Abstract
Introduction
The impact of telemedicine on the access and quality of paediatric emergency care remains largely unexplored because most studies to date are focused on adult emergency care. We performed a systematic review of the literature to determine if telemedicine is effective in improving quality of paediatric emergency care with regards to access, process measures of care, appropriate disposition, patient-centred outcomes and cost-related outcomes.
Methods
We developed a systematic review protocol in accordance with PRISMA (Preferred Reporting Items for Systematic Review) guidelines. We included studies that evaluated the impact of synchronous and asynchronous forms of telemedicine on patient outcomes and process measures in the paediatric emergency care setting. Inclusion criteria were study setting, study design, intervention type, age, outcome measures, publication year and language.
Results
Overall, 1.9% (28/1434) studies met study inclusion and exclusion criteria. These studies revealed that telemedicine increased accuracy of patient assessment in the pre-clinical setting, improved time-to disposition, guided referring emergency department (ED) physicians in performing appropriate life-saving procedures and led to cost savings when compared to regular care. Studies focused on telepsychiatry demonstrated decreased length of stay (LOS), transfer rates and improved patient satisfaction scores.
Discussion
Our comprehensive review revealed that telemedicine enhances paediatric emergency care, enhances therapeutic decision-making and improves diagnostic accuracy, and reduces costs. Specifically, telemedicine has its most significant impact on LOS, access to specialized care, cost savings and patient satisfaction. However, there was a relative lack of randomized control trials, and more studies are needed to substantiate its impact on morbidity and mortality.
Introduction
The American Telemedicine Association defines telemedicine as ‘the use of medical information exchanged from one site to another via electronic communications to improve patients' health status’. 1 Despite the fact that the emergency department (ED) serves as the entry point into the healthcare system for many patients and for some, as their only access to healthcare, the state of emergency care in the United States is uneven, with one of the reasons being inadequate access to high-quality care. Telemedicine has improved access, reduced disparities and enhanced quality of emergency/acute care to remote settings; however, most existing data on effectiveness involve the adult population, and its effectiveness in the context of paediatric emergency care remains largely unexplored. 2 3 A 2016 review of telemedicine in paediatric emergency care by Gattu et al. focused on challenges regarding implementation; however, patient-centred outcomes were not explored. 4 A 2018 review by Butterfield focused on telepsychiatry care in the emergency setting, but did not consider telemedicine provision from other medical specialties. 5 A 2018 review by Nadar et al. analyzed telemedicine's effects on clinical outcomes and quality of care, but did not focus on cost-related outcomes or include asynchronous (store-and-forward) telemedicine applications in its inclusion criteria. 6 And a 2018 review by Sheikhtaheri and Kermani reported on the application of telepaediatrics across various medical specialties, but did not analyze particular outcomes across the included studies. 7 To address these existing gaps, we performed a systematic review to determine if telemedicine is effective in improving quality of paediatric emergency care with regards to access, process measures of care (length of stay, time-to-intervention), appropriate disposition (transfer rates) and patient-centred outcomes (satisfaction, morbidity and mortality rates) and cost-related outcomes.
Methods
We developed the study protocol in accordance with PRISMA (Preferred Reporting Items for Systematic Review) guidelines and registered it in PROSPERO (The International prospective register of systematic reviews) (supplementary Appendix 1). A preliminary search for existing systematic reviews on this subject was conducted using the Cochrane Database of Systematic Reviews, PubMed, Proquest and PROSPERO databases.
Eligibility criteria
We included cohort studies, trials, randomized control trials, comparative studies, scoping reviews, systematic reviews, and research letters on children between the ages of 1 day to 18 years implementing telemedicine in the pre-hospital or ED context that studied patient outcomes, process measures and cost-related outcomes. We limited our search to English language studies from the year 2000 to 2019. Most commonly used forms of telemedicine became widely accessible in the year 2000, which marks the time period when cellular network capabilities such as 3G became available. We excluded opinion pieces and case reports and studies with settings outside of the pre-hospital or emergency care context, such as intensive care unit-based studies.
