1 Introduction

The emergency department (ED) is often the most crucial and busiest unit within a hospital, as it is where urgent medical care is delivered to patients in need of immediate attention; it also commonly serves individuals experiencing difficulty accessing primary care [1, 2]. However, delivering high-quality and timely care in the ED may be limited by available resources, such as beds, medical personnel, nursing staff, laboratories, and imaging facilities [3].

Many hospitals have implemented an electronic medical information system (EMIS), which is defined as technology used to collect, organize, store, analyze, and report health information [4, 5], in an attempt to optimize the delivery of ED care [6, 7]. In addition, hospitals have also adopted the use of specific EMIS tools, such as push notifications to alert of specific events, dashboards to assist with the coordination of care services, and real-time analytics to predict the likelihood of defined outcomes (such as admission or revisit) [8, 9]. To date, existing literature on EMIS use in the ED has primarily focused on perceived benefits and downfalls of implemented technologies [10,11,12]; reported findings specific to timeliness of care and patient outcomes are diverse and sometimes conflicting.

Although use of an EMIS in ED settings may improve time-based patient outcomes, such as time to treatment, time to diagnosis, and length of stay [13,14,15], there remains a paucity of information on this topic and additional research is required to determine the effects of these technologies on time-based ED metrics. Therefore, the primary objective of our scoping review was to identify and describe interventions aimed at improving timeliness of ED care that are based on the use of an EMIS. Our secondary objective was to summarize the effects of examined interventions on the timeliness of ED care.

2 Methods

Our review complies with the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist [16].

2.1 Studies and participants

Primary studies examining the effects of an electronic medical information system (EMIS) or EMIS tool (individually and collectively referred to as EMIS hereafter) on the timeliness of patient care in the emergency department (ED) were eligible for inclusion in this review. This consisted of observational studies, pre-post studies, time-series trials, and randomized control trials (RCTs). Systematic reviews, case series, case reports, anecdotal reports of provider experiences, commentaries, and letters to the editors were excluded. Eligible participants comprised all patients receiving care at an ED. Participants were not restricted by age, sex/gender, diagnosis, or ED characteristics (such as location, size, or type).

2.2 Eligible interventions and comparison/control

Examined interventions included the use of an EMIS, which may comprise of electronic medical record (EMR) systems and/or specific programs, modules, or applications within an EMIS. Interventions had to have been compared with non-use, or to the use of a different EMIS. Studies that compared EMIS upgrades relative to an existing EMIS (without the upgrade) were also eligible for inclusion, since such studies allow for the measurement of effects of specific EMIS upgrades on timeliness of ED care.

2.3 Outcome measures

Outcomes of interest were quantitative measures of timeliness of care in the ED. These included, but were not limited to, time to triage, first physician assessment, first procedure or medication, or discharge. Studies reporting solely on outcomes that are not explicit measures of ED care timeliness were not included in this review.

2.4 Electronic searches

We searched the following four bibliographic databases to identify all potentially eligible studies: MEDLINE, PubMed, Web of Science, and the Cochrane Database of Systematic Reviews. Searches were designed with the assistance of a research librarian and executed on August 25, 2021; search results were limited to English language studies published after 2014 to capture the most relevant reports of time-based outcomes in the ED. Our search strategies are presented in Appendix 1 of the Supplementary Material.

2.5 Selection of studies

Citations retrieved from database searches were imported into DistillerSR (Evidence Partners Incorporated, Ottawa, Canada). Following the removal of duplicates, the title and abstracts of remaining citations were screened by two independent reviewers. Inclusion of the citation by at least one reviewer at this stage was necessary in order for the citation to advance to full-text screening. Full-text screening was performed by two independent reviewers. Disagreements regarding study eligibility at this stage were resolved via consensus and/or through consultation with a third, independent reviewer.

2.6 Data extraction and charting

Data extraction was performed by two independent reviewers using a custom piloted form. Data extracted from each study included: (1) study characteristics (such as authors, title, date of publication, study design, and setting); (2) population characteristics (such as size, mean and/or median age, and sex distribution); (3) intervention and comparator characteristics (such as details about the EMIS); (4) timeliness of care outcome(s) (such as mean or median time to receipt of care or discharge); and (5) reported study limitations. A 20% sample of included studies was extracted in duplicate by both reviewers and compared to ensure data integrity. Following extraction form validation, the remaining study data were extracted by a single reviewer.

Study characteristics and key findings of included studies were reported. Where possible, reported outcomes were grouped by similarity and summarized using descriptive statistics.

3 Results

3.1 Search results

The database search identified a total of 9043 citations. After removal of duplicates, a total of 4386 citations were subject to title and abstract screening (Fig. 1). Of these, 114 studies underwent full-text review and were assessed for eligibility. Twenty-five studies reporting on the effects of an EMIS on the timeliness of patient care in the ED met the eligibility criteria.

Fig. 1
figure 1

PRISMA flow diagram

3.2 Excluded studies

A list of the 89 studies that were excluded following full-text review is provided in Supplementary Table 1. Overall, studies were excluded for not satisfying the following inclusion criteria: population (23), intervention (27), comparison (10), outcome (20), or study type (9).

3.3 Characteristics of included studies

Key characteristics of included studies are provided in Table 1. The majority of studies were single-centre studies, with six multicentre studies that recruited participants from EDs located in Canada, the United States, Germany, South Korea, and Japan [17,18,19,20,21,22]. Studies were published between 2015 and 2021. Participants included both males and females, with few restrictions according to age or severity of visit.

Table 1 Characteristics of included studies

3.4 Types of studies

Of the 25 studies included in this review, 13 were retrospective studies, seven were pre-post studies, three were randomized controlled trials (RCTs), one was a crossover study, and one was a time series trial.

3.5 Types of interventions and comparisons

Interventions reported by all studies were directly related to the implementation of an EMIS. They included evaluations of a new or updated EMIS [23,24,25,26,27,28], EMIS tools related to disease diagnosis and/or management [17, 19, 20, 22, 29,30,31,32,33,34,35,36], electronic triage tools [37,38,39], result availability alerts [18, 40], health information exchange systems [41], and computerized provider order entry tools [21]. Included studies compared interventions with relevant comparator groups, such as prior versions of an EMIS, absence of an EMIS, pen-and-paper documentation, and/or communication via telephone and fax.

3.6 Outcomes assessed

Reported time-based outcomes varied among included studies and were categorized into six distinct groups: (1) time to provider (n = 3 studies); (2) time to order (n = 4 studies); (3) time to medication (n = 6 studies); (4) time from result to disposition (n = 2 studies); (5) ED length of stay (n = 17 studies); and (6) other time-based outcomes (n = 3 studies) (Table 2). Ten studies reported on multiple time-based outcomes of interest.

Table 2 Effects of examined interventions on the timeliness of emergency department care

3.7 Time to provider

Three studies examined the effects of an EMIS on the time to care provider [21, 26, 37]. Two of the three studies suggest that the implementation of an EMIS increased the length of time that patients waited for a healthcare provider [21, 37]. An emergency severity index tool within the EMIS used during triage of sickle cell disease patients slightly increased both door-to-nurse and door-to-attending time by approximately one minute [37]. Similarly, the implementation of computerized provider order entry increased the wait time for first physician assessment to 83 min in comparison with 78 min with paper-based order entry [21]. In contrast, a single study comparing an EMIS to paper documentation reported a slightly shorter wait time to first physician assessment from 4.2 min pre-EMIS to three min post-EMIS [26].

3.8 Time to order

Four studies evaluated the effects of implementing an EMIS on time to order [17, 18, 24, 31]. These articles evaluated a variety of time intervals and generally found that a reduction of time to order may be attributed to the implementation of an EMIS compared with the absence of an EMIS. Koziatek et al. [18] reported an eight-minute decrease in median order time between pre-implementation and post implementation cohorts when comparing an EMIS with notifications to those without. Another study found that implementing an EMIS containing Wells score criteria decreased time to order a venous duplex ultrasound from 226 min pre-implementation to 165 min post-implementation [31]. Furthermore, time from ED arrival to X-ray order was considerably decreased from 36.5 min to four minutes when searchable chief complaint, best practice advisory, and order sheet features were added to an EMIS for children presenting to the ED with coin-shaped foreign body ingestion [24]. When evaluating time to hydrocortisone order, a decrease of 13.2 min was observed for the critical information note-present group compared with the critical information note-absent group [17]. Lastly, Koziatek et al. [18] reported no meaningful decrease in time from hemoglobin result availability to the administration of a blood transfusion following the implementation of notifications within an EMIS.

3.9 Time to medication

Six studies reported on outcomes related to the timely administration of medication in the ED [17, 30, 34,35,36,37]. Most concluded that a tool embedded within an EMIS decreased time to medication, with improvements ranging from 30.5 to 91 min. For instance, three studies implemented some version of a tool within an EMIS to aid in the early detection and management of sepsis [30, 34, 36]. Collectively, findings demonstrated improved time to antibiotics and time to a bolus of fluid, as well as a greater proportion of patients receiving antibiotics within the target time of 60 min [30, 34, 36]. Another study implemented electronic pediatric asthma guidelines in the EMIS and found that time to steroid administration was reduced from 196 min pre-intervention to 105 min post-intervention [35]. In contrast, two of the six studies did not report any notable improvements in timeliness of patient care [17, 37]. The incorporation of an outpatient critical information note within an EMIS minimally decreased time to the administration of hydrocortisone for pediatric adrenal insufficiency from 121.9 to 112.7 min [17]. Furthermore, the implementation of a triage tool actually led to a slight increase in time to analgesia, from 106.7 to 115.2 min, for sickle cell disease patients presenting to the ED in vaso-occlusive crisis [37].

3.10 Time from result to disposition

Two studies reported on the time from result to disposition [18, 40]. Both concluded that there was a decrease in median time from result availability to physician decision-making owing to the implementation of an alert system within an EMIS. One of these studies reported on four distinct results relating to this primary outcome [18]. When evaluating time from chest X-ray result availability to disposition, there was a 24-min decrease in median time among the push-notification cohort (56 min) compared with the no push-notification cohort (80 min). Similarly, the median time from various laboratory results to physician decision was shortened considerably following the addition of the push-notifications to the EMIS. For example, median time from basic metabolic panel, urinalysis, and respiratory pathogen panel results to disposition decreased by 12, 10, and 43 min, respectively. Moreover, Verma et al. [40] reported that the median time from last troponin result to discharge was decreased from 94.3 to 68.5 min when push notifications for all troponin results were added to an existing EMIS.

3.11 Length of stay

We found that ED length of stay (LOS) was the most commonly examined time-based outcome within included studies. Seventeen distinct studies reported on outcomes related to LOS and yielded highly variable results [17, 19,20,21,22,23, 25,26,27,28,29, 33, 37,38,39,40,41]. The majority of studies reported an increase in ED LOS following the implementation of an updated EMIS. Three of these studies explored the effects of a tool embedded within an EMIS and reported modest increases in LOS ranging from five to 18 min [17, 21, 29]. Two other studies involved the implementation of an emergency severity index triage tool and reported an overall increase in LOS of 12.1 min and 18 min, respectively [37, 39]. However, the latter study also reported decreased triage time and a greater number of high acuity patients seen within recommended timeframes [39]. Following the introduction of a new or updated EMIS, four studies reported an increase in ED LOS ranging from 1.8 min to 35 min [23, 25, 26, 28]. A post-hoc analysis performed by one of these studies showed that LOS decreased by four minutes for patients admitted to hospital [28].

Six studies reported a decrease in ED LOS, ranging from 6.7 to 54 min, following the implementation of an updated EMIS [27, 30, 34,35,36, 38]. One of these studies found that incorporating the Korean triage and acuity scale into an existing EMIS decreased LOS by 54 min [38]. Another study implemented a health information exchange system that permitted providers to access patient records from outside the ED. When compared to telephone and fax-based systems for acquiring records, LOS was decreased by 26.9 min [41]. A third study involving smartphone push-alert notifications for troponin results decreased both total ED LOS and time from triage to discharge decision by 17 min [40]. Similarly, when the HEART (history, electrocardiogram, age, risk factors, and initial troponin) pathway accelerated diagnostic protocol was integrated into an EMIS, LOS was decreased by 24 min alongside improvements in hospitalization rates, mortality, and myocardial infarction [20]. Finally, two studies implemented newly updated EMIS and compared their performance with an original EMIS and paper-based documentation [27, 28]. Findings from Inokuchi et al. [27] demonstrated that the new EMIS with advanced electronic features decreased LOS for all patients by 6.7 min. Additional time savings were observed when groups were stratified according to level of urgency. Tall et al. [28] reported that an EMIS decreased LOS by four minutes for patients who were ultimately admitted to hospital from the ED, when compared with paper-based documentation.

Three of the 17 studies examined ED LOS as either a secondary outcome or balancing measure and, despite positive effects observed for primary outcomes, did not report any difference in LOS following implementation of an electronic tool [19, 22, 33]. For instance, Atabaki et al. [33] and Kharbanda et al. [22] reported a reduction in computed tomography use and associated radiation exposure for pediatric head trauma and appendicitis, respectively, while Delahanty et al. [19] demonstrated that use of an electronic tool was considerably more specific and sensitive for the early detection of sepsis.

3.12 Other timeliness outcomes

Three studies reported on six distinct outcomes that are related to timeliness of care, including time to flagging for sepsis, time to electrocardiogram acquisition, triage interval, disposition-to-admit time, door-to-bed time, and provider-to-disposition time [26, 32, 39]. Lloyd et al. [32] reported that, by integrating an electronic sepsis screening tool within an EMIS, the time to flagging for sepsis resulted in sepsis being identified 68 min earlier compared with a paper-based screening tool. Lavin et al. reported a significant decrease in median time from ED arrival to completion of X-ray (from 59 to 41 min) after implementing an EMIS-based intervention targeting children with coin-shaped foreign body ingestion [24]. Another study found that implementing a stepwise triage tool in an EMIS reduced the triage interval from 5.9 to 2.8 min, and slightly increased time to ECG (electrocardiogram) acquisition from two minutes to 4.5 min [39]. Rupp et al. [26] explored the effects of converting from paper charts to an EMIS system on disposition-to-admit time, door-to-bed time, and provider-to-disposition time. Results were variable, with disposition-to-admit time slightly decreased by 7.2 min, door-to-bed time increased by 3.8 min, and provider-to-disposition time increased by 9.8 min.

4 Discussion

We completed a scoping review to identify recent peer-reviewed studies that described the implementation of an EMIS and examined its effects on timeliness of patient care in the ED. Our primary findings were that the nature of EMIS implemented, as well as the timeliness of patient care outcomes examined, varied considerably among the 25 studies included in this review. There were no clear trends on the impact of the EMIS, as reported findings varied for the same time-based outcome and across examined outcomes. Noteworthy secondary findings include: (1) ED LOS was most commonly examined as a timeliness of care outcome, with moderate changes observed in the overall LOS (ranging from 54 min decrease to 35 min increase); (2) a number of studies examined time to medication delivery, with most reporting decreases in overall time to event; and (3) other important markers of timeliness of care studied within ED settings included time to provider consult, time to intervention request (such as imaging, medication, or procedure), and time from test result to discharge.

The most frequently examined ED time-based outcome was LOS. Reported findings were highly variable among studies. The range of reported findings varied from a 54-min decrease to a 35-min increase in LOS [23, 38]. Additionally, a number of studies found no change in LOS, but reported improvements in other patient care outcomes such as a reduction in unnecessary radiation exposure [19, 22, 33]. These findings are consistent with those reported in systematic reviews that examined patient care efficiencies and outcomes associated with EMIS implementation [15, 42, 43]. Notable improvement in ED LOS of up to 54.5 min with reductions in unnecessary diagnostic tests and/or imaging have previously been reported [15, 42]. Inconsistencies observed within the literature may be attributed to EMRs and related tools being relatively new, with their implementation accelerating over the last decade [15]. As a result, their use in the ED has yet to be thoroughly explored. Nevertheless, improvements in ED LOS may be attributed to EMIS and related tools providing faster access to patient information, reducing time to diagnosis and management, aiding in the proper distribution of resources, and shortening documentation time [15, 27, 38, 41, 42]. Conversely, increased LOS may be associated with system complexities, steep learning curves, and poor system implementation [21, 23, 25, 26, 28, 29, 39].

Time to medication was the second most commonly reported time-based outcome that we observed within our review of the literature. With the exception of a single study, reported findings suggest that an EMIS substantially decreases time to medication administration, with improvements reaching up to 91 min [35, 37]. Interestingly, all studies reporting a shortened time to medication implemented an EMIS tool, such as a push-alert or management guideline, as an intervention within an existing EMIS [17, 30, 34,35,36]. Our findings are consistent with a prior review examining the effects of automated EMIS alerts on ED personnel response to sepsis [13]. Similarly, we observed evidence for improved process-of-care markers, including decreased time to antibiotic delivery [13]. Future studies are required to determine whether such findings apply to the administration of all ED medications and to determine whether additional time-based efficiencies in the delivery of ED medications may be achieved.

Other time-based outcomes identified from our literature review were diverse and included measures such as time to flagging for sepsis and disposition to admit time. Although findings were generally inconsistent across studies, some consistent timeliness of care benefits were observed for time to provider and time to order outcomes, with time savings ranging from eight to 61 min [26, 31]. These findings suggest that the delivery of essential urgent health services may be improved with the adoption of an EMIS. Future studies are necessary to determine whether observed timeliness benefits in the ED extend to all ED settings and patient health outcomes.

Inconsistencies in time-based indicators of patient flow and departmental performance across EMIS implementations make it challenging to generalize regarding the potential benefits of such systems and tools, which may in-turn adversely affect their respective uptake or perceived utility. As such, it is important to comprehensively understand factors that may account for the observed heterogeneity in reported time-based metrics associated with EMIS deployment. It is reasonable to presume that variations in observed effects are attributed to non-standardized EMIS implementation across hospitals, as well as differences in offered training and overall comfort of ED personnel with new health technology. Furthermore, time-varying hospital-level factors that are difficult to control for, such as the number of vacant beds for admissions from the ED and the availability of nurses and other staff, may partially explain the variability in reported findings to date. It is also possible that the lack of standardized time-based outcome definitions and guidelines regarding when and how best to measure the impact of an implemented EMIS render it difficult for hospitals to report similar findings across comparable outcomes, especially for shorter time-based metrics. Future studies should therefore focus on elucidating the underlying factors that may explain observed differences across EMIS deployments, and aim to standardize research parameters specific to EMIS studies.

Our study has a number of strengths. To our knowledge, our study is the first to complete a scoping review of the literature on the impact of the use of an EMIS on time-based outcomes in the ED. We employed a systematic approach in searching for and identifying eligible studies and did not restrict our findings by geography. As such, this work maps the emerging evidence from studies that have evaluated the impact of implementing electronic tools in ED settings on the timeliness of care. Findings from this review may therefore be used to inform future systematic reviews that seek to quantitatively synthesize the impact of these tools on a specific outcome.

A number of limitations should be considered when interpreting our findings. Since the adoption of an EMIS is relatively new within ED settings, we limited our search to articles published after 2014 in English. Therefore, relevant findings published prior to 2015 or those published in another language are not reflected in our review. Furthermore, “Electronic Medical Information System”, “Electronic Medical Record” and “Emergency Department” are broad terms for which there may be many synonyms. To account for this, our searches were developed in consultation with a research librarian and included select synonyms for these terms; though we may have unintentionally missed some relevant records. Notwithstanding these limitations, our study meaningfully adds to the paucity of information on the effects of an EMIS on the timeliness of ED care.

In conclusion, many ED settings have implemented an EMIS in effort to improve timeliness of care outcomes, or plan to do so in the future. The impact of these systems and tools has been evaluated by recent studies, with a particular focus on length of stay. Generally, reported findings are inconsistent and varied within and across time-based outcomes. Despite this, reported findings show promise in improving timeliness of care in the ED, which may ultimately be associated with improved health outcomes for patients. Future studies should explore barriers to the adoption of such systems and tools in the ED, examine the long-term effects of such technologies on important time-based indicators of quality care (such as revisit), and assess provider and patient satisfaction with these ever-changing health information systems.