Patterns in the Use of Emergency Telemedicine Support in the Management of Traumatic Road Crashes

Abstract

Objective:

The introduction of emergency telemedicine care models is a common theme in health jurisdictions that include rural and remote populations. How the availability of these models influences the way clinicians manage traumatic road crashes is not yet fully understood. This study seeks to compare road crashes where telemedicine was and was not used and to identify any variables that may increase the likelihood of telemedicine usage by treating clinicians.

Methods:

Road crashes reported in the state Department of Transport and Main Roads (Queensland, Australia) crash database between January 1, 2019, and November 30, 2020 (n = 23,734) were compared to videoconferencing call logs to determine which crashes resulted in treatment that was supported by telemedicine (n = 204). Analysis was performed to examine differences in characteristics related to the crash depending on whether telemedicine support was requested.

Results:

Road crashes where telemedicine support was requested on average involved more casualties (1.6 vs. 1.41; t(11,287) = −3.26, p < 0.001, relative risk = 1.13). Crashes that occurred in rural settings accounted for most requests for telemedicine (65.68%; X ² = 159.2, p < 0.001) and a greater percentage of crashes in remote locations (3.36% vs. 2.35%; X ² = 256.97, p < 0.001, relative risk = 1.43). The use of telemedicine support for crashes was associated with a 13% increase in the mean number of casualties, compared to crashes where telemedicine support was not used.

Conclusion:

Telemedicine support is requested by clinicians providing emergency treatment in the management of road crashes that produce more severe injuries, involve multiple casualties, and take place in more rural settings or remote locations.

Introduction

Despite a growing body of evidence that telemedicine is effective in the management of unscheduled hospital presentations in rural and remote locations, levels of usage vary greatly.^{1

–12} Even in evaluations of emergency telemedicine models that describe positive effects such as improved clinical effectiveness, speed of care, or reduced transfers, telemedicine was used in as few as 0.8% of all possible opportunities.⁹ Investigations have typically focused on the impact of telemedicine in instances where it was used and less frequently compare the circumstances between when telemedicine was and was not used. Understanding these differences may provide practical contributions to clinical guidelines and how clinicians are trained in the use of telemedicine technology.

Telemedicine support of emergency presentations has been used within Queensland to respond to the needs of Australia's most decentralized mainland population since 2006. Today, over 110 public hospitals and multipurpose health services throughout regional and remote Queensland are connected to emergency clinicians operated by Retrieval Services Queensland (RSQ) with high-definition video cameras. This technology allows frontline clinicians working in isolated health services to link with consultants from tertiary centers for assistance with an unscheduled patient presentation. Critical care continues to be provided until such time that the patient stabilizes or is retrieved.

Evaluating the use of telemedicine support for road crash victims is challenging because each event is a unique combination of many variables making their comparison difficult. Clinical users of emergency telemedicine support models have been asked for their opinions of its efficacy,^13
–15 and studies have produced evidence of improved system outcomes such as reductions in emergency wait times of 12 min.^15

–18 The authors, however, are not aware of any attempt to systematically compare specific road crashes grouped by whether telemedicine models of care were used or not.

While RSQ has developed several documents guiding the notification of severe trauma and the use of telemedicine in emergency settings (Figure 1), there are no prescriptive indications for the use of telemedicine in severe trauma. This project aimed to determine if there are patterns in the use of telemedicine among clinicians in the absence of any requirement to use of telemedicine. This may indicate that under certain circumstances, clinicians are independently determining that emergency telemedicine support would assist them in the management of their patients.

Fig. 1.

Criteria for early notification of trauma for interfacility transfer.

Methods

The Department of Transport and Main Roads (Queensland, Australia) was contacted for an extract from their crash database of all crashes reported on public roads in Queensland between January 1, 2019, and November 30, 2020, the most recent validated data available at the time of analysis. The Human Research Ethics Committee at The Royal Brisbane and Women's Hospital determined the retrospective nature of this research involved no more then minimal risk to the subjects. Crashes in which occupants were uninjured or resulted in injuries that did not require hospitalization were removed to create a subset of crashes that could potentially require remote telemedicine support.

