Calling an ambulance for non‐emergency medical situations: Results of a cross‐sectional online survey from an Australian nationally representative sample

Abstract Objective To investigate the Australian general public's perception of appropriate medical scenarios that warrants a call to an emergency ambulance. Methods An online survey asked participants to identify the likely medical treatment pathway they would take for 17 hypothetical medical scenarios. The number and type of non‐emergency scenarios (n = 8) participants incorrectly suggested were appropriate to place a call for an emergency ambulance were calculated. Participants included Australian residents (aged >18 years) who had never worked as an Australian registered medical doctor, nurse or paramedic. Results From a sample of 5264 participants, 40% suggested calling an emergency ambulance for a woman in routine labour was appropriate. Other medical scenarios which were most suggested by participants to warrant an emergency ambulance call was ‘Lego in ear canal’ (11%), ‘Older person bruising’ (8%) and ‘Flu’ (7%). Women, people aged 56+ years, those without a university qualification, with lower household income and with lower emotional wellbeing were more likely to suggest calling an emergency ambulance was appropriate for non‐emergency scenarios. Conclusions Although emergency healthcare system (EHS) capacity not increasing at the same rate as demand is the biggest contributor to EHS burden, non‐urgent medical situations for which other low‐acuity healthcare pathways may be appropriate does play a small role in adding to the overburdening of the EHS. This present study outlines a series of complaints and demographic characteristics that would benefit from targeted educational interventions that may aid in alleviating ambulance service attendances to low‐acuity callouts.


Introduction
Ambulance demand continues to rise at a rate higher than population growth. Between 2008 and 2015 in Victoria, Australia, ambulance demand rose by 29.2%. 1 Increased ambulance utilisation contributes towards ED overcrowding, ambulance ramping and lowered access to care and ambulances in the community, 2 each with the capacity to lead to diminished patient outcomes. 3 These factors place substantial and ongoing burden on emergency healthcare workers.
Increased demands on emergency healthcare services (EHS) have been attributed to a growing elderly population and increasing population comorbidities, 4 along with public hospital acute bed capacity not increasing at the • Calls for emergency ambulance utilisation for non-emergency conditions contributes to the overburdening of emergency healthcare systems. • Women, older people (56+ years), those without a university qualification, lower household income and lower emotional wellbeing were factors increasing the likelihood of suggesting a call would be placed for an emergency ambulance for non-emergency situations. • A woman going through routine labour was by far the nonemergency scenario members of the general public most frequently suggested warranted engagement with emergency ambulance services.
same level as demand. 5 Although evidence suggests Australians can appropriately self-refer to the ED, 6 and overcrowding is overwhelmingly attributable to a lack of system capacity to meet increasing demand, 7 non-urgent presentations that could be effectively managed via alternate primary healthcare pathways can divert EHS resources from patients with serious/ acute medical emergencies. 8,9 Cases attended by Victorian paramedics requiring no intervention from paramedics increased by on average 6.7% annually between 2008 and 2015. 1 Further, 21.2% of secondary telephone triage cases between 2009 and 2012 were considered not suitable for transport to ED. 10 Patients frequently perceive urgency of their medical conditions to be greater in comparison to medical practitioners. 11,12 The present research aims to extend upon a previous investigation 13 exploring public understanding of appropriate medical response to non-emergency situations not requiring immediate emergency intervention, whereby a call placed to Triple Zero (000) for an emergency ambulance would be unwarranted.

Study design
Cross-sectional via online survey.

Participants
Prospective participants included any Australian resident aged >18 years who was not currently nor had ever before worked as an Australian registered medical doctor, nurse or paramedic. Participants were recruited through an online market research company Pureprofile.

Ethics approval
Ethics approval was granted by the Edith Cowan University Human Research Ethics Committee (#2020-01958).