Intervention
Telemedicine interventions included synchronous and asynchronous formats such as videoconferencing, remote transmission of images, store-and-forward technology and remote monitoring of vitals. We excluded audio-only phone and smartphone application-based telemedicine.
Outcomes
Our primary outcomes of interest were morbidity and mortality rates, time-to-intervention, time-to-diagnosis, time-to-provider, changes in management, length of stay (LOS) and patient satisfaction. We specifically looked at the percent reduction or increase between the intervention group and control group and did not include subjective outcomes or those collected in a non-standardized manner because of the inability to measure them objectively. Secondary outcomes of interest included transfer rates and cost-related outcomes.
Search strategy
We collaborated with a health sciences librarian with expertise in systematic reviews and performed a comprehensive literature search using terms similar to ‘telemedicine’, ‘telehealth’ and ‘telemonitor’, amongst others, to ensure sensitivity. The search was performed on electronic databases including CINAHL (Cumulative Index to Nursing and Allied Health Literature), Ovid Medline, Embase, Scopus and Web of Science. The full search strategies for all five databases are listed in supplementary Appendix 2.
Study records
We compiled all identified studies into an electronic reference library (EndNote) and then uploaded the references to a web platform (Covidence) that facilitates systematic review, 8 allowing reviewers to screen each article independently while being blinded to the other reviewer’s decision to either ‘include’, ‘exclude’, or ‘maybe include’. All three reviewers first screened the same 200 abstracts to determine the inter-rater reliability and the Fleiss’s kappa coefficient between them. 9 The remaining abstracts were then independently screened by just two reviewers, with cases of non-concordance being adjudicated by the third reviewer. Articles included for full-text review were similarly screened by two reviewers and disagreements adjudicated by the remaining reviewer. Reasons for exclusion in the full-text review were assigned to each article based on which criterion was met first in the following order: irrelevant study setting, irrelevant telemedicine intervention or no intervention, irrelevant study design, irrelevant outcome(s), non-paediatric study population, study published before the year 2000, non-English language, duplicate study, full-text unavailable.
We developed a standardized data abstraction form based on PRISMA guidelines and recorded information on the author(s), year of publication, study location, study setting, sample size, patient characteristics, study aim, telemedicine intervention type, comparator and details of these, primary/secondary outcomes and key findings that relate to the systematic review objective. For any data that were missing or unclear in a study, we contacted the author(s) and sought clarification. We assessed studies for risk of bias using the Newcastle–Ottawa scale (NOS) that examines methodological quality and rigour of non-randomized studies. 10
Finally, at the completion of our review, a repeat search was performed to capture relevant studies published in the interim.
Results
Selection of articles
Our search generated 2478 articles. After de-duplication and screening, 20 original studies eventually met inclusion criteria and were included in the systematic review analysis (Figure 1). At the completion of our review, a repeat search was performed to capture relevant articles published in the interim, resulting in an additional 141 articles. These additional articles were reviewed and yielded two additional papers. The Fleiss’s kappa coefficient, between all three reviewers prior to the start of the screening process, was 0.91.
Flow diagram of article selection process.
Results were prioritized and analyzed based on primary and secondary outcomes of interest as listed below. Table 1 provides a summary of the patient outcomes and process measures reported in each study.
Data extraction of included studies.
