Article Text

Download PDFPDF

Secondary triage in prehospital emergency ambulance services: a systematic review
  1. Kathryn Eastwood1,2,3,
  2. Amee Morgans3,
  3. Karen Smith1,2,
  4. Johannes Stoelwinder1,2
  1. 1Department of Epidemiology and Preventative Medicine, The Alfred Centre, Monash University, Melbourne, Victoria, Australia
  2. 2Ambulance Victoria, Melbourne, Victoria, Australia
  3. 3Department of Community Emergency Health and Paramedic Practice, Monash University, Melbourne, Victoria, Australia
  1. Correspondence to Kathryn Eastwood, Department of Epidemiology and Preventative Medicine, The Alfred Centre, Monash University, 99 Commercial Road, Melbourne, VIC 3004, Australia; Kathryn.Eastwood{at}


Objective Secondary telephone triage to divert low-acuity patients to alternative non-ambulance services before ambulance arrival has been trialled in the UK and USA as a management strategy to cope with the increase in ambulance demand. The objective of this systematic review was to examine the literature on the structure, safety and success of secondary triage systems.

Methods For inclusion in the study, the telephone triage system had to be a secondary process, receiving referred patients who had already been categorised as low priority by a primary triage process. Two independent reviewers conducted the search to identify relevant studies. Six articles and one report were identified.

Results The major theme of the papers was the safety and accuracy of secondary telephone triage in identifying low-acuity patients. Two studies also discussed patient satisfaction. There was a low incidence of adverse events, as expected as these patients had already been subjected to primary telephone triage. In the studies identifying ambulance dispatch as a potential final disposition, at least half of the patients were diverted away from ambulance dispatch. In the studies that identified self/home care as a final disposition, a maximum of 31% of patients were categorised to this outcome. Otherwise all patients were recommended for assessment by a healthcare professional other than ambulance clinicians. Patients appeared to be satisfied with secondary telephone triage on follow-up.

Conclusions These results suggest that, while secondary triage of these patients is safe, further research is required to determine its most appropriate structure and its effect on ambulance demand.

  • Triage
  • Telephone
  • Emergency medical service communication systems
  • Ambulance

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


This systematic review was conducted to identify the reported ambulance-based secondary telephone triage systems for triaging low-priority patients in terms of the structure, safety and efficacy of such systems.

The demand for ambulance services has been increasing in many countries1–7 and is predicted to continue to increase, driven mainly by a disproportionate growth in demand from those with chronic illness, particularly the elderly.1 ,8 A portion of this workload is made up of patients categorised as low priority by primary telephone triage who have no, or minimal, physiological derangement, yet still access ambulance resources.2 ,3 ,9–13 The concern expressed in the literature is that the use of ambulance resources by this low-priority caseload is potentially delaying responses, thus endangering the lives of those in need of more urgent medical attention13 and contributing to both emergency department overcrowding and the mounting cost of ambulance services.9–14 Strategies such as secondary telephone triage to divert ambulance requests from patients categorised as low priority to alternative non-ambulance transport services could be an important part of coping with these challenges.

Many ambulance services use computer-based priority dispatch systems to undertake primary triage and allocate a priority level of service on receiving the initial call for assistance.14–19 Ambulance clinicians of varying skill level are dispatched, depending on this computer-based prioritisation and categorisation, to provide assessment and treatment, where necessary, followed by the option to transport to an appropriate hospital-based emergency department. This is an expensive and time-consuming response to situations that may not always require it.

In attempting to address this issue, some ambulance services have implemented secondary telephone triage services using various healthcare professionals to further assess the low-priority cases. These services attempt to provide these patients with more appropriate alternative care options to meet their immediate healthcare needs, rather than the resource-intensive ambulance response.12 ,15 ,19–22 This strategy of low-priority patient management has the potential to significantly affect the way ambulance services manage their growing demand and could contribute to ameliorating emergency department overcrowding. Concerns have been raised, however, about the safety of such systems.15 ,22

Materials and methods

Search strategy

This systematic review was conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) criteria.23 The databases searched included the seven EBM Reviews databases, Books@Ovid, Journals@Ovid Full Text, PsycARTICLES Full Text, The Joanna Briggs Institute EBP Database, AMED (Allied and Complementary Medicine), Ovid MEDLINE(R), Maternity and Infant Care, Mental Measurements Yearbook, PsycBOOKS, PsycCRITIQUES, PsycINFO, Transport Database, CINAHL and Embase from their beginning until the end of November 2012.

The prehospital search strategy used was a simplified version of the search strategy developed by the Cochrane Collaboration's Prehospital and Emergency Health Field.24

The medical subject headings (MeSHs) and keywords used in the prehospital search were: emergency medical service, emergency medical technician, ambulance, air ambulance, military medicine, prehospital, pre-hospital, out-of-hospital, out of hospital, paramedic, emergency health service, paramedical personnel, emergency services, emergency technician, emergency practitioner, emergency dispatch, emergency despatch, first responder, ems and emt.

Then a second search containing the following MeSHs and keywords was conducted: telenursing, NHS 24 referral, telephone telemetry, telephone triage, telephone assessment, secondary telephone assessment, secondary telephone triage, telephone medicine. These two searches were then combined and duplicates were removed.

Study selection

Two reviewers independently screened the articles for their eligibility for inclusion. To be included, the telephone triage system had to be a secondary process, where the patient had already been categorised as low priority. The secondary telephone triage system had to be part of, or feed into, an ambulance service directly, where cases could be transferred for an ambulance response. Articles about primary telephone triage, out-of-hospital triage of patients seeking general practitioner (GP) appointments or with emergency departments were excluded. Editorials and letters were also excluded. All screening conflicts were resolved by consensus. The reference lists were reviewed to identify additional articles. A flow diagram of the phases of the systematic review was produced according to the PRISMA statement (figure 1).23

Figure 1

Flow diagram of study selection for this review.

Assessment of study quality

Methodological quality of the studies was assessed using a previously described five-level modified instrument that has been applied to clinical trials, descriptive studies and surveys.25 Quality levels 1 and 2 consisted of prospective studies with random or consecutive sampling and clearly defined outcome measures. The sample sizes in studies meeting quality level 1 are larger, allowing narrower CIs and increased generalisability. Quality level 3 studies met the criteria for level 1 and 2 studies, except that they were retrospective. Quality level 4 studies used convenience sampling or other techniques that could potentially allow the introduction of bias, and quality level 5 did not have clearly defined or validated outcome measures.25


The prehospital search resulted in 41 470 English references in Ovid MEDLINE, 97 596 in Embase, and 73 464 in CINAHL. The second search specifically relating to telephone triage returned 510 references in Ovid Medline, 1005 in Embase, and 5641 in CINAHL. Once the prehospital and telephone triage searches were combined, grey literature and reference list searches were included and the duplicates were removed, 1607 references were returned for title and abstract review. After title and abstract review by KE and AM, 1544 articles were found to be irrelevant, as they were not related to secondary triage, ambulance dispatch and low-acuity patients, leaving 63 articles for manuscript review.

After manuscript review, six articles and one report were identified specifically relating to secondary triage processes and alternative responses for callers to emergency ambulance health services. Two of these articles related to the same study of telephone triage in the UK (figure 1).

After methodological quality review, three studies were found to meet quality level 2, one study each met quality level 3, 4 and 5, and one paper was more of a commentary on a study that had occurred than a report of the results. The seven papers referred to six secondary triage systems and varied in what they reported. One paper outlined a pilot study; however, only minimal quantitative analysis was provided. Four papers reported some demographical data, three papers included follow-up surveys on the patients’ condition, and six papers made some statement about the safety of secondary triage systems. Four studies reported on the patient disposition at the end of the secondary triage, and three papers investigated whether there was a financial benefit in having secondary triage systems in place. A summary of these papers is presented in table 1.

Table 1

Secondary telephone triage of low-priority ambulance patients

The major theme the papers discussed was the ability and safety of the secondary triage process to identify low-acuity patients. Patient satisfaction was also discussed in two of the studies.

It was expected that the secondary triage systems used in each of the studies would have some level of variation. Five studies had preformulated questioning algorithms for patients, three of which were identified as computer-based.12 ,14 ,15 ,20 ,26 Two studies included either paramedics or emergency medical technicians (intermediate) conducting secondary triage,12 ,15 ,22 and five used nurses.15 ,19 ,20 ,22 ,26 The secondary telephone triage systems used in the various studies safely differentiated the low-acuity patients.14 ,19 ,20 ,22

The 2004 study by Dale et al22 was conducted because the authors identified that the use of hospital admission as an outcome measure in their 2003 study was not an accurate means of determining the safety or the success of the secondary triage process and the decision to send an ambulance or not. After reanalysing their results, they found there were only two cases (0.3%) where a panel of experts believed a patient should have received emergency ambulance transport to avoid potential harm. One other potentially adverse event (0.2%) was reported in another paper, where a patient allocated to a non-ambulance transport disposition was subsequently admitted to an intensive care unit.19 No details about this patient were given in the paper with regard to their condition at the time of secondary triage, ambulance attendance or emergency department assessment. Another study involving 133 patients had no adverse events occur as a result of secondary telephone triage.20 Conversely, one study reported four cases (0.2%) that were identified as requiring immediate ambulance dispatch (three for potential cardiac complaints and one for intubation), which would otherwise have remained coded as low priority.14

As stated above, using hospital admission as an outcome measure does not provide an accurate picture of the safety of secondary telephone triage, as many patients who require admission may be best transported by private vehicle (eg, elective surgery), just as many patients appropriate for ambulance transport (eg, simple dislocations or fractures) may be discharged from the emergency department. Analysis of the success of such systems should be limited to differentiating between patients who require ambulance intervention and/or transport and patients who do not. The need for emergency department intervention should not carry the weight some studies have attributed to it in assessing the triage system. While the 2004 study by Dale et al22 identified the flaw in using hospital admission, this paper, along with the remaining papers included in this study, did not clearly identify the importance of limiting the analysis of the secondary triage system to the need for ambulance intervention and/or transport.12 ,14 ,15 ,19 ,20 ,22 ,26

Five of the six studies reported the final disposition following secondary telephone triage. Three studies specified which cases were recommended for ambulance dispatch following secondary triage, which ranged from 17% to 48% of the calls. The remainder of the calls within each study were categorised into dispositions recommending other levels of care. Three studies reported the number of patients categorised to the home care or self-care disposition.14 ,15 ,20 This was the only category, to the authors’ knowledge, that did not result in an assessment from a healthcare professional. The number of cases categorised to this disposition ranged from 2% to 31%.14 ,15 ,20 This low proportion of cases going without any face-to-face assessment may suggest a conservative approach to patient triage by those conducting the secondary triage and/or the triage algorithms themselves. It also demonstrates that a large amount of the ambulance and emergency department workload is being directed to more appropriate healthcare services—bypassing both ambulance clinician attendance and transport to hospital—and emergency department assessment and referral.

Two studies included follow-up surveys, with the patients involved being asked about their experiences.14 ,20 In one study, 96% of the respondents were satisfied with the outcome.20 This study reported that many of the patients did not want an ambulance response or assessment in an emergency department. Instead, they were seeking information and direction to appropriate services.20 Another study found that 75% of patients were satisfied with the service, and 20% would have liked advice about alternatives to ambulance attendance.14 This study did, however, find that this satisfaction level was lower in the group who underwent secondary triage than in a control group where no secondary telephone triage intervention was used (85%).14


While there is a large amount of literature on telephone triage, only six studies were identified that specifically addressed secondary telephone triage of calls to emergency services. These articles are of varying quality and date back to 1981, indicating that the need for alternative healthcare options to manage low-acuity ambulance demand was identified long ago.12 There was considerable variation among the systems and studies, making comparison difficult. Four studies were prospective, two were retrospective, and one paper was a narrative describing a study. Sample sizes were relatively small, ultimately affecting our ability to rate the studies as quality level 1 using the quality instrument used.25 Despite this, three of the six studies were able to be classified as quality level 2.

Studies have found primary telephone triage software for ambulance dispatch to be safe8 ,16; however, the safety of nurse- or paramedic-led telephone triage has been questioned, because of the higher level of subjectivity involved in the decision-making and algorithm usage.15 ,27 ,28 The present systematic review has found that generally the advice given and disposition assigned was safe, appropriate and reflective of the likelihood of hospital admission.12 ,14 ,15 ,19–22 ,27 ,29 ,30 We have found that the weight of concerns raised in the literature about the risks associated with the loss of a face-to-face assessment31 ,32 is reduced, as those conducting the triage referred many of the patients to further medical assessment rather than home/self-care. These patients were then being managed by more appropriate healthcare professionals for their current needs, rather than simply having ambulance clinicians sent. This delegation of care will result in a reduction in patients presenting to the emergency department by ambulance, contributing to a reduction in emergency department workload and time for the patient to reach suitable definitive management.

As expected, the number of adverse events was small, as all cases had already been put through a primary triage process and categorised as low priority. The occurrence of potentially adverse events was further mitigated by the fact that these patients were then subjected to a secondary telephone triage by a healthcare professional, providing a second opportunity for indicators of higher acuity to be identified. In fact, more cases with potentially high-priority indicators were found that were missed by primary triage than those that resulted in potentially adverse outcomes as a result of secondary triage. Overall, five of the six studies concluded that secondary telephone triage was safe.14 ,19 ,20 ,22 ,26

Ambulance services have moved beyond simply responding to a call for help and transporting patients to medical attention. Ambulance clinicians now provide highly skilled medical care, and, based on this, ambulance services now to need to consider delegating the right care to the right patient to appropriately manage their resources. Ambulance services need to review their service delivery models and decide whether simply sending an ambulance is fulfilling their responsibility to answer the call for help. When doing this, they need to consider the increasing demand on emergency medical services, including emergency departments, and consider whether they need to respond with a broader range of medical response options.

The potential benefits of secondary telephone triage identified in this systematic review include appropriate, more timely medical attention for patients seeking medical intervention, increased patient satisfaction, a reduction in demand on ambulance and emergency departments, cost savings to ambulance services, improved resourcing for cases requiring urgent medical attention, decreased response times, and potentially improved outcomes for emergency patients who will receive the care they require sooner.12 ,14 ,15 ,19–21 However, the effectiveness of secondary triage is limited by the range of options available for those conducting the triage. We are aware of at least one ambulance service that offers home nursing and GP visits, hospital response team visits, direct transfer to other health-management telephone lines such as poisons’ advice, community psychiatric services response teams and non-emergency ambulance response as an alternative service.33 This is in contrast with the studies reported here, where advice and a recommendation to self-access further medical advice or attention were the only alternative outcome options. Providing such alternative options may reduce the reported high level of referral back to ambulance transport.


Most ambulance services have evolved around the population they service and the resources available. As a result, they can be very different in their general operation, including the receiving and processing of initial calls for assistance. The secondary triage software or algorithms used in each of the studies was not identified. There is likely to be variation among call centres in the primary telephone triage software they use, which can result in different primary triage outcomes. As a result, there is likely to be a difference in the cohorts of patients identified as suitable for secondary triage within each of the studies.

Only one of the systems had an alternative service provider identified. This was a private conveyance service. The lack of any linked alternative service providers may have resulted in higher referral to ambulance dispatch to ensure patients received a face-to-face assessment and lower patient satisfaction in terms of service provision. Health professionals conducting the secondary triage could only refer patients to self-access other healthcare resources, resulting in a potential for non-compliance from the patients. This could have also potentially influenced the decision-making of those conducting the triage.

The sample sizes in most of the studies were relatively small, and the hours of operation were limited in some studies to times when there was generally no access to other community-based medical services . The final patient recommendation or disposition was not uniform between the studies, so a clear picture of the volume of patients potentially suitable for each category could not be ascertained. Finally, response rates for the follow-up surveys were low, so there may be some selection bias in the results from this part of the studies.


Further research is required to more accurately measure the appropriateness of each disposition outcome recommended by the secondary triage process. To do this, appropriate outcome measures must first be identified. Follow-up surveys of patients to determine whether they adhered to the advice given, whether they received any intervention at the disposition level to which they were assigned, and whether any escalation of medical services was required within a given timeframe would also help to more accurately measure the safety and accuracy of such systems.

Further research should be carried out into those conducting the secondary triage to determine what level of skills and education are required and what decision-support strategies are best used to ensure consistent, appropriate and safe patient care.

Finally, the primary triage process will need to be reviewed with respect to the new service delivery model to determine whether more discrimination needs to occur at the point of first contact.


The literature does indicate a potential positive effect in terms of financial viability for ambulance services, resource supply and demand management, and—most importantly—patient outcomes. Owing to the limited number of studies, small sample sizes and variation among secondary triage systems, the findings from this systematic review would be difficult to generalise. Further comprehensive studies need to be conducted into these systems before the safety can be truly established.


View Abstract


  • Contributors KE: study conception, devised and implemented the search strategy, analysed data and wrote the paper. AM: discussed core ideas to study, was involved in search strategy and study selection, and edited the paper. KS: discussed core ideas to study and edited the paper. JS: discussed core ideas to study and edited the paper; JS is KE's primary PhD supervisor.

  • Competing interests KE is an intensive care paramedic who intermittently works for the Ambulance Victoria Referral Service (secondary telephone triage service). JS is the Chair of the Board of Ambulance Victoria.

  • Ethics approval Monash University Human Research Ethics Committee and Ambulance Victoria Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.