Journal of the New Zealand Medical Association, 29-June-2012, Vol 125 No 1357
The use of troponin in general practice
Sally Aldous, Peter Gent, Graham McGeoch, Denise Nicholson
International guidelines recommend serial cardiac troponin (cTn) measurement in patients presenting with symptoms suggestive of acute coronary syndrome (ACS).1,2 Cardiac troponin is the gold standard biochemical criterion for the diagnosis of acute myocardial infarction (AMI)1 and investigations and treatments guided by cTn results have been shown to influence outcomes.2–6 As such, cTn is not only a diagnostic tool but is also highly effective in risk stratification.
The primary care physician or general practitioner (GP) faces the challenge of identifying patients with ACS and therefore those at risk of adverse cardiac events such as death, AMI and heart failure. It is often the task of the GP to initiate further diagnostic procedures in a timely manner in those with suspected ACS. However, although symptoms such as chest pain are common in general practice, they are due to acute coronary artery disease in only a minority of cases. It is therefore also the role of the GP to protect patients from over-diagnosis and inappropriate treatment, and prevent overcrowding of specialist and emergency services.7–10
Cardiac troponin tests are available in the community and GPs are making increasing use of these tests to help triage such patients.7
Concerns with the use of cTn in the community are:
A retrospective audit was performed of all cTn requests referred to Canterbury Health Laboratories from General Practice in Christchurch, New Zealand, in 2010. Canterbury Health Laboratories is one of three laboratory groups in the region that received cTn requests during this time period.
Patient data was collected by means of a regional and then national health events search (which identifies any hospital attendance using an alpha numeric identifier unique to that patient). Admissions to hospitals outside of New Zealand were not sought. Data was collected regarding whether admission occurred after index cTn testing and the subsequent diagnosis according to the discharge notice from the admitting team.
Patients were followed for 6 months for adverse events including death, non-fatal AMI, revascularisation (percutaneous coronary intervention or coronary artery bypass surgery) and admission for heart failure with diagnoses again according to the discharge diagnosis of the admitting team. These were analysed in those above and those below 75 years of age as an arbitrary (but not ideal) marker of comorbidity.
The cTn assay utilised by Canterbury Health Laboratories is Abbott Architect Troponin I (99th percentile 0.028 mcg/L, 10% coefficient of variation 0.032 mcg/L, limit of detection 0.010 µg/L, decision cut-point as per manufacturer >0.03 µg/L)
There were 2662 patient episodes in 2575 patients during 2010 in which ≥1 sample for cTn testing were sent to Canterbury Health Laboratories. The median age was 63 (interquartile range 51 to 74) and 1186 (44.6%) were male. Other patient characteristics were not available. There were 321 (12.1%) patient episodes in which serial samples were taken.
There were 223 patients episodes (8.4%) in which ≥1 cTn results were elevated above the decision threshold, median age 73 (63–83), 131 (58.7%) male. Twenty-one (9.4%) had serial measurements, of which 11 were elevated on the second sample only. 184 (82.5%) were admitted to hospital for further evaluation, median age 71 (62–81), 111 (60.3%) male, median cTn 0.14 (0.06–0.61) µg/L. The discharge diagnoses of those who were admitted are shown in Figure 1.
Figure 1. Discharge diagnosis of patients admitted following elevated troponin in general practice
STEMI – ST elevation myocardial infarction, NSTE ACS – non ST elevation acute coronary syndrome. Arrhythmia=atrial fibrillation, supraventricular tachycardia, complete heart block, Respiratory disease=chronic obstructive pulmonary disease, pulmonary embolus, pneumonia.
Those with a diagnosis of myocardial infarction had higher median cTn levels, 0.27 (0.10–1.65) µg/L than those with other diagnoses, 0.08 (0.05–0.22) µg/L. Those not admitted had a median cTn of 0.05 (0.03–0.09) µg/L.