Primary outcomes
Morbidity and mortality
We found one study that assessed the risk of mortality, and no study specifically evaluated mortality outcomes. Dayal et al. measured the impact of telemedicine on patient outcomes in paediatric patients transferred to the paediatric intensive care unit (PICU) from EDs with and without telemedicine by calculating the paediatric risk of mortality III (PRISM III) score on admission, and they found that the patients managed using telemedicine were significantly less sick at admission than those not managed using telemedicine suggesting that paediatric expertise provided through telemedicine enabled earlier stabilization. 11 Another study conducted by Dharmar et al. determined that the quality of care for critically ill children who presented to rural EDs was higher for those receiving telemedicine consultations versus those receiving telephone consultations or no consultations. 12
Some authors sought to evaluate predictors or influential factors of morbidity such as diagnostic accuracy and rate of errors when using telemedicine. In a study conducted by Harvey et al., paediatric intensivists who provided consultations to rural EDs concluded that their assessments were more accurate when using telemedicine compared to when using a telephone. There was a statistically significant difference in accuracy between telemedicine and telephone reports of the patient, with 6.8% of telephone consultations rated as poor compared to 0.0% of telemedicine consultations (p < 0.001). 13 Furthermore, use of telemedicine reduced ICU-level transfers without any increase in transfers to higher-level care within 24 h. 13 Another study which examined the accuracy of store-and-forward technology in smartphones for infectious disease consultations for children with rash in the ED found that the initial diagnosis was identical to the final diagnosis in 96.3% of the cases when using store-and-forward technology and was the first of its kind to evaluate the utility of smartphones for the consultation and diagnosis of paediatric rash in the ED. 14 Finally, Dharmar et al. found a reduction in physician-related medication errors when using telemedicine, which were significantly fewer than the telephone consultation group or the no consultations group (3.4% vs 10.8% and 12.5%, respectively; p< 0.05) for paediatric critical care consultations. 15
In summary, studies revealed telemedicine can lower the risk of mortality in critically ill children, improve diagnostic accuracy and quality of care, and decrease physician-related medical errors.
Change in management
Given the enhanced audio-visual capabilities of telemedicine, multiple studies aimed to determine if telemedicine altered patient management, including level of acuity, diagnosis, treatment and disposition. A retrospective analysis performed by Dharmar et al. found that referring ED physicians in rural EDs changed their diagnosis 47.8% vs 13.3% (p < 0.01) of the time when consultations for critically ill children were provided using telemedicine rather than telephone, respectively. 12 Telemedicine also led to a change in therapeutic interventions 55.2% vs 7.1% (p < 0.01) of the time, respectively. Similarly, Heath et al. conducted a prospective analysis of a rural telemedicine programme for critically ill children and found decisions surrounding intubation were changed in 30% of the cases, thus demonstrating the ability of telemedicine to alter life-saving therapeutic decisions. 16 Notably, 11% of the patients were intubated after the recommendation was made via telemedicine, and in 19% of the cases, intubation was avoided. A study comparing usual in-person nursing triage to a physician-operated robot in triage showed that the physician-placed diagnostic orders were more aligned with the primary ED provider than with the nursing-placed orders. 17 Regarding diagnosis, Roberts et al. showed that telepsychiatry consultations led to a higher proportion of patients diagnosed with adjustment disorder (22% vs 8.3%, p = 0.004) compared to in-person consultations. 18
Many studies specifically looked at changes in disposition. Fugok et al. determined that the telemedicine consultation prior to paediatric transport improved patient assessment and therefore disposition. Over a 1-year period, they found that ED utilization decreased from 57% to 27% (p < 0.0001) and PICU admissions increased from 11% to 34.4% (p < 0.0001). 19 Of those directed to the ED, 58.6% were admitted to the regular inpatient floor for further care, 13.7% to PICU for escalation of care and 24.1% were discharged. Of those directly admitted to the PICU, 90.4% had care escalation and 9.6% were observed in the PICU without escalation. 19 Similarly, Johnson et al. showed that telemedicine-equipped emergency medical services (EMS) and fire units enhanced pre-hospital triage and reduced the frequency of low-acuity transports and ED visits. Results showed that 88% (95% CI 86–90%) of the patients included were triaged for non-EMS transport, and of these, 22% (95% CI 20–25%) were diverted to non-ED destinations. 20 A study conducted by Harvey et al. showed a decrease in PICU admissions when critical care telemedicine consultations were provided instead of telephone consultations to rural EDs. The odds of telemedicine triage to a non-ICU setting were 2.55 times higher than with telephone consultations. 13 In 95.7% of cases where patients were transferred to non-ICU settings after a telemedicine consultation, referring ED providers felt that telemedicine influenced their decision (p < 0.001). 13 Yang et al. demonstrated similar findings, which showed acutely ill children receiving critical care telemedicine consultations in rural EDs had significantly lower admission rates (59.5% vs 87.5%, p <0.05) and were more likely to get discharged (29.7% vs 6.3%, p <0.05) compared to those receiving telephone consultations. 21 Lastly, Patel et al. conducted a prospective pilot study of telemedicine use in pre-hospital transport compared to telephone alone and found no statistically significant change in disposition or orders, but noted the study was underpowered for this aim, as they were primarily evaluating feasibility and ease of use of telemedicine. 22
In summary, telemedicine can substantially alter remote decision-making, at times leading referring providers to perform life-saving measures, while avoiding unnecessary invasive procedures at other times. Telemedicine can also reduce the frequency of EMS transports, ED visits and admission rates, and improve disposition determinations through an enhanced ability to assess patients.
