ORIGINAL ARTICLE

A high proportion of prehospital emergency patients are not

transported by ambulance: a retrospective cohort study in
Northern Finland
M. Hoikka1,2, T. Silfvast3 and T. I. Ala-Kokko1,2
1
Medical Research Center, Research Group of Surgery, Anesthesia and Intensive Care, University of Oulu, Oulu, Finland
2
Division of Intensive Care, Department of Anesthesiology, Oulu University Hospital, Oulu, Finland
3
Emergency Medical Service, Department of Emergency Medicine, Helsinki University Central Hospital and University of Helsinki, Oulu, Finland

Correspondence Background: The number of missions in the emergency medical

M. Hoikka, Department of Anesthesiology, services (EMS) has increased considerably in recent years. People
Division of Intensive Care, Oulu University
are requesting ambulance for even minor illnesses and non-medi-
Hospital, PO BOX 21, 90029 OUH, Finland
E-mail: [email protected]
cal problems, which is placing financial and resource burdens on
the EMS. The aim of this study was to determine the rate of non-
Funding transportation missions in Northern Finland and the reasons for
This study was funded by an EVO grant from these missions.
Oulu University. Methods: All ambulance missions in two hospital districts in
Northern Finland during the 6-month period of January 1 through
Conflicts of interest
June 30, 2014, were retrospectively evaluated from the EMS
All authors declare that they have no conflicts
of interests.
charts to identify missions in which the patients were not trans-
ported by the EMS. The non-transportation rates and reasons
Submitted 2 March 2017; accepted 4 March were calculated and expressed as percentages.
2017; submission 18 August 2016. Results: In 41.7% of the 13,354 missions, the patient was not
transported from the scene by an ambulance. After a medical
Citation assessment and care was provided by the EMS, 48.2% of these
Hoikka M, Silfvast T, Ala-Kokko TI. A high
non-transport patients were evaluated as not needing further
proportion of prehospital emergency patients
are not transported by ambulance: a
treatment in the emergency department and were directed to con-
retrospective cohort study in Northern Finland. tact the municipal health care center during office hours. There
Acta Anaesthesiologica Scandinavica 2017 was no need for any medical care in 39.9% of non-transportation
missions.
doi: 10.1111/aas.12889 Conclusion: This study showed a high rate of EMS missions
resulting in non-transportation in two hospital districts in North-
ern Finland. In the majority of these missions there was no need
for emergency admission to an emergency department or for any
medical care at all. These findings indicate that an improvement
in the dispatch process and primary care resources might be of
benefit.

Editorial Comment
Ambulance resources are under pressure due to a growing number of requests for services. This
study reports a high frequency of ambulance missions not resulting in patient transportation from
two districts in Northern Finland. Optimization of patient triage precision, ambulance dispatch,
and sufficient patient access to local primary health care resources are needed to maintain ambu-
lance availability for appropriate missions.

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 549
 M. HOIKKA ET AL.

systems. Secondarily, we wanted to explore

The number of missions in the emergency medi-
whether there is a relationship between the
cal services (EMS) has increased considerably in
availability of, and the distance to, health care
recent years.1,2 People increasingly contact emer-
services and the use of EMS.
gency medical communication centers (EMCCs)
and request an ambulance for even minor ill-
nesses or injuries, as shown by a recent review.3 Methods
Adequate risk assessment and emergency dis-
This retrospective registry study was carried out
patching play key roles in EMS utilization,
in two similar hospital districts and EMS sys-
EMS resource management and costs.4 Triage at
tems (Kainuu and L€ansi-Pohja) in Northern Fin-
dispatch has to be sufficiently sensitive to iden-
land (Fig. 1).
tify true life-threatening situations, but over-
In Finland, there is a national dispatch
triaging and merely sending ambulances on
authority with six regional EMCCs. The com-
request may lead to an inappropriate use of
mon European emergency number 112 is used
available EMS resources.
for all emergencies requiring urgent assistance,
In recent decades, the rate and nature of non-
whether from police, rescue, medical care or
transportation EMS missions have changed.
social services. EMCC personnel receive a
Several EMS providers in the UK did not
national 18-month formalized training without
previously allow EMS crews to leave patients
an official status as health care professionals.
non-transported, unless the patient refused
transportation.5 Such a policy may lead to over-
crowding of the emergency departments; for
example, in the United States, 70% of trans-
ported patients are discharged home from the
emergency departments.6 Because the health
care system is imposing a widened range of
tasks and interventions to be performed by the
EMS, the scope of EMS practice has changed,
and guidelines for non-transportation situations
have been implemented.5,7 In Finland, guideli-
nes enable EMS providers to make liberal non-
transportation decisions based on the patient’s
actual needs, which may lead to a high propor-
tion of missions in which transportation by
ambulance is not indicated.
In Finland, municipal health care centers are
responsible for urgent outpatient medical care
during office hours (Mon–Fri 8:00 am to
4:00 pm), whereas care outside office hours is
centralized to the regional hospitals. The hospi-
tal network is currently undergoing a reform,
which creates larger but fewer units for the pro-
vision of emergency care overall. For many
patients, this means longer distances to the clos-
est outpatient clinic or hospital. In Northern
Finland this will result in distances longer than
100 km for some patients, which may affect the
use of EMS services.
The main objective of this study was to deter-
mine the rate and causes of non-transportation
missions in two Finnish emergency medical Fig. 1. Map of Northern Finland and the hospital districts.

