Emergency Medicine Australasia

(2005)

17

, 387–391

Blackwell Science, LtdOxford, UKEMMEmergency Medicine Australasia1035-68512005 Blackwell Publishing Asia Pty Ltd 2005174387391Miscellaneous

Helicopter retrieval coordinationS Cameron

et al.

Correspondence: Dr Steve Cameron, Department of Emergency Medicine, Gosford Hospital, PO Box 361, Gosford, NSW 2250, Australia.

Email: stevecameron1@aapt.net.au

Stephen Cameron, MBChB, Emergency Registrar; Peter Pereira, MB BS, FACEM, Director of Emergency Medicine; Richard Mulcahy, MBBCh,

BAO, FACEM, Staff Specialist; Jamie Seymour, BSc (Hons), PhD, Senior Lecturer.

P

REHOSPITAL

C

ARE

Helicopter primary retrieval: Tasking who should do it?

Stephen Cameron,

1

Peter Pereira,

2

Richard Mulcahy,

2

and Jamie Seymour

3

1

Department of Emergency Medicine, Gosford Hospital, Gosford, New South Wales,

2

Department of Emergency Medicine, Cairns Base Hospital,

3

School of Tropical Biology, James Cook University, Cairns,

Queensland, Australia

Abstract

Background:

Cairns Base Hospital utilizes a helicopter retrieval system, which until 2001 had beentasked and staffed by emergency physicians. Since 2001, the ambulance service hasassumed the role of both tasking and staffing the helicopter with intensive care paramed-ics. The present study examines whether the change has resulted in different activationpatterns and patient outcomes.

Method:

A retrospective chart review over 4 consecutive years, comparing the two groups, wascarried out examining 30 day mortality, length of in-hospital stay, transfer rates, theRevised Trauma Score where appropriate and rates of discharge directly from the ED.

Results:

A total of 374 patients were retrieved (211 patients in the emergency physician group from1 April 1999 to 31 March 2001 and 163 in the ambulance group from 1 April 2001 to 31March 2003) over the 4 year period. The demographics of the two groups were similar.Fifty-four patients in the ambulance group (33.1%) were discharged from the ED withoutadmission while 31 (14.7%) were discharged from the physician group. This was statisti-cally significant (

P

=

0.0001). There were no other significant differences between the twogroups. The subgroup of patients admitted also did not show any significant differencein outcomes.

Conclusion:

The similarities in outcomes for admitted patients support the view that both groups havesimilar tasking criteria for high-acuity patients and suggest that paramedics are as effi-cacious as physicians in delivering prehospital care in this group of patients. However,for lower-acuity patients, there is a statistically significant higher rate of clinically unnec-essary taskings by the ambulance group. Given the recent fatal aeromedical accidents inQueensland (Thursday Island 1998, Rockhampton 1999 and Mackay 2003), it would seemprudent to reduce clinically unnecessary retrievals through clinical coordination withappropriately qualified emergency physicians.

Key words:

coordination

,

helicopter

,

primary

,

retrieval

,

tasking

.

S Cameron

et al.

388

Introduction

The Cairns Base Hospital (CBH), Cairns, Australia, isthe major regional referral centre for Cape York Penin-sula with all medical services except neurosurgery, car-diothoracic and paediatric surgery, burns, hyperbaric orspinal units.

The state-run Bell 412 rescue helicopter is availablefor primary retrievals (not from a Queensland Healthfacility) such as roadside and boat retrievals, and forsecondary retrievals that are interhospital transfers(between Queensland Health facilities). These facilitiesrange from outlying hospitals staffed by doctors toremote clinics staffed by nurses alone.

Until 2001, the tasking of the helicopter for all retriev-als in the Cairns District was at the discretion of theemergency physician (EP) at CBH. Since 2001 theresponsibility for tasking for primary retrievals hasrested with the Queensland Ambulance Service (QAS)communications centre (Reefcom) and QAS intensivecare paramedics (ICP) have taken over the role ofthe EP and flight nurse team in attending primaryretrievals.

The EP and flight nurse team on a retrieval canperform all the usual treatment and monitoring facilitiesthat they can provide in the ED. ICP are limited in thatthey can establish peripheral i.v. access, administer i.v.fluids (non-blood products), opiate analgesia, benzodi-azepines in seizures, bronchodilators, antiemetics andbasic drugs for cardiac arrest. They can perform needlethoracostomy, ventilation via mask, laryngeal airwayor ETT (‘cold intubation’). They are unable to performrapid sequence induction, advanced airway proceduresor invasive monitoring of the patient.

Prior to 2001, calls requesting activation of the heli-copter were received by the EP from QAS, QueenslandPolice Service, Royal Flying Doctor Service of Australiaand also the lay public (reefboat operators, islandresorts, commercial fishing vessels etc.). Upon receivinga request for activation, the EP took extensive clinicaldetails before making a decision regarding mode oftransport based on the clinical judgement of the EP.

When the QAS receives a request, via the ‘000’ emer-gency number, for tasking of the helicopter, it utilizesthe same activation process that it uses for road retriev-als. This involves a computer-aided despatch system(RIGHTCAD and AMPDS) operated by a non-clinicalemployee who is supervised by a clinical QAS staffmember.

There is potential for different tasking outcomesgiven the different tasking systems and clinical experi-

ence involved. To examine this, the authors examinedprimary aeromedical taskings carried out over a totalof 4 years by the two groups. The aim was to examinewhether there were any differences in tasking or out-comes between the two groups.

Method

A period of two consecutive years for each group wasexamined in a retrospective chart review. The EP groupwas studied from 1 April 1999 to 31 March 2001 andthe QAS group from 1 April 2001 to 31 March 2003.

The sole criterion for inclusion was any primary task-ing of the helicopter. The cases were identified by the‘mode of arrival’ field recorded in real time on the EDinformation system used in the CBH ED. This fieldspecifies which patients arrive by helicopter and alsowhether the retrieval was carried out by an EP or ICP.

The variables studied were:1. 30 day mortality rate.2. Hospital length of stay (LOS). A day is defined as

bed occupation at midnight of an admitted patientas per the Queensland Health coding guidelines.

3. Those patients transferred to another facility.4. Direct discharge from the ED (including those

patients who self-discharged against medicaladvice).

5. For trauma patients, a Revised Trauma Score (RTS)

1

was calculated using the initial clinical observations(GCS, respiratory rate and systolic blood pressure)in the medical records and ambulance report forms.An emergency registrar was trained as a chart

abstractor to study the charts for the above objectiveobservations and clinical data.

One-way analysis of variance (

ANOVA

) using the Sta-tistical Package for Social Services (

SPSS

), Version 10.0,was undertaken and

P

-values and a

c

2

-test with Yate’scorrection were calculated where appropriate. Non-normally distributed data were normalized using appro-priate transformations when analysed via

ANOVA

.A second trained reviewer collected 40 charts (10.7%)

at random and reabstracted the data while blinded tothe original reviewer’s data. The kappa value was cal-culated using

SPSS

to check for inter-observer reliability.

Results

The ages, sex and proportion of trauma cases in bothgroups were statistically similar (Table 1).

Helicopter retrieval coordination

389

Measured outcomes (Table 2) such as 30 day mortal-ity, hospital LOS and percentage transferred to a facilitywith services not available at CBH (as previously dis-cussed) were also similar. The only statistically signif-icant difference in outcomes related to the greaternumber of patients discharged from the ED in the QASgroup.

The number of trauma cases and RTS were alsosimilar between the two groups (Table 3).

In the subgroup of patients admitted, there was nosignificant difference in the 30 day mortality, LOS,number transferred to another facility, number oftrauma cases or the RTS in the latter (Table 4).

To validate the accuracy of data extraction, a kappavalue was calculated using

SPSS

. The calculated kappavalue of 0.937 shows an excellent level of inter-observerreliability.

Discussion

There is little in the medical literature comparing thetasking of helicopters for primary retrievals. Thosepapers that have been published examine tasking pat-terns of either physician or ambulance retrievals, but do

Table 2.

Measured outcomes

EP QAS Statistics

No. retrievals 211 163

30 day mortality (%) 6 (2.8) 4 (2.5)

P

=

0.817,

c

2

=

0.54

Mean hospital length of stay (range) (days) 2 (1–79) 1 (1–65)

P

=

0.296,

F

=

1.090

No. transferred (%) 19 (9) 13 (8)

P

=

0.724,

c

2

=

0.125

No. discharged from ED (%) 31 (14.7) 54 (33.1)

P

=

0.0001,

c

2

=

17.799

EP, emergency physician; QAS, Queensland Ambulance Service.

Table 3.

Trauma subgroup

EP QAS Statistics

No. trauma cases (%) 113 (53.6) 90 (55.2)

P

=

0.749,

c

2

=

0.102

Mean RTS 7.722 7.730

P

=

0.909,

F

=

0.013

No. trauma patients with maximum RTS (%) 101 (89.4) 81 (90.0)

P

=

0.886,

c

2

=

0.210

No. trauma patients admitted (%) 98 (86.7) 62 (68.9)

P

=

0.002,

c

2

=

9.547

No. non-trauma patients admitted (%) 82 (83.7) 47 (64.4)

P

=

0.004,

c

2

=

8.402

EP, emergency physician; QAS, Queensland Ambulance Service, RTS, Revised Trauma Score.

Table 4.

Admitted patients subgroup

EP QAS Statistics

No. patients admitted 180 109

30 day mortality (%) 3 (1.7) 3 (2.8)

P

=

0.530,

c

2

=

0.394

Mean hospital length of stay (range) (days) 7.12 (1–79) 7.27 (1–65)

P

=

0.919,

F

=

0.010

No. transferred (%) 19 (10.6) 13 (11.9)

P

=

0.719,

c

2

=

0.130

No. trauma cases (%) 98 (54.4) 62 (56.9)

P

=

0.686,

c

2

=

0.163

Mean RTS 7.707 7.696

P

=

0.902,

F

=

0.015

No. trauma patients with maximum RTS (%) 87 (88.8) 58 (93.5)

P

=

0.707,

c

2

=

0.141

EP, emergency physician; QAS, Queensland Ambulance Service, RTS, Revised Trauma Score.

Table 1.

Population characteristics

EP QAS Statistics

No. retrievals 211 163

Mean age

(median)

37 (34) 37 (33)

P

=

0.795,

c

2

=

1.674

Males (%) 135 (64.0) 107 (65.6)

P

=

0.739,

c

2

=

0.111

EP, emergency physician; QAS, Queensland Ambulance

Service.

S Cameron

et al.

390

not compare and contrast the tasking and outcomes ofthe different groups.

For instance, Coats and Newton found that placing aparamedic from the Helicopter Emergency Medical Ser-vice (HEMS) based in London, UK in the ambulancecontrol room reduced the number of unnecessaryretrievals and increased the number of appropriatetaskings.

2

They also stated that the ambulance controlsystems in London were not refined enough to task theHEMS appropriately. The process of helicopter activa-tion in London at that time has many similarities to thatcurrently used by the QAS.

Wills

et al.

reported on a paramedic-staffed, protocol-driven primary retrieval service from John Hunter Hos-pital in New South Wales, Australia.

3

In contrast withHEMS, the majority of retrievals were for minor injuriesand there was seldom benefit for tasking less than35 km from the hospital.

Interestingly, Coats

et al

. found that the HEMS ser-vice with the addition of an appropriately trainedretrieval doctor to the crew, was able to appropriatelyselect patients who would benefit from transportdirectly to a trauma centre.

4

In Queensland the sole responsibility for primaryactivations now (since 2001) rests with the QAS.

The populations in the present study had statisticallysimilar demographics and outcomes. Both taskinggroups have similar activation patterns for patientswho require in-hospital treatment, presumably patientswith higher acuities. The similarity in mortality rates,hospital LOS and transfer to a facility that has theservices that CBH lacks as previously discussed,strongly suggests that despite different training, expe-rience and clinical capabilities, this higher-acuity popu-lation group is equally served by both tasking agencies.

There was a higher number of taskings in the EPgroup. There is a continuing reduction in the numberof retrievals. There is no clear reason for this but itmight be a result of more judicious tasking by bothagencies. Some other factors that might account for thisare the increasing safety restrictions on the divingindustry and the successful public campaigns regard-ing the prevention of jellyfish envenoming in tropicalQueensland.

Although it can be argued that discharge from EDdoes not necessarily equate to lower acuity, it is likelyto be the case. The significantly higher rate of dis-charges of QAS retrievals from the ED strongly sug-gests that there are more retrievals for lower-acuitypatients where an EP is not involved. This suggeststhat EP are more capable of activation according to the

clinical requirements of the patient. Given the risingnumbers of fatalities associated with rotary wing pri-mary retrievals in Queensland (Thursday Island 1998,Rockhampton 1999 and Mackay 2003), this finding isincreasingly pertinent.

A possible explanation for this disparity relates tothe considerable experience required to reject a requestfor activation for clinical reasons. An EP has a distinctclinical advantage in this area allowing them to suggestsafer modes of transport. This is because the EP is notprotocol-driven and, unlike the QAS, does not have torespond to all requests for transport. The EP coordinat-ing the tasking is experienced in retrieval medicine andhas the clinical experience to decide if retrieval isrequired or if the patient can wait to be transported bysome other means or be treated in the field with phoneadvice. QAS is severely limited in this regard as theirprotocols (RIGHTCAD and AMPDS) only provide first-aid advice while despatching. QAS also has an obliga-tion to dispatch a retrieval within a specified timeframe.

Another disparity that needs to be addressed is thatalthough the trauma patients had similar RTS betweenthe two tasking groups, there was a significant differ-ence in the admission rates of these patients. This dis-crepancy is also reflected in the significant numbers ofnon-trauma cases admitted between the two groups.

Although unlikely, it is conceivable that EP arebiased in admitting patients whom they have retrieved.It is highly unlikely that an EP would admit a patientwho doesn’t require it given the chronic problems ofaccess block.

The hospital admission policy has been consistentover the 4 year study period and so it is not a factor.

It has been questioned whether the ICP give bettercare before and during the retrieval, thus allowing morepatients to be discharged directly from the ED. This isunlikely as ICP are much more limited in the clinicalcare they can give as compared with an EP.

A more likely reason for the discrepancy is that theRTS used was not an accurate predictor of what typesof patients required admission. This is supported by thefact that 101 out of the 113 (89.4%) trauma cases in theCBH group and 81 out of the 90 (90%) trauma cases inthe QAS group had the maximum RTS value possible.This equates to a survival probability of 98.8%.

5

Theauthors suggest that another system such as a modifiedApache 2 score (measuring the initial parameters thatare available at presentation) might have been a betterpredictor as to which patients required admission.Unfortunately, the data available to the authors did not

Helicopter retrieval coordination

391

include all the variables to use such a system and theauthors suggest that future prospective studies shouldutilize this.

The similarities in both groups in admitted patientoutcomes support the view that for the majority ofpatients it is appropriate to send an ICP to primaryretrievals rather than an EP. This is not the conclusionreached in the recent literature review by Garner inwhich only 2 out of 12 papers showed no difference inoutcome between physician and paramedic staffing ofprehospital helicopter retrievals.

6

The other 10 papersdemonstrated a survival benefit with physicians staff-ing the helicopter. However, it is acknowledged that themeasures used in the present study are crude and aprospective evaluation using more focused clinicalparameters is indicated to shed further light on thiscontroversial topic

7–10

as also advocated recently byRashford and Myers.

11

It would also have been beneficial for the dischargedfrom ED patient group to have been followed up as totheir outcome. Unfortunately, this was outside the scopeof the present study but should be addressed by futureprospective studies.

The authors acknowledge that this was a retrospec-tive study, which did not include requests for retrievalsthat were declined in the two groups. It would also beenuseful to be able to have abstracted other clinical vari-ables that would have allowed a more comprehensiveseverity scale to have been used.

Conclusion

There are three major conclusions that can be drawnfrom this study:1. Helicopter tasking for primary retrievals can be per-

formed by ambulance or medical personnel.The results show that the activation capabilities arethe same for sick patients and that there is no differ-ence in admitted patient outcomes when an ICP issent on a primary retrieval instead of an EP. Theauthors suggest that the tasking of an ICP to pri-mary retrievals is appropriate safe practice for themajority of patients.

2. There is no difference in outcome when prehospitalcare is performed in a helicopter by paramedics ordoctors. However, there is a statistically significantincrease in the number of patients who are able tobe discharged directly from the ED when a helicopterretrieval service is tasked by an ambulance servicerather than an EP.

3. Tasking of a helicopter into the prehospital environ-ment is more safely carried out by appropriatelytrained medical staff.To reduce the need for unnecessary and potentiallydangerous missions the authors suggest that anappropriately trained EP should coordinate all aeromedical retrievals.

Acknowledgements

Thanks to Carl Houlihan of the QAS for helping sourcemissing QAS ambulance report forms.

Competing interests

None declared.

Accepted 16 March 2005

References

1. Champion HR, Sacco WJ, Copes WS et al. A revision of thetrauma score. J. Trauma 1989; 29: 623–9.

2. Coats TJ, Newton A. Call selection for the helicopter emergencyservice: implications for ambulance control. J. R. Soc. Med. 1994;87: 208–10.

3. Wills VL, Eno L, Walker C et al. Use of an ambulance-basedhelicopter retrieval service. ANZ J. Surg. 2000; 70: 506–10.

4. Coats TJ, Wilson AW, Cross FW. On-scene medical decisionmaking and overtriage. Br. J. Surg. 1993; 80: 1291–3.

5. Trauma Scoring, Revised Trauma Score. Available from URL:http://www.trauma.org/scores/rts.html [Accessed 24 August2003].

6. Garner AA. The role of physician staffing of helicopter emer-gency medical services in prehospital trauma response. Emerg.Med. Australas. 2004; 16: 318–23.

7. Baxt WG, Moody P. The impact of a physician as part of theaero medical prehospital team in patients with blunt trauma.JAMA 1987; 257: 3246–50.

8. Bartolacci RA, Munford BJ, Lee A et al. Air medical sceneresponse to blunt trauma: effect on early survival. Med. J. Aust.1998; 169: 612–16.

9. Cameron PA, Flett K, Kaan E et al. Helicopter retrieval of pri-mary trauma patients by a paramedic helicopter service. ANZ J.Surg. 1993; 63: 790–7.

10. Hamman BL, Cué JI, Miller FB et al. Helicopter transport oftrauma victims: does a physician make a difference? J. Trauma1991; 31: 490–4.

11. Rashford S, Myers C. Optimal staffing of helicopter emergencymedical services is controversial. Emerg. Med. Australas. 2004;16: 269–70.