Figure 2. Six month event rates according to troponin result and whether patient was admitted at index testing
(a) Troponin positive/Admitted
(b) Troponin positive/Not admitted
(c) Troponin negative/Admitted
(d) Troponin negative/Not admitted
Of the 184 admitted, 5 (2.7%) underwent stress testing, 86 (46.7%) underwent coronary angiography and 49 (26.6%) were revascularised, 4 with coronary artery bypass surgery. Events post discharge are shown in Figure 2. The 6-month event rates of the 39 with elevated cTn who were not admitted, median age 81 (66–88), 19 (48.7%) male, are also shown in Figure 2. The median cTn of those with (any) 6-month event was 0.10 (0.05–0.58) µg/L compared with those without events, 0.12 (0.05–0.41) µg/L.
Figure 3. Discharge diagnosis of patients admitted following normal level troponin in general practice
STEMI – ST elevation myocardial infarction, NSTE ACS – non ST elevation acute coronary syndrome. Other cardiac diseases=peri/myo/endocarditis, heart failure, valvular disease, Arrhythmia=atrial fibrillation, supraventricular tachycardia, complete heart block, Respiratory disease=chronic obstructive pulmonary disease, asthma, pulmonary embolus, pneumonia, influenza, malignancy, Gastrointestinal disease=cholecystitis/cholangitis/pancreatitis, ileo/colitis, appendicitis, peptic ulcer disease.
There were 2439 patients in whom all cTn measurements were below the decision cut-point, median age 62 (50–73), 1056 (43.1%) male. Two hundred and ninety seven (12.2%) of these had serial troponin measurement. Three hundred and forty four (14.1%) were admitted to hospital for further evaluation, median age 65 (54–78), 136 (39.5%) male. The discharge diagnoses of those who were admitted are shown in Figure 3.
Of the 344 admitted, 69 (20.1%) underwent stress testing, 31 (9.0%) underwent coronary angiography and 13 (3.8%) were revascularised, 5 with coronary artery bypass surgery. Events post discharge are shown in Figure 2. The 6 month event rates of the 2095 with normal level cTn who were not admitted, median age 62 (50–73), 917 (43.8%) male, are also shown in Figure 2.
This audit investigates the use of cTn testing in an urban community setting over a 1 year period and shows that such a test is highly utilised.
The utility of cTn in the community should be to:
Previous studies have shown that if suspicion for ACS is low, GPs are more likely to either order cTn and wait for the result or manage medically. If suspicion is intermediate, GPs tend to either refer for admission without cTn results (especially when patients present early after symptom onset) or order cTn and wait for the result. In contrast, if the suspicion is high, patients are referred without cTn results.11
This study shows that the vast majority of patients with elevated cTn were admitted. This appears highly appropriate as those with elevated cTn were much more likely to come to harm in the short term with higher 6 month rates of death (8.5% versus 1.1%), AMI (2.2% versus 1.2%), revascularisation (1.8% versus 0.7%) and heart failure (3.1% versus 1.0%) than those with normal cTn levels. Those with and without events had similar median cTn levels.
Those with elevated cTn who were not admitted had higher event rates than those who were admitted (12.8% versus 7.6% for death, 7.7% versus 1.1% for AMI, 2.6% versus 1.6% for revascularisation and 7.7% versus 2.2% for heart failure). Although not all comorbidities were known, the median age of these patients was 81, suggesting that admission may not have been appropriate and that community treatment was possibly medical/palliative. However, 2.6% of these patients were revascularised within 6 months suggesting that this was not the case in all patients. Figure 2 shows how those ≥75 (a surrogate but not ideal marker for comorbidity) were more at risk of events, as expected.
The median cTn levels in these patients showed only borderline elevations which may also contribute to the decision not to admit, unfortunately we do not have data regarding the level of probability for AMI assigned by the General Practitioner.
It is likely that a significant number of patients would have been referred for admission irrespective of the cTn results, for example the 13 patients with STEMI and the 14.4% of those with negative cTn results who were still admitted. If admission is inevitable, it is unlikely that measurement of cTn in the community is necessary and may have economic implications.11 It can be seen that a negative result for cTn in the community indicates low risk of subsequent adverse cardiac events. As such, it seems reasonable to suggest that current management of these patients in the community is appropriate.