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

550 ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
 NON-TRANSPORTATION IN EMS

The same EMCC dispatcher works as a call han- regions was obtained from the National Institute
dler as well as a dispatcher for all authorities for Health and Welfare, which produces a range
throughout the emergency call. All incoming of statistics in the fields of social welfare and
calls for medical emergencies are assessed health care to support decision-making, devel-
according to a criteria-based, nationally stan- opment and research. As a statistical authority,
dardized dispatch protocol. The EMS calls are it is responsible for the maintenance and devel-
prioritized in four categories, A, B, C, and D, opment of statistical and register resources.
where A indicates an evident or suspected The study design was observational and no
life-threatening situation, B other high-risk clinical interventions were performed. There-
situation, C other urgent situation, and D a fore, according to the local policies, approval of
non-urgent situation. the local ethics committee was waived. Permis-
The study areas consist mostly of suburban sion to carry out the study was obtained from
and rural populations. A total of 140,000 inhabi- the Hospital Districts (March 20, 2014; April 8,
tants, representing 2.6% of the Finnish popula- 2014) and the Office of Data Protection
tion, live in these areas with a population Ombudsman (dnro 719/4225/2014).
density of 4.7 inhabitants per square kilometer.
There are six to eight municipal health care cen-
Statistical analyses
ters and one regional hospital in both districts.
Both districts are covered by a single EMCC Statistical analyses were performed with SPSS
(Oulu) which dispatches a total of 35,000 EMS Statistics, version 22 (IBM Corp., Armonk, NY).
missions in the study areas annually. The dis- Data were expressed as mean with standard
tricts share similar three-tier EMS systems, orga- deviation, unless otherwise stated. Categorical
nized by the hospital districts. The EMS variables were expressed as percentages, and
providers have written guidelines for document- Fisher’s exact test was used for statistical com-
ing the reasons for non-transportation situa- parison. A two-tailed P value of < 0.05 was con-
tions. The decisions to not transport the patient sidered statistically significant. The Spearman’s
are made after examining the patient and correlation coefficient was used to evaluate the
excluding disturbance of vital functions, and association between distance to the emergency
after consultation with an on-call emergency department and rate of non-transportation mis-
department physician according to protocol. sions.
The data for this study were collected retro-
spectively from all EMS charts in the two study
Results
districts between January 1 and June 30, 2014.
In the Kainuu region, the main author manually A total of 13,354 EMS missions fulfilled the
transferred the data from the paper EMS charts inclusion criteria during the 6-month study per-
to the statistical program, whereas in the L€ansi- iod (Fig. 2). Three-quarters (10,332) of the mis-
Pohja region the data were electronically trans- sions were due to illness or disease, while the
ferred from the EMS database (Merlot Medi, remaining missions were related to various trau-
CGI, Canada). All dispatched EMS missions mas. Of the patients, 50.7% were male, and the
within every priority category were included, median age of all patients was 68 years.
whereas secondary (inter-facility) transports and The rates and the reasons for non-transporta-
missions of home care assistance were excluded. tion missions are presented in Table 1. There
The EMS missions that were cancelled before were 5570 missions (41.7%) that did not lead to
patient contact were also excluded. Collected transportation by ambulance. In almost half of
data included priority and dispatch code, demo- them (48.2%), the patient was evaluated by the
graphic data and the non-transportation code. EMS to not require acute treatment in the emer-
Patients with multiple non-transport missions gency department and was instructed to contact
were also identified. Distances to hospital were the municipal health care center during office
calculated using the municipal population cen- hours. In another 1891 patients (33.9% of the
troid, not the patient’s actual home address. non-transported patients), there was no need for
General information on health care use in the medical care at all, and 509 patients (9.1% of
Acta Anaesthesiologica Scandinavica 61 (2017) 549–556
ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 551
 M. HOIKKA ET AL.