A data matching exercise was performed to compare video call history logs that closely match in both time and location to a road crash that required hospitalization (a video call made to a hospital within 2 h and in the same Local Government Area of a reported crash was proposed to be in response to that crash). Proposed matches were then validated by the review of patient-level records retained by RSQ by an emergency physician to confirm that treatment provided was consistent with that which would be provided to a road crash victim (Fig. 2).

Fig. 2.

Selection of 204 cases for further analysis from the original dataset.

A total of 23,734 road crashes occurring within Queensland during the study timeframe were reduced to a subset of 11,287 when crashes not producing injuries that require hospitalization were removed. From this list, 211 (1.87%) potential instances of RSQ making a video call to a regional hospital in response to the presentation of road crash victim(s) were identified. A validated list of 204 (1.81%) matches were produced after reviewing available medical records.

Variables were collected on each crash including

Remoteness. The Accessibility and Remoteness Index of Australia (ARIA) divides the nation into five classes of remoteness based on a measure of relative access to services (major cities, inner regional, outer regional, remote, or very remote).

Setting. Significant Urban Areas (SUAs) represent Urban Centres, or groups of Urban Centres that contain a population of 10,000 or more. A crash occurring within an SUA is considered urban otherwise classed as a rural setting.

Severity. The severity of injuries to all occupants in a road crash is recorded and includes minor injury or medically treated where paramedic or hospital staff provide care to a person, but they are not admitted to hospital, hospitalized when they are admitted to a hospital or a fatality, where a person dies within 30 days of a crash as a result of injuries received in the crash.

Crash Nature. Determine by the initial event in any sequence of events in a road traffic crash. Crash Nature categories include angle, fall from vehicle, head-on, hit animal, hit object, hit parked car, hit pedestrian, overturned, rear-end, and sideswipe.

Crash Type. A logical category into which one or more similar crash natures is classified. Crash natures can be grouped into the following crash types: Single vehicle (hit parked vehicle, hit fixed obstruction or temporary object, overturned, and fall from moving vehicle), Multivehicle (angle, sideswipe, rear-end, and head-on) or Hit Pedestrian.

A series of Two Sample t-test or Chi-Square tests (developed in R ver 4.0.5) used the newly combined dataset containing road crash variables (crash severity, crash nature, crash setting, and telemedicine usage) to determine association between telemedicine usage and (1) number of casualties per crash, (2) remoteness, and (3) setting (Table 1).

Table 1.

Comparison of Crashes by Telemedicine Use

	USED	%	NOT USED	%	TOTAL
Crash severity (crashes)
Fatality	7	1.7	413	98.3	420
Hospitalized	197	1.8	10,650	98.0	10,867
Medical treatment	0	0	9,093	100	9,093
Minor injury	0	0	3,354	100	3,354
Remoteness (ARIA)
Major cities	0	0	15,158	100	15,158
Inner regional	91	2.0	4,564	98.0	4,655
Outer regional	90	2.8	3,125	97.2	3,215
Remote	17	3.9	415	96.1	432
Very remote	6	2.2	268	97.8	274
Crash nature (crashes)
Angle	49	0.7	6,486	99.3	6,535
Fall from vehicle	9	1.0	928	99.0	937
Head-on	10	1.3	736	98.7	746
Hit animal	3	1.4	205	98.6	208
Hit object	74	1.6	4,505	98.4	4,579
Hit parked vehicle	5	0.7	714	99.3	719
Hit pedestrian	4	0.4	1,021	99.6	1,025
Other	1	2.9	33	97.1	34
Overturned	20	2.2	878	97.8	898
Rear-end	27	0.4	6,743	99.6	6,770
Sideswipe	2	0.2	1,281	99.8	1,283
Crash setting (crashes)
Urban	70	0.4	19,003	99.6	19,073
Rural	134	2.9	4,527	97.1	4,661
Occupants
Total	327		31,059		31,386
Average per accident	1.6		1.32

ARIA, Accessibility and Remoteness Index of Australia.

Results

Road crashes that prompted the use of a telemedicine support model to assist with management produced a higher average number of casualties in the dataset used by this study compared to those that did not (1.6 vs. 1.41; t(11,287) = −3.26, p < 0.001, relative risk = 1.13). A strong association between telemedicine use and setting was observed with 65.68% of total crashes prompting telemedicine support occurring rurally (X ² = 159.2, p < 0.001). Similarly, crashes occurring in remote locations compared to regional locations (3.36% vs. 2.35%; X ² = 256.97, p < 0.001, relative risk = 1.43) were more likely to initiate a request for telemedicine support. Finally, Crash Type (X ² = 16.878, p < 0.001) and Crash Nature (X ² = 31.89, p < 0.001) were found to be associated with the use of telemedicine support. A visualization of how different crash variables increases the predicted number of casualties in a crash is presented in Figure 3.

Fig. 3.

Average casualties per crash by crash variable (telemedicine used vs. not used).

Discussion

Traumatic road crashes are a well-suited subset of hospitalizations to conduct this retrospective study due to the large number of details about each crash that are routinely collected for law enforcement, insurance, and for ongoing safety improvements to the state road network. The collection of crash severity details, for example, makes it possible to establish a correlation between increased complexity and increased likelihood of telemedicine use which previously had not been conclusively established.³

The results generated in this study support previously observed findings regarding the likelihood of casualties based on the nature of a road crash^19
–21 as well as published literature that describes clinician attitudes toward the value of emergency telemedicine support in the management of complex presentations.^13
–15 The use of emergency telemedicine support in response to a traumatic road crash is associated with a 13% increase in the predicted mean number of casualties. Although a modest effect compared to other crash variables, in particular crash nature, the increased reliance of this telemedicine model in the management of road crashes that involve multiple, severe casualties and take place in rural settings is a strong indication that the model provides benefits to the remote clinical workforce. This benefit is most acutely felt when an emergency presentation threatens to overwhelm staff such as a traumatic road accident managed by a small, rural health service.

Care must be exercised to determine the correct temporal order of this study. Unlike the other variables relating to the severity, nature, and setting of the road crash, the increase in predicted casualties caused using emergency telemedicine models is not the result of individuals being placed at increased risk before the road crash occurring. It is likely that the result of the inclination of clinicians to request this service in the management of road crashes that produce more severe injuries, involve multiple casualties, and take place in more rural settings after they occur. Alternatively put, although a head-on car crash and the use of telemedicine support in the management of a crash are both associated with increased casualties, the former is due to a more dangerous crash nature, while the latter is due to clinicians being more likely to request support for more severe crashes that have occurred.

Emergency telemedicine models of care have been demonstrated to be broadly effective in the management of critical and noncritical presentations as well as delivering additional benefits including reductions in patient retrievals and wait times. This study establishes that in an environment such as a rural health service which experiences frequent low acuity presentations and infrequent high acuity presentations, clinicians do seek support via an emergency telemedicine model to manage a road crash that produce more severe injuries and involve multiple casualties.

Between 1% and 2% of all road crashes that require treatment in Queensland are supported by telemedicine. This number increases when only crashes that occur in remote locations are examined; however, most crashes do not prompt a request for telemedicine support. This may be due to clinicians believing that telemedicine support would not be helpful for all but the most serious crashes or a perception among clinicians that treating injured occupants of a road crash is a situation that they and their colleagues should be independently capable of. If the latter applies, the authors hope that in sharing the results of this study in clinical guidelines and during training on the use of telemedicine will reduce hesitancy to request support when it may be helpful.

Conclusion

Predicting the use of emergency telemedicine support in response to a road crash is a difficult task. Remotely located and isolated clinicians providing initial assessment and treatment to road crash victims may seek collegial support in response to high acuity presentations such as a road crash. This includes the use of emergency telemedicine support when it is available; however, this is not consistently requested. Sharing the patterns of use of emergency telemedicine support models among the rural workforce during telemedicine training and awareness exercises should reduce hesitancy to request telemedicine support when it may be helpful.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Tsou

, Robinson

, Boyd

, et al. Effectiveness of telemedicine in rural and remote emergency departments: Systematic review. J Med Internet Res, 2021; 23(11):e30632.

Du Toit

, Malau-Aduli

, Vangaveti

, et al. Use of telemedicine in the management of non-critical emergencies in rural or remote emergency departments: A systematic review. J Telemed Telecare, 2019; 25(1):3–16.

Wesson

, Kupperschmidt

. Rural trauma telemedicine. J Trauma Nurs, 2013; 20:199–202.

Ricci

, Caputo

, Amour

, et al. Telemedicine reduces discrepancies in rural trauma care. Telemed J E Health, 2003; 9(1):3–11.

Galli

, Keith

, McKenzie

, et al. TelEmergency: A novel system for delivering emergency care to rural hospitals. Ann Emerg Med, 2007; 51(3):275–284.

Herrington

, Zardins

, Hamilton

. A pilot trial of emergency telemedicine in regional Western Australia. J Telemed Telecare, 2013; 19:430–433.

Blackwell

, Kelly

, Lenton

. Telemedicine ophthalmology consultation in remote Queensland. Med J Aust, 1997; 167:583–586.

Armstrong

, Haston

. Medical decision support for remote general practitioners using telemedicine. J Telemed Telecare, 1997; 3:27–34.

Rosengren

, Blackwell

, Kelly

, et al. The use of telemedicine to treat ophthalmological emergencies in rural Australia. J Telemed Telecare, 1998; 4:97–99.

10.

Saurman

, Lyle

, Kirby

, et al. Use of a mental health emergency care-rural access programme in emergency departments. J Telemed Telecare, 2014; 20(6):324–329.

11.

Ward

, Ullrich

, Mueller

. Tele-emergency utilization: In what clinical situations is tele-emergency activated?. J Telemed Telecare, 2016; 22(1):25–31.

12.

Natafgi

, Shane

, Ullrich

, et al. Using tele-emergency to avoid patient transfers in rural emergency departments: An assessment of costs and benefits. J Telemed Telecare, 2018; 24(3):193–201.

13.

Kyle

, Aitken

, Elcock

, et al. Use of telemedicine for patients referred to a retrieval service: Timing, destination, mode of transport, escort level and patient care. J Telemed Telecare, 2012; 18(3):147–150.

14.

Matthews

, Elcock

, Furyk

. The use of telemedicine to aid in assessing patients prior to aeromedical retrieval to a tertiary referral centre. J Telemed Telecare, 2008; 14(6):309–314.

15.

Sharpe

, Elcock

, Aitken

, et al. The use of telemedicine to assist remote hospital resuscitation and aeromedical retrieval tasking: A 12-month case review. J Telemed Telecare, 2012; 18(5):260–266.

16.

Fairchild

, Ferng-Kuo

, Rahmouni

, et al. Telemedicine decreases rural emergency departments wait times for behavioral health patients in a group of critical access hospitals. Telemed J E Health, 2019; 25(12):1154–1164.

17.

Ward

, Jaana

, Natafgi

. Systematic review of telemedicine applications in emergency rooms. Int J Med Inform, 2015; 84:601–616.

18.

Mohr

, Harland

, Chrischilles

EA et al

. Emergency department telemedicine is used for more severely injured rural trauma patients, but does not decrease transfer: A cohort study. Acad Emerg Med, 2017; 24:177–185.

19.

Winston

, Senserrick

. Risk factors for death among older child and teenaged motor vehicle passengers. Arch Pediatr Adolesc Med, 2008; 162(3):253–260.

20.

Liu

, Pressley

. Side impact motor vehicle crashes: Driver, passenger, vehicle and crash characteristics for fatally and nonfatally-injured rear-seated adults. Inj Epidemiol, 2016; 3:23.

21.

Mitchell

, Bambach

, Foster

et al. Risk factors associated with the severity of injury outcome for paediatric road trauma. Injury, 2015; 46:874–882.