Emerg Med J 30:883-887 doi:10.1136/emermed-2012-201852
  • Review

Ambulance demand: random events or predicable patterns?

  1. Karen Smith1,2,5
  1. 1Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia
  2. 2Ambulance Victoria, Doncaster, Victoria, Australia
  3. 3Burnet Institute, Melbourne, Victoria, Australia
  4. 4Department of Community Emergency Health and Paramedic Practice, Monash University, Melbourne, Victoria, Australia
  5. 5Emergency Medicine Department, University of Western Australia, Western Australia, Australia
  1. Correspondence to Kate Cantwell, Department of Epidemiology and Preventive Medicine, Monash University, The Alfred Centre, 99 Commercial Road, Melbourne, Victoria 3004; Australia; Katharine.cantwell{at}
  • Accepted 25 October 2012
  • Published Online First 26 November 2012


Background Occupational, social and recreational routines follow temporal patterns, as does the onset of certain acute medical diseases and injuries. It is not known if the temporal nature of injury and disease transfers into patterns that can be observed in ambulance demand. This review examines eligible study findings that reported temporal (time of day, day of week and seasonal) patterns in ambulance demand.

Methods Electronic searches of Medline and Cumulative Index of Nursing and Allied Health Literature were conducted for papers published between 1980 and 2011. In addition, hand searching was conducted for unpublished government and ambulance service documents and reports for the same period.

Results 38 studies examined temporal patterns in ambulance demand. Six studies reported trends in overall workload and 32 studies reported trends in a subset of ambulance demand, either as a specific case type or demographic group. Temporal patterns in overall demand were consistent between jurisdictions for time of day but varied for day of week and season. When analysed by case type, all jurisdictions reported similar time of day patterns, most jurisdictions had similar day of week patterns except for out-of-hospital cardiac arrest and similar seasonal patterns, except for trauma. Temporal patterns in case types were influenced by age and gender.

Conclusions Temporal patterns are present in ambulance demand and importantly these populations are distinct from those found in hospital datasets suggesting that variation in ambulance demand should not be inferred from hospital data alone. Case types seem to have similar temporal patterns across jurisdictions; thus, research where demand is broken down into case types would be generalisable to many ambulance services. This type of research can lead to improvements in ambulance service deliverables.


Over the past 20 years, there has been an increase in demand for emergency ambulance services across the developed world,1–4 placing significant strain on ambulance resources. However, it is not known if demand is constant across different times of day, days of the week or months of the year.

Occupational, social and recreational routines follow temporal patterns, as does the onset of certain acute medical diseases and injuries.5–12 It is therefore important for ambulance service planning to know if the temporal nature of injury and disease transfers into patterns that can be observed in ambulance demand.

Temporal patterns in hospitalisations have been examined in detail.7–9 13–16 However, ambulance service users differ from inhospital populations in important ways. For example, not all people who use ambulance services are transported to hospital and not all hospital patients arrive by ambulance. Previous studies have found significant differences between the two populations in trauma and drug related cases.17 ,18 These differences mean that it is important to examine temporal patterns in ambulance demand in their own right and not infer trends from hospital-based studies.

In this review, we examine the published evidence on temporal patterns in ambulance demand. Describing and understanding these patterns will aid in designing ambulance demand management strategies.


The objectives of this paper are to examine temporal patterns in ambulance demand.


A systematic search of published literature was conducted using online databases. Medical databases searched included Ovid Medline and the Cumulative Index of Nursing and Allied Health Literature. Terms used to search the databases were: ambulance, EMS, emergency medical service, paramedic, prehospital, pre-hospital, demand, workload, utilization, epidemiology, 999 calls, 911 calls, 000 calls, ambulance attendance, temporal.

The search terms were based on the prehospital search filter for the Cochrane Library to identify research conducted in the prehospital setting.19 This search was supplemented by an internet-based search through Google and Google Scholar which revealed government documents and reports from the USA, Europe, Great Britain and Australia. Contact was initiated with authors in these countries to locate further government and service provider reports that other publications suggested may be of relevance to this review.

Inclusion criteria for publications


This review examined the patient population transported by road-based emergency medical service systems in developed countries.

Variables of interest

The dependent variable investigated was ambulance demand. Papers were sorted into categories depending on the aspect of ambulance demand that was reported: overall service demand or demand for a specific subset of cases. Case subsets were defined by either a specific demographic group, such as the elderly or children or a group of similar case types such as trauma or overdose. Case type is a classification term used to group cases instead of diagnosis, as often in the prehospital setting the tools needed for an exact diagnosis are not available. Medical cases were often classified by the body system affected.20 In trauma settings, cases were often grouped according to the external cause of injury as opposed to the body system affected; examples of these are ‘falls’ ‘road traffic injuries’ and ‘assault’.17 ,21–23

The explanatory variable examined was trend or temporal pattern with temporal referring to time of day, day of week, month of year or season.