Excluded missions 2823

Total missions Inter-facility transports 2220
16 177 Home care missions 408
Missions aborted by EMCC 195

Evaluated missions
13354

Priority A Priority B Priority C Priority D

642 missions 3412 missions 5962 missions 3338 missions

197 (30.7%) 1189 (34.8%) 2555 (42.9%) 1629 (48.8%)

non-transported non-transported non-transported non-transported

Fig. 2. Flowchart of study selection and distribution of missions by priority categories and non-transportation rates. Priority A, evident or
suspected life-threatening situation; priority B, other high-risk situation; priority C, other urgent situation; priority D, non-urgent situation.

Table 1 The rates and reasons for non-transportation missions categorized according to the time of the emergency medical service missions
(during office hours and outside office hours) and the urgency (priority groups A, B, C and D). The total number of missions was 13,354.

The reasons for non-transportation missions

N (percent within non-transport)

Patient
Number of Patient handed
non-transported Non-transportation Medical care found Patient over to
missions/all rate provided at No need of Transport by Patient dead at not other
missions Percent (95% CI) the scene medical care other vehicle refusal the scene found authority

5570/13,354 41.7 (40.9–42.5) 2686 (48.2) 1891 (33.9) 509 (9.1) 194 (3.5) 123 (2.2) 93 (1.7) 72 (1.3)
Time of the mission
During office 1157/3943 29.3 (27.9–30.7) 525 (45.4) 363 (31.4) 123 (10.6) 53 (4.6) 48 (4.1) 21 (1.8) 23 (2.0)
hours
Outside office 4414/9411 46.9 (45.9–48.0) 2161 (49.0) 1528 (34.6) 386 (8.7) 141 (3.2) 76 (1.7) 72 (1.6) 49 (1.1)
hours
Urgency
Priority A 197/642 30.7 (27.0–34.5) 90 (45.7) 15 (7.6) 24 (12.2) 10 (5.1) 54 (27.4) 2 (1.0) 2 (1.0)
Priority B 1189/3412 34.8 (33.2–36.4) 632 (53.2) 318 (26.7) 70 (5.9) 59 (5.0) 66 (5.6) 25 (2.1) 19 (1.6)
Priority C 2555/5962 42.9 (41.6–44.2) 1295 (50.7) 840 (32.9) 245 (9.6) 87 (3.4) 2 (0.1) 41 (1.6) 45 (1.8)
Priority D 1629/3338 48.8 (47.1–50.5) 669 (41.1) 718 (44.1) 170 (10.4) 38 (2.3) 1 (0.1) 25 (1.5) 8 (0.5)

the non-transported patients) were considered 29.3% (95% CI: 27.9–30.7%) and outside office
to need assessment or treatment in the emer- hours it was 46.9% (95% CI: 45.9–48.0%)
gency department but were directed to use vehi- (P < 0.001). Figure 3 shows the NTR in relation
cles other than ambulance (e.g., taxi, a relative’s to the time of the EMS mission. The NTR out-
car). side office hours increased with longer distances
During office hours (Mon–Fri 8:00 am to to the emergency department (Spearman
4:00 pm) the non-transportation rate (NTR) was q = 0.656, P = 0.008) (Fig. 4). The NTR in the

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

552 ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
 NON-TRANSPORTATION IN EMS

Fig. 3. Non-transportation rate (bars) and number of missions (line) in relation to the time of the emergency medical service mission.

Fig. 4. Non-transportation rate in relation to the distance between the municipal population centroid and the emergency department during
office hours and outside office hours. Spearman’s correlation coefficient q = 0.220, q = 0.656 (P = 0.431, P = 0.008).

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 553
 M. HOIKKA ET AL.

