QJM Advance Access originally published online on April 8, 2005
QJM 2005 98(5):337-342; doi:10.1093/qjmed/hci052
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Usage of troponin in the real world: a lesson for the introduction of biochemical assays
From the Departments of 1Cardiology and 2Clinical Chemistry, Charing Cross Hospital, London, UK
Address correspondence to Dr K. Rajappan, Department of Cardiology, Charing Cross Hospital, Fulham Palace Road, London W6. e-mail: krajappan{at}hhnt.org
Received 18 August 2004 and in revised form 7 March 2005
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Studies have demonstrated economic and clinical effectiveness using troponin as a risk stratification tool in chest pain patients. Those with a positive result are investigated invasively, whilst those with a negative result and ECG are promptly mobilized, facilitating discharge.
Aim: To determine whether our use of troponin I (cTnI) in routine clinical practice conforms to ideal standards.
Design: Audit study.
Methods: Data were collected from 93 laboratory request forms for cTnI measurement on 72 patients with matched available patient records.
Results: Eighty requests had no information regarding timing of blood sample in relation to the clinical event; 39% gave no clinical indication. Only 71% of results were available within 12 h. An admission diagnosis of acute coronary syndrome (ACS) was made in 25%. Fifteen had typical cardiac chest pain with a negative cTnI: 6 of these had an exercise treadmill test before discharge. Nine had a positive cTnI, but only two had coronary angiography. Of patients with negative cTnI and possible ACS, 84% were in hospital for >4 days.
Discussion: The introduction of troponin assays into widespread use requires careful assessment. cTnI requests and subsequent patient management remain below expected standards. Ideally, the laboratory should provide an accurate result within a reasonable time frame, while physicians need to request cTnI at a suitable time-point and use the result appropriately. Lessons from the introduction of cTnI measurement may be useful for the introduction of future new tests in other areas of cardiology and medicine.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patients presenting with chest pain account for 20–30% of all acute medical admissions via the accident and emergency department.1 Of these, 50% are potentially cardiac in nature, and the immediate need is to identify those that are suitable for thrombolysis or primary angioplasty. This is done through rapid assessment of the 12-lead electrocardiogram (ECG) in the context of a history of chest pain. However, if the patient does not require thrombolysis or angioplasty, it is necessary to further investigate whether the cause of the pain is cardiac ischaemia. These patients are normally admitted to hospital whilst further tests are performed and perhaps more importantly, the risk of cardiac events in the imminent future is assessed. Measurement of cardiac troponins (especially troponin T) has become established as a strategy for the identification of high-risk patients,2,3 providing a means of determining risk of subsequent events, at least within the subsequent 42 days.4 Troponin assays, both bedside and laboratory based, are therefore seen as useful tools in expediting the discharge of patients with chest pain, and are now in widespread use. Published guidelines5 now include troponin measurement within care pathways. The local audit standard used for this study is based on these guidelines (Figure 1). However, in almost all patients studied within randomized controlled trials, blood samples for troponin measurement were appropriately requested and correctly timed, results were promptly reported back to the clinical team, and the result was then acted upon appropriately as per clinical guidelines.6,7 We audited our hospital practice to see if the use of cTnI in routine clinical situations conformed to these standards, specifically with regard to cTnI requests and subsequent patient management.
|
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Data relating to 104 consecutive cTnI requests over two separate weeks at our hospital were collected. CTnI was measured by automated immunoassay (Abbott AxSym). This assay has been in routine use at our institution since March 2002. cTnI samples were batched and run twice daily (1000 and 1600 h), as the assay reagents were not sufficiently stable to allow continuous access of samples. Local hospital guidelines state the information required on request forms and the recommended timing of blood samples in relation to the suspected cardiac event (12 h after an event or 12 h after hospital admission). A positive cTnI result is >0.5 µg/l.
Data were collected from request forms (requesting department; clinical indication for cTnI; and timing of specimen in relation to clinical event), and from patient medical records (clinical indication for cTnI; onset/duration of symptoms; admission diagnosis; history of smoking, diabetes mellitus, hypertension, hyperlipidaemia, and family history of ischaemic heart disease (IHD); and potentially ischaemic features on the 12-lead ECG). Results of cardiac investigations including exercise treadmill tests (ETT), dobutamine stress echocardiography (DSE), perfusion scans and coronary angiography were recorded. The results below include 93 cTnI requests (72 patients), representing 89% of the total sample, for which complete information was available. Eleven requests on nine patients had no medical records available at the time of the audit.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Requests
Requests for cTnI measurement were received from the following departments: cardiology (25%), other general medical specialties (25%), renal medicine (19%), care of the elderly medicine (13%), general surgery (10%), vascular surgery (5%), accident and emergency (2%), and oncology (1%).