Language and timeframe

The search was limited to papers in English. The time limit chosen for the search was 1980–2011 because modern ambulance services offering assessment and treatment as well as transport only began in the early to mid-1970s. Therefore, papers published before 1980 related to ambulance systems are not comparable with ambulance systems today.

Exclusion criteria for publications

Papers were excluded if ambulance workload was described by response times or response coding, as these studies only investigated the time taken to get to a case and did not give a description of the types of cases attended or key variables such as age or gender.

Papers about extremes of temperature were also excluded as, although associated with time of day and season, these are special events not necessarily repeatable patterns.

Results and discussion

Relevant papers identified

In all, 37 studies were found that reported on temporal patterns in ambulance demand (figure 1).

Figure 1

Search findings. Access the article online to view this figure in colour.

Total ambulance demand

Five studies reported temporal variations in total demand but there was no consistent method of reporting; one study reported only calls to ambulance help lines, three studies reported ambulance responses, and one study reported calls and responses. All of the studies reported on the types of cases found in overall demand but each study used a different method to group the case types, often not well described. Data collection periods ranged from 1 week to 2 years. The temporal unit of focus in all five studies was time of day. Four studies also reported on day of week and one study reported on month of year or season.

Subsets of ambulance demand

Overall, 32 studies reported on temporal patterns for a subset of ambulance demand. Of these, 30 were descriptive studies, one was a case-control and one a cohort study.

The areas of ambulance demand covered by the 32 studies were:

  • high acuity cases (2)

  • cardiac arrest cases (10)

  • cardiovascular cases (2)

  • heroin, alcohol and other drug related cases (11, 4 of which were from the same database)

  • aspects of trauma (4, one study examined violent injury only)

  • paediatrics (2)

  • elderly patients (1)

The method of reporting and analysis for each study is summarised in table 1.

Table 1

Reviewed studies

Indepth comparison of results across studies was not possible because of major variations in analysis and reporting of temporal variables. Nevertheless, it is clear that previous research shows that there are temporal trends in overall ambulance demand, as well as in subsets of specific cases. Further, the temporal patterns in ambulance demand examined appear influenced by age and gender. Temporal patterns in demand are considered below in relation to time of day, day of week and month.

Time of day

Time of day was reported in 33 studies. Studies of overall or total ambulance demand showed a bimodal distribution with a first peak in calls that started at 08:00 that reached the maximum around midday and the second peak from 19:00 to just before midnight.20 ,25 The lowest number of calls occurred around 05:00 to 07:00. All studies showed similar temporal patterns.4 ,20 ,24–26

Compared with total ambulance demand, specific case type trends showed different temporal patterns. Cases involving heroin overdose were more common in the late afternoon to early evening than in the morning.18 ,42–44 ,48 ,49 Alcohol related cases showed an increased frequency of cases between 20:00 and 02:0020 ,46 ,47 with a similar pattern evident for intentional injuries and assaults which also occurred in the late evening and early hours of the morning.17 ,50

Cases of out-of-hospital cardiac arrest also show a bimodal pattern with peaks in the frequency of cases occurring around 08:00–09:00 and 18:00–19:00.25 ,29–33 ,35 One study from Belgium compared the temporal patterns of ambulance attended out-of-hospital cardiac arrests with inhospital cardiac arrests. The patterns for the two types of events differed with inhospital cardiac arrests evenly distributed across the day and night and out-of-hospital cardiac arrest having a bimodal distribution with a nadir at night.35

Time of day patterns has also been shown to vary by age. Ambulance demand among the elderly has been shown to be the highest in the morning or early afternoon.53 In contrast, ambulance demand among paediatric patients has been shown to be the highest in the afternoon and evening.51 ,52

Day of week

Three studies examined patterns of total ambulance demand by day of the week. In contrast to the similar time of day patterns detailed above, day of week demand patterns differed across the studies. Saturday was the busiest day in a study in Madrid, and responses on Friday, Saturday and Sunday accounted for around 47% of total demand.20 Fridays and Saturdays were also the busiest days for the London Ambulance Service.4 ,26 However, in a study in Singapore, demand was the greatest on a Monday, with Friday and Saturday having the least number of calls.25 Only the studies of Madrid and London services reported the percentages of case types that contributed to total workload. This means that no comparisons between types of workload could be conducted across cities in order to examine the basis of the differences between the Singapore study and the studies in the other cities.20 ,26

Day of week patterns for case types were similar across the studies.17 ,20–22 ,46 ,47 Alcohol related cases were more common on Friday and Saturday nights, especially among young male subjects.20 ,46 ,47 Trauma related cases were higher on weekends than weekdays also occurring most frequently among young male subjects.17 ,21 ,22 Elderly trauma was more commonly found to occur on Wednesdays and Thursdays.21

Data on day of week were reported in four studies of out-of-hospital cardiac arrest. While studies of services in Singapore, Sweden and Berlin found that Monday was the most common day for cases of out-of-hospital cardiac arrest,29 ,32 ,37 a study of over 9000 patients from nine regional centres across the USA and Canada found that there was no significant differences across the different days of the week.30

Month of year, season

In all, 17 out of the 38 studies reported either month of year or seasonal trends. Winter months were associated with increased frequency of out-of-hospital cardiac arrest,29 ,30 ,32 ,38 high acuity cases and elderly workload in general.27 ,53 Summer months were associated with increases in the frequency of heroin and alcohol related cases.46 ,48 Trauma cases occurred most frequently in the summer months in studies in the northern hemisphere,17 ,22 but trauma cases in the southern hemisphere occurred most frequently during autumn (March, April, May); however, this was limited to a single study.21

Only one out of the five studies examining total ambulance workload reported month of year or seasonal patterns. In the study of cases from Madrid, June (early summer) was the busiest month with August (late summer) being the quietest month.20 The researchers suggested that this finding resulted from people leaving the city during the annual vacation period. No comparisons could be made with other ambulance services due to a paucity of published research.


This review has two limitations. Only English language publications were included and there was limited searching of ‘grey’ literature (unpublished studies and reports) as our search strategy was limited to available published research; there may be other data and reports used in ambulance service internal planning that we were unable to access.


Ambulance temporal patterns are an important source of information. People requesting and using ambulance services are different populations to those found in inhospital studies. Therefore, it is important to examine ambulance temporal patterns and not infer trends from hospital-based studies. A better understanding of temporal patterns in ambulance demand is required so that ambulance services can make more informed decisions about how best to use the limited ambulance resources that are available.

Time of day patterns in total demand were similar across the ambulance services studied. Day of week patterns for total demand were not consistent across the reporting ambulance services; however, only two studies reported the percentages of case types that contributed to total demand so no comparisons between types of demand could be undertaken to explain differences. Only one study reported on demand by month of year.

There were strong temporal patterns in the small number of ambulance case types examined and these patterns were influenced by age and gender. These case type patterns were consistent even though the methods used to assign cases to case types were different across studies and often not well described. For example, trauma cases in the 15–34 year age group were likely to occur over the weekend period while the 75+ age group were more likely to have a traumatic incident requiring ambulance attendance on a Wednesday or Thursday.21 Most case types have similar time of day and day of week temporal patterns across ambulance services and countries. As these patterns are consistent across many jurisdictions, more studies should be undertaken which examine ambulance demand by case types that are assigned using valid and comparable methods as the results are relevant to a range of ambulance services.

Information obtained from this research can benefit ambulance services. The ambulance service in Singapore used the results obtained from time of day pattern research to change the deployment strategy of ambulances to out-of-hospital cardiac arrest cases.36 This relatively low-cost change was associated with significantly reduced response times to cardiac arrest cases.36 However, further research is needed to better understand patterns across a larger range of case types. Distributions of case types might give an explanation of patterns seen in overall demand. Important explanatory variables such as age and gender should also be investigated as a limited number of studies have shown that these impact on demand in relation to certain case types. This information will provide ambulance services and health policy makers with the knowledge needed to make appropriate and responsible development of strategies for future management of ambulance services. The results should also encourage ambulance services to examine their own patterns of demand, as differences were observed across jurisdictions. Importantly, we would encourage the publication of findings from these types of studies so that innovations in service planning and delivery can be shared across ambulance services.


  • Contributors The manuscript forms the foundation of a doctoral research project. KC undertook the search of the peer-reviewed and grey literature. She collated the articles and synthesised the literature for the manuscript. In their capacity as supervisors and mentors, AM, PD and KS raised issues for discussion pertaining to the first author's interpretation of the findings in the review, and contributed to the preparation of the final manuscript.

  • Funding None.

  • Competing interests None.

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


Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of EMJ.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.