (31%) showed predefined ST-segment/T-wave changes in the resting 12-lead ECG. CGM yielded a sensitivity of 84% and specificity of 81% and the ECG yielded a sensitivity of 29% and specificity of 67% when compared with the coronary angiography.

Conclusions:

CGM is a non-invasive technique recently developed for quantitative three-dimensional vectorial analysis of myocardial activity and detection of ischemia and infarction. This technique is clearly more sensitive and more specific than a standard resting 12-lead ECG.

Emergency physician estimates of the probability of acute coronary syndrome in a cohort of patients enrolled in a study of coronary computed tomographic angiography

INTRODUCTION

Little information exists regarding how accurately emergency physicians (EPs) predict the probability of acute coronary syndrome (ACS). Our objective was to determine if EPs can accurately predict ACS in a prospectively identified cohort of emergency department (ED) patients who met enrolment criteria for a study of coronary computed tomographic angiography (CCTA) and were admitted for a "rule out ACS" protocol.

METHODS

A prospective observational pilot study in an academic medical centre was carried out. EPs caring for patients with chest pain provided whole-number estimates of the probability of ACS after clinical review. This substudy was part of the now published Rule Out Myocardial Infarction/Ischemia Using Computer Assisted Tomography (ROMICAT) study, a study of CCTA and admission of patients for a rule out ACS protocol after a nondiagnostic evaluation. Predictions were grouped into probability groups based on the validated Goldman criteria. ACS was determined by an adjudication committee using American Heart Association/American College of Cardiology/European Society of Cardiology guidelines.

RESULTS

A total of 334 predictions were obtained for a study population with a mean age of 54 (SD 12) years, 63% of whom were male. There were 35 ACS events. EPs predicted ACS better than by chance, and increasingly higher estimates were associated with a higher incidence of ACS (p = 0.0004). The percentage of patients with ACS was 0%, 6%, 7%, and 17%, respectively, for very low, low, intermediate, and high probability groups. EPs' estimates had a sensitivity of 63% using a > 20% probability of ACS to define a positive test. Lowering this threshold to > 7% to define a test as positive increased the sensitivity of physician estimates to 89% but lowered specificity from 65% to 24%.

CONCLUSION

Our data suggest that for a selected ED cohort meeting eligibility criteria for a study of CCTA, EPs predict ACS better than by chance, with an increasing proportion of patients proving to have ACS with increasing probability estimates. Lowering the estimate threshold does not result in an overall sensitivity level that is sufficient to send patients home from the ED and is associated with a poor specificity.

Accuracy of interpretation of preparticipation screening electrocardiograms

OBJECTIVE

To evaluate the accuracy of pediatric cardiologists' interpretations of electrocardiograms (ECGs).

STUDY DESIGN

A series of 18 ECGs that represented conditions causing pediatric sudden cardiac death or normal hearts were interpreted by 53 members of the Western Society of Pediatric Cardiology. Gold-standard diagnoses and recommendations were determined by 2 electrophysiologists (100% concordance).

RESULTS

The average number of correct ECG interpretations per respondent was 12.4 ± 2.2 (69%, range 34%-98%). Respondents achieved a sensitivity of 68% and a specificity of 70% for recognition of any abnormality. The false-positive and false-negative rates were 30% and 32%, respectively. Based on actual ECG diagnosis, sports participation was accurately permitted in 74% of cases and accurately restricted in 81% of cases. Respondents gave correct sports guidance most commonly in cases of long QT syndrome and myocarditis (98% and 90%, respectively) and least commonly in cases of hypertrophic cardiomyopathy, Wolff-Parkinson-White syndrome, and pulmonary hypertension (80%, 64%, and 38%, respectively). Respondents ordered more follow-up tests than did experts.

CONCLUSIONS

Preparticipation screening ECGs are difficult to interpret. Mistakes in ECG interpretation could lead to high rates of inappropriate sports guidance. A consequence of diagnostic error is overuse of ancillary diagnostic tests.

Agreement in electrocardiogram interpretation in patients with septic shock

OBJECTIVE

The reliability of electrocardiogram interpretation to diagnose myocardial ischemia in critically ill patients is unclear. In adults with septic shock, we assessed intra- and inter-rater agreement of electrocardiogram interpretation, and the effect of knowledge of troponin values on these interpretations.

DESIGN

Prospective substudy of a randomized trial of vasopressin vs. norepinephrine in septic shock.

SETTING

Nine Canadian intensive care units.

