JRSM >
Volume 97, Number 5 >
Pp. 230-234
1 Public Health Genetics Unit, Strangeways Research Laboratory, Worts Causeway,
Cambridge CB1 8RN
2 South Cambridgeshire Primary Care Trust, Fulbourn, Cambridge CB1 5EE
3 Division of Cardiovascular Medicine, Department of Medicine, University of
Cambridge Addenbrooke's NHS Trust, Cambridge CB2 2QQ
4 Centre for Applied Statistics, Institute of Public Health, Cambridge CB2
2SR
5 East Anglian Ambulance NHS Trust, Norwich NR6 5NA, UK
Correspondence to: Mark Kroese
E-mail:
mark.kroese{at}srl.cam.ac.uk
 | SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The best predicted local health benefit from the proposed prehospital thrombolysis service is that, if 45 minutes can be cut off the call-to-needle time, 61 cases of acute myocardial infarction need to be treated to save one additional life at 35 days.
By use of published research data, the health benefits of prehospital thrombolysis can be estimated for a local population. Variables in the treatment population and ambulance service will influence the size of the health benefit that can be achieved.
| INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
One way to reduce treatment delay, suggested in the NHS Plan and in a recent Department of Health review of thrombolysis,5,6 is to give thrombolytic agents before the patient reaches hospital. Such a service was proposed by our local ambulance trust, whereby prehospital thrombolysis would be available for patients more than 15 minutes' ambulance travelling time from pick-up point to hospital. Paramedics would first transmit an electrocardiogram (ECG) by telemetry to the receiving hospital and discuss the clinical findings with hospital doctors. The final decision on whether to treat would be made by hospital doctors and the attending paramedic would administer the thrombolytic. The local commissioning group requested an evaluation of the possible health impact of such a prehospital thrombolysis service, to inform future investment policy.
Background
Table 1 shows that, although
the major trials favour prehospital administration of thrombolytics,
individually their results were not statistically significant. In a
meta-analysis of the three trials presented, the pooled result was
statistically significant in favour of prehospital thrombolysis. The odds
ratio (OR) for the pooled results was 0.84 (95% confidence interval [CI]
0.70–0.99).7
|
Considerable evidence shows that prehospital thrombolysis is feasible.8-20 Studies also show that the side-effect and complication rates differ little between prehospital and in-hospital thrombolysis.8,9,21
Whilst it is possible to show the mortality benefit of prehospital thrombolysis in a study population, it is more difficult to estimate what the likely mortality benefit will be in a local population. For example, the impact of the intervention will depend on median time delay from symptom onset to calling for professional help, median transfer time to hospital, and median overall ‘symptom onset to treatment’ time. A mathematical model is needed which will incorporate local population variables, and one such was identified.3
Describing the relationship between treatment delay and absolute mortality derived from randomized studies, it is based on eight studies comparing prehospital and hospital thrombolysis including the three already presented. The data for a total study population of 6634 people were used. A weighted regression line was fitted to the data and a linear regression line was found to best fit. The equation describing the regression line was not presented in the paper but was calculated as y=21.3x+45, where y is the absolute mortality per 1000 treated patients and x is the treatment delay (from onset of symptoms). The equation is valid only for reductions from a maximum of 4 hours to a minimum of 1.3 hours ‘symptom onset to treatment’ delay because this is the range for which there are research data.
| METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
Analysis
The time interval for each stage of the clinical care pathway was
calculated for the group of cases fulfilling the selection criterion of over
15 minutes' travelling time from the pick-up point to hospital. The
‘symptom onset to call’ time was available for only half the
cases, so the health benefit was calculated for a range of values for this
time—1 hour, 2 hours, 3 hours and 4 hours. This range of values was
chosen to match the median values found in studies where the ‘onset of
symptom to call’ time was
measured.22-26
As has been shown, the expected reduction in the treatment delay due to prehospital thrombolysis varies between studies. A range of values was therefore used—20, 33 and 45 minutes. 33 minutes was chosen because this was the median value obtained from the MITI study8 and this was the only trial using a paramedic provider similar to the service proposed by the ambulance trust. 20 minutes was chosen in case the local experience was that the reduction was less than 33 minutes; 45 minutes was chosen because this was the value of the median time reduction obtained from a meta-analysis of several trials excluding the GREAT study.7
From the time data obtained in the first stage of this analysis, the health benefit in terms of the absolute reduction in 35-day mortality was estimated from the model. This involved using the actual recorded delay from ‘symptom onset to thrombolysis’ for a particular case and then repeating the same calculation with the expected shorter delay achieved with prehospital thrombolysis. The predicted reduction in mortality for prehospital thrombolysis for that particular case was the difference between the two values. Once the individual health benefit was obtained, this could be aggregated to derive a population benefit for a particular reduction time.
