European Journal of Trauma and Emergency Surgery

, Volume 37, Issue 1, pp 73–78

Prehospital triage, discrepancy in priority-setting between emergency medical dispatch centre and ambulance crews

  • A. Khorram-Manesh
  • K. Lennquist Montán
  • A. Hedelin
  • M. Kihlgren
  • P. Örtenwall
Original Article

DOI: 10.1007/s00068-010-0022-0

Cite this article as:
Khorram-Manesh, A., Lennquist Montán, K., Hedelin, A. et al. Eur J Trauma Emerg Surg (2011) 37: 73. doi:10.1007/s00068-010-0022-0

Abstract

Background

The timely provision of emergency medical services might be influenced by discrepancies in triage-setting between emergency medical dispatch centre and ambulance crews (ACR) on the scene resulting in overloaded emergency departments (ED) and ambulance activities. The aim of this study was to identify such discrepancies by reviewing ambulance transports within a metropolitan city in the western region of Sweden.

Methods

All data regarding ambulance transports in Gothenburg, Sweden, during a 6-month period were obtained and analyzed by reviewing the available registry.

Results

There was a discrepancy between emergency medical dispatch centre and ACR in priority setting, which may result in a number of unnecessary transports to the hospital with consequent overloading of ED and a negative impact on ambulance availability.

Conclusion

Appropriate ambulance use is one important part of emergency preparedness. Overuse results in decreased emergency medical services (EMS) availability and ED-overcrowding. Several factors, such as an imprecise triage system and increased public demands, may influence such overutilization. Improving the triage system and comprehensive public education on appropriate use of ambulances are two important steps toward a better use of national EMS resources.

Keywords

Ambulances Emergency Medical Services Triage 

Introduction

Recently, it was reported that the hospital-related incidents within the western region of Sweden (Region Västra Götaland), with over 1.5 million inhabitants, have increased more than 15 times from four incidents in 2006 to 65 incidents in 2008 [1]. One reason for such increase may be overloaded emergency departments (ED), which in turn may be due to increased inflow of patients, who in part are transported to ED by ambulances. The timely provision of emergency medical services (EMS) is one of the key factors in the management of emergency situations. Better planning and integration of EMS, at all levels (regional, national) is essential for maintaining a secure and reliable public healthcare system [1, 2]. ED-overcrowding, which is also linked to a hospital’s resource constraints, may result in ambulance diversions and overloading of ambulance capacity and endanger patients’ safety and disaster preparedness [1, 3, 4, 5, 6, 7, 8].

In Sweden, the majority of persons calling the emergency medical dispatch centre (SOS Alarm) for help are provided with ambulance transport. In earlier published reports, the appropriateness of ambulance transports has been evaluated based on EMS personnel triage on the scene or the physicians’ diagnosis at the ED [9, 10]. However, an ambulance transport actually starts at the emergency medical dispatch centre (EMDC), where the dispatchers set the priorities and decide on transport options, based on the information given through the telephone call. Many of these transports end at the hospitals ED, but are not always medically justified. There are also reports indicating an increasing number of ambulance transports, substituting taxi-transports, used by patients that are not critically ill and, as mentioned before, many ambulances may also be diverted to nearby hospitals by EMDC when the target hospital is overloaded [1, 10, 11]. Thus, one key factor influencing the activities of ambulances and overloading of ED is a correct triage coordinated between EMDC and ambulance crews (ACR) to minimize discrepancies between the first priority setting at the EDMC, at the scene and in the ED.

The aim of this study was to identify such discrepancies by reviewing ambulance transports within a metropolitan city in the western region of Sweden during a 6-month period.

Materials and methods

General

All registered ambulance transports in Gothenburg during a 6-month period (2008) were obtained by reviewing Ambulink™ [12]. AmbuLink is a web-based ambulance medical file, in which all information about patients’ medical conditions and all activities during the transports are registered. The study population consists of normal population in Gothenburg (population around 0.5 million people, age > 1 year). To obtain exact age of the study population, thorough study of each medical file is necessary as the registry itself does not offer this information. For the present study the following variables were extracted and studied;
  • Total number of transports

  • Type of transports

  • Priorities set by the emergency dispatch centre

  • Priorities set by the ambulance personnel on site

  • Severity and grade of the patient’s condition at the time of arrival in the ED

  • The distribution of different diagnosis and specialties

Triage methods

The first priority is set by the ambulance dispatcher (EMDC). All ambulance dispatchers have formal training in triaging and are also able to give treatment directions before ambulances reach the scene. A printed “index” is used as adjunct to perform a structured interview of the caller, to set a priority and give a treatment direction. The index was initially produced in the USA in the late 1970s and was applied in many states. It was later adapted by Norwegian prehospital authority for use in Norway, and around 1995 it was reviewed and reproduced by national medical specialist groups and committees in Sweden for national use. To the best of our knowledge the index has not been validated internally in Sweden, but there are publications from other countries dealing with its accuracy [13, 14]. Three priorities are used for ambulance transports; (1) P1: Acute life threatening condition; immediate dispatch (90 s) with blue lights and sirens. (2) P2: Acute but not life (ANL) threatening; dispatch as soon as an ambulance becomes available. (3) P3: Other missions; can wait up to 90 min. (4) P4: A forth category is also used for patients needing a litter transport, but without any need for medical care or monitoring (“stretcher taxi”). Based on the priorities and in consequent orders, different units are dispatched.

The second priority is set by ACR. All units are staffed with at least one registered nurse (RN). Some units have additional RNs (specialist nurses in anesthesiology) and physicians (anesthesiologist). The same triaging criteria were used by ACR when assessing the patients on the scene. If possible both of these priorities were compared with the patient’s condition on arrival at ED.

Ambulance types

There were four different transport types; OLA, an ambulance equipped and staffed for the most severe cases. Ordinary emergency ambulances (EA), usually having one nurse, specialist in emergency medicine or anesthesia and an ambulance technician. Patient transport vehicles (PT) deal mainly with patients who do not need special medical care in transit. AD is a special transport accompanied by a physician and can offer more treatment on the scene and during transport.

Statistic

Data was transferred to Excel (Microsoft Corp, USA). No statistical analyses were performed.

Results

Type of transports versus priorities made by EMDC versus ACR

A total number of 35,893 transports were identified. Transports registered twice or unclear were excluded. The remaining 27,318 transports (patients) were prioritized by EMDC as P (priority) 1; 9,208, P2; 7,918, P3; 3,154, and P4; 7,038. The reported priorities made by ACR on site are shown in Table 1. Of 9,208 patients with P1, 2,512 (27%) were found to have priority 1 by ACR. More than 50% of P1 patients were found to be P2. Additionally 301 patients were prioritized as P1 of ACR, but were primarily prioritized as P2–4 by EMDC [27% (P1), 53% (P2), 82% (P3), and 99% for P4].
Table 1

Distribution between type of transports and each priority

*

Total

P1 = 9,208

P2 = 7,918

P3 = 3,154

P4 = 7,038

**

P1

P2

P3

P4

P1

P2

P3

P4

P1

P2

P3

P4

P1

P2

P3

P4

OLA

1,227

658

296

148

74

4

11

1

0

3

5

25

1

1

0

0

0

EA

19,161

1,843

4,629

1,207

327

271

5,334

1,779

497

18

168

2,540

363

0

2

12

171

PT

6,887

4

1

0

6

0

1

4

0

0

0

2

20

4

2

31

6,812

AD

43

10

7

1

0

1

8

4

3

0

1

7

1

0

0

0

0

Total

27,318

2,512

4,933

1,356

407

276

5,354

1,788

500

21

174

2,574

385

4

4

43

6,987

% Right priority

 

27

54

   

53

18

   

82

    

99

It also shows the total number primarily prioritized as P1–4 by the emergency dispatch center (* first row) and the secondary priorities based on the outcome **. P1 acute life threatening conditions. P2 acute but not life threatening conditions. P3 other missions; can wait up to 90 min. P4 patients needing a litter transport, but without any need for medical care or monitoring

OLA advanced emergency ambulances, EA emergency ambulance, PT patient transport vehicle, AD ambulance with a doctor

Priorities 1–4 (Table 1)

The number of patients prioritized as P1 by EMDC was 9,208 out of 27,318 (33.8%) patients. ACR prioritized 2,813 (10.3%) patients as P1; 2,512 in P1, 276 in P2, 21 in P3 and 4 in P4 groups. In total 27% of priority-setting in P1 made by EMDC was in agreement with ACR (73% over-triage). Another 301 patients prioritized as P1 by ACR were grouped under P2–4 (1% of all patients). The number of P2 prioritized by EMDC was 7,918 out of 28,318 (28.9%) patients. ACR prioritized 10,465 (38.3%) patients as P2; 4,933 in P1, 5,354 in P2, 174 in P3 and 4 in P4 groups. In total 53% of priority-setting in P2 made by EMDC was in agreement with ACR and another 5,111 patients were grouped as P1, P3 and P4 (18.0% over-triage and 6.0% under-triage). The number of P3 prioritized by EMDC was 3,154 out of 28,318 (11.6%) patients. ACR prioritized 5,761 (21.1%) patients as P3; 1,356 in P1, 1,788 in P2, 2,574 in P3 and 43 in P4 groups. Around 82% of priority-setting in P3 made by EMDC was in agreement with ACR and another 3,187 patients were grouped as P1, P2 and P4 (1.5% over-triage and 0.1% under-triage). The number of P4 prioritized by EMDC was 7,038 out of 28,318 (25.7%) patients. ACR prioritized 8,279 (30.3%) patients as P4; 407 in P1, 500 in P2, 385 in P3 and 6,987 in P4 groups. Around 99% of priority-setting in P4 made by EMDC was in agreement with ACR (1.0% under-triage). Another 51 patients were grouped as P1, P2 and P3.

Severity and grade of disease versus priority and definitive priority outcome (Table 2)

All patients were prioritized at the arrival to the ED as; AL, acute and life threatening (P1); ANL, acute and not life threatening (P2), and NOA, not acute (P3 and P4). The table shows all prioritized patients and their definitive priority on admission to the ED. Information could be obtained in 21,421 patients. A total number of 2,023 patients were classified as AL, 10,755 as ANL and the remaining 8,643 as NOA. There was a discrepancy between priorities in all three groups. During this period different triage scoring systems were used in the regions different hospitals. However, the condition of all patients on arrival at ED was registered as AL. ANL, and NA at all hospitals (see “Materials and methods”).
Table 2

Severity and grade of disease versus priority and definitive priority outcome

Priority emergency dispatch centre

Priority made by ambulance crews

Total number

AL

ANL

NOA

1

1

2,308

1,717

575

16

8,422 patients

2

4,801

47

4,332

422

3

1,257

3

174

1,080

4

56

0

0

56

2

1

274

210

63

1

9,814 patients

2

5,280

28

4,789

463

3

3,955

4

473

3,478

4

305

0

4

301

3

1

21

11

7

3

3,041 patients

2

169

3

153

13

3

2,536

0

179

2,357

4

315

0

1

314

4

1

1

0

1

0

144 patients

2

3

0

3

0

3

13

0

0

13

4

127

0

1

126

Total

 

21,421

2,023

10,755

8,643

AL acute life-threatening, ANL acute not life threatening, NOA not acute

Priorities, disease severity versus institutional diagnosis (Table 3)

The distribution of different priorities and the diagnosis on admission to the hospital for 21,418 patients are presented in this table. Data for three patients were not available. Most of the P1 patients were acute (life threatening or not). Patients in need of surgical treatment and those with circulatory diseases were over-represented. All patients were admitted irrespective of their priorities and the severity of their disease.
Table 3

Relation between first priority-setting, patient’s condition and diagnosis obtained at the emergency departments

P

Status

NO

AIR

CIRC

TOX

GYN

INF

SURG

MED

NEURO

ORT

PSYC

TRAU

OTHER

Sum

1

AL

1,767

166

507

230

28

16

90

42

300

34

3

320

31

1,767

ANL

5,062

611

1,070

332

111

105

477

346

894

172

65

768

111

5,062

NOA

1,588

171

221

106

31

42

150

155

243

58

79

215

117

1,588

Sum

8,417

948

1,798

668

170

163

717

543

1,437

264

147

1,303

259

 

2

AL

241

39

69

19

3

1

45

8

28

6

1

16

6

241

ANL

5,330

519

638

227

82

145

1,071

328

646

722

46

735

171

5,330

NOA

4,246

256

324

127

55

143

833

463

518

571

119

477

360

4,246

Sum

9,817

814

1,031

373

140

289

1,949

799

1,192

1,299

166

1,228

537

 

3

AL

14

4

2

0

0

0

2

1

1

0

0

1

3

14

ANL

340

24

44

2

5

21

63

42

31

51

6

26

25

340

NOA

2,686

76

426

25

28

140

478

311

167

387

130

97

421

2,686

Sum

3,040

104

472

27

33

161

543

354

199

438

136

124

449

 

4

AL

0

0

0

0

0

0

0

0

0

0

0

0

0

0

ANL

5

1

2

0

0

0

1

1

0

0

0

0

0

5

NOA

139

5

7

1

1

3

13

7

4

16

11

3

68

139

Sum

144

6

9

1

1

3

14

8

4

16

11

3

68

 

Total

21,418

             

All patients transported by ambulances are admitted to the hospital irrespective of their disease severity or their primary priority-setting

AIR airway symptoms, CIRC circulatory symptoms, TOX intoxication, GYN gynecological condition, INF infectious disease, SURG surgical conditions, MED medical condition, NEURO neurological conditions, ORT orthopedic conditions, PSYC psychiatric conditions, TRAU trauma, AL acute life threatening, ANL acute not life threatening, NOA not acute

Discussion

Our data show that there is a discrepancy between EMDC and ACR in current priority setting. Consequently, a number of transports to the hospital are unnecessary and may cause overloading of ED and negative impact on ambulance availability. In addition over-triaging patients lead to waste of resources, while under-triaging may endanger patient safety and cause unfavorable outcome. Readily available EMS units are one of the important key factors for national and regional preparedness in case of major incidents/disaster. Overloading ED and inappropriate use of ambulances reduce such functionality [1, 15, 16].

As shown in the current study, a discrepancy between EMDC and ACR in setting the priorities is one reason for improper use of resources. This discrepancy could be due to a substantial safety margin in the priority assessment made by EMDC operators. Although over-triage is primarily accepted by ACR, it consequently results in 30% of unnecessary transports [10]. In our study, the difference between EMDC and ACR in priority-setting varied between 70% in P1 and 1% in P4. Although the national expert panel on field triage in the USA has reported a normal rate of under- and over-triage of 1–5 and 25–50%, respectively, the goal must be to improve the selection process to downgrade priorities whilst maintaining patient safety [17, 18]. An improved system should also enable transport of patients directly from the scene to different healthcare centers based on their needs (“short tracks”). Current policy in our region indicates that almost all patients transported by an ambulance should be delivered to the ED. Many patients are not in need of the hospital facilities. Some might need hospital admission, but don’t need the pre-admission work-up and may bypass ED [11]. Thus, ACR should have the mandate to evaluate the patient and change the priority or refuse a medically unjustified transport. This kind of decision-making is, however, not easy and needs training, practice and legislative changes [19, 20, 21]. Adequate triage setting is also important for the choice of transport. As shown in our study, many over-triaged patients were transported by high quality EMS units. Besides being unnecessary, these are costly alternatives to cheaper and more adequate transports, if the right decision is made early.

Another reason for inappropriate use of ambulances is the increased public demands for being transported to the healthcare facilities. Different studies have shown that there is a discrepancy between healthcare providers' and the public's beliefs and definition of what an appropriate indication for ambulance transport is [9]. A person’s decision to call an ambulance in a non-emergency situation may be influenced by various socioeconomic factors, e.g. age, gender, household income, and possession of a car. On the contrary, other factors, e.g. hesitation to use an ambulance and the knowledge of nearest hospital’s location may limit ambulance overuse [22]. There are other reports indicating that many patients using an ambulance for transport to the hospital, would be willing to consider other means of transport to the ED, if one was offered [23]. In another report a directed campaign to the public for appropriate ambulance use, reduced the ambulance transports during the study period [24]. There is thus enough evidence to believe that public demands may be altered by adequate information and improvement within the metropolitan infrastructure and transport system.

An interesting question is if the presence of a physician (e.g. anesthesiologist) may increase the accuracy of triaging in the prehospital setting. Earlier experience of using physician at dispatch centres has shown the importance of better protocols and not the physician’s presence. However, there are reports indicating that physician staffed ambulances increases the chance of survival and reduces mortality in patients with severe conditions [25, 26, 27]. In our study (Fig. 1), physician staffed ambulances were rarely used. The reason for the scarce use of these units is unknown, but one reason may be the fact that within the urban areas the principal of “load and go”, and not “stay and play”, is often advocated.
Fig. 1

Differences between different types of transport based on the priority-setting by EMDC and ACR. OLA advanced emergency ambulances, EA emergency ambulance, PT patient transport vehicle, AD ambulance with a doctor, EMDC ambulance dispatch centre, ACR ambulance crews, P1 acute life threatening condition, P2 acute but not life threatening conditions, P3 other missions, P4 patients needing a litter transport, but without any need for medical care or monitoring

In conclusion, appropriate ambulance use is one important part of emergency preparedness. Overuse results in decreased EMS availability and ED-overcrowding. Several factors, such as an imprecise triage system and increased public demands, may influence such overutilization. Improving the triage system and comprehensive public education on appropriate use of ambulances are two important steps toward a better use of national EMS resources.

Conflict of interest statement

None.

Copyright information

© Urban & Vogel 2010

Authors and Affiliations

  • A. Khorram-Manesh
    • 1
  • K. Lennquist Montán
    • 1
  • A. Hedelin
    • 1
  • M. Kihlgren
    • 1
  • P. Örtenwall
    • 1
  1. 1.Prehospital and Disaster Medicine CentreGothenburgSweden

Personalised recommendations