PATIENTS

Adults with septic shock requiring at least 5 μg/min of norepinephrine for 6 hrs.

INTERVENTIONS

Twelve-lead electrocardiograms were recorded before study drug, and 6 hrs, 2 days, and 4 days after study drug initiation.

MEASUREMENTS

Two physician readers, blinded to patient data and group, independently interpreted electrocardiograms on three occasions (first two readings were blinded to patient data; third reading was unblinded to troponin). To calibrate and refine definitions, both readers initially reviewed 25 trial electrocardiograms representing normal to abnormal. Cohen's Kappa and the φ statistic were used to analyze intra- and inter-rater agreement.

RESULTS

One hundred twenty-one patients (62.2 ± 16.5 yrs, Acute Physiology and Chronic Health Evaluation II 28.6 ± 7.7) had 373 electrocardiograms. Blinded to troponin, readers 1 and 2 interpreted 46.4% and 30.0% of electrocardiograms as normal, and 15.3% and 12.3% as ischemic, respectively. Intrarater agreement was moderate for overall ischemia (κ 0.54 and 0.58), moderate/good for "normal" (κ 0.69 and 0.55), fair to good for specific signs of ischemia (ST elevation, T inversion, and Q waves, reader 1 κ 0.40 to 0.69; reader 2 κ 0.56 to 0.70); and good/very good for atrial arrhythmias (κ 0.84 and 0.79) and bundle branch block (κ 0.88 and 0.79). Inter-rater agreement was fair for ischemia (κ 0.29), moderate for ST elevation (κ 0.48), T inversion (κ 0.52), and Q waves (κ 0.44), good for bundle branch block (κ 0.78), and very good for atrial arrhythmias (κ 0.83). Inter-rater agreement for ischemia improved from fair to moderate (κ 0.52, p = .028) when unblinded to troponin.

CONCLUSIONS

In patients with septic shock, inter-rater agreement of electrocardiogram interpretation for myocardial ischemia was fair, and improved with troponin knowledge.

Interobserver variability in antroduodenal manometry

Interobserver variability affects investigations involving assessment of complex visual data, such as histopathology, radiology and motility. This study assessed interobserver variation for interpretation of antroduodenal manometry (ADM), as this has not been previously investigated. Thirty-five ADM recordings from children aged 0.3-18 years were independently evaluated by five experienced paediatric gastroenterologists who were blinded to cases' clinical histories. Intra-class correlation (ICC) was analysed for detection and measurement of phase three of the migrating motor complex (MMC) and Cohen's kappa statistic was calculated between observer pairs for detection of specific motility features and final diagnosis. Observers were unanimous on the differentiation of normal and abnormal motility in 63% of cases. There was excellent interobserver agreement for the number of phase three of the MMC in fasting (ICC = 0.82, P < 0.0001) and for measurements of phase three of the MMC (ICC = 0.9999, P < 0.0001). Detection of other normal and abnormal motility patterns varied more. Objective findings such as the presence of phase three of the MMC correlated more closely than findings that involved the integration of several variables, such as final diagnosis. However, these data overall indicate that agreement between expert observers for the distinction of normal and abnormal antroduodenal motility compares favourably with other standard medical assessments.

Differences between local hospital and core laboratory interpretation of the admission electrocardiogram in patients with acute coronary syndromes and their relation to outcome

In the prospective, multinational Global Registry of Acute Coronary Events (GRACE), patients diagnosed with non-ST-elevation acute coronary syndromes had their admission electrocardiogram independently evaluated by a central core laboratory, and its interpretation by the core laboratory and enrolling site were compared. One in 6 of these patients had clinically important features of left-bundle branch block or ST-segment deviation diagnosed by the core laboratory that were apparently not recognized at the local sites; this subgroup of patients was less likely to undergo risk stratification and revascularization. Importantly, failure to recognize these features as confirmed by the core laboratory in routine clinical practice was independently associated with higher mortality and recurrent myocardial infarction at 6 months (adjusted odds ratio 1.41, 95% confidence interval 1.01 to 1.96, p = 0.043). In conclusion, these findings underscore an urgent need to promote more accurate interpretation of electrocardiograms in contemporary clinical practice to bridge treatment gaps and improve patient outcome.

The prognostic value of the admission and predischarge electrocardiogram in acute coronary syndromes: the GUSTO-IIb ECG Core Laboratory experience

BACKGROUND

Prior research suggests that patients may be entered into clinical trials with different electrocardiographic (ECG) findings than specified by study protocol criteria; the extent and impact of this variability in a large-scale trial have not been previously described.