SPSS 10.0 was used for statistical analyses. This study was approved by the local research ethics committee.
| RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
47 cases fulfilled the criterion of more than 15 minutes' travelling time from the pick-up point to hospital. For this group, the male/female ratio was 3.7/1. The median age was 61 years (mean 62.8, range 40–87). The care pathway time results for the 47 cases are shown in Figure 3. The health benefit was calculated from the time information in these cases. The best predicted health benefit is that, if 45 minutes can be cut off the call-to-needle time, 61 cases of AMI need to be treated to save one additional life at 35 days (Table 2).
|
|
For a fixed reduction in treatment delay (20 minutes, 33 minutes or 45 minutes) there is a fixed health benefit, as shown by the linear relationship of the model. The ‘symptom onset to call’ delay did not affect the absolute 35-day mortality reduction although the total mortality was higher with longer delays.
Ambulance time records for 34 cases eligible for analysis could not be found. Statistical testing revealed no significant difference in terms of age or sex between this group and the group of cases for which there were records.
| DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
For the purposes of this study, all the cases fulfilling the 15-minute criterion were included in the analysis. This overestimates the health benefit of prehospital thrombolysis. This is because studies involving paramedic provision of prehospital thrombolysis8,11 and pilot studies already being conducted in the UK use treatment protocols with exclusion criteria including age, history of stroke/surgery, ST elevation < 2 mm and chest pain of more than 6 hours. In addition not all AMI cases will have classic symptoms or ECG features. Both these factors will reduce the number of cases eligible for prehospital thrombolysis and the health benefit as a result of this intervention. For example, use of the age exclusion criterion of > 75 years would reduce the study group from 47 cases to 38 cases. If a combination of criteria is applied then the group of eligible patients will be smaller still. The MITI trial showed that, following screening and ECG review, fewer than half of cases receiving thrombolysis in hospital would be eligible for prehospital thrombolysis.8
Mortality at 35 days was chosen as the outcome measure in the absence of other suitable published models that estimate the health benefit of prehospital thrombolysis. Additional health measures might have included myocardial infarction size, ventricular function measures, Q wave myocardial infarction frequency, and complications such as heart failure.8,9,16,27,28 It is possible that examination of longer-term health measures would reveal greater health benefits from prehospital thrombolysis.
In the calculation of health benefit, several assumptions had to be made—for example, that the results of studies involving different thrombolytics and different models of care can be pooled and that the results of this meta-analysis are generalizable to the population in the UK; that the model equation derived from the meta-analysis adequately describes reality in the local population when the average symptom onset to initiation of therapy time was much longer in the local population than in the studies; that all eligible patients will be treated with prehospital thrombolysis; and that the experience of a historical cohort of cases reflects what will occur in the future.
When evaluating the impact of prehospital thrombolysis it is important to consider the health benefit already achieved with hospital thrombolysis. The added health benefit of reducing the treatment delay by introducing prehospital thrombolysis may be less than expected in some areas where the ‘call to door’ and ‘door to needle’ times are short. The introduction of thrombolysis nurses and the increasing number of patients with AMI who are receiving thrombolysis in accident and emergency departments will reduce the ‘door to needle’ times. There is evidence that hospitals in the UK have achieved such improvements in performance.29
| CONCLUSION |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
We recommend that the pilots now in progress in the UK should be formally evaluated to provide the information on clinical effectiveness, cost effectiveness and service models for different populations. This is required to inform national implementation of prehospital thrombolysis.
| REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J.A.L. Smith, K.P. Jennings, E.A. Anderson, P. Green, and G.S. Hillis Reducing call-to-needle times: the critical role of pre-hospital thrombolysis QJM, October 1, 2004; 97(10): 655 - 661. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
