Accident and Emergency Nursing (2007) 15, 27–33

Accident and

www.elsevierhealth.com/journals/aaen

EmergencyNursing

Guideline implementation fails to improvethrombolytic administration

Leigh Kinsman CCRN, BHSc, MHSc (Senior Research Fellow) a,*,Kathleen Tori CCRN, BHSc, Grad Dip VET, MHSc (Nurse PractitionerCandidate) b, Ruth Endacott RN, DipN(Lond), MA, PhD (Professor ofClinical Nursing) c, Matthew Sharp RN, BN(Hons), PGDipCritCare, MBus(Management) (Director of Primary Care & Quality Management) d

a School of Rural Health, Monash University, P.O. Box 666, Bendigo, Vic. 3552, Australiab Bendigo Health, Bendigo, Australiac School of Nursing and Midwifery, La Trobe University, Bendigo, Australiad Echuca Regional Health, Echuca, Australia

Received 4 July 2006; received in revised form 29 October 2006; accepted 9 November 2006

SummaryBackground: International randomised controlled trials conducted over the last twodecades have consistently demonstrated improved mortality and morbidity resultingfrom thrombolytic therapy for patients with acute myocardial infarction (AMI). Sub-sequently, evidence-based guidelines have been designed and implemented to opti-mize thrombolytic delivery. The effect of evidence-based clinical guidelines onclinical practice is heavily influenced by strategies used to develop, disseminateand implement those guidelines.Aims: This study evaluated the impact of a collaborative, multifaceted implemen-tation strategy for AMI management guidelines on thrombolytic usage in the LoddonMallee Region, Victoria, Australia.Intervention: The multi-faceted implementation strategy included an inter-disciplinary team representing all treating venues contributing to the contentof the ‘‘Guidelines for the Early Management of Acute Myocardial Infarction’’ fol-lowed by education sessions that coincided with the dissemination of the guide-lines.

KEYWORDSAcute myocardialinfarction;Thrombolytic therapy;Guidelineimplementation

0d

e

965-2302/$ - see front matter �c 2006 Elsevier Ltd. All rights reserved.oi:10.1016/j.aaen.2006.11.004

* Corresponding author. Tel.: +61 3 5440 9022; fax: +61 3 5440 9080.E-mail addresses: leigh.kinsman@med.monash.edu.au (L. Kinsman), ktori@bendigohealth.org.au (K. Tori), r.endacott@latrobe.

du.au (R. Endacott), msharp@erh.org.au (M. Sharp).

28 L. Kinsman et al.

Methods: A retrospective medical records audit 12 weeks before and 12 weeks afterthe intervention was used to evaluate the impact on proportion of those patients eli-gible and receiving a thrombolytic and door-to-needle time. Variables of treatingvenue, age, gender, type of AMI, and type of transport to hospital were also mea-sured to determine their impact on results.Results and conclusions: A retrospective audit of 170 medical records found that theintervention appeared to have had no impact on the proportion of patients eligibleand receiving a thrombolytic (74.2% vs. 62.5%: p = 0.275), and door-to-needle time(67.7 min vs. 60.5 min: p = 0.759). Venue specific influences produced a variety ofpatterns in thrombolytic delivery that require further exploration. This suggests thata single solution approach across multiple venues will have limited impact.

�c 2006 Elsevier Ltd. All rights reserved.

Introduction

Compelling evidence from large randomised trialsconducted into the benefits of thrombolytic ther-apy has overwhelmingly confirmed that rapid resto-ration of patency of an occluded artery is the keyto preserving myocardial muscle and survival out-comes (ISIS-2, 1988; GUSTO Angiographic Investiga-tors, 1993; GUSTO Investigators, 1993; Roetigg andTanbae, 2000). That timely, appropriate and effi-cient delivery of thrombolytic drugs to individualssuffering an acute myocardial infarction (AMI) sig-nificantly improves mortality and morbidity isincontrovertible (Letovsky and Allen, 1996; Gibleret al., 2002; Clare and Bullock, 2003; Heathet al., 2003; Quinn et al., 2003; Corfield et al.,2004; Jones, 2005). However, evidence also sug-gests that thrombolytics are not being adminis-tered according to international recommendationsin a significant proportion of AMI cases (Jancin,2000; Priglinger and Huber, 2000; White, 2000;Savage and Channer, 2002; Boland et al., 2003).

Clinical guidelines have become a common toolused to implement evidence into practice, includ-ing recommendations to support the administrationof thrombolytics for AMIs, but have often proven tobe weak or ineffective in influencing clinical prac-tice (Haines and Donald, 1998; Mowatt et al.,2001; Mountain et al., 2002; Savage and Channer,2002; Jowett and Thompson, 2003; Quinn et al.,2003; Schunemann et al., 2004; Eccles et al.,2005).

In order to promote optimal thrombolyticdelivery for AMIs within the Loddon Mallee Regionthe Loddon Mallee Region Emergency and CriticalCare Consultative Committee (LMRECCCC) de-signed a resource folder entitled ‘‘Guidelines forthe Early Management of Acute Myocardial Infarc-tion’’ (Loddon Mallee Region, 2002, Section 3,

p. 1). A multi-disciplinary team (ambulance offi-cers, nursing, medical and management staff)representing all treating venues across the regioncontributed to the planning, content and develop-ment of the guidelines which were designed to beused by ambulance officers, nurses and doctorsdirectly involved in the early management of anAMI. The resource folder contained succinct infor-mation on the diagnosis of an AMI, thrombolyticadministration criteria and protocols, and a sum-mary of the evidence underpinning the contentsof the resource. Education sessions were offeredto all treating venues in the region at the timeof distribution of the guidelines. These sessionswere conducted in three out of the four treatingvenues and attended by the majority of the med-ical staff.

All participating venues had 24-hour emergencynursing cover but no venue utilised nurse-initiatedthrombolysis. All venues had a system of nursingassessment followed by resource and triage alloca-tion. These nurses played a pivotal role establishingthe urgency of each case and judging whether theAMI guidelines should be implemented. There werevarying degrees of emergency medical cover be-tween participating venues.

The Loddon Mallee Region is located in thenorth-west corner of the state of Victoria, Austra-lia, covers 58,965 square kilometers and has a pop-ulation of approximately 294,000 (Department ofHuman Services, 2005). There are four hospitalsresourced to treat AMIs in the region and therewas no capacity to offer percutaneous coronaryinterventions such as angioplasty and coronarystenting at the time of this study.

The specific criteria promoted in the guidelinesfor delivery of a thrombolytic drug were as follows:

‘‘Typical ischaemic chest pain for >20 min and<12 h, plus

Guideline implementation fails to improve thrombolytic administration 29

ST elevation P1 mm in any two ofleads I, II, III, aVL and aVf,

or ST elevation P2 mm in any two or more ofleads V2–V6,

or new left bundle branch block (LBBB),or true posterior infarct,or right ventricular infarct.

Then give thrombolytic drug as soon as possi-ble.’’(Loddon Mallee Region, 2002)

Subsequently, the aim of this before- and -afterstudy was to evaluate the impact of the implemen-tation of the guidelines on the appropriate andtimely administration of thrombolytic drugs forAMIs in the Loddon Mallee Region. The timely andappropriate administration of thrombolytic drugswas also compared by the variables of age, gender,type of AMI, treating hospital and type of transportto hospital.

Methods

Design

A retrospective before- and -after medical recordsaudit (12 weeks before and 12 weeks after) ofthrombolytic usage in AMI patients in 2003 wasconducted.

Inclusion criteria

Following identification of the medical recordsaccording to the ICD code for an AMI, inclusion cri-teria were that the patient had an elevated serumtroponin level indicative of an AMI, the patient’sdischarge diagnosis had been confirmed as an AMIby the treating Medical Officer, and the patientwas treated in the Loddon Mallee Region.

Data collection

Data were collected utilizing a de-identified formand medical records accessed to obtain variablessuch as type of AMI, gender, time period, age,participating venue, type of transport to hospital,patient eligibility for thrombolysis, whether con-traindications existed for thrombolytic therapyand time from arrival at the hospital to commence-ment of thrombolytic. ECG criteria for patienteligibility, as previously described in the introduc-tion, was checked by the primary researcheragainst the physician’s ECG interpretation. Data

were stored and analysed using the StatisticalPackage for Social Sciences (Version 10.5).

Measurements

Compliance with clinical guidelines was measuredin two ways:

1. The proportion of AMI patients meeting the cri-teria (eligible) and receiving a thrombolytic.

2. Door-to-needle time when a thrombolytic wasadministered.

Ethics approval to access medical records wasreceived from the appropriate university and par-ticipating hospitals’ ethics committees. Individualhospitals are not identified.

Results

Sample

One hundred and ninety five medical records whichidentified patients treated for AMI during the studyperiod, were accessed. Of those, 25 were found notto have the principal diagnosis of an AMI. There-fore, data were collected from 170 confirmed AMIcases. No cases were identified where a thrombo-lytic drug was given without meeting the specificcriteria for delivery. Seventy-five cases were eligi-ble for thrombolysis.

The sample for this study being majority male(59%), a mean age of 69 years and with femaleshaving a greater mean age than males (73.9 yearsvs. 65.5 years) appears to be representative ofthe broader population of cases treated for anAMI in Australia (Mountain et al., 2002). Thebefore- and -after samples appeared sufficientlyevenly balanced in terms of numbers anddemographic profile to allow comparison (Table 1).

Comparison of before and afterimplementation of guidelines

There was no statistically significant difference be-tween the before- and -after samples when com-pared by the proportion of those eligiblereceiving a thrombolytic drug (74.2% vs. 62.5%;p = 0.275). There was no significant differencefound in the before- and -after samples whendoor-to-needle time was compared (67.7 min vs.60.5 min; p = 0.759). The results are provided inTable 2.

Table 1 Sample Characteristics (n = 170)

Variable Before n = 81 After n = 89 Overall n = 170

Mean age (years) 69.1 68.8 68.9 (SD ± 14.0)(SD ± 13.5) (SD ± 14.6) Range 34–96

Gender

Male 47 53 n = 100Mean age 62.5(SD ± 11.2)

Female 34 36 n = 70Mean age 71.8(SD ± 13.7)

Type of AMI

Inferior 22 15 n = 37 (22%)Anterior 20 35 n = 55 (32%)Sub-endocardial 18 16 n = 34 (20%)Unspecified 21 23 n = 44 (26%)

Treating hospital

A 24 40 n = 64 (38%)B 4 7 n = 11 (6%)C 26 20 n = 46 (27%)D 16 9 n = 25 (15%)E 11 13 n = 24 (14%)

Met criteria for thrombolytic without contraindications 35 40 n = 75

Table 2 Comparison of pattern of delivery of thrombolytic drugs before and after the implementation process(n = 75)

Before After p Value

Number eligible for thrombolytic drug 35 40Number receiving thrombolytic drug 26 (74.2%) 25 (62.5%) 0.275

Mean time to administration of thrombolytic drug 67.7 min 60.5 min 0.759Range 9–430 min Range 5–450 min(SD ± 79.2) (SD ± 86.8)

Note: The ‘‘after’’ mean time is based on 24 cases as time to administration could not be ascertained in one case.

30 L. Kinsman et al.

Relationships between variables andthrombolytic delivery

Age and genderIndependent sample t-test analysis demonstrated anegative relationship between age and the likeli-hood of a thrombolytic being ordered. That is,the mean age of cases meeting the criteria andhaving a thrombolytic ordered was significantly lessthan when a thrombolytic was not ordered (64.0years [SD ± 13.8] vs. 71.9 years [SD ± 12.4];p = 0.008). There was no statistical relationship be-tween age and time to delivery of thrombolytic.

Measures comparing thrombolytic delivery bygender did not reach statistical significance.

Type of AMIThere was no significant difference according tothe type of AMI for the proportion of those eligiblereceiving a thrombolytic or door-to-needle time.These results are provided in Table 3.

Treating venuesThe differences between hospitals in pattern ofthrombolytic delivery did not reach statistical sig-nificance. These results are provided in Table 4.

Transport to hospitalWhether an ambulance or private vehicle was usedas transport to hospital did not appear to influenceeither ordering of a thrombolytic when compared

Table 3 Comparison of thrombolytic delivery by type of AMI

Type of AMI Proportion receiving ofthose eligible

Percentage (%) Mean time to thrombolyticdelivery (min)

Inferior 16/26 72 47.1 (SD ± 17.8)Anterior 31/39 79 79.8 (SD ± 102.5)Sub-endocardial 1/5 20 18Unspecified 3/5 60 16.3 (SD ± 15.5)

Note: None of the differences in the table above reached statistical significance <0.05.

Table 4 Comparison of thrombolytic delivery by treating hospital

Treating hospital Proportion receiving ofthose eligible

Percentage (%) Mean time to thrombolyticdelivery (min)

A 20/28 71 37.1 (SD ± 22.0)B 2/5 40 54.0 (SD ± 8.5)C 12/23 52 88.1 (SD ± 115.0)D 8/8 100 105.1 (SD ± 46.8)E 9/11 82 61.3 (SD ± 46.8)

Note: None of the differences in the table above reached statistical significance <0.05.

Table 5 Comparison of thrombolytic delivery by type of transport to hospital

Type of transport Proportion receiving ofthose eligible

Percentage (%) Mean time to thrombolyticdelivery (min)

Ambulance 35/50 70 70.2 (SD ± 98.9)Private vehicle 15/24 62 50.7 (SD ± 8.5)Other 1/1 100 65

Note: None of the differences in the table above reached statistical significance <0.05.

Guideline implementation fails to improve thrombolytic administration 31

by v2 analysis (70% vs. 62%; p = .304) or time todelivery when compared by independent t-test(70.2 min vs. 50.7 min; p = .453). These resultsare provided in Table 5.

Discussion

Despite the instigation of local initiatives withinthe Loddon Mallee Region, incorporating collabora-tion and education to optimise awareness of theevidence-based guidelines for early managementof acute myocardial infarction, our study foundthat the implementation strategy appeared to havehad no immediate impact on the timely and effi-cient administration of thrombolytic drugs forAMIs. The lack of impact could not be explainedby the before and after samples exhibiting demo-graphic or disease variables that would impact onthe rate of thrombolytic usage.

Furthermore, variation between treating venuesexisted. These differences in whether a thrombo-lytic was ordered and door-to-needle time did notreach statistical significance, but the differences

between venues would produce differing clinicaloutcomes and appear to be process related. Forexample, venue A had the shortest mean time todelivery (37.1 min) with the proportion of thoseeligible receiving a thrombolytic being 71%. Con-versely, venue D had an extended mean door-to-needle time of 105 min, but 100% of those eligiblereceived a thrombolytic. Further exploration of var-iation between the processes at venues A and Dfound that venue A had protocols supporting theordering of thrombolytics by relatively junior med-ical officers, whilst the protocol at venue D directedsignificant communication with off-site senior med-ical officers before a thrombolytic was ordered.This appears to have contributed to venue A sup-porting quicker, but less accurate decision makingcompared to venue D. This would suggest that indi-vidual influences on AMI management at each par-ticipating venue require further explorationbefore implementation of guidelines, rather thana single solution or intervention across the region.

Age influenced whether or not a thrombolyticwas ordered, despite age not being considered partof the exclusion criteria or contraindications

32 L. Kinsman et al.

provided in the guidelines and education sessions(White, 2000; Priglinger and Huber, 2000; Savageand Channer, 2002). This is further evidence thatthe guidelines disseminated in this region, andthe associated education sessions, had little or noimpact on clinical beliefs and practice.

Study limitations

Data collection was reliant on accurate documen-tation in the medical records and was performedby one investigator only. Strategies to overcomethese limitations included using the ICD code toidentify an AMI, elevated serum troponin level con-firming an AMI, documented discharge diagnosisalso confirming an AMI, and ECG criteria beingchecked independently by the investigator againstthe treating doctor’s documented interpretation.

The relatively small number of patients receiv-ing a thrombolytic precluded the use of time seriesanalysis and means that other influences or pat-terns of thrombolytic delivery may not be detectedby the simple before and after method. A largersample would also have improved external validityof the results.

Conclusions

This study indicated that the implementation pro-cess for dissemination of the guidelines and the re-lated education sessions appeared to have noimpact on the administration of thrombolytic drugsfor AMIs in the Loddon Mallee Region. An explora-tion of local issues impacting on the ordering ofthrombolytics for AMIs is recommended to under-pin future guideline implementation strategies.

The study also demonstrated that the appropri-ate and timely administration of thrombolyticdrugs for AMIs in the region could be improved, asevidenced by:

� Thrombolytics were often not ordered (32% ofreviewed cases) for eligible patients.� There were delays in time to delivery of throm-bolytic drugs.

Recommendations

Local issues obviously have an impact on the ef-fect of strategies used to implement evidence intopractice and need to be rigorously investigated if

guidelines are to positively impact on nursingpractice and clinical outcomes. Future implemen-tation strategies need to consider influences onthrombolytic delivery at each venue rather thanproviding a single solution across multiple emer-gency services.

Acknowledgements

The authors thank the LaTrobe University, Bendigo,Outside Studies Program Committee for their sup-port for this study.

References

Boland, A., Dundar, Y., Bagust, A., Haycox, A., Hill, R., MujicaMota, R., Walley, T., Dickson, R., 2003. Early thrombolysisfor the treatment of acute myocardial infarction: a system-atic review and economic evaluation. Health TechnologyAssessment 7 (15), 1–136.

Clare, C., Bullock, I., 2003. Door to needle time bull’s eye orjust bull? The effects of reducing door to needle times on theappropriate administration of thrombolysis: implications andrecommendations. European Journal of Cardiovascular Nurs-ing 2 (1), 39–45.

Corfield, A.R., Graham, C.A., Adams, J.N., Booth, I., McGuffle,A.C., 2004. Emergency department thrombolysis improvesdoor to needle times. Emergency Medicine Journal 21, 676–680.

Department of Human Services, Victoria (online). http://www.dhs.vic.gov.au (Accessed 6 June 2005).

Eccles, M., Grimshaw, J., Walker, A., Johnston, M., Pitts, N.,2005. Changing the behaviour of healthcare professionals:the use of theory in promoting the uptake of researchfindings. Journal of Clinical Epidemiology 58, 107–112.

Gibler, B.W., Armstrong, P.W., Ohman, E.M., Weaver, W.D.,Stebbins, A.L., Gore, J.M., Newby, L.K., Califf, R.M., Topol,E.J., 2002. Persistence of delays in presentation and treat-ment for patients with acute myocardial infarction: theGUSTO-I and GUSTO-III experience. Annals of EmergencyMedicine 39 (2), 123–130.

GUSTO Angiographic Investigators, 1993. The effects of rTPA, SKor both on coronary artery patency, ventricular function andsurvival after AMI. New England Journal of Medicine. 329,1615–1622.

GUSTO Investigators, 1993. An international randomized trialcomparing thrombolytic strategies for AMI. New EnglandJournal of Medicine. 329, 673–682.

Haines, A., Donald, A., 1998. Making better use of researchfindings. British Medical Journal. 317, 72–75.

Heath, S.M., Bain, R.J., Andrews, A., Chida, S., Kitchen, S.I.,Walters, M.I., 2003. Nurse initiated thrombolysis in theaccident and emergency department: safe, accurate andfaster than fast track. Emergency Medicine Journal 20 (5),418–420.

ISIS-2, 1998. In: Second International Study of Infarct SurvivalCollaboration Group, 1988. Randomised trial of IV SK, oralaspirin, both or neither – 17,187 cases of AMI, TheLancet 349–360.

Jancin, B., 2000. Reduced door-to-needle time boosts survival inacute MI- heart attack. Family Practice News 30, 6.

Guideline implementation fails to improve thrombolytic administration 33

Jones, I., 2005. Thrombolysis nurses: time for review. EuropeanJournal of Cardiovascular Nursing 4 (2), 129–137.

Jowett, N.I., Thompson, D.R., 2003. Comprehensive CoronaryCare, third ed. Bailliere Tindall, China.

Letovsky, E., Allen, T., 1996. Initiating thrombolytic therapy foracute myocardial infarction: whose job is it anyway? Cana-dian Medical Association Journal 154 (4), 509–511.

Loddon Mallee Region Emergency Critical Care ConsultativeCommittee (LMRECCCC) 2002. Guidelines for the EarlyManagement of Acute, Myocardial Infarction. Bendigo,Department of Human Services.

Mountain, D., Jelinek, G.A., O’Brien, D.L., Ingarfield, S.L.,Jacobs, I.G., Lynch, D.M., 2002. Thrombolysis for acutemyocardial infarction in Australasia 1999. Emergency Medi-cine 14 (3), 267–274.

Mowatt, G., Grimshaw, J., Davis, D.A., Mazmanian, P.E., 2001.Getting evidence into practice: the work of the CochraneEffective Practice and Organization of Care group (EPOC).Journal of Continuing Education in the Health Professions 21,55–60.

Priglinger, U., Huber, K., 2000. Thrombolytic therapy in acutemyocardial infarction. Drugs and Aging 16 (4), 301–312.

Quinn, T., Allan, T.F., Birkhead, J., Griffiths, R., Gyde, S.,Gordon Murray, R., 2003. Impact of a region-wide approachto improving systems for heart attack care: the WestMidlands thrombolysis project. European Journal of Cardio-vascular Nursing. 2 (2), 131–139.

Roetigg, M.L., Tanbae, P., 2000. Emergency management ofacute coronary syndrome. Journal of Emergency Nursing 26(6), S1–S42.

Savage, M.W., Channer, K.S., 2002. Improving the managementof acute myocardial infarction. British Medical Journal 325,1185–1186.

Schunemann, H.J., Cook, D., Grimshaw, J., Liberati, A.,Heffner, J., Tapson, V., Guyatt, G., 2004. Antithrombolyticand thrombolytic therapy: from evidence to application.Chest: Cardiopulmonary and Critical Care Journal 126 (3),688–696.

White, H.D., 2000. Thrombolytic therapy in the elderly. TheLancet 356, 2028–2032.