Only approximately 12% of patients had serial testing. Current International Guidelines recommend serial cTn measurement at presentation and again at 8–12h after symptom onset2 or 6–9h post presentation.1 Previous audit data at Christchurch Hospital has shown that the cTn used in this study is reliably negative by 10 hours post symptom onset.
Unfortunately the time from symptom onset to presentation was unknown. Although a negative cTn in the community appears to indicate low risk, serial cTn would still be recommended in those presenting less than 10 hours from symptom onset. It was shown that of the 21 patients with elevated cTn undergoing serial measurements, only 10 had elevations on the first test.
A previous community audit performed elsewhere in New Zealand showed that 12% of negative tests were performed in patients presenting less than 10 hours after the onset of symptoms.7 It may be therefore, that most of the early presenters in this study did infact have serial measurements.
The indication for cTn testing in all patients was unknown as clinical information at the time of the test was not collected. Therefore this did not allow an assessment of the clinical appropriateness of testing. However, the discharge diagnosis of the patients admitted, demonstrated a wide variety of conditions.
Previous audit data in Christchurch Hospital shows that in patients admitted for rule out of ACS, approximately 20–25% have elevated cTn and approximately 90–95% of these have a diagnosis of AMI. In contrast, this audit shows that only 8.3% have elevated cTn in the community and only just over 50% of these had a diagnosis of AMI.
This illustrates the wide variety of conditions that can lead to cTn elevation although those without AMI had lower median cTn levels than those with AMI. It should therefore be emphasised that cTn is to only be measured in patients with suspected ACS. Previous studies have shown that patients with symptoms such as chest pain presenting in the community are most likely to have musculoskeletal conditions and a significant number of patients have psychogenic disorders, respiratory infections and gastrointestinal diseases.8,9 Only a minority have ACS,8,9 1 study had an ACS rate of 3.6%9 compared with our 5.4%. The American Heart Association guidelines2 have suggested that symptoms suggestive of ACS include acute chest, epigastric, neck, jaw or arm pain or discomfort or pressure without apparent non-cardiac source. Symptoms such as fatigue, shortness of breath, syncope or arrhythmia can in a minority of patients be secondary to AMI but are not symptoms that automatically indicate the need to measure cTn if ACS is otherwise not suspected. While it is accepted that many patients present atypically, this highlights perhaps that cTn testing in the community may not be measured in the appropriate clinical context in a significant proportion of patients.
A previous study showed that only 25–40% of GPs could correctly identify other causes of cTn elevation.11 Other studies suggest that community cTn testing is indicated to determine the need for acute referral in those with intermediate suspicion of ACS and not those with a high likelihood (who should be referred without cTn testing)7,9 and is also not for reassurance of the patient in whom there is low suspicion of ACS.7
In conclusion, this study suggests that the use of cTn in the community appropriately triages patients regarding the need for admission. However, the indication for testing only in cases of suspected ACS and the use of serial cTn measurement in early presenters should be emphasised.
A clinical pathway for management of suspected ACS in the community is currently being developed. Each patient is categorised as to whether their likelihood of ACS is high, intermediate, low or unlikely using predefined criteria to help determine the best management. If high risk, immediate referral of patients without prior cTn testing should occur. Low risk patients should be managed in the community with cTn testing, including serial measurements in patients presenting less than 10 hours after onset of symptoms. If unlikely to be ACS, cTn testing is not advised.
However, there is always a limit to how specific such an algorithm can be and clinical judgment will always be required, especially in cases of greater uncertainty. It is our intent to repeat this audit in 1 year following the instigation of this pathway.
Competing interests: None declared.
Author information: Sally Aldous, Cardiologist, Cardiology Department, Christchurch Hospital, Christchurch; Peter Gent, General Practitioner, Christchurch; Graham McGeoch, General Practitioner, Christchurch; Denise Nicholson, General Practitioner, Clinical Editor Health Pathways, Canterbury District Health Board, Christchurch
Acknowledgements: We thank all the staff at Canterbury Health Laboratories for troponin assays.
Correspondence: Sally Aldous, Cardiology Department, Christchurch Hospital, Riccarton Ave, Christchurch, New Zealand. Fax: +64 (0)3 3641415; email: email@example.com
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals