Primary angioplasty versus thrombolysis for acuteST-elevation myocardial infarction: an economicanalysis of the National Infarct Angioplasty project

Allan Wailoo,1 Steve Goodacre,1 Fiona Sampson,1 Monica Hernandez Alava,2

Christian Asseburg,3 Stephen Palmer,4 Mark Sculpher,4 Keith Abrams,5

Mark de Belder,6 Huon Gray7

ABSTRACTObjective To estimate the cost-effectiveness of primaryangioplasty compared with thrombolysis for acute STelevation myocardial infarction.Design Cost analysis of UK observational database,incorporated into decision analytical model.Methods Patients receiving treatment withina comprehensive angioplasty service were comparedwith control patients receiving thrombolysis-based care.The treatment costs and delays to treatment ofthrombolysis and angioplasty were estimated. Theseestimates were then incorporated into an existing modelof cost-effectiveness that synthesises evidence from 22randomised trials to estimate health outcomes measuredby quality-adjusted life years (QALYs).Main outcome measures Costs from a health serviceperspective and outcomes measured as quality adjusted.Results The mean cost of the initial treatment was£3509 for thrombolysis at control sites, £5176 forangioplasty in usual working hours at National InfarctAngioplasty Project sites and an additional £245 ifundertaken out of hours. Angioplasty-based care had anincremental cost of £4520 per QALY gained and 0.9probability of being cost-effective at a threshold of£20 000 per QALY gained. This probability was >0.95 ifpatients were directly admitted to the cardiac catheterlaboratory, 0.75 if admitted via the emergencydepartment or coronary care unit and 0.38 if transferredto the angioplasty centre from another hospital.Conclusions Overall, primary angioplasty-based care ishighly likely to be cost-effective at an assumed thresholdof £20 000 per QALY gained. It is more likely to be cost-effective if patients are admitted directly to the cardiaccatheter laboratory rather than via other hospitaldepartments, or if transferred from another hospital.

INTRODUCTIONAlthough current NHS policy for the treatmentof ST elevation myocardial infarction (STEMI)recommends rapid thrombolytic drug treatment,1

evidence from a number of trials dating from theearly 1990s suggests that primary angioplasty maybe more effective in reducing mortality, reinfarctionand stroke.2 However, the benefits from either typeof treatment diminish as the time from symptomonset to treatment increases. Therefore, primaryangioplasty must be performed in a manner thatminimises the additional time to treatment in orderto be clinically effective compared with thrombol-ysis. Furthermore, the requirement for specialist

staff and facilities presents practical challenges toproviding primary angioplasty cost-effectively.A recent cost-effectiveness analysis (the ‘York

model’) compared the two treatment options ina UK setting.3 The analysis was based on meta-analysis of clinical trial data and suggested thatprimary angioplasty is likely to be considered cost-effective provided that the additional time delayassociated with primary angioplasty does notexceed an hour. However, the extent to which someof the values used in this analysis apply to NHSpractice remains unclear, particularly in relation tocosts and time to treatment.The National Infarct Angioplasty Project (NIAP)

was set up as the result of a recommendation fromthe prime minister ’s delivery unit in June 2003 thatthe Department of Health should develop a clearpolicy for expanding primary angioplasty and drawconclusions about the feasibility of a nationalroll-out of this service. The secretary of statesubsequently announced funding of up to £1million to carry out a feasibility study, which wasused to collect data from a number of angioplastypilot sites, selected on the basis of their ability tocollect the required data and to reflect a rangeof geographical settings and service delivery types.This initiative provided the opportunity to assessthe cost of the initial treatment episode and timedelays of primary angioplasty and thrombolysisin the NHS.We aimed to estimate the cost-effectiveness of

services based on comprehensive primary angio-plasty versus thrombolytic-based services by furtherdeveloping the York model with ‘real-life’ cost andtreatment delay estimates derived from NIAP data.