Process measures
Several studies evaluated telemedicine’s impact on ED process measures such as time-to-provider, time-to-diagnosis and LOS. Two studies evaluated the use of telepsychiatry in the paediatric ED compared to in-person consultations and found a statistically significant reduction in LOS on average 92 h less per month compared to previous months without telepsychiatry (p = 0.032) and reduced travel time for in-person evaluations by 75%. 23 Findings supported a prior study by Thomas et al. that showed a significantly shorter mean ED LOS compared to usual in-person care (5.5 h vs 8.3 h, respectively, p < 0.01) and concluded that telemedicine improved clinical and operational efficiency at each of its operational sites. 24
A study by Bucak et al. aimed to determine if notification of abnormal lab results via text messaging shortened LOS for patients in the paediatric ED observation unit. In contrast to previous studies, results showed that this telemedicine intervention did not reduce LOS for patients ultimately discharged from the unit, but did reduce LOS by an average of 66.3 minutes for patients who required more rapid hospitalization (p<0.001). 25 Devrim et al. identified the potential for shorter LOS using store-and-forward technology in smartphones for infectious disease consults in the ED in comparison to in-person consultations, particularly on night shifts, though this was not explicitly measured. 14
Marconi et al. studied the effects of a robot with remote-controlled tele-presence technology with full audio and video capabilities, including a built-in stethoscope, on triage times. The more advanced form of telemedicine technology allowed a remote physician to see and interact with patients and staff while triaging. Their results showed a statistically significant increase in triage times by an average of 21 seconds in comparison to usual nursing triage. 17 Results also showed an increase in accuracy of triage scores with use of the robot. Nurse triage scores were 71% accurate compared to physician telemedicine scores, which were 95% accurate (accuracy determined by the in-person paediatric emergency physician eventually treating the patient). 17 Lastly, Patel et al. determined videoconferencing during EMS transport resulted in a shorter duration of calls by 47 seconds compared to the use of telephone alone. 22
Most studies involving videoconferencing or store-and-forward forms of telemedicine demonstrated a reduction in LOS. One study found longer triage times; however, this difference was small and involved a newer and more advanced form of technology in the form of a robot that physicians were not accustomed to using. 17
Patient satisfaction
Dharmar et al. reported higher parent satisfaction scores of telemedicine consultations in the ED compared to telephone consultations. 12 Within child and adolescent mental health, Roberts et al. found higher patient satisfaction scores with telepsychiatry consultations in rural and remote EDs via the Ontario Telemedicine Network compared to travel to an urban ED for face-to-face consultation. 18 The increase in scores was attributed largely to reduced travel time and cost for patients. 18 Similarly, Thomas et al. studied patient and caregiver satisfaction of telepsychiatry consultations at in-network EDs compared to EMS-to-ED transfers for in-person consultations. 24 Average ratings were above four (on a 5-point scale) and patients and caregivers specifically noted the benefit of not having to travel to the main hospital. 24 Respondents were overall satisfied with the effectiveness and efficiency of telepsychiatry, which was largely due to the statistically significant reduction in LOS and total patient charges (p < 0.001). 24 Lastly, one study reported patient and caregiver satisfaction with use of a high-functioning robot in triage. Patients were more likely than their parents to request the robot again at their next ED visit, which was thought to be due to children’s regular exposure to new technologies and therefore willingness to accept new technologies. 17
Secondary outcomes
Cost-related outcomes
Five studies evaluated cost-related outcomes of telemedicine. In a recent study conducted by Thomas et al., telepsychiatry, in comparison to in-person consultation, resulted in cost savings of $5118 on average, per patient. 24 Similarly, 1 year prior to this study, Roberts found that the implementation of telepsychiatry enabled patients and their families to gain access to specialized care without having to travel long distances, which allowed them to save time and money. 18 Telepsychiatry ultimately increased patient satisfaction scores as well. Apart from telepsychiatry, studies performed in rural EDs receiving critical care telemedicine consultations also demonstrated cost savings. Dharmar et al. surmised that since telemedicine consultations for critically ill children presenting to rural EDs in northern California led to significantly fewer physician-related errors, telemedicine would, in turn, decrease health care costs. 15 Their hypothesis was supported by the findings of Yang et al. who conducted an economic evaluation of telemedicine consultations provided by paediatric intensivists to children in rural EDs in California. Findings showed an annual cost saving of $4662 per child and a return on investment (ROI) of 1.28. 26 Lastly, Dharmar et al. showed that the average hospital revenue per year increased from $2.4 million pre-telemedicine to $4 million post-telemedicine in 16 rural hospitals. 27
All studies that evaluated the impact of telemedicine on cost-related outcomes showed that telemedicine could lead to cost savings not only for the hospital or health system, but also for patients and their families.
Transfer rates
As telemedicine improves access to specialized care, multiple studies aimed to determine if it affected transfer rates. Yang et al. found that critical care telemedicine consultations given to children in rural EDs reduced transfer rates by 31%. 21 Similarly, in a study of rural communities in Ireland, Holt et al. showed that remote consultation via robotic technology reduced use of paediatric transport services and allowed two-thirds of patients originally thought to need transfer to instead be treated at their community hospital. 28 Shifting to Australia, multiple rural EDs in Queensland collectively avoided 47% of transfers by utilizing telemedicine consults instead of telephone consults. 29 None of the patients required transport in the remainder of their course due to clinical deterioration, and no adverse events occurred.
Overall, telemedicine consults compared to telephone consults led to a decrease in transfer rates among rural communities in the United States, Ireland and Australia.
Discussion
Our updated and comprehensive review of paediatric telemedicine literature reveals that despite heterogeneity, telemedicine improves patient outcomes in the paediatric emergency care setting across multiple domains including timeliness and accuracy of diagnosis, access to appropriate care with timely transfers or avoidance of transfers all together, reduction of medication errors, cost savings, patient satisfaction and overall quality of care.
In comparison to previous scoping reviews of synchronous telemedicine, our review focused on the paediatric population, included both synchronous and asynchronous forms of telemedicine, and evaluated a more broad set of patient outcomes and process measures. 30 Furthermore, we included an additional 4 years of literature in comparison to previous reviews. 4
Recent literature suggests there is an increase in accuracy of evaluation and management of patients via telemedicine. 19 The ability to remotely and more clearly determine the clinical status of patients is largely due to the audio-visual capabilities of telemedicine that augment remote decision-making. Critical care consultants can guide referring ED providers and advise them on the need to perform or delay invasive but potentially life-saving measures such as intubations. 16 Teleconsultation can also positively impact patient management in terms of assessment and non-procedural therapeutic interventions, thus providing a higher quality of care. 12 19
Though 20% of paediatric patients in the United Staes reside in rural communities, only 3% of paediatric subspecialists practice in rural settings. 16 Our review finds that telemedicine consultations are especially meaningful for patients in rural areas who otherwise have limited access to specialized care, serving as a vital virtual bridge between rural communities and specialized acute care facilities. Telemedicine reduces inter-facility transfer rates for paediatric patients and leads to more efficient resource utilization and substantial cost savings. 21 26 28 29
Similarly, telemedicine has the potential to improve resource allocation by preventing low-acuity EMS transports and diverting them from the ED altogether. 20 In addition to lessening the burden of travel, telemedicine capabilities have decreased LOS, especially in telepsychiatry where there are substantial gaps in access to adequately trained paediatric mental health professionals. 23 24 Patients have reported higher satisfaction rates, cost savings and willingness to use telemedicine again for future visits. 12 24 These developments suggest the economic benefits of telemedicine and its ability to provide specialized care without associated increase in travel time and cost. 27
Recent data show that 76% of US hospitals could connect with patients via telemedicine and over half had implemented remote patient monitoring capabilities. 31 Driven by changes in Medicare and Medicaid policies, telemedicine uptake has rapidly increased across US hospitals. Most recently, the coronavirus disease 2019 (COVID-19) pandemic has led to new and unforeseen demands of healthcare delivery that have further accelerated the use of telemedicine. As people within the United States and around the world are forced to socially distance, clinics and hospitals have rapidly developed and re-modelled infrastructure to support telemedicine capabilities.
As technology advances, telemedicine’s potential to impact patient care grows in parallel and can be especially meaningful in communities where there is a gap between specialized medical care needs and available resources. In particular, audio-visual capabilities were shown to have a greater impact in several studies compared to telephone use alone.12131921–32 As such, access to this technology is paramount. While the effect of telemedicine on morbidity and mortality rates is less studied, the literature does suggest that telemedicine can decrease the risk of mortality given studies have shown an increase in accuracy of assessment, significantly altered management plans and improved process measures, which may all contribute to improved patient outcomes. 11 13 Further studies more directly examining the impact of emergency telemedicine on paediatric morbidity and mortality are needed.
Limitations
There were several limitations to our review. First, our search strategies required MeSH terms and keywords that were specific to paediatrics and therefore may have excluded studies that involved telemedicine in both the paediatric and adult populations where the paediatric population was not specified in the title, abstract or MeSH terms. To address this, we screened all references of studies included for review to ensure capture of all relevant studies. Second, our keyword searches related to emergency medicine were limited to the title field (rather than title and abstract) in Embase and Scopus to mitigate thousands of irrelevant results when including the abstract search. Third, our review excluded studies in a non-English language, but we found only three studies in total (two in French and one in German) that were excluded for this reason. Lastly, while access to telemedicine in post-discharge care can be important for improving longer term access to care, given it does not directly influence management decisions in the emergency setting, we considered it out of scope for this review.
Conclusion
Our systematic review has shown that telemedicine can play an essential role in paediatric emergency care delivery by improving patient outcomes and process measures, reducing harm from unsafe events, reducing costs and improving the overall quality of care. Given its potential to link rural and remote paediatric patients to specialized care and facilitate appropriate and life-saving management, health systems should prioritize enhancing their telemedicine capabilities with a minimum of videoconferencing capabilities.
Supplemental Material
sj-pdf-1-jtt-10.1177_1357633X211010106 - Supplemental material for Telemedicine in paediatric emergency care: A systematic review
Supplemental material, sj-pdf-1-jtt-10.1177_1357633X211010106 for Telemedicine in paediatric emergency care: A systematic review by Aditi Mitra, Rubina Veerakone, Kathleen Li, Tyler Nix, Andrew Hashikawa and Prashant Mahajan in Journal of Telemedicine and Telecare
Footnotes
Author note
Aditi Mitra is now affiliated with St. Joseph Mercy Hospital, USA. She is also affiliated with Michigan State University, USA.
Acknowledgements
We acknowledge Alexander Rogers MD, Jim Cranford PhD and Whitney Townsend for their efforts in conceptualizing the methodology and specifically acknowledge Dr. Cranford for performing the statistical analysis to determine the inter-rater reliability.
Contributors’ statement
Mitra conceptualized and designed the study, designed the data collection instruments, collected data, carried out the initial analyses, drafted the initial manuscript, and reviewed and revised the manuscript. Ms. Veerakone and Dr. Li reviewed the studies, designed data collection instruments, collected data, and critically reviewed and revised the manuscript for important intellectual content. Mr. Nix designed the search strategy and performed the search in addition to reviewing and revising the manuscript. Dr. Hashikawa conceptualized and helped design the study and in addition, critically reviewed and revised the manuscript. Dr. Mahajan conceptualized and designed the study and critically reviewed and revised the manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