METHODS

We evaluated the relationship between case report form (CRF) categorization of the admission ECG and a Core Laboratory and subsequent outcome in a retrospective analysis of a trial of patients with acute ischemia and a broad spectrum of ECG changes (the GUSTO-IIb trial).

RESULTS

In 11,037 patients with CRF information and an interpretable ECG, there was agreement in 89.1% of ST-elevation and 81.9% of non-ST-elevation cases. Among patients designated as having no ST elevation on the CRF, 1-year mortality rates were significantly higher in the subgroup of patients with Core Laboratory-determined ST elevation as compared with those where both the CRF and Core Laboratory classification were in agreement (8.8% vs 6.8%, P = .0093). Among patients designated as having ST elevation by the CRF, 1-year mortality rates were similar in both the subgroup of patients with and without Core Laboratory agreement (7.7% vs 8.2%, P = .72).

CONCLUSIONS

These findings have important implications for clinicians in routine practice because even a simple evaluation (presence or absence of ST elevation) on the admission ECG was often discordant and was associated with adverse clinical outcome.

The ubiquitin-proteasome system in cardiac physiology and pathology

The ubiquitin-proteasome system (UPS) is the major nonlysosomal pathway for intracellular protein degradation, generally requiring a covalent linkage of one or more chains of polyubiquitins to the protein intended for degradation. It has become clear that the UPS plays major roles in regulating many cellular processes, including the cell cycle, immune responses, apoptosis, cell signaling, and protein turnover under normal and pathological conditions, as well as in protein quality control by removal of damaged, oxidized, and/or misfolded proteins. This review will present an overview of the structure, biochemistry, and physiology of the UPS with emphasis on its role in the heart, if known. In addition, evidence will be presented supporting the role of certain muscle-specific ubiquitin protein ligases, key regulatory components of the UPS, in regulation of sarcomere protein turnover and cardiomyocyte size and how this might play a role in induction of the hypertrophic phenotype. Moreover, this review will present the evidence suggesting that proteasomal dysfunction may play a role in cardiac pathologies such as myocardial ischemia, congestive heart failure, and myofilament-related and idiopathic-dilated cardiomyopathies, as well as cardiomyocyte loss in the aging heart. Finally, certain pitfalls of proteasome studies will be described with the intent of providing investigators with enough information to avoid these problems. This review should provide current investigators in the field with an up-to-date analysis of the literature and at the same time provide an impetus for new investigators to enter this important and rapidly changing area of research.

Diagnostic conclusions from the EASI-derived 12-lead electrocardiogram as compared with the standard 12-lead electrocardiogram in children

BACKGROUND

Fewer electrodes on more easily located places would facilitate electrocardiogram (ECG) recording. To investigate the possibility of simplifying ECG recording in children, we compared the diagnostic conclusions when interpreting standard versus EASI-derived 12-lead ECGs. Our hypothesis was that the variation of the interpretation of standard versus EASI-derived 12-lead ECGs was not greater than the intrareader variation of the interpretation of standard ECGs.

METHODS

The study included 221 children. The 2 lead systems were recorded simultaneously. Two experienced pediatric cardiologists interpreted the ECGs. First, the reader interpreted a set of 221 ECGs with randomly allocated standard and EASI-derived 12-lead ECGs. Next, the reader interpreted the complementary ECG set without having access to the first set. Finally, the reader reinterpreted the standard ECGs from 98 children.

RESULTS

The variation of the interpretation of standard versus EASI-derived 12-lead ECGs was only slightly larger than the intrareader variation of the interpretation of standard ECGs.

CONCLUSIONS

For most of the electrocardiographic diagnoses, the conclusions from EASI-derived 12-lead ECGs were similar to those from standard ECGs. These findings support the suggestion that the EASI lead system is a potential alternative to the standard ECG in children.