METHODSSource of dataTen hospitals providing primary angioplastybecame NIAP pilot sites (see supplementary onlineappendix 2 for details) and collected detailed dataon the initial treatment episode and follow upinformation to 1 year for all patients with STEMIadmitted between 1 April 2005 and 31 March 2006.These hospitals varied in the times primary angio-plasty was provided (not all offered 24/7 provisionduring the study period) and whether they hadarrangements with hospitals unable to provideangioplasty for patients to be transferred forprimary angioplasty. These hospitals providedinformation on 2083 patients.

< An appendix is publishedonline only. To view these filesplease visit the journal online(http://heart.bmj.com/content/vol96/issue9).1School of Health and RelatedResearch, University ofSheffield, Sheffield, UK2Department of Economics,University of Sheffield, Sheffield,UK3Swedish Institute for HealthEconomics, Lund, Sweden4Centre for Health Economics,University of York, York, UK5Department of HealthSciences, University ofLeicester, Leicester, UK6The James Cook UniversityHospital, Middlesborough, UK7Southampton UniversityHospital, Southampton, UK

Correspondence toDr Allan Wailoo, HealthEconomics and DecisionScience, School of Health andRelated Research, University ofSheffield, Regent Court, 30Regent Street, Sheffield S14DA, UK;a.j.wailoo@sheffield.ac.uk

Accepted 26 May 2009Published Online First8 June 2009

668 Heart 2010;96:668e672. doi:10.1136/hrt.2009.167130

Acute coronary syndromes

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from

As part of NIAP, one hospital that did not provide primaryangioplasty also collected data for the study period. Wesupplemented this control site information by obtaining equiv-alent data for the same time period from four other hospitalsthat did not provide primary angioplasty for their patients withSTEMI. A total of 919 control patients were included.

NHS resource use and costThe analysis sought to estimate the cost of the initial treatmentfor patients treated either by thrombolysis or by primaryangioplasty from an NHS perspective. The NIAP databaseprovided information on cardiac drugs (including thrombolytictype), consumables (including number, type and make of stent),tests and the length of stay for each patient and similar infor-mation was obtained from each of the control sites for throm-bolysed patients. It was not possible to obtain detailedinformation on the small number of patients treated by primaryangioplasty in the control hospitals. All drugs were includedwhether they were administered before arrival at hospital, at thefirst hospital for transferred patients or at the subsequentprimary angioplasty centre. Exercise tests, echocardiography,radionuclide studies and coronary angiography were reported foreach patient, although we assumed that every patient under-going primary angioplasty also received coronary angiography.

A separate survey of 50 primary angioplasties at five NIAPhospitals was used to estimate the mean duration of theprocedure, the typical staffing of the catheter laboratory andhow this differed according to whether a procedure occurred inor out of usual operating hours.

Patients treated at primary angioplasty centres may betransferred from feeder hospitals, resulting in additional ambu-lance journeys. These were defined as transfers of patientsbetween non-primary angioplasty hospitals and primary angio-plasty hospitals for any reason, and transfers of patients fromprimary angioplasty hospitals back to other hospitals. Whereclinical staff accompanied the ambulance crew, this wasrecorded.

Subsequent coronary artery bypass grafts (CABG) or percu-taneous coronary interventions (PCI) up to 1 year were identi-fied although cost differences estimated using this data wereonly incorporated into the cost-effectiveness analysis as a sensi-tivity analysis. The base-case cost-effectiveness estimates arebased on the estimated costs from the rates of CABG and PCIfound in the clinical trials.

Unit costs for 2006e7 were obtained from national sourceswhere available (see online appendix 3). In the absence ofnationally published data on the unit costs of consumables orthe operating costs for the catheter laboratories, we surveyed theNIAP hospitals for these items.

Time to treatmentPresentation delay, defined as the time between the first call forprofessional help and arrival at hospital, was estimated forthrombolysis patients. For all patients, the time between thefirst call for professional help and treatment (needle for throm-bolysis patients, and inflation of balloon for angioplastypatients) was calculated.

