Methods

Ambulance services in England and Wales that had conducted a recent audit of aspirin administration were identified through the National Clinical Effectiveness Programme for the Ambulance Service Association.

Nine services out of a total of 35 had collected appropriate data and made these available to us. The data that had been recorded differed considerably from service to service. Relevant comparable data were extracted where possible from each services response.

The survey was completed in December 2000.

Results

Table 1 shows the results of the audits conducted by each ambulance service. Several organisations supplied figures from more than one study, but only the data from the most recent audit are given.

The proportion of patients receiving pre-hospital aspirin from a paramedic or alternative health provider ranged from 19% to 78%. It is estimated that 15% to 74% of patients who should have been given aspirin before hospital admission did not receive it.

Table 2 shows the range of indications and contraindications to aspirin administration for each ambulance service.

Discussion

It is widely accepted that aspirin is of value after MI. However, it is of even greater efficacy in the treatment of unstable angina. Platelet aggregation and embolisation are the important factors in this condition.¹⁰ Evidence suggests that the reduction in relative risk for subsequent non-fatal MI or death after aspirin administration is greater for unstable angina than it is for MI (51%,^11,12 compared with 25%²).

While the proportion of patients eligible for aspirin in hospital who receive the drug is high, there remains room for further improvement. The gap between the need for aspirin and the proportion of patients who actually receive it could, therefore, be reduced via pre-hospital administration by paramedics. A 1992 to 1993 inhospital study of aspirin administration to 10 108 patients with MI and aged 65 or over found treatment rates of only 61% in the first two days. However, it also determined that aspirin administration was significantly associated with a lower mortality (odds ratio 0.78; 95% confidence intervals 0.70 to 0.89).¹³ Other authors have reported inhospital aspirin administration rates ranging from 45.1% (4052 elderly Medicare patients, USA, 1992/1993)¹⁴ to 87.8% to 91.3% (1737 patients, USA, 1997).¹⁵ European authors have reported similar aspirin administration rates after MI.¹⁶ But administration rates of aspirin after episodes of unstable angina are much lower. A review of practice against guidelines in nine European countries reported antiplatelet administration rates of only 70.2%.¹⁷ While inhospital compliance rates are high after MI, pre-hospital administration of aspirin by paramedics does, therefore, have the potential to help reduce the number of eligible patients who do not receive it and who forgo its benefits. It is not possible to definitively diagnose MI in the pre-hospital arena according to standard criteria, and consequently administration of aspirin to all patients with chest pain suggestive of ischaemic heart disease seems to be a reasonable recommendation.

The data we present come from ambulance services that had conducted an audit of aspirin administration to patients with chest pain suspected to be associated with ischaemic heart disease. The proportion of patients who received aspirin from any source varied from 19% to 78%. These figures are disappointing, but the reasons for these low levels of compliance to protocol are unclear. The proportion of subjects with chest pain for whom aspirin is contraindicated is likely to be around 5%–7%.¹⁸ Typically, therefore, 93% of patients with chest pain arising from suspected ischaemic heart disease might safely be given aspirin. This implies that, in the samples reported in this study, from 15% to 74% of suspected ischaemic chest pain patients who were potentially eligible for aspirin did not receive it.

The protocols followed by each ambulance services differed, but the description of patients to whom aspirin is to be given is typically “a patient presenting with signs and symptoms of ischaemic chest pain”. One service gives a more detailed list of signs and symptoms and required evidence of infarction from an ECG before aspirin was given. This is of some concern, as it has been shown that early in the infarction process diagnostic ECG changes may not be present initially, but might develop later.¹⁹ Perhaps even more importantly, aspirin is also of benefit (and is recommended) for patients suffering from a range of acute coronary syndromes, not just those with confirmed MI.²⁰

The protocols from each ambulance service all listed criteria that defined patients to whom aspirin should not be given, but these differed greatly. Allergy to aspirin, current use of anticoagulants, haemophilia and other bleeding disorders were present in all the protocols. But in two protocols the taking of aspirin within the previous 24 hours constituted a contraindication. The wisdom of this criterion can be questioned as, even if a prophylactic dose of aspirin had been taken earlier, fresh platelets may have entered the circulation since this time and it is important to reduce the proportion of un-acetylated platelets and maintain these at the lowest possible level. A further dose of 300 mg of aspirin may therefore be beneficial and is unlikely to do harm; it should not be withheld. It seems reasonable to argue that contraindications should be kept to an absolute minimum in a situation where a relatively safe drug could be life saving.

The variability identified between the various services indicates the need for an UK standard protocol based on the available evidence from trials. In formulating this it should be borne in mind that a rigid approach, restricting aspirin only to those patients with definitive evidence of acute MI may deprive many others of the benefits of aspirin treatment. Patients who report having had chest pain but who are later judged not to have had an infarct do, in fact, have an increased mortality.²¹ Undoubtedly some of these will indeed not have had an infarct, but even in those in whom the pain is muscular aspirin may have a useful role. In a small proportion the chest pain may arise from a stomach lesion, and there is a risk that the questions which paramedics ask may not reveal this. Nevertheless, while aspirin would be best withheld in such patients, it could be argued on the basis of relative risks that where the diagnosis is uncertain, aspirin should be given.

Further UK wide standardised audits of aspirin administration should be conducted, as poor compliance to this aspect of paramedic treatment protocols may indicate poor compliance to other aspects of the treatment of chest pain. While early administration of aspirin is important there is no strong evidence to suggest that its effectiveness is increased by giving it at the earliest possible time. Other forms of treatment may be considerably more time sensitive (such as the administration of fibrinolytics). The Royal College of Physicians (RCP) has developed a core dataset for documenting the treatment of MI, and has collaborated with the Joint Royal Colleges Ambulance Liaison Committee and the Ambulance Service Association in defining its ambulance related aspects. The RCP is also conducting a national audit of MI (MINAP). Ambulance services must audit and report every administration of a thrombolytic by a paramedic, and may be required to audit other treatments in the future. These initiatives may well meet the need for ongoing audit.