Clinical chemistry
Of the 93 cTnI requests, 16 results were >0.5 µg/l. Eighty request forms (86%) had no information regarding timing of the blood sample in relation to the clinical event, and 39% had no clinical indication given for the cTnI request. When given, ‘chest pain’ and ‘possible IHD’ were the most common indications (33%). Other indications are shown in Table 1. Where an indication was not given, it was possible to determine from their medical records that this was likely to be cardiac in up to 52%. The onset of symptoms could be determined from the patient medical records in 46/93 requests (49%), but in only 17/93 (18%) could the duration of the symptoms be ascertained.
|
Figure 2 shows the number of cTnI measurements available from the laboratory within various time frames. Results for 71% of requests were available within 12 h. In four cases, the result of the assay was not available within 24 h after the sample was sent. Mean turnaround time was 10 h 5 min ± 9 h 45 min.
|
Clinical management
In 12 patients, two cTnI requests were made during the same admission, and in four patients, three were made. All patients who had cTnI measured on more than one occasion had the same result on every sample. Admission diagnosis of an acute coronary syndrome (ACS) was made in 18/72 patients (25%). The remaining diagnoses varied widely from other cardiac causes, such as aortic stenosis and pericarditis, to less obviously cardiac-related diagnoses, such as urinary tract infection and whiplash. The 12-lead ECG showed potentially cardiac ischaemic changes in 34/72 patients (47%). In two patients, no 12-lead ECG or report of its findings could be found.
In 15 patients, the clinical notes suggested typical cardiac chest pain associated with a normal cTnI. In nine of these, the admission diagnosis was an acute coronary syndrome, and was associated with a potentially ischaemic ECG. Six of these 15 patients had an ETT performed, of whom three had coronary angiography before discharge. Three other patients went on to have coronary angiography prior to discharge, without any functional testing.
A positive cTnI result was obtained in nine patients. In those patients with a positive cTnI, the request form was similarly incomplete with regards to clinical information and timing of the clinical events. Ischaemic ECG changes were present in five patients, no ETTs were performed, and one patient had a DSE. Only two patients had coronary angiography performed. Of the remaining cTnI-positive patients, one declined angiography, and three were not referred to a cardiologist during their hospital stay.
Length of stay was 10 ± 27 days (median ± SD) for all patients. For the patients with positive cTnI results, length of stay was 12 ± 21 days. For those with a negative cTnI, length of stay was 10 ± 28 days (p = 0.8 vs. patients with positive cTnI result). These data include a large number of patients with a non-cardiac diagnosis, and are skewed by long-stay patients. For the 19 patients with an admission diagnosis of acute coronary syndrome and/or typical cardiac chest pain and negative cTnI results, length of stay was 8 ± 40 days (Figure 3). Some 84% of our patients stayed in hospital for longer than the average length of stay (3.6 days) for a similar patient population in the literature.8
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The widespread introduction of troponin measurement to our trust was supported for clear clinical and financial reasons based on the available published data. However, our audit suggests that this has not demonstrated any advantage to patients or the NHS in terms of selecting patients for further cardiac investigation and/or reducing length of hospital stay. Several reasons outlined below may account for these findings, all of which are likely to apply to many UK institutions.
With any laboratory assay, the patient population being sampled will affect the predictive value of the test.9 It was often impossible to determine the appropriateness of requesting of the cTnI from the clinical information provided, either on the laboratory request forms or in the patients' medical records. Encouraging the physician to provide more clinical data could be done in several ways: (i) by highlighting the result given as being unaccompanied by sufficient clinical data; (ii) by requiring physicians who have provided incomplete forms to contact the department for the result; or (iii) by refusing to process samples with incomplete clinical information on the request form. All of these options have advantages and disadvantages, some likely to be more effective but having considerable time and governance implications. Reviewing the patients' medical records revealed that for the most part the indication for requests pertained to cardiac disease. There were still some requests that cannot be explained. The presence of positive results in patients in whom the test was inappropriate in the first instance may create several problems: reducing specificity, adding expense, increasing patient anxiety, and potentially resulting in damage to patient health (if they have a complication from prolonged stays, treatments or further tests).