Medical scenarios
Participants were presented with 17 hypothetical medical scenarios (Table 1) and asked what healthcare  pathway (from a list of nine options;  Table 2) would they likely undertake if presented with this scenario in real life. Among these response options was an 'Other' option allowing participants to type in open-ended responses if they felt their response would differ from one of the eight prompted response options. Participants could only choose one response option. Open-ended responses were coded as 'Call 000 for an Ambulance' if participants indicated they would call for an ambulance in the first instance. All other legible responses were coded as not calling 000 for an Ambulance.
The 17 hypothetical scenarios were re-purposed from a previous investigation 13 exploring the general public's ability to correctly categorise emergency vs non-emergency medical scenarios. A panel of experienced registered paramedics (n = 5) reached 100% consensus on their interpretation of whether medical scenarios were of sufficient risk or severity to warrant an emergency call to 000 for an ambulance or not. This process involved the panel meeting with members of the research team as a group, and was presented with each of the 17 scenarios. Following presentation of a scenario, panel members were asked to (confidentially) write down whether they felt the scenario should or should not warrant a call for an emergency ambulance. Where responses were not 100% unanimous across the panel, research team members were to facilitate discussion until consensus was reached. Of the 17 scenarios, nine were identified as emergencies warranting a call to 000 for an ambulance with the remaining eight deemed as non-emergencies. For the present research study, the nine 'emergency' scenarios were hidden among 'non-emergency' scenarios and displayed to study participants as 'red herrings' to make distinguishing the non-emergency medical scenarios less obvious. Scenario display order was universal across study participants; however, all 17 scenarios were randomly mixed to provide a finalised scenario display order.
Given incorporation of images alongside text can improve comprehension of information being presented, particularly among people with low literacy skills, 14 graphical depictions of medical scenarios were provided to complement scenario text (images provided as Appendix S1). Photos were either original creations taken with a 12-megapixel wide-angle camera found on an Apple iPhone X (Cupertino, CA, USA), or sourced from stock photos (Dreamstime.com).

Demographics
In the online questionnaire participants initially completed a demographics section. The demographics section included questions on age, sex, identification as Aboriginal or Torres Strait Islander, highest level of education, metropolitan or rural residence, employment, annual household income, number of children and whether or not they suffered from any chronic illness or had a disability. Participants were also asked to complete the Brief Emotional Experience Scale (BEES) as a measure of emotional wellbeing. The BEES comprises of three positive (Happy, Calm, Confident) and three negative (Worried, Sad, Afraid) emotional adjectives rated on a 4-point response scale: (1) Not at all; (2) A little bit; (3) Quite a bit; and (4) A lot. An overall emotional wellbeing score is calculated by summing across the positive and negative adjectives separately, and then subtracting the negative score away from the positive score. The overall score can range from À9 to +9 where a higher score indicates greater selfreported emotional wellbeing.

Procedure
Participants were sent an invitation to participate in the research through their online Pureprofile account. The online survey was active from 19 November 2019 to 2 December 2020, facilitated through the Qualtrics survey platform. Upon completion of the survey, Pureprofile facilitated financial reimbursement for participant's time. Estimated time to complete the survey was 20 min.

Analysis
For the eight non-emergency scenarios, participants were coded as either

10
Alcohol intoxication A 22-year-old male is conscious, not injured and has drunk a substantial amount of alcohol on a night out.

11
Woman in labour A 33-year-old woman is 9 months pregnant and goes into early stages of labour. Her waters have broken, and she feels uncomfortable.

12
Back pain A 40-year-old man with a 6-month history of back pain wakes up in the middle of the night with a sore back and has run out of pain killers. The man is in quite a bit of pain.
14 Cut finger A 42-year-old man has cut his finger while chopping vegetables, and the bleeding is controlled with pressure.
Emergency scenarios 1 Box Jellyfish sting While in Northern Queensland, a boy is stung by a Jellyfish while swimming at the beach, and large welts appear on his arm.

Snake bite (unidentified)
A 50-year-old woman has been bitten on her ankle by an unidentifiable snake.

5
Mild chest pain A 40-year-old woman is experiencing mild chest pain. She does not think it is indigestion or a strained muscle.
8 Stroke A 67-year-old man is slurring his words; he has not drunk any alcohol.

9
Severe chest pain A 52-year-old man has severe chest pain, is sweating and grey in colour.

13
Paracetamol overdose A 32-year-old female has taken 10 regular paracetamol tablets in the last 12 h, and is feeling extremely unwell. She has abdominal pain and feels nauseous.

15
Child head haematoma A 3-year-old boy has fallen off the couch and bumped his head. He began crying immediately and a golf-ball size lump with a bruise promptly appears. 16 Potential meningococcal A 4-year-old girl has woken up with a high temperature, feels hot to touch, has a really sore neck and a headache which Panadol is not relieving.

17
Older person hip pain A 80-year-old woman feel out of bed, is now unable to get up and is complaining of hip pain on her right side.
(incorrectly) calling 000 for an ambulance or choosing any other healthcare pathway. The number of non-emergency scenarios participants incorrectly coded as warranting a call to 000 for an ambulance was calculated (scored out of 8). Descriptive statistics were calculated and significant differences within groups determined using ttests and one-way ANOVAs. Generalised linear modelling assuming binomial distribution was used to study the relationships between key demographic variables and the number of non-emergency scenarios correctly identified as not warranting a call to 000 for an ambulance.

Results
A total of 6723 individuals began the online survey. Of these, 109 participants did not proceed passed the first page containing a detailed participant information letter. A further 30 participants were screened out for identifying as under 18 years of age, a further 112 for not being an Australian resident, and a further 752 for suggesting they had previously worked in Australia as a registered doctor, nurse or paramedic. Last, a total of 89 participants were further screened out as they completed demographic information only. This left a final sample of 5631 eligible participants. A total of 5264 participants completed all 17 scenarios. Given no significant differences were noted across any demographic factors (e.g., age, sex, income, BEES score) for those who did and did not provide responses to all 17 medical scenarios, missing data was deemed missing completely at random. Demographics for the final sample are outlined in Table 3.
Missing data was associated with some demographic variables where participants chose not to disclose information. These are not reflected in Table 1. These include sex n = 14 (0.27%); Aboriginal/Torres Strait Islander status n = 37 (0.70%); relationship status n = 203 (3.86%) and household income n = 517 (9.82%).
Suggesting non-emergency scenarios warrant a call to 000 The mean score (out of 8) for the number of non-emergency scenarios for which participants incorrectly suggested a call to 000 for an emergency ambulance was 0.84 (SD = 1.23). By far the most common non-emergency scenario for which participants suggested an emergency ambulance was warranted was the 'Woman in labour' scenario (40.6% suggesting they would call an ambulance), followed by 'Lego in ear canal' (10.5%), 'Older person bruising' (7.5%) and 'Flu' (7.3%) ( Table 4).
Demographic factors influencing likelihood of calling an emergency ambulance for nonemergency scenarios Table 5 depicts the results of the multivariate analysis demonstrating the impact of demographic variables on suggestions of calling 000 for an ambulance for non-emergency scenarios. Compared to males, females were 33% less likely to call for an emergency ambulance for nonemergency scenarios. Similarly, those aged 18-35 years were 21% less likely to call for an emergency for ambulance for non-emergency scenarios compared to those aged 56+ years.
Those without a university degree were 13% more likely to call for an emergency ambulance for nonemergency scenarios, and there was a decreasing trend in the likelihood to call an emergency ambulance for non-emergency scenarios for income; higher earning individuals were less likely to suggest they would call for an ambulance than lower income individuals. Further, those with negative mental health scores measured via the BEES were 11% more likely to call for an ambulance for nonemergency scenarios than those with positive mental health scores.

Summary of findings
Risk aversion is common when it comes to personal health, with a preference to request emergency medical intervention and not need it than require emergency medical intervention and not have it. 15 Whereas this should not necessarily change, an appropriate balance needs to be found between risk aversion and overreliance/overburdening already stretched EHS.
Our findings are not the first to suggest members of the general public can miscategorise non-emergency scenarios as emergencies warranting ambulance attendance. For example, a woman going into labour has often been miscategorised. 13,16 Although data from the present study suggested men were more likely to perceive routine labour as an 'emergency' warranting a call for an emergency ambulance compared to women, the difference was only small (52% vs 48%, respectively, P < 0.001).
Scenario wording expressed no reason to suspect labour complications, suggesting that even for normally progressing labours, a substantial proportion of the public's first notion would be to call for an ambulance.
Overall, our data suggests women are less likely to call for an ambulance for non-emergency scenarios. This finding seems contentious in consideration of corresponding literature, suggesting either women are more likely to call for an ambulance 17 or attend EDs 18,19 for nonemergency medical situations, no differences between men and women 20,21 or that (as was the case  16 Further researchboth among the general population (who can call for an emergency ambulance at any time) and retrospective analysis of actual emergency ambulance usersmay be warranted to definitively ascertain between-sex differences. Our data also suggested people aged 56+ years were more likely to suggest non-emergency scenarios warranted a call for an emergency ambulance than those aged 18-35 years. This is counter to the majority of pre-existing literature suggesting either young people are more likely to inappropriately engage with EHS 18,23 or little to no differences across different age groups. 20 Previous research suggests younger people are more likely to directly seek EHS attention, often because of the added convenience EHS offers over other nonemergency healthcare pathways. 18,24 Interestingly, another study presenting hypothetical non-emergency scenarios also found older people more likely to call for an emergency ambulance than younger people. 25 We noted those who had completed a university degree and had higher household income were less likely to call an ambulance for nonemergency scenarios (previous research also links heightened education with more appropriate ambulance use, 26 as were people with higher emotional wellbeing as measured via the BEES. Although several studies have investigated the impact of patients utilising EHS for acute mental health issues, few have investigated the association between poor mental and/or emotional wellbeing and potentially unnecessary use of EHS. One study found heighted ambulance use for minor conditions among those with a psychiatric disorder, 27 with another review article suggesting some interventions targeting social/ emotional issues among patients can reduce unnecessary EHS use. 28

Implications
Patients may choose to access EHS where it may not be entirely necessary because of limited confidence in other healthcare pathways, convenience, perceived urgency of their condition, or a perception their condition may require resources and/or facilities not available through other healthcare pathways. 24 Undoubtedly, there are other aspects outside of an individual's knowledge of what should and should not warrant EHS engagement that contribute to the decision of which healthcare pathway should be taken. For example, a perception of not being able to obtain a timely appointment with ones GP can lead to increased EHS engagement. More globally, although wider-reaching public health interventions addressing poverty, homelessness and support for childcare will reduce impact on EHS, undoubtedly consideration of increased system capacity in alignment with increased demand is key. 7 Nonetheless, to address knowledge at the individual level which may have some small capacity to alleviate system pressures, educational initiatives targeting reducing delay seeking help when needed, services provided by GPs, ambulance services and EDs, and guidance about the clinical urgency of symptoms (and the most appropriate healthcare pathway for managing these) are suggested. 29

Strengths and limitations
Inherent strengths of this present study include: (1) our ability to leverage trialled study materials (including medical scenarios) from a similar previously published investigation, (2) the additional incorporation of graphics alongside text to improve contextualisation of medical scenarios, and (3) the representative nature of the Australian adult population from which data was derived.
However, this present study is not without limitations. For example, participants were recruited via an online market research company which did include some small   Data collection occurred in November/December 2020. Although national COVID-19 infection rates were (comparatively) low during this period, data was collected in the midst of a global pandemic, whereby it has been suggested people have been less willing to engage with EHS out of fear of exposure to SARS-CoV-2. 30 It is unclear the extent to which attitudes changing in retaliation of the COVID-19 pandemic may have impacted on study findings.
Last, defining unnecessary ambulance use is complex and often subjective. For example, it can sometimes be necessary for paramedics to transport patients to hospital EDs for non-clinical reasons. 'Unnecessary' use is not always a deliberate misuse, particularly as research suggests many individuals are hesitant to engage with EHS. 29 Although our classification of nonemergency scenarios not warranting an emergency ambulance response came from a panel of experienced paramedic perspectives based on information present in medical scenarios, it is acknowledged binary judgement of appropriateness of emergency ambulance engagement lacks nuance and consideration of some individual circumstances.

Conclusions
Emergency ambulance use for low acuity conditions continues to contribute to the stretched service capacity of healthcare systems around the world. Until service capacity and integrated healthcare pathways are improved, enhancing understanding among the general public of the situations that warrant emergency ambulance intervention (and those that do not) will play a small role in easing burden on jurisdictional ambulance services, EDs and their staff.