Kainuu region was 46.2% (95% CI: 45.1–47.3%) aim at identifying acute illness or injury, and in
and that in the L€ansi-Pohja region was 32.3% most non-transported patients the nature of the
(95% CI: 30.9–33.6%) (P < 0.001). problem (chronic disease or a social issue) was
During the 6-month study period, 28.8% of less urgent. The protocol used may not be suffi-
the non-transportation patients had multiple ciently accurate to identify such problems. In
contacts with EMS resulting in non-transporta- patients with high-risk symptoms (e.g., chest
tion missions. The highest count of non-trans- pain, shortness of breath), however, it is reason-
portation missions for one patient was 45. able to dispatch EMS to evaluate the patient’s
condition at the scene, even if in retrospect the
mission did not require transportation of the
Discussion
patient.
This study showed that four out of 10 EMS mis- On the other hand, the dispatcher’s education
sions did not lead to ambulance transportation plays a significant role. The current national 18-
of the patient. In almost half of these cases the month dispatcher training in Finland may need
patients could stay at home after the medical to be refocused on the changing needs in the
assessment by the EMS crew and contact their evaluation of requests for EMS, as the advisory
municipal health care center later with a non- and assessment tasks in less urgent situations
urgent matter, while one-third of the non-trans- appear to become increasingly in demand. As
portation patients did not need any medical care the current Finnish EMCC dispatchers are not
at all. The NTR was higher outside office hours health care professionals and therefore not cov-
and it increased with the distance to the emer- ered by the legal framework the same way as
gency department. These findings necessitate a health care professionals are, they may experi-
discussion as to the dispatch process and the ence a fear of prosecution, and this in turn may
reasons why people with minor illnesses lower the threshold for dispatching an ambu-
request an ambulance. lance. There are differences between the Nordic
The limitations of this study relate to the fact countries with regard to the organization of
that the study was retrospective, and therefore emergency medical dispatch systems as well as
the registries used were not specifically the use of dispatch protocols.8 The level of edu-
designed to explore those issues under examina- cation of the EMCC personnel varies from a few
tion. The available data, however, which were weeks’ training to that of health care profession-
initially documented for administrative and als—nurses or paramedics—who handle medical
medico-legal purposes, were of good quality emergency calls. A study by Forslund et al.
and had no data missing. Human factors in the showed that when health care professionals pro-
EMS system may distort the results, as the com- vided informational support to the dispatchers
pliance of EMCC dispatchers and EMS provi- who were dealing with non-urgent calls to the
ders with the predefined dispatching protocols dispatch center, the number of missions leading
and non-transportation guidelines could not be to non-transportation was reduced.9
documented during the study. The strengths of It is unclear why people with minor illnesses
this study pertain to the large cohort from two or non-medical problems request an ambulance.
different hospital districts. This inappropriate utilization of ambulance ser-
The finding that more than 80% of the non- vices has been a worldwide problem,1 and the
transported patients could be left at the scene phenomenon has been known since the 1970s.10
and instructed to contact their primary health In our series there was no need for medical care
care center during office hours, or did not need or interventions in one-third of the non-trans-
any health care at all, highlights the difficulty in portation missions. These missions represented
assessing a patient’s condition over the phone. 14.2% of all EMS missions. In addition, 3.8%
It may be difficult to obtain adequate informa- of all patients were directed to use vehicles
tion from the caller, but one must also ask other than an ambulance. In a Swedish study,
whether the current dispatch protocols, the dis- ambulance personnel estimated that the usage
patchers’ education and dispatch centers’ rate of taxis or other vehicles could be even
resources are optimal. The dispatch protocols higher, up to 31% of the missions.11 One reason
Acta Anaesthesiologica Scandinavica 61 (2017) 549–556
554 ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
 NON-TRANSPORTATION IN EMS