The predictive value of troponin assays is also affected by timing.10 In this instance, it is not possible from our data to be certain that our cohort were sampled at an appropriate time (i.e. 12 h after onset of symptoms or after presentation if symptom onset is uncertain) and this is again because of incomplete recording of information either on the request form or in the patients' notes. This is not to say that the timing was incorrect in all of these tests, but without this information, repeated samples are more likely to be required, further delaying appropriate management and discharge.
The prompt return of results is essential if they are to speed up and improve clinical decision-making. With the turnaround times of our audit, even a patient presenting in the early hours of the morning would not have their assay result available that working day. If the result is negative, further testing (e.g. ETT) could then only be performed the following day, and discharge would be delayed. A simple phone survey of three institutions that refer patients for invasive cardiac procedures to our Trust suggested that none of these were achieving turnaround times for troponin assays of <1 h.
With regard to clinical management, it has been clearly shown in patients with ACS that several factors are associated with an increased risk of future events. These include: (i) severity of symptoms—increased risk with new-onset, accelerating, rest or prolonged angina; (ii) symptoms that fail to settle on medical therapy; (iii) previous MI, PTCA, CABG; (iv) age and co-morbid diseases, e.g. renal failure; (v) risk factors—diabetes mellitus, hypertension, hyperlipidaemia; (vi) clinical manifestations—especially haemodynamic disturbance during pain; (vii) accompanying signs and symptoms—new mitral regurgitation, pulmonary oedema, hypotension, third heart sound; (viii) abnormal 12-lead ECG; and (ix) abnormal CK-MB, troponin I, troponin T, C-reactive protein.11
Using a scoring system that includes the troponin level, enables risk of death, myocardial infarction, recurrent ischaemia, or need for revascularization in the following 2 weeks to be assessed.12 Within the framework of these assessments, the subsequent management of these patients remains an area of interest, especially with the economic implications of prolonged hospital stays. Many studies have now demonstrated the benefit of early revascularization in patients with ACS, focusing on the need in those with a positive troponin.6,7 The audit performed suggests that this is not being achieved. Whether this was for valid reasons (e.g. the significant comorbidity associated with a revascularization procedure) is not clear from our data, but it may be that this decision process never occurred in several patients. Furthermore, it must be emphasized that some form of risk stratification should be performed before discharge in a patient with suspected ACS with any of the high-risk features mentioned previously, irrespective of troponin value, as a negative troponin result only indicates lower risk. Most commonly this would be an ETT, but again the data suggest that this was not the practice of those audited. It is more likely that these patients were discharged from hospital after a negative troponin.
It seems that although troponin assays have reduced length of hospital stay (albeit non-significantly in our audit), it is likely they are not improving outcomes and costs, as would have been projected from the multicentre randomized controlled trials. Similar findings have been shown in assessing other interventions such as door-to-needle times for thrombolysis and the use of ACE inhibitors in secondary prevention, where performance initially falls below expectation,13 but may be improved substantially after appropriate audit.14 Many of the reasons for this could equally apply to other recently introduced assays (e.g. B-natriuretic peptide in heart failure15 or D-dimers in thromboembolic disease),16 and suggest there is much room for improvement. Education as always has a key role to play in the successful introduction of an assay. The biochemist (‘provider’) needs to appreciate that if this assay is to provide maximal benefit in management of these patients, it is not only the result that is important, but also the time in which this is provided. Equally, the physician (‘requester’) needs to provide sufficient information and make appropriate requests, while using the result in a manner in which it has been proven to be valid, not simply as an economic tool. It is then essential that assessment be performed on a regular basis (including audits such as our own) to ensure that standards that are set out at initiation of the assay are maintained. Ongoing education will be necessary.
One proposed alternative to the system used in this audit is near-patient testing with a bedside assay (kept in the coronary care unit for example). Quicker results, direct control of assay requesting, and better control on outcomes are all advantages of such a system, however these must be weighed against the probable lack of adequate quality control of the assay itself, the lack of a central record of the result for future referral, and the inevitable issue of funding resource allocation.
This study has certain limitations. One group of patients not considered in this work is those with suspected ACS but no cTnI performed. Assessing their outcomes, and the effect of not performing the assay, is an important part of auditing ACS management, but not within the scope of our work. Furthermore, some aspects of this study are unique to the specific type of assay used in our institution. However, overall a strong economic and clinical case exists, based on published studies, for the introduction of troponin assays as part of the care pathway for patients with suspected acute coronary syndromes. This audit has demonstrated that in our institution (and possibly in others within the UK), we are not meeting the standards of practice which are required to reproduce the published outcomes, and therefore it is likely that the intended benefits are being compromised. This audit provides important lessons, both to us and to all those who introduce troponin assays and other new tests into their institutions.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Capewell S, McMurray J. ‘Chest pain please admit‘: is there an alternative? Br Med J 2000; 320:951–2.