Comparative accuracy of manual versus computerized electrocardiographic measurement of J-, ST- and T-wave deviations in patients with acute coronary syndrome

Accurate and rapid electrocardiographic interpretation is of crucial importance in acute coronary syndrome (ACS). Computerized electrocardiographic algorithms are often used in out-of-hospital settings. Their accuracy should be carefully validated in ACS, particularly in ST-elevation myocardial infarction. This study evaluated the comparative accuracy of lead-specific computer-based versus manual measurements of the J-point, ST-segment, and T-wave deviations in standard 12-lead electrocardiograms (ECGs) (excluding lead aVR). Sixty-nine consecutive patients with suspected ACS were included. The interobserver reliability in the determination of ST-segment deviation>or=0.2 mV in leads V2 and V3 was very good (kappa=0.94 and 0.93, respectively). Agreement between a cardiologist and the computer regarding ST elevation>or=0.2 mV in lead V2 was moderate (kappa=0.72) and in V3 was very good (kappa=0.85). For ST depression or elevation>or=0.05 mV in lead LIII, agreement was good and moderate (kappa=0.79 and 0.51, respectively). Bland-Altman analysis demonstrated clinically acceptable limits of agreement comparing measurements of the J point and the T wave, but clinically inadequate limits of agreement with respect to ST-segment deviation, between the electrocardiographer and the computer. The optimal cut-off points were 0.115 mV (sensitivity 89%, specificity 98%) for the computer program to detect ST elevation>or=0.2 mV and 0.045 mV (sensitivity 74%, specificity 99%) for revealing ST elevation>or=0.1 mV. It was found that automatically measured ST-segment deviations were smaller than those manually measured. In conclusion, a correction should be performed to obtain optimal results in the automated analysis of ECGs, because the results have important implications for clinical decision making.

Biomarkers in drug discovery and development: from target identification through drug marketing

Biomarkers of disease play an important role in medicine and have begun to assume a greater role in drug discovery and development. The challenge for biomarkers is to allow earlier, more robust drug safety and efficacy measurements. Their role in drug development will continue to grow for the foreseeable future. For biomarkers to assume their rightful role, greater understanding of the mechanism of disease progression and therapeutic intervention is needed. In addition, greater understanding of the requirements for biomarker selection and validation, biomarker assay method validation and application, and clinical endpoint validation and application is needed. Biomarkers need to be taken into account while the therapeutic target is still being identified and the concept is being formulated. Biomarkers need to be incorporated into a continuous cycle that takes what is learned from the discovery and development of one series of biomarkers and translates it into the next series of biomarkers. Optimum biomarker development and application will require a team approach because of the multifaceted nature of biomarker selection, validation, and application, using such techniques as pharmacoepidemiology, pharmacogenetics, pharmacogenomics, and functional proteomics; bioanalytical method development and validation; disease process and therapeutic intervention assessments; and pharmacokinetic/pharmacodynamic modeling and simulation to improve and refine drug development. The potential for biomarkers in medicine and drug development will be limited by the least effective component of the processes. The team approach will minimize the potential for the least effective component to be fatal to the rest of the process. As scientific/regulatory foundations for biomarkers in medicine and drug development begin to be established, successes and applications will need to be effectively communicated with all of the stakeholders, including not only internal and external drug developers and regulators but also the medical community, to ensure that biomarkers are totally integrated into drug discovery and development as well as the practice of medicine.

Standards for the function of an academic 12-lead electrocardiographic core laboratory

An academic 12-lead electrocardiogram (ECG) core laboratory aims to provide the highest possible quality ECG recording, measurement, and storage to aid clinicians in research into important cardiovascular outcomes and to maximize the credibility of scientific results based solely, or in part, on ECG data. This position paper presents a guide for the structure and function of an academic ECG core laboratory. The key functional aspects are: 1) Data collection, 2) Staff composition, 3) Diagnostic measurement and definition standards, 4) Data management, 5) Academic considerations, 6) Economic consideration, and 7) Accreditation. An ECG Core Laboratory has the responsibility for rapid and accurate analysis and responsible management of the electrocardiographic data in multicenter clinical trials. Academic Laboratories, in addition, provide leadership in research protocol generation and production of research manuscripts for submission to the appropriate peer-review journals.

Controlling for chance agreement in the validation of medical expert systems with no gold standard: PNEUMON-IA and RENOIR revisited

In the validation of medical expert systems, agreement among different human specialists on a random sample of cases may be taken as a substitute to a missing gold standard. Distance measures between pairs of experts, extensively described in previous studies, do not take into account the influence of chance-expected agreement. A weighted kappa index, with three different weighting schemes, is proposed as an alternative to be applied in the general situation of N cases assessed by E experts about K possible diagnoses, each of them qualified with one of G ordinal categories. A hierarchical cluster analysis, applied to the kappa matrices generated, allows for the classification of the expert system among clinical specialists, providing a relative assessment of its diagnostic ability. The above methodology is applied to the validation of two medical expert systems, PNEUMON-IA and RENOIR.

Interobserver agreement in the electrocardiographic diagnosis of acute myocardial infarction in patients with left bundle branch block

STUDY OBJECTIVE

To determine the interobserver agreement between cardiologists and emergency physicians in the ECG diagnosis of acute myocardial infarction (AMI) in patients with left bundle branch block (LBBB) using the ECG algorithm previously described by Sgarbossa et al.

METHODS

Using the Sgarbossa ECG algorithm, 4 cardiologists and 4 emergency physicians independently interpreted a test set of 224 ECGs with LBBB, of which 100 ECGs were from patients with an evolving AMI. A subset of 25 ECGs was reinterpreted by each reader to test intraobserver agreement for AMI as well as interobserver agreement for the degree of ST-segment deviation. Agreement rates for AMI were estimated using the kappa statistic. In addition, the sensitivity and specificity for diagnosing AMI were determined for each reader, using the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO I) enzyme criteria for AMI as the gold standard. The study was conducted at 3 university-affiliated medical centers. The test set contained ECGs from 100 patients enrolled in the GUSTO I trial with LBBB on their initial ECG and an evolving AMI confirmed by serum cardiac enzyme changes, and 124 control patients from the Duke Databank for Cardiovascular Disease who had stable, angiographically documented coronary artery disease and LBBB.

RESULTS

There was excellent interobserver agreement (kappa=0.81, 95% confidence interval [CI] 0.80 to 0.83) between cardiologists and emergency physicians for diagnosing AMI. Intraobserver agreement kappa values for AMI diagnosis by cardiologists and emergency physicians were 0.81 (95% CI 0.67 to 0.94) and 0.71 (95% CI 0.54 to 0.89). The median sensitivity for diagnosing AMI by cardiologists and emergency physicians was 73% (range 66% to 80%) versus 67% (range 61% to 75%); median specificity was 98% (range 97% to 99%) versus 99% (range 98% to 99%). Spearman rank correlation coefficients for the degree of ST-segment deviation in all 12 leads was 0.86 (95% CI 0.85 to 0.87) among all readers.

CONCLUSION

There is excellent interobserver agreement between cardiologists and emergency physicians for diagnosing AMI when applying the Sgarbossa ECG algorithm to patients with LBBB. Emergency physicians should be able to reliably use this algorithm when evaluating patients.

[The impact of the clinical data and the interhospital agreement in the interpretation of myocardial perfusion tomography]

INTRODUCTION AND OBJECTIVES

The aim of this study is to analyze the impact of clinical data and the interhospital agreement in the interpretation of myocardial perfusion single photon emission tomography (SPECT) images and polar mapping.

METHODS

150 patients from 5 hospitals were studied. Each center contributed with tomographic images and polar maps of 99mTc-tetrofosmin exercise SPECT and clinical reports of 30 patients. Thus, 300 images (150 of tomographic images and 150 of polar maps) were interpreted by each center without knowledge of clinical data of the patient ("blinded" report).

RESULTS

90 (60%) out of 150 patients had a coronary stenosis > or = 50%. Sensitivity and specificity of "non blinded" report were 91% and 86%, respectively. Sensitivity determined by majority decision (three or more centers) was 82% for tomographic images and 83% for polar maps (p = 0.002 and p = 0.03, respectively, regarding the "non-blinded" report). Specificity was 88% for tomographic images and 79% for polar map (p = 0.05 with respect to tomographic images). Interhospital agreement was good not only for tomographic images (kappa: 0.625) but for polar maps (kappa: 0.7) as well.

CONCLUSIONS

Sensitivity of clinical or "non blinded" report of myocardial perfusion SPECT is significantly higher than the "blinded" report. Specificity of the "blinded" report of polar mapping is lower than that of tomographic images. A good interhospital agreement in interpretation of both types of images was observed.

Observer variation in measured ST-segment elevation

STUDY OBJECTIVE

ST-segment elevation is used to make early decisions about using thrombolytic therapy in patients with suspected myocardial infarction. This study was performed to assess interobserver and intraobserver variation in subjects' measurements of ST-segment elevation in isolated ECG complexes.

METHODS

We performed a masked, paired-sample experiment. Emergency physicians, emergency medicine residents, and senior medical students were asked to measure ST-segment elevations in a packet of 40 isolated single ECG complexes from patients with proven myocardial infarction. Each packet consisted of 20 pairs of randomly ordered ECG complexes. Subjects were masked to the duplicate nature of the pairs of complexes. Estimates of each ST-segment measurement pair were subtracted; summary statistics were calculated for these differences. Agreement between first and second ST-segment estimates was analyzed by using weighted kappa values. Tests for differences in the paired ST-segment estimations among groups were carried out with 1-factor analysis of variance.

RESULTS

Fifty-two subjects completed the study and measured a total of 2,070 ST segments (1,035 pairs). The mean difference in segment height among all groups was 0.28 mm (95% confidence interval [CI], 0.27 to 0.30). The median was 0.2 mm, the 80th percentile was 0.5 mm, and the 95th percentile was 0.9 mm. Statistical agreement between paired ST-segment measurements was very good (kappa=0.85; 95% CI, 0.83 to 0.87). However, by using a threshold of at least 2.0 mm for ST-segment elevation, 143 (14%) of 1,035 paired estimations (95% CI, 12% to 16%) produced inconsistent classifications. Physicians, residents, and students had similar median differences (P =.77). Some subjects exhibited more intraobserver variation than did others, and some complexes were associated with greater intraobserver variation than others (P <.001).

CONCLUSION

One fifth of the time, intraobserver measurements of paired ST-segment elevations differed by more than half a millimeter. Independent interpretations of the same ST segment by the same reader as greater than or equal to 2.0 mm or not were different 14% of the time. This could result in misclassification of candidates for thrombolytic therapy.

Admission standard electrocardiogram for early risk stratification in patients with unstable coronary artery disease not eligible for acute revascularization therapy: a TRIM substudy. ThRombin Inhibition in Myocardial Infarction

OBJECTIVES

The purpose of this study was to determine the prognostic capacity of a single electrocardiogram (ECG) obtained early after admission to the hospital in patients suspected of non-Q-wave myocardial infarction and unstable angina pectoris.

METHODS

Six hundred twenty-nine patients from the TRIM study were included. The patients were divided into subgroups on the basis of ST-segment changes in the inclusion ECG. Death, acute myocardial infarction, or refractory angina (despite treatment) were registered during a follow-up period of 30 days.

RESULTS

Patients with ST depression had a significantly higher event rate compared with patients with other ECG changes or with normal ECG results. The difference in event rates between patients with ST depression and patients without ST depression regarding the composite of death and acute myocardial infarction was highly significant (P =.0008). A significant association between the magnitude of the ST-segment depression (in millimeters) and the risk of cardiac events was also demonstrated. Multivariate analysis proved ST depression early after admission to be an independent predictor of high risk.

CONCLUSION

In patients with unstable coronary artery disease, ST-segment depression at admission is a strong predictor of early (30 days) cardiac events and the extent of ST depression carries important prognostic information as well.

Very early risk stratification using combined ECG and biochemical assessment in patients with unstable coronary artery disease (A thrombin inhibition in myocardial ischemia [TRIM] substudy). The TRIM Study Group

BACKGROUND

The diagnostic capability of troponin T (TnT), troponin I (TnI), myoglobin, and creatine kinase (CK)-MB mass for detection of myocardial injury seems evident. Newer studies have found these sensitive markers to carry independent prognostic information in patients with unstable coronary artery disease as well. ST-segment depression in the admission ECG is known to be an important indicator of poor outcome in these patients. The present study investigates the prognostic capacities of the ECG in combination with biochemical admission measurements in 516 patients admitted to hospital with unstable coronary artery disease.

METHODS AND RESULTS

Baseline ECG recordings and blood samples were collected for central analysis. The patients were followed up for 30 days, and predefined end points, ie, death, myocardial infarction, and refractory angina, were registered as end points. By univariate analysis, ST-segment depression, inverted T waves in >/=5 leads, TnT >/=0.1 microg/L, TnI >/=0.5 microg/L, myoglobin >/=40 microg/L, female sex, and age >/=65 years were predictors of death and myocardial infarction at 30 days. By multivariate analysis, female sex, ST-segment depression at randomization, or inverted T-waves in >/=5 leads were the only independent predictors of death or myocardial infarction. On the basis of baseline ECG ST-T changes and CK-MB mass/TnT/TnI/myoglobin levels, the patients were divided into 3 subgroups at high (14% event rate), intermediate (6%), and low (3%) risk of early death/myocardial infarction.

CONCLUSIONS

The present study found the combination of baseline values of TnT, TnI, CK-MB mass, and ST-T changes in the ECG to be effective for early risk stratification in patients with unstable coronary artery disease.