Statistical analysisTo estimate the mean treatment cost and associated variance,regression techniques were used with a range of patient cova-riates to adjust for case mix. Generalised linear multilevel modeltechniques were used to account for the skewed nature of thecost data and the grouping of patients within individual hospi-

tals. A g distribution with an identity link was employedtogether with correlated random coefficients for the variables ‘notreatment’, ‘thrombolysis’ and ‘primary angioplasty ’.Median time to treatment and binomial exact confidence

intervals are reported.The original York model comprises short-term and long-term

elements in the comparison of cost-effectiveness betweenthrombolysis and primary angioplasty.3 We retained the prin-cipal features of the model but made several adaptations. Thefirst of these was to compare primary angioplasty-based serviceswith thrombolysis-based services, where different proportions ofboth treatments are performed, rather than a comparison oftreatment types themselves. The patient characteristics werethose of the NIAP population and all costs were updated to2006e7 values. The short-term element of the model estimatesthe probability of death, non-fatal myocardial infarction,non-fatal stroke and revascularisations at 6 months according tothe type of treatment and the additional time delay associatedwith primary angioplasty-based on a statistical synthesis of 22randomised controlled trials. We made adjustments to thecalculation of mortality rates for thrombolysis patients, basedon pooled trial data,4 to reflect the effectiveness of thrombolysisas a function of the observed distribution of presentation delayfor these patients. We also incorporated the additional timedelay associated with primary angioplasty as observed in thepilot year, defined as the difference between the median time toneedle for thrombolysed patients and the time to inflation ofballoon for primary angioplasty patients, in place of the timedelay seen in the clinical trials. The costs up to 6 months arethen calculated with treatment costs updated for this analysis.Long-term costs and benefits are estimated in the second part

of the model by extrapolating using the Nottingham HeartAttack Register. The model expresses benefits as quality-adjustedlife years (QALYs) by adjusting the weight allocated to the timeperiod the average patient is alive according to the probability ofbeing in each of the health states. Weights were applied to strokeand myocardial infarction based on published literature. Fulldetails of the cost-effectiveness model are available elsewhere.3

Sensitivity analysesThe cost-effectiveness model reflects the uncertainty in modelinputs using probabilistic sensitivity analysis5 to present theprobability that primary angioplasty is cost-effective for a givencost-effectiveness thresholddhere £20 000 per QALY is used,reflecting the value below which the National Institute forHealth and Clinical Excellence would typically consider a tech-nology to be acceptable.6 In addition, we incorporated coronaryevents up to 1 year after the initial treatment in place of thetrial-based estimates used in the base-case analysis.

RESULTSPatient characteristicsA total of 3002 patients were included in the analysis, 2083treated at NIAP hospitals and 919 controls (table 1).More patients at NIAP hospitals received primary angioplasty

than thrombolysis (67.1% vs 15.8%), whereas in the controlhospitals most patients received thrombolysis (72.6% vs 4.2%).The remainder (17.0% and 23.2%, respectively) received noreperfusion therapy. Of the 667 thrombolysis patients from thecontrol sites only 9% had it administered pre-hospital. Overall,patients at the NIAP hospitals were younger, more ethnicallydiverse, had a higher prevalence of previous coronary heartdisease, and a higher prevalence of hypertension, peripheralvascular disease and diabetes.

Heart 2010;96:668e672. doi:10.1136/hrt.2009.167130 669

Acute coronary syndromes

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from

Time to treatmentMedian time from call to needle was shorter than time from callto balloon (CTB) (table 2).

Call to needle time was shortest for patients receivingthrombolysis in the control hospitals (67 min, 95% CI 64 to 69).For patients treated with thrombolysis in the NIAP hospitalsthis was longer (75 min, 95% CI 68 to 87). The median CTB inthe NIAP hospitals was 131 min (95% CI 129 to 135). Forpatients transferred to a NIAP hospital from a non-primaryangioplasty hospital the time from CTB was longer (167 min,95% CI 159 to 174) compared with those who were takendirectly to the primary angioplasty hospital (120 min, 95% CI115 to 124). Patients who were taken directly to the catheterlaboratory once they had arrived at the PCI hospital hada significantly lower median CTB time (123 min, 95% CI 116 to130) than those who were not taken directly to the catheterlaboratory (140 min, 95% CI, 136 vs 145).

Cost of initial treatment episodeAlternative model specifications were compared. Using thedeviance as a measure of fit, a g distribution with an identitylink was chosen to transform the expected costs. Explanatory

variables were only deleted from the model where they weregrossly insignificant with t values well below one. A quadraticeffect of age on total cost was found to perform better thaneither using age on its own or the logarithm of age. The resultsfrom the preferred specification are reported in table 3.The coefficients for no reperfusion treatment, thrombolysis

and primary angioplasty provide the estimated episode costacross all hospitals according to treatment given, after control-ling for other covariates. For patients treated with neitherthrombolysis nor primary angioplasty, there was no significantdifference between NIAP hospitals and control hospitals andthis final version of the model, therefore, does not include anyvariable to distinguish location for these patients. For patientstreated by thrombolysis, the cost was £3509 (95% CI £2202 to£4817) but was higher in NIAP sites by £852 (95% CI £75 to£1628) after controlling for other covariates. For patients treatedby primary angioplasty the cost was £5176 (95% CI £4002 to£6350) for those treated within working hours and was higherfor patients treated out of hours by £245 (95% CI �£41 to £531).A number of other patient covariates were included in the

analysis. Age was significantly correlated with treatment cost,with minimum costs occurring at approximately 40 years of age.

Table 1 Patient characteristics

NIAP (n[2083) Control (n[919)

No % No %

Treatment type PPCI 1398 67.1 39 4.2

Lysis 330 15.8 667 72.6

None 355 17.0 213 23.2

Mean age in years (SD) 63.5 (14.0) 66.1 (13.3)*

Ethnic group Caucasian 1407 80.5 624 98.0

Previous CHD

Comorbidities Hypertension 931 45.9 327 39.3*

Hypercholesterolaemia 789 40.3 343 43.3

Peripheral vascular disease 87 4.3 12 2.0*

Cerebrovascular disease 129 6.4 44 5.1

Asthma or COPD 238 11.8 97 11.4

Chronic renal failure 50 2.5 19 2.1

Diabetes 339 16.6 101 11.2*

LVEF Good 547 55.6 151 58.3

Moderate 327 33.3 89 34.4

Poor 109 11.1 19 7.3

CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; LVEF, left ventricular ejection fraction; NIAP, National Infarct Angioplasty Project; PPCI, primary percutaneouscoronary intervention.*p<0.01.

Table 2 Time to treatment (min)

Median

95% CI IQR

Low High Low High

Call to needle

Control (n¼522) 67 64 69 52 89

NIAP (n¼184) 75 68 87 54 104

Call to balloon (NIAP hospitals)

All PPCI patients (n¼1172) 131 129 135 98 169

Transferred from another hospital(n¼368)

167 159 174 134 224

Not transferred from another hospital(n¼838)

120 115 124 90 154

Admission directly to catheterlaboratory (n¼588)

123 116 130 87 170

Admission not directly to catheterlaboratory (n¼618)

140 136 145 111 179

IQR, interquartile range; NIAP, National Infarct Angioplasty Project; PPCI, primary percutaneous coronary intervention.

670 Heart 2010;96:668e672. doi:10.1136/hrt.2009.167130

Acute coronary syndromes

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from

Previous coronary heart disease (any of previous myocardialinfarction, angina, PCI or CABG) was positively correlated withtreatment cost as were all other comorbidities. Variables withparticularly large coefficients were peripheral vascular diseaseand cerebrovascular disease. The mean cost was lower wherepatients died before discharge. Finally, although the procedurecosts of non-primary angioplasty and CABG were not compo-nents of the initial treatment costs calculated here, these didcontribute to costs, principally through their impact on lengthof stay. CABG during the index admission had a large effect(£4630) on the initial treatment cost.

Cost-effectiveness analysis: base caseResults of the cost-effectiveness analysis are shown in table 4. Inthe base-case analysis, the mean cost of patients treated in NIAPsites was £829 higher and a mean 0.18 additional QALYs weregenerated. The incremental cost-effectiveness ratio was there-fore £4520 and assuming a cost-effectiveness threshold of£20 000 per QALY the probability that NIAP is cost-effective is0.90. Two issues were explored in the sensitivity analysis: theadditional treatment delay associated with primary angioplastyand the rate of revascularisations. For treatment delay weconsidered the impact of inter-hospital transfer and of emer-gency department or coronary care unit bypass for primaryangioplasty patients.

For transferred patients the point estimate in this evaluationwas that treatment in a primary angioplasty-based service wasdominated by thrombolysis-based caredthat is, it costed £664

more (95% CI �£324 to £1390) and generated 0.085 fewerQALYs (95% CI �0.83 to 0.34). The probability that primaryangioplasty-based care was cost-effective for these patients wasapproximately 0.38 at the £20 000 threshold and did not exceed0.45 at any threshold.For patients who were not transferred from another hospital

and for those who were admitted directly to the catheter labo-ratories, the cost-effectiveness of primary angioplasty-based carewas marginally improved compared with the base case. Forpatients who were not admitted directly to the catheter labo-ratory, and in the scenario where we used observational data toinform the probability of subsequent CABG or angioplasty, thecost-effectiveness of the primary angioplasty-based servicedeteriorated but remained well below £10 000 per additionalQALY gained.

DISCUSSIONUsing observational data relating to actual UK NHS practice in2005e6, this analysis estimates that primary angioplasty isapproximately £1700 more expensive to provide than throm-bolysis in a standard setting, after adjusting for observed differ-ences between patients. The additional treatment delayassociated with primary angioplasty in the NIAP hospitals wasslightly in excess of 1 hour but this varied substantially betweenpatients who were transferred from another hospital andpatients who were taken directly to the catheter laboratory.Overall, the NIAP hospitals provided a service that would be

Table 3 Generalised linear multilevel model of total cost (2006e7 £s)

Independent variable Coefficient SE p Value 95% CI

No treatment 3343 643 0.000 2083 4604

Thrombolysis 3509 667 0.000 2202 4817

Primary PCI 5176 599 0.000 4002 6350

NIAP thrombolysis 852 396 0.032 75 1628

Age in years/10 �592 199 0.003 �982 �203

(Age in years/10)2 75 17 0.000 42 107

Previous CHD 203 107 0.057 �6 412

Hypertension 111 91 0.220 �67 289

Hypercholesterolaemia 202 95 0.033 16 387

Peripheral vascular disease 925 288 0.001 361 1489

Cerebrovascular disease 777 230 0.001 326 1228

Diabetes 245 129 0.057 �7 497

Dead at discharge �702 190 0.000 �1073 �330

Primary PCI out of hours 245 146 0.093 �41 531

Non-primary PCI in this admission 251 178 0.158 �98 601

CABG at this admission 4630 965 0.000 2738 6523

CABG, coronary artery bypass grafts; CHD, coronary heart disease; NIAP, National Infarct Angioplasty Project; PCI, percutaneous coronary interventions.

Table 4 Mean costs, QALYs and ICERs: NIAP versus control

Control cost (£) NIAP costs (£) Control QALYs NIAP QALYs ICER (£) Prob<£20000

Base case 10700 (8010, 16100) 11600 (8810, 17200) 6.4 (5.82, 6.96) 6.58 (6.01, 7.15) 4520 0.90

Sensitivity analyses

Transferred patients 10700 (8010, 16100) 11400 (8710, 16800) 6.4 (5.82, 6.96) 6.32 (5.45, 7.902) Dominated* 0.38

Non-transferred patients 10,700 (8010, 16100) 11600 (8820, 17300) 6.4 (5.82, 6.96) 6.64 (6.08, 7.20) 3635 0.95

Directly to catheter laboratory 10700 (8010, 16100) 11600 (8820, 17300) 6.4 (5.82, 6.96) 6.63 (6.07, 7.19) 3817 0.95

Not directly to catheter laboratory 10700 (8010, 16100) 11500 (8790, 17100) 6.4 (5.82, 6.96) 6.53 (5.93, 7.11) 6112 0.75

Observational data for subsequentCABG/PCI

10300 (7580, 15600) 11600 (8840, 17300) 6.4 (5.82, 6.96) 6.58 (6.01, 7.15) 7070 0.85

Numbers in brackets are 95% credible intervals.CABG, coronary artery bypass grafts; ICER, incremental cost-effectiveness ratio; NIAP, National Infarct Angioplasty Project; PCI, percutaneous coronary interventions; QALYs, quality-adjustedlife years.*‘Dominated’ refers to a situation where NIAP is both more costly and less effective than the control.

Heart 2010;96:668e672. doi:10.1136/hrt.2009.167130 671

Acute coronary syndromes

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from

considered to be cost-effective compared with the thrombolysis-based service offered in control sites. Cost-effectiveness isimproved for the subgroup of patients admitted directly to thecatheter laboratories. For the subgroup of patients who weretransferred from non-primary angioplasty hospitals to a primaryangioplasty hospital, the additional treatment delay of 100 minworsens expected outcomes to the extent that thrombolysis-based care is likely to be more effective as well as being less costly.

Strengths and limitationsThis analysis draws on detailed observations of resource use inthe UK NHS setting to estimate the costs of primary angio-plasty and thrombolysis. The analysis reflects actual duration ofhospital stay, the specific types of drugs, including thrombolyticagents used, the consumables that were used including stentsand drug-eluting stents by type and by manufacturer, whereasprevious estimates of cost-effectiveness relied on assumedtypical patterns of usage by procedure.

We examine a pragmatic question by comparing systems ofcare which make use of both primary angioplasty and throm-bolysis to differing degrees, rather than comparing individualtreatments.

The study has several potential limitations. It was not possibleto consider patients who received no reperfusion treatment in thecost-effectiveness analysis, although the cost analysis doessuggest that costs for these patients do not differ between theNIAP and control hospitals. It therefore seems unlikely that thislimitation would significantly influence the results.

National agreements on staffing for out of hours work havebeen used in this analysis but this will vary in practice acrosssites. Current arrangements may not be considered sustainablein the long term, which might increase costs beyond thoseattributed here.

We do not have detailed information on the cost of primaryangioplasty in the control sites where PCI is performed andtherefore assume an equal cost to that at the NIAP sites. Giventhat there are substantial differences in the cost of thrombolytictreatment between the two systems there may also be differ-ences between primary angioplasty patients, although this is notnecessarily the case since the main reason for the difference inthrombolysis cost is length of stay.

Observed rates of angiography, PCI and CABG up to 1 yearwere relatively low in thrombolysed patients. There is evidencethat the rate of early angiography has now risen to approxi-mately 60% in these patients in the UK (J Birkhead, personalcommunication), which may lead to higher costs and improvedoutcomes that would need to be reflected in updated economicanalysis.

Finally, the rate of pre-hospital thrombolysis in the controlcohort was relatively low. It could therefore be argued that wedid not compare the NIAP hospitals with the best alternative,although we did compare them with a common ‘real-world’alternative. If more widespread use of pre-hospital thrombolysisreduced call to needle times in thrombolysis-based practice thenthe potential effectiveness and cost-effectiveness of primaryangioplasty would be reduced. In the CAPTIM trial subgroupanalysis, those presenting in the first 2 h of appearance of

symptoms who were treated with pre-hospital thrombolysisseemed to have better outcomes than those taken directly forprimary PCI.7 This needs further evaluation and a pre-hospitalthrombolysis strategy for the early presenter, combined withmandated rescue or early angiography, is being compared withprimary angioplasty in the current Strategic Reperfusion EarlyAfter Myocardial Infarction trial.

CONCLUSIONPrimary angioplasty-based services that operated in the NHS inEngland in 2005e6 did so cost-effectively. Service arrangementssuch as direct catheter laboratory admissions that reduce theprimary angioplasty time delay further should be promoted,while particular attention should be given to patients outsidethe immediate catchment area of a primary angioplasty hospital.To enable primary angioplasty to be given in a timely andthereby effective and cost-effective manner to these patientsrequires either rapid transfer or alternative arrangements such asambulance bypass of non-angioplasty centres, both of whichwill depend on local arrangements and geography. Where this isnot feasible, thrombolysis in the local hospital or communitymay be the more appropriate treatment option.

Acknowledgements We thank all those who helped in the collection of data forthe NIAP and from the control hospitals. A full list of names is given in onlineappendix 1.

Funding This work was supported by the National Institute for Health ResearchService Delivery and Organisation Programme. The views expressed in this publicationare those of the author(s) and not necessarily those of the NHS, the NIHR or theDepartment of Health. Project reference number (SDO120/2006). The funder had norole in any part of this analysis, reporting of results or the decision to submit the articlefor publication.

Competing interests MS has received consultancy fees and research funding frommanufacturers of interventional cardiology equipment and pharmaceuticals. MDB hassat on advisory boards for companies manufacturing interventional cardiologyequipment as well as a manufacturer of a thrombolytic agent. All other authorsdeclare no competing interests.

Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES1. Department of Health. NHS Coronary Heart Disease National Service Frameworks

2000. http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4094275 (accessed 10 June 2008).

2. Asseburg C, Bravo Vergel Y, Palmer S, et al. Assessing the effectiveness of primaryangioplasty compared with thrombolysis and its relationship to time delay: a Bayesianevidence synthesis. Heart 2007;93:1244e50.

3. Bravo Vergel Y, Palmer S, Asseburg C, et al. Is primary angioplasty cost effectivein the UK? Results of a comprehensive decision analysis. Heart 2007;93:1238e43.

4. Boersma E. and the PCAT-2 Trialists’ Collaborative Group. Does time matter? Apooled analysis of randomized clinical trials comparing primary pertcutaneous coronaryintervention and in-hospital fibrinolysis in acute myocardial infarction patients. EurHeart J 2006;27:779e88.

5. Briggs A, Claxton K, Sculpher M. Decision modelling for health economic evaluation.Oxford: Oxford University Press, 2006.

6. National Institute for Health and Clinical Excellence (NICE). Guide to theMethods of Technology Appraisal. London: NICE, 2008. http://www.nice.org.uk/aboutnice/howwework/devnicetech/technologyappraisalprocessguides/guidetothemethodsoftechnologyappraisal.jsp?domedia¼1&mid¼B52851A3-19B9-E0B5-D48284D172BD8459 (accessed 23 July 2008).

7. Bonnefoy E, Lapostolle F, Leizorovic A, et al. For the CAPTIM study group.Comparison of angioplasty and pre-hospital thrombolysis in acute myocardialinfarction. Lancet 2002;360:825e9.

672 Heart 2010;96:668e672. doi:10.1136/hrt.2009.167130

Acute coronary syndromes

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from

doi: 10.1136/hrt.2009.167130 2010 96: 668-672 originally published online June 8, 2009Heart

 Allan Wailoo, Steve Goodacre, Fiona Sampson, et al. Angioplasty projecteconomic analysis of the National Infarct

anacute ST-elevation myocardial infarction: Primary angioplasty versus thrombolysis for

http://heart.bmj.com/content/96/9/668.full.htmlUpdated information and services can be found at:

These include:

Data Supplement http://heart.bmj.com/content/suppl/2010/04/09/96.9.668.DC1.html

"Web Only Data"

References

http://heart.bmj.com/content/96/9/668.full.html#related-urlsArticle cited in:  

http://heart.bmj.com/content/96/9/668.full.html#ref-list-1This article cites 4 articles, 3 of which can be accessed free at:

serviceEmail alerting

the box at the top right corner of the online article.Receive free email alerts when new articles cite this article. Sign up in

CollectionsTopic

(201 articles)Health policy   � (7898 articles)Drugs: cardiovascular system   �

(2490 articles)Acute coronary syndromes   � (2671 articles)Interventional cardiology   �

(4371 articles)Clinical diagnostic tests   � Articles on similar topics can be found in the following collections

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on September 25, 2014 - Published by heart.bmj.comDownloaded from