for the high NTR could be that the primary health care professionals in EMCCs reduces the
health care is unable to meet the needs of the number of missions that do not necessitate an
patient. During the study, there was a severe EMS response. The feasibility, safety and cost-
shortage of primary care physicians in the study effectiveness of forwarding non-urgent requests
districts (Kainuu 20.8% and L€ ansi-Pohja 8.3%), and missions to telephone health advisors, home
and 96% of the population in Kainuu experi- care, or paramedicine providers, for example,
enced difficulties in getting a doctor’s appoint- should be studied. Involving professionals other
ment on a weekly basis.12 Moreover, several than EMS professionals for the evaluation of
ambulance visits to the same patient without patients on the scene in less urgent situations is
the need for transportation may indicate that the becoming increasingly common in Finland, cor-
patient’s problems have not been adequately responding with reports from several programs
dealt with in primary health and social care. An on community paramedicine.14 Although the
association between a lack of primary care results cannot be generalized to other health care
resources and the regional unfocused use of systems, our study indicates that there is a need
EMS may be a sign to decision makers that pri- for studies evaluating the overall provision of
mary care must be strengthened. The high NTR EMS in health care systems.
implies that the EMS compensates for the lack
of primary care resources in some regions,
Conclusion
although this has not originally been the inten-
tion in the health care system. This study showed a high rate of non-transpor-
Another reason for the high NTR might be tation missions in two Finnish EMS systems. In
the distances involved in these regions. We half of these non-transportation missions, the
found a positive correlation between the dis- EMS personnel assessed that there was no need
tance to the emergency department and NTR for emergency admission to an emergency
(Fig. 4). Especially outside office hours, when department, while in another third there was no
municipal health care centers are closed, people need for any medical care at all. These findings
may prefer to call for an EMS unit rather than indicate that an improvement in the dispatch
make their way to a distant emergency depart- process and primary care resources might be of
ment. Similar findings were found in a Norwe- benefit, especially during out-of-office hours.
gian study in which increased distance was
associated with lower rates of all contact types
Acknowledgements
to casualty clinics except telephone consulta-
tions by a doctor.13 This correlation between The EMS personnel who collected the data
NTR outside office hours and the distance to the while taking care of the patients are highly
emergency department was an interesting find- acknowledged. The statistical help of Mr. Pasi
ing which may be difficult to interpret. Could it Ohtonen, MSc, during the design and analysis
be that the threshold for transporting a patient a of the study is appreciated. This study was
long distance is higher at night than during funded by an EVO grant from Oulu University.
daytime? If so, can such a pattern be ascribed to
the fact that an EMS unit is away from its
Authors’ contributions
response area for a lengthy time and the reluc-
tance to transportation is therefore higher? Or is All authors designed the study, analyzed and
the threshold for dispatching an ambulance interpreted the data, drafted and critically
lower at night? revised the manuscript. MH collected and
The results of this study indicate a need for fur- extracted the data. All authors have read and
ther studies—most importantly, studies that look approved the final version of the manuscript.
into patient safety issues in an EMS system with
such a high NTR. Data as to whether these non-
References
transport decisions are correct or affect patient
outcome are lacking. Studies are also needed to 1. Lowthian JA, Cameron PA, Stoelwinder JU, Curtis
evaluate the extent to which involvement of A, Currell A, Cooke MW, McNeil JJ. Increasing

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd 555
 M. HOIKKA ET AL.

utilisation of emergency ambulances. Aust Health 8. Langhelle A, Lossius HM, Silfvast T, Bj€ ornsson
Rev 2011; 35: 63–9. HM, Lippert FK, Ersson A, Søreide E. International
2. Pittet V, Burnand B, Yersin B, Carron P. Trends of EMS systems: the Nordic countries. Resuscitation
pre-hospital emergency medical services activity 2004; 61: 9–21.
over 10 years: a population-based registry analysis. 9. Forslund K, Kihlgren M, Sorlie V. Experiences of
BMC Health Serv Res 2014; 14: 380. adding nurses to increase medical competence at an
3. Booker MJ, Shaw ARG, Purdy S. Why do patients emergency medical dispatch centre. Accid Emerg
with ‘primary care sensitive’ problems access Nurs 2006; 14: 230–6.
ambulance services? A systematic mapping review 10. Morris DL, Cross AB. Is the emergency ambulance
of the literature. BMJ Open 2015; 5: e007726. service abused? Br Med J 1980; 281: 121–3.
4. Ornato JP. Science of emergency medical dispatch. 11. Hj€
alte L, Suserud B, Herlitz J, Karlberg I. Why are
Circulation 2009; 119: 2023–5. people without medical needs transported by
5. Snooks HA, Dale J, Hartley-Sharpe C, Halter M. ambulance? A study of indications for pre-hospital
On-scene alternatives for emergency ambulance care. Eur J Emerg Med 2007; 14: 151–6.
crews attending patients who do not need to travel 12. The National Institute for Health and Welfare.
to the accident and emergency department: a review Statistical database 2015. 2015.
of the literature. Emerg Med J 2004; 21: 212–5. 13. Raknes G, Hansen EH, Hunskaar S. Distance and
6. Hettinger AZ, Cushman JT, Shah MN, Noyes K. utilisation of out-of-hours services in a Norwegian
Emergency medical dispatch codes association with urban/rural district: an ecological study. BMC
emergency department outcomes. Prehosp Emerg Health Serv Res 2013; 13: 222.
Care 2013; 17: 29–37. 14. Choi BY, Blumberg C, Williams K. Mobile
7. Gray JT, Wardrope J. Introduction of non-transport integrated health care and community
guidelines into an ambulance service: a paramedicine: an emerging emergency medical
retrospective review. Emerg Med J 2007; 24: 727–9. services concept. Ann Emerg Med 2016; 67: 361–6.

Acta Anaesthesiologica Scandinavica 61 (2017) 549–556

556 ª 2017 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
Copyright of Acta Anaesthesiologica Scandinavica is the property of Wiley-Blackwell and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.