2. Heeschen C, Hamm CW, Goldmann B, Deu A, Langenbrink L, White HD, for the PRISM Study Investigators. Troponin concentrations for stratification of patients with acute coronary syndromes in relation to therapeutic efficacy of tirofiban. Lancet 1999; 354:175.[CrossRef][Web of Science][Medline]
3. Hamm CW, Heeschen C, Goldmann B, Vahanian A, Adgey J, Miguel CM, Rutsch W, Berger J, Kootstra J, Simoons ML. Benefit of abciximab in patients with refractory unstable angina in relation to serum troponin T levels. c7E3 Fab Antiplatelet Therapy in Unstable Refractory Angina (CAPTURE) Study Investigators. N Engl J Med 1999; 340:1623–9.
4. Antman EM, Tanasijevic MJ, Thompson B, Schactman M, McCabe CH, Cannon CP, Fischer GA, Fung AY, Thompson C, Wybenga D, Braunwald E. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. New Engl J Med 1996; 335:1342–9.
5. Bertrand ME, Simoons ML, Fox KA, Wallentin LC, Hamm CW, McFadden E, De Feyter PJ, Specchia G, Ruzyllo W. Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2002; 23:1809–40.
6. Cannon CP, Weintraub WS, Demopoulos LA, Vicari R, Frey MJ, Lakkis N, Neumann FJ, Robertson DH, DeLucca PT, DiBattiste PM, Gibson CM, Braunwald E. TACTICS (Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy)-Thrombolysis in Myocardial Infarction 18 Investigators. Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med 2001; 344:1879–87.
7. FRagmin and Fast Revascularisation during InStability in Coronary artery disease (FRISC II) Investigators. Invasive compared with non-invasive treatment in unstable coronary-artery disease: FRISC II prospective randomised multicentre study. Lancet 1999; 354:708–15.[CrossRef][Web of Science][Medline]
8. Zarich S, Bradley K, Seymour J, Ghali W, Traboulsi A, Mayall ID, Bernstein L. Impact of troponin T determinations on hospital resource utilization and costs in the evaluation of patients with suspected myocardial ischaemia. Am J Cardiol 2001; 88:732–6.[CrossRef][Medline]
9. Altman DG, Bland JM. Statistics notes: diagnostic tests 2: predictive values. Br Med J 1994; 309:102.
10. Wu AH, Apple FS, Gibler WB, Jesse RL, Warshaw MM, Valdes R. National Academy of Clinical Biochemistry Standards of Laboratory Practice: Recommendations for the use of cardiac markers in coronary artery diseases. Clin Chem 1999; 45:1104–21.
11. Théroux P, Fuster V. Acute Coronary Syndromes: Unstable Angina and Non–Q-Wave Myocardial Infarction. Circulation 1998; 97:1195–206.
12. Antman EM, Cohen M, Bernink PJ, McCabe CH, Horacek T, Papuchis G, Mautner B, Corbalan R, Radley D, Braunwald E. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000; 284:835–42.
13. Birkhead J. National Audit of Myocardial Infarction Project. Where are we today? Early results from MINAP, the National Audit of Myocardial Infarction Project. Heart 2003; 89(Suppl. 2):ii 13–15 and 35–7.[Medline]
14. Birkhead JS, Walker L, Pearson M, Weston C, Cunningham AD, Rickards AF. National Audit of Myocardial Infarction Project Steering Group. Improving care for patients with acute coronary syndromes: initial results from the National Audit of Myocardial Infarction Project (MINAP). Heart 2004; 90:1004–9.
15. Schmitt M, Cockcroft JR, Frenneaux MP. Modulation of the natriuretic peptide system in heart failure: from bench to bedside? Clin Science 2003; 105:141–60.[Medline]
16. Perrier A. D-dimer for suspected pulmonary embolism: whom should we test? Chest 2004; 125:807–9.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Lippi, G.L. Salvagno, M. Montagnana, and G.C. Guidi How many troponins should we measure to get a clinically significant result? QJM, June 1, 2007; 100(6): 389 - 390. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||