Key result
Point-of-care and contemporary troponin assays demonstrated poor analytical concordance compared with a high-sensitivity assay at the 99th percentile, with false-negative rates up to 33%.
Answers are grounded in this paper and cite their sources.
Analytical concordance between POC/contemporary troponin assays and hs-cTn assays is poor at the 99th percentile, emphasizing the need for assay-specific cutoffs and caution against interchangeable interpretation.
Cardiac troponins (cTn)3 I and T are pervasively used to rule in or rule out acute myocardial infarction and/or assess for myocardial injury, with an abnormal result defined as a troponin concentration greater than the assay-specific 99th percentile (1, 2). High-sensitivity cTn assays (hs-cTnI, hs-cTnT) are being implemented globally and offer refined precision at low concentrations, which allow for enhanced clinical and analytical sensitivity and use of rapid rule-out strategies and algorithms. However, although hs-cTn assays continue to improve from an analytical perspective, the clinical performance gap continues to widen between automated/central laboratory hs-cTn assays vs point-of-care (POC) cTn assays. POC cTn testing has demonstrated lower clinical sensitivity with a positive predictive value around 50% (3), thereby resulting in a greater number of patients being undiagnosed, inadequately treated, or discharged in the acute care setting (4, 5). Nonetheless, POC cTn testing may be used in situations when turnaround time requirements cannot be met with central laboratory assays. Inappropriate utilization of POC cTn testing frequently occurs when POC cTn is ordered simultaneously or interpreted serially with automated cTn/hs-cTn assays, which is particularly problematic because of the lack of harmonization and standardization between cTn assays and analytical platforms (6). In this issue of JALM, Tsui et al. (7) explore the potential use of a single harmonized cutoff using POC cTnI assays and central laboratory contemporary cTnI assay vs a comparator hs-cTnI assay. The study evaluated the analytical concordance at various cTn cutoffs using plasma and whole blood specimens collected from a cohort of 102 patients admitted to a single coronary care unit. An acceptable analytical concordance of >90% between the reference hs-cTnI method and the test method was defined. Three notable findings from this work are apparent: (a) The analytical concordance at the recommended 99th percentile for contemporary and POC assays compared with the hs-cTnI assay was poor, specifically with regard to analytical sensitivity, for which the false-negative rates were 19% (Alere Triage Troponin I), 33% (Radiometer AQT90 and Abbott i-STAT), 7% (Beckman Coulter DxI AccuTnI+3 heparin), and 20% (Beckman Coulter DxI AccuTnI+3 EDTA); (b) the 99th percentiles derived from the manufacturers' package inserts are suboptimal; (c) the false-positive rate for the contemporary Beckman Coulter cTnI assay was unusually high for both lithium heparin (up to 16%) and EDTA plasma (up to 12%) and greater than any of the POC cTn assays. The authors demonstrated improvement in analytical concordance when cutoffs other than the 99th percentile were used, particularly for the Abbott i-STAT, in which the true-positive and false-negative rates at the 99th percentile (0.08 μg/L) were 67% and 33%, respectively, but at a cutoff of 0.03 μg/L, the true-positive and false-negative rates were 98% and 3%, respectively (7). Although analytical concordance is one factor to evaluate when comparing results between different laboratory assays, for troponin the value does not lie solely with analytical concordance but is also highly dependent upon the clinical performance between assays. The Fourth Universal Definition of Myocardial Infarction continues to emphasize the requirement for serial sampling protocols, with myocardial injury defined as an increased cTn above the 99th percentile (1). Serial samples define whether the injury is acute (rise and/or fall of cTn) or chronic (persistent elevations of cTn above the 99th percentile), and the same cTn assay must be used due to lack of assay harmonization or standardization (8). Although analytical concordance studies provide insight as to the cTn concentrations one might encounter when implementing or comparing assays across different patient populations, clinical studies with outcomes are needed to truly assess the implications to patient care if a variety of troponin assays and/or cutoffs are used or going to be used within a healthcare system. Debate and confusion around the troponin 99th percentile still exists, and most manufacturers' package inserts do not include relevant information about the reference interval cohort, inclusion/exclusion criteria, or the statistical approach used to establish the assay and platform-specific 99th percentile. However, it is clear from guidelines and consensus documents that the 99th percentile should be derived from a normal healthy population—even though the intended-use population differs, i.e., emergency department or hospitalized patients. Inclusion of older healthy individuals and those with chronic conditions (e.g., diabetes, anemia, borderline chronic kidney disease, asymptomatic heart failure) in a cTn reference interval study can significantly skew calculation of the 99th percentile, which in turn negates the sensitivity of hs-cTn assays (9, 10). Furthermore, it is well documented that any detectable troponin, particularly above the 99th percentile, is predictive of adverse events, and the magnitude of the elevation correlates with cardiovascular morbidity and mortality (11–13). Therefore, manufacturers should strive to yield consistent data for their reference interval studies and remain transparent about how the cutoffs were derived. For laboratories unable to meet turnaround time goals and clinical expectations, POC troponin assays provide a mechanism to achieve appropriate patient care needs. As laboratory professionals, we understand that rapid testing cannot be achieved at the expense of diagnostic performance, and that using the same cTn cutoff between assays does not imply equality. However, there are other external pressures that attempt to imply that cTn assays can be interpreted interchangeably or clinical results from different cTn assays can be conjointly reported in the electronic medical record—both notions that are false. There should be visually discernable differences between use of POC/contemporary and hs-cTn assays, owing to the recommendations on reporting units between the (μg/L vs ng/L, respectively) (14, 15). Until hs-cTn POC assays are developed and rigorously validated, comparing the analytical and clinical performance of POC cTn testing will continue to weaken, particularly when compared with hs-cTn. In addition, we must be cautious not to use hs-cTn 99th percentiles or abnormal results that simply mirror the performance of a contemporary or POC cTn test; to do so would obfuscate the improved sensitivity of hs-cTn assays and mitigate any benefit to patients in diagnosis of acute myocardial infarction and myocardial injury. The findings of Tsui et al. highlight an important issue to reconcile when assessing the analytical performance of troponin assays and limitations of simply adopting the manufacturers' 99th percentile troponin reference interval. cardiac troponin high sensitivity point of care.
No takes yet. Share an insight, caveat, or question.
Amy K. Saenger (2018) conducted an editorial in Coronary care unit admission (n=102). Point-of-care and contemporary cTnI assays vs. High-sensitivity cTnI (hs-cTnI) assay was evaluated on Analytical concordance >90% between reference hs-cTnI method and test method. Point-of-care and contemporary troponin assays demonstrated poor analytical concordance compared with a high-sensitivity assay at the 99th percentile, with false-negative rates up to 33%.
Synapse has enriched 10 related papers. Consider them for comparative context: