Journal of the New Zealand Medical Association, 29-June-2012, Vol 125 No 1357
Availability of troponin testing for cardiac patients in New Zealand 2002 to 2011: implications for patient care
Mohammad Latif, Chris Ellis, Alexei Chataline, Greg Gamble, Cam Kyle, Harvey White
The American College of Cardiology (ACC) and the European Society of Cardiology (ESC) have published a joint consensus statement redefining acute myocardial infarction (MI).1 The cornerstone of the new definition of MI is the elevation of cardiac troponin (T or I) concentration in the appropriate clinical context and, due to the range of other causes of troponin elevation, evidence of a rise and/or fall.2,3 Troponins are not just useful for the diagnosis of MI but also have been shown to be prognostically important.4–6
Over the past 2 decades, immunoassays have been developed and refined for cardiac troponins T and I which are increasingly sensitive and specific for cardiac muscle injury. A large number (10–20) of manufacturers have produced assays for troponin I. Only Roche Diagnostics has produced and marketed a troponin T assay, although it has licensed the assay for one point of care analyser marketed by another company (Radiometer AQT90).
Different manufacturers use their own antibodies raised against different antigen epitopes on the cardiac troponin I protein, and also use different calibrator and control materials.7 While there has been progress in standardising troponin I
assays using an international reference preparation, this has not resolved the inherent difficulties in comparing troponin results between different methods for an individual patient. It is very difficult, and may be dangerous, to try to diagnose a rise and/or fall in troponin when different troponin assays and platforms are used.
Use of different methods and different assay platforms could lead to diagnostic confusion and inconsistencies between centres and/or regions, affect treatment, and could confound health statistics. Hence we conducted an audit looking at the various troponin assays (including point of care), providers, analyser types and the local cutpoints used across New Zealand, in hospitals and private providers and reviewed how this has changed over time (2002, 2007 and 2011).
We reviewed the troponin tests, analysers and cutpoints available at hospital laboratories across New Zealand which admitted patients with a probable acute coronary syndrome (ACS) in 2002, 2007 and 2011. These hospitals also participated in the comprehensive Cardiac Society of New Zealand Audit in 2002 and 2007 which represented all hospitals admitting ACS patients in New Zealand.
The development of the New Zealand Acute Coronary Syndrome (NZACS) Audit Group and the methodology for the national audits has been published previously.8–11 A few smaller hospitals, not routinely admitting ACS patients, had troponin testing facilities and these were also included in this survey.
Data were collected from the hospital laboratory staff and/or clinicians by e-mail or telephone in order to establish which assays and analysers were being used, as well as the local cutpoints for a positive test. Community laboratory staff were contacted to provide similar data.
We consulted representatives from a number of major New Zealand troponin testing providers to aid in data collection, to enquire about manufacturer cutpoints, and identify other sites where point of care analysers measuring troponin were being used for patient management.
Distribution of assays used in hospital laboratories—In 2011, the 43 hospitals across New Zealand which admit possible ACS patients used 10 different troponin analysers provided by five companies. In comparison there were 11 analysers in use in 2007 and 8 analysers in 2002 (Table 1). Additionally, the Siemens Centaur assay is in use by a private laboratory (Labtests) in 2011, making a total of 11 different analysers in current use.
Troponin analysers from 2002, 2007 and 2011 are shown in Table 2 by geographical distribution. In 2011, 58% of the hospitals (25/43) use troponin T assays and 42% (18/43) use troponin I assays as their first-line method. Troponin T assays also slightly predominated in 2002 (68%) and 2007 (70%).
In 2011 Ortho and Bayer are no longer present in New Zealand, and there are two new commercial providers: Siemens and Beckman Coulter. Siemens now provide a reformulated troponin test on the Centaur (previously Bayer) platform, in addition to the Dimension and Vista analysers. Radiometer provides a bench top analyser, the AQT90, which is standardised against the Roche Troponin T assays and uses an identical antibody configuration.
Of hospital laboratories measuring troponin I, the Abbott Architect assay is the most widely used in 2011. Abbott provided 78% (14/18) of the assays measuring troponin I as the first-line test, mostly using the Architect assay. Other troponin I assays are also provided in regional North Island locations; Siemens Vista assay at North Shore and Waitakere, and Beckman Coulter DXI and Access platforms in Bay of Plenty and Whakatane.
Table 1. Major providers, number and different types of analysis methods in use at hospitals admitting acute coronary syndrome patients across New Zealand in 2002, 2007 and 2011
* Brackets () denote number of times instrument is also used as backup instrument
** One other method (Siemens Centaur assay) is being used in the Auckland community by Labtests.
Table 2. Analysers and cutpoints across New Zealand in 2002, 2007 and 2011×
Figure 1A shows the geographical distribution of troponin use across the 20 New Zealand district health boards (DHBs) in 2011. Even within the same DHB, different troponin assays and analysers are sometimes used.
Distribution of assays used in private laboratories—There is an even distribution of troponin T and I users among private laboratory providers across New Zealand. Almost half of them, (8/15) use the Roche troponin T assays in their main laboratories, whereas the rest use Abbott (4/15), Beckman Coulter Access (2/15) and Siemens (1/15) troponin I assays.
Table 3 and Figure 1B show the analysers and cutpoints of the private laboratory providers across New Zealand. Note that some peripheral hospitals are also run by private providers, and may have assays that differ from the main laboratory. For example, Southern Community Laboratories (SCL) runs the service for Queenstown, Oamaru and Clyde-Dunstan, all of which have Abbott i-stat instruments, whereas their main Dunedin and Christchurch labs use Roche Troponin T.
Introduction of Roche high sensitivity (5th generation) assay—Since the introduction of the 5th generation high sensitivity assay (hs Troponin T), from May 2010 onwards there has been a rapid uptake among existing Roche users, initially in the lower North Island, then in Auckland/Northland and later in the South Island. By February 2011 all Roche laboratory analyser platforms had switched to the hs Troponin T test.
Despite the readiness to accept this improved assay, the changeover was delayed in some places through uncertainty among clinical staff on the diagnostic and workflow implications of the new test, and there were concerns over a possible significant increase in “abnormal” results and therefore emergency department presentations, before diagnostic and management algorithms were established.
Both 4th and 5th generation tests were typically available for a short time to enable clinicians to become more familiar with the more sensitive 5th generation test during the changeover. In some regions, such as the upper North Island, the introduction was coordinated between laboratories to occur at the same time.
Figure 1A. Distribution of troponin T, troponin I and mixed assays by District Health Board Hospital laboratories as at 2011 (Chatham Islands [troponin T] not shown)
Figure 1B. Distribution of troponin T and troponin I assays used by community laboratories as at 2011
Table 3. Main testing platforms of private community laboratories: 2007 and 2011
Laboratory reporting practices—With the introduction of the Roche hs Troponin T method, the reporting units have changed from micrograms per litre (ug/L) to nanograms per litre (ng/L), to reflect greater sensitivity and minimise confusion. However, troponin I tests are still reported in ug/L by most laboratories in New Zealand.
The exception is Auckland where all four laboratories measuring troponin I (Middlemore, North Shore, Waitakere, Labtests) report results in ng/L, effectively by multiplying the analyser result by a thousand using their laboratory information system.
The number of different ‘cutpoints’ used by different NZ hospitals in 2002, 2007 and 2011 was reviewed. Table 4 shows the recommended manufacturer and laboratory cutpoints at these times. During these periods there were differences in quoted cutpoints for major assays in use. For example in 2007 three cutpoints (0.03, 0.04 and 0.15ug/L) were reported for the Architect platform. There were also three cutpoints for the Roche Troponin T assay (0.03, 0.04 and 0.01ug/L).
By February 2011, all laboratories using Roche hs Troponin T had the same assay cutpoint, with all but one reporting <14ng/L with their report (Kaitaia reported 0-13 ng/L as their reference interval). Importantly, however, a different cutpoint is still currently used for troponin T point of care instruments (0.03ug/L).
Table 4. Quoted cutpoint levels for New Zealand laboratories and recommended manufacturer cutpoints 2002, 2007 and 2011×
Troponin I assays also have different cutpoints in different laboratories. For example, for the Abbott i-stat machine four different cutpoints are quoted (0.03, <0.04, 0.08 and <0.09ug/L).
Point of care testing—Point of care testing (POCT) is widely used in a number of rural hospitals both for initial patient management and also as backup when laboratories cannot be viably staffed for a full 24 hours. Table 1 includes laboratories where POCT analysers are available after hours. The Abbott i-stat machine is a common choice with 7 rural labs using this instrument for routine first-line testing and 5 using it as a backup.
POCT is also used in some general practice settings where obtaining a laboratory result urgently is otherwise impractical, especially after hours, e.g. Waiheke Island (Cardiac reader), and in rural practice settings, e.g. Hawera, Waipukurau (i-stat) (data not shown).
Results reported by POCT are less precise, and therefore less sensitive, than laboratory analysers. Therefore laboratory reports using POCT analysers have a comment emphasising that the test is helpful to diagnose myocardial damage if positive, but a negative result does not exclude it, i.e. a positive test is useful to ‘rule in’ myocardial injury in the right clinical setting, but a negative result also does not rule it out.
POCT methods using both Troponin T (Cardiac reader, Radiometer AQT) and troponin I (Abbott i-stat) currently continue to be expressed in ug/L, although the Roche POCT analysers will be calibrated in ng/L in the near future.
The quoted sensitivity cutpoints for POCT analysers measuring troponin T (Roche cardiac reader/H232, Radiometer AQT) are similar at 0.03ug/L, although the Cardiac reader provides a semi-quantitative (‘low’) result in the lower abnormal range (0.03-0.1ug/L), while the AQT instrument provides a numerical result. For the i-stat machine, the manufacturer (Abbott Diagnostics) quotes two different cutpoints, either 0.08ug/L corresponding to the 99th population percentile, or 0.03ug/L, representing the 97.5th population percentile.
The individual choice of POCT instrument depends on a range of factors, including clinical preference, cost (of analyser, reagents and staff time/training) and alignment with other laboratory needs such as the ability of the POCT analyser to measure other analytes besides troponin. For example, the Abbott i-stat, Roche H232 analyser and Radiometer
AQT instruments can measure other tests, making them useful options in an emergency department or rural/community setting. Thus the i-stat is used in Ashburton and Buller, while in the Waikato the Radiometer AQT is used as the first-line method in Tokoroa and Te Kuiti, and as a backup in Thames and Taumarunui.
In some centres (e.g. Rawene), the wide menu of other tests offered by the i-stat means that this instrument is used to measure troponin I, even though the referral hospitals (Whangarei, Auckland) measure hs Troponin T.
Troponin measurement is pivotal in the diagnosis, risk stratification and management of ACS. The definition of MI1 requires a rise and/or fall in cardiac troponin concentration. Strictly, this means measurement by the same method, in the absence of close correlation between different assays. The magnitude of troponin elevation correlates with risk of death or non-fatal MI following ACS.12–14 Patients with elevated troponin levels also benefit more from antithrombotic therapy, glycoprotein 2b/3a receptor antagonists and revascularisation than patients with normal troponin levels.15–17
Troponin cutpoints—In 2000 the ESC/ACC Consensus group recommended that the appropriate cutpoint for troponin levels for diagnosis of MI should be the 99th percentile of a healthy population.18 They also stated that the assay reproducibility (the imprecision measured as coefficient of variation) should be 10% or less at that concentration. The International Federation of Clinical Chemistry (IFCC) Committee on Standardisation of Markers of Cardiac Damage (C-SMCD) also supports these guidelines.19
Since 2000 there have been major advances in assays. With improved assay precision and sensitivity, recommended cutpoints have changed progressively from CKMB-derived thresholds (using the now outdated WHO definition of MI), to those based on 10% assay precision, to the 99th population percentile. Until recently, no assay in clinical use had sufficient sensitivity to satisfy this 99th percentile cutpoint. The Roche hs Troponin T and Siemens ‘Ultra’ now have this precision.
Rather than provide a single ‘cutpoint’, it is widespread practice for manufacturers to provide a summary of their own ‘in-house’ data (usually validated by independent published studies), showing the 10% CV and 99th population percentiles for their assay. There is routinely a recommendation that each laboratory validate a reference range for its own population, which is a difficult task, particularly for small rural laboratories with staffing and cost-constraints.
These reporting difficulties are also compounded as laboratories generally report ‘rounded’ results generated by their analysers to two decimal places (e.g. 0.032 becomes 0.03ug/L, while 0.037 becomes 0.04ug/L). Some laboratories also reported a reference range with an upper normal limit, while others report a ‘less than’ number (e.g. <0.04ug/L).
Some manufacturers still quote a threshold that is “unequivocal” or “diagnostic” of MI, referring back to comparisons with previous CKMB results. For example Abbott uses an ‘unequivocal’ threshold of 0.3ug/L and Beckman a threshold of 0.5ug/L. Both are about 10 times higher than ranges routinely quoted by NZ laboratories, although some still refer to this “unequivocal” threshold with a comment in their reports. While hard to quantify, anecdotal evidence suggests local clinician preferences played a role in some cases in influencing laboratory cutpoints.
Interpretation and standardisation of troponin results—Despite significant assay improvements, there are persisting problems with interpretation and standardisation of serum troponin measurements. This is more so for cardiac troponin I as different manufacturers use different antibodies raised against different epitopes on the cardiac troponin I protein.21
These different antibodies vary in their ability to bind troponin I. It is also important to note that troponin I can exist in multiple forms in the serum:
There are potentially several additional forms that also exist for each of these three forms, representing N- and C-terminal degradation products, oxidised and reduced forms, and phosphorylated forms.22
Different assays do not react on an equimolar basis with these different molecular species. Therefore, different assays do not produce equivalent concentration results and comparisons of absolute troponin I concentrations cannot be reliably made.23
There have been considerable efforts towards international standardisation of troponin I assays.7 However, troponin I results can still differ by at least several-fold, sufficient to make direct comparisons of results between assays unhelpful and potentially dangerous when serial samples are taken in an ACS setting.24,25
By contrast assay standardisation issues are minimal with troponin T because the same antibodies are used—even when the assay has been released for use on another manufacturer’s platform (e.g. Radiometer AQT).
Given the heterogeneity of troponin assays, Apple26 has proposed a cardiac troponin assay scorecard. This lists the various assays according to the total imprecision at the 99th percentile and whether the assays are “clinically usable” or “guideline acceptable”.
This scorecard may be useful in providing some guidance to clinicians and laboratories regarding the strengths and weakness of various assays. He suggests only two assays are guideline compliant (Siemens 'Ultra' troponin I and Roche hs Troponin T) and, of the two assays available in New Zealand only one, hs Troponin T, is classified as high sensitivity. However, a recent comparison of several 'sensitive' troponin assays suggested comparable performance in the diagnosis of MI in patients presenting acutely with ACS but superiority to the 4th generation troponin T assay.27
This study included three of the assays currently in use in New Zealand, the Abbott Architect troponin I, Roche hs Troponin T, and Siemens troponin I Ultra assays. Receiver operator curve (ROC) analysis showed high diagnostic accuracy with identical areas under the curve of 0.96 for these three assays. A separate recent comparison study also indicated that the Beckman Coulter assay had excellent diagnostic sensitivity.28
Use of these high sensitivity troponin tests allows for earlier diagnosis, or exclusion of MI. The recent update of the Cardiac Society of Australia and New Zealand ACS Management Guidelines suggest that clinicians now take specimens 3 hours apart, with 1 at least 6 hours after symptom onset.29
Troponins and clinical management—Although laboratories have provided education to users as well as interpretive comments with their results, clinicians have taken time to become familiar with the newer assays and the implications of elevated levels, especially in the low range.30 Assay 99th population percentiles quoted by manufacturers are based on a ‘healthy’ (typically relatively young) population. However, many elderly patients have small degrees of troponin elevation above this ‘healthy’ cutpoint for multiple reasons. This can cause diagnostic difficulty if symptoms are non-specific and accompanied by “mild elevation” on a first test.
Diagnosis in such patients requires follow-up tests performed using the same assay platform to confirm or exclude acute myocardial injury. In New Zealand, based on biological and analytical variability considerations, a hs Troponin T initial level of 14-49 ng/L and a change of 50% or more in serial blood tests is currently required for the diagnosis of myocardial infarction in the appropriate clinical setting. If the initial hs Troponin T level is ≥50 ng/L (at some sites 53 ng/L), a change of 20% or more is required for diagnostic purposes.31
In some referral networks there has been a trend towards similar methodology as instrument platforms are updated. The Northland regional centres (Kaitaia, Kawakawa, Dargaville and Whangarei) which routinely send their patients to Auckland City Hospital have since 2007 changed their assays to Roche troponin T which is also being used at LabPlus in Auckland.
Peripheral and regional hospitals in Waikato and the far North also use troponin T, similar to their major referral hospital. Conversely, in Whakatane hospital troponin testing has moved from troponin T to Beckman Coulter troponin I, also using the same test performed in the local community.
However, despite uniformity in some areas, in a number of regions different troponin assays and analysers are being used so that patients frequently have troponin measurements by more than one (even several) methods. Since results from different methods cannot be directly compared this has implications for ACS patients transferred from one centre to another.
Extra costs are incurred in repeating troponin tests and hospital stay may be prolonged while repeat tests are performed. The same applies to patients who have had troponin tests done in the community or by a private laboratory and are subsequently referred to the public hospital. Table 5 illustrates the potential transfer pathways of cardiac patients across New Zealand and the troponin assays in use among different centres.
Changes in troponin testing are related to multiple other factors besides the choice of the troponin assay itself. In the past 10 years national and regional district health board initiatives with tendering for bulk-funded laboratory services have resulted in significant consolidation of laboratories in New Zealand, but with little coordination across district health boards as each has pursued individual solutions for laboratory services.
Troponin tests in some provincial hospitals are routinely performed by their local private laboratories (e.g. Palmerston North, Whakatane), as part of a contract to run the hospital laboratory. Conversely some private laboratories in the community send their troponin requests on to their local public hospital (e.g. Taranaki). Further, there has been little attention to the impact of different troponin tests on patient management across New Zealand's 5 major Cardiothoracic Regions which collaborate to manage ACS patients who require invasive management at the Regional Centre9, 11.
For cost and efficiency reasons the choice of assay has often been determined not just by clinician and pathologist preference, but by what is considered to be the most cost-effective overall analyser configuration, due to the need to purchase large and expensive analyser platforms that perform many different tests.
Table 5. Examples of transfer pathways for cardiac patients across New Zealand and various troponin assays used 2011
R=Roche, Ra= Radiometer, A=Abbott, BC=Beckman Coulter, S = Siemens
I=Troponin I, T=Troponin T
^ Usually a community laboratory or peripheral hospital. In these sites the same laboratory usually handles both community and hospital work. In some cases patients are referred directly from community labs or peripheral hospitals to referral hospitals. In others they may be referred to regional hospitals, and then sent on to referral hospitals for intervention or further evaluation.
* In Hawke's Bay some GP practices get troponin T (Roche) through Southern Community Laboratories (SCL), while others get Abbott troponin I (Hawke's Bay DHB laboratory). After hours all GPs get troponin I through the hospital laboratory.
The difficulties of providing a 24 hour service for this important test have also led to individual solutions to provide backup testing. The constraints and cost-implications of doing this have not always been well understood by funders. Point of care testing (POCT) is a common solution, depending on the staffing needs of the hospital, the range of other platforms available, and distance from other laboratories able to also measure troponin.
Clinical trials are increasingly using troponin values as part of their enrolment as well as outcome criteria. Use of different assays could lead to diagnostic inconsistencies between different centres and/or regions, potentially confounding trial results. This also can affect the accuracy of community wide disease rates and compromise the ability to apply trial results to the general population.
Mechanisms of troponin release—Our understanding of the underlying mechanisms of troponin release has also increased. Very small amounts of troponin may be released by means other than gross myocyte destruction (infarction). Recently White20 has suggested six possible major pathobiologic mechanisms for troponin elevations. Troponin elevations could be due to myocyte necrosis, apoptosis, normal myocyte cell turnover, cellular release of proteolytic troponin degradation products, increased cellular wall permeability or passage of membranous blebs.
Study limitations—A potential limitation of this study is that the data regarding ‘cut points’ was obtained via personal communication, and repeated phone and email questioning of laboratory and clinical staff. Much of the data obtained were also retrospective from up to 10 years ago.
The implications of varied laboratory testing platforms, in terms of the cost of retesting troponin results on different platforms for transferred patients, would require further study. In addition, the risk to a patient where a cut point has been inappropriately applied, and the frequency with which errors of interpretation are related to different platforms, would also require further exploration.
Troponin assays have improved throughout New Zealand in the last 10 years reflecting international improvements in assays and clinical practice. However, there remain differences in the troponin tests and reporting practices across New Zealand hospitals.
Point of care troponin tests are widely used in smaller regional hospitals and in some rural and community settings. Care is required in interpretation, especially when comparing with follow-up by a more sensitive test.
The use of different assay platforms, units and sometimes different cutpoints for the same assay can result in difficulties in interpretation of changes when referring patients from the community or peripheral hospitals to tertiary referral hospitals. This is a particular problem when diagnosis requires clinical interpretation based on a change in troponin results.
Clinicians and District Health Board management need to be aware of these issues. Physicians and pathologists need to collaborate in establishing acceptable and uniform criteria for troponin assays and their use. Some standardisation of available tests, thresholds and testing protocols is likely to improve patient management and the accuracy of community wide disease rates.
A more coordinated national approach may result in better use of medical resources and improve patient care.
Competing interests: None declared.
Author information: Mohammad Latif, Cardiology Registrar, Green Lane Cardiovascular (CVS) Service, Auckland City Hospital, Auckland; Chris Ellis, Cardiologist , Green Lane CVS Service, Auckland City Hospital, Auckland; Alexei Chataline, Medical Officer, Green Lane CVS Service, Auckland City Hospital, Auckland; Greg Gamble, Statistician, University of Auckland, Auckland; Cam Kyle, Chemical Pathologist, LabPlus and Diagnostic Medlab; Harvey White, Cardiologist and Director of the Coronary Care Unit & Green Lane CVS Research Unit, Green Lane CVS Service, Auckland City Hospital, Auckland
Acknowledgements: We acknowledge and thank all of the local laboratory staff and clinicians at the participating centres for helping us with collaboration of this data; the many laboratory company representatives for their assistance in this study; and all of the doctors and nurses who contributed to the New Zealand Cardiac Society Acute Coronary Syndrome Audits for 2002 and 20078-11 which facilitated the collection of the initial data for this study.
Correspondence: Dr Chris Ellis, Cardiology Department, Green Lane CVS Service, Level 3, Auckland City Hospital, Grafton, Auckland 1023, New Zealand. Email: email@example.com
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals