Original Research
A Usability Evaluation of a Touchscreen Workstation on Wheels in a Simulated Emergency Department Workflow

Saad Razzaq ¹, Omar Idrissi ¹, Jake Rose ¹, Santiago Marquez ², Jose Correa ³ Antony Robert ⁴
Published online: June 2025

¹Faculty of Medicine & Health Sciences, McGill University, Montréal, Québec, Canada
²Division of Centre of Outcomes Research and Evaluation (CORE), Research Institute of the MUHC, Montréal, Québec, Canada
³Department of Mathematics and Statistics, McGill University, Montréal, Québec, Canada
⁴Department of Emergency Medicine, McGill University Health Centre, Montréal, Québec, Canada

Corresponding Author: Saad Razzaq. Email Address: saad.razzaq@mail.mcgill.ca

DOI: 10.26443/mjm.v22i1.1150

Introduction

Emergency Departments (EDs) serve as crucial entry points to hospitals, providing immediate care to patients with diverse medical needs. However, the high patient volume often leads to overcrowding, resulting in delayed care and emergency access blocks, the number one safety concern in first-world EDs (1–3). Timely care in the ED is essential for improving patient outcomes and reducing mortality rates (4–8). A recent study conducted in 2019 that analyzed ED data from 25 Canadian hospitals found that improvements that can increasing tasks performed at the bedside and reducing tasks at the stationary workstation have been observed to reduce workflow interruptions, decrease errors of omission, and increase time with patients and their families, which improved patient satisfaction and thus optimized provider service (9). Though technological innovations and information technology (IT) solutions have the potential to improve healthcare delivery, limited research exists on the specific mechanisms by which these solutions impact ED operations.

Touchscreens have become ubiquitous in our daily lives, offering a comfortable and natural human-technology interactive experience (10,11). A prospective pilot study conducted in a large level 1 ED introduced touchscreen tablet computers at the bedside, reducing physicians' time away from patients by 38 minutes per shift (12). Physicians reported positive perceptions of these devices, considering them clinically useful, efficient, easily portable, easy to disinfect, and improved overall patient care. While portable tablets show promise, their adoption can be limited regarding universal accessibility, operational capacity, security, fragility, interface quality, and compatibility with EMRs and associated software. Furthermore, the specific aspects of these interventions, such as portability or touchscreen functionality, that contribute to improved workflow efficiency require further investigation.

Our study aims to assess workflow efficiency and perceived usability of a touchscreen interface compared to a mouse-based interface on a standard workstation on wheel (WOW) with common tasks performed on two existing electronic medical record (EMR) systems in the emergency department (ED).

Methods

Study Design

We performed a randomized, controlled, single-center, 2-intervention-2-period crossover study. The study subjects included a convenience sample of 12 resident physicians and 12 staff emergency physicians of the McGill University Health Center (MUHC), a large tertiary and quaternary academic teaching center in Montreal, Canada, who regularly use the existing MedUrge and Oacis EMRs for clinical care.

The study involved the completion of the following EMR tasks in tandem with a ten-minute break in between interventions:

1. Open a triage note and identify blood pressure via MedUrge
2. Open most recent patient chart via Oacis
3. Order CBC and CHEM7 bloodwork via Oacis
4. Order chest x-ray imaging via Oacis
5. Order acetaminophen medication via MedUrge
6. Request a cardiology consult via MedUrge
7. Discharge patient home via MedUrge

In keeping with a crossover design, all participants were evaluated for the completion of the same tasks using each of the two interventions. One intervention (A) involved completing the tasks with a non-touchscreen WOW (non-touchscreen monitor, keyboard, mouse) whereas the other intervention (B) involved completing the tasks with a touchscreen WOW (touchscreen monitor, keyboard, no mouse). Twelve subjects were randomly selected to perform the tasks in intervention A and then the tasks in intervention B (sequence AB), while the other twelve subjects completed the tasks in the opposite order (sequence BA). No washout period between interventions was introduced given that these are commonly performed EMR tasks and we assumed minimal to no carry-over effect (i.e., that the effect on time to perform the tasks of the intervention used in the first period did not alter the effect on time to perform the tasks of the intervention used in the second period).

User Interface

While both the touchscreen-based and non-touchscreen-based WOWs use the same EMRs user interface, differences exist in their interactions. The EMRs were designed more than a decade ago with the mouse and keyboard in mind, and thus, did not have user interface design considerations for touchscreens. Interaction difference between the mouse action versus touch actions include: 1) dragging of mouse vs movement of finger to find interface of interest, 2) mouse double left-click vs double touch to confirm specific selected options, 3) mouse right-click vs touch and hold to open patient care menus. Other differences may exist between interactions and the EMRs user interfaces but were not observed nor tested in the tasks assigned to participants. It is important to note that in this simulated environment, we did not have the opportunity to modify the design of EMRs user interface.

Data Collection and Processing

An asynchronous observation study design was used to collect data during testing; asynchronous in that the interventions were video recorded and analyzed at a later point in time. Research assistants were present before and after the interventions to collect pre- and post-intervention surveys for participant demographic characterization and to assess perceptions of physician users, respectively. Pre- and post-intervention surveys were electronically collected using the Google Forms. To ensure survey completion, the survey responses were confidential although not anonymous.

Outcome Measures

The primary outcome of this study was the combined time to complete all tasks sequentially for each of the interventions. Secondary outcomes include the time to complete each individual task as well as all participant’s perceptions of touchscreen utility, efficiency, portability, reliability, capacity to improve care, satisfaction, and ease of use.

Primary Data Analysis

Descriptive statistics are reported as counts and percentages for categorical variables. For continuous variables, we reported means and standard deviation (SD) if there was evidence that the distribution of values followed a normal distribution, and median and interquartile range otherwise.

Intervention effect was investigated by comparing the average time to complete each task and the average total time to complete all tasks with the touchscreen and the non-touchscreen WOWs. Data was analyzed using standard methods for a 2-treatment, 2-period cross-over design for continuous data (15,16). For each comparison, we used a mixed linear regression model with the type of intervention as the main independent variable, adjusting for period of measurement and the sequence of intervention, as fixed effects, and subject level variability as a random effect, to consider that subjects are observed under both types of intervention. We did not account for carry over effect as it was assumed to be negligible in our study design. Furthermore, the analysis adjusted for age, sex, level of training, experience with computers, touchscreen devices, touchscreen computers, Medurge and Oacis.

On each model, assumptions on the model errors (randomness, normality, and homogeneity of variances) and the presence of possible influential observations or outliers were assessed with diagnostic plots of the model residuals. Robust standard errors (SE) were used to adjust for violations of the homogeneity of variances assumption, if applicable. Statistical tests of hypothesis were two-sided and with a 5% level of significance. Results of the linear models are reported as estimated adjusted means and SE, as well as differences in adjusted means between interventions with 95% confidence interval (CI). (13) All analyses were conducted in R version 4 (R Core Team 2020).

We have estimated that 24 participants would be sufficient to detect a difference in the mean total time to complete both interventions in tandem by 30 seconds between the touchscreen WOW and the non-touchscreen WOW workflow with 80% power and a significant level of 5%. Considering there are no previous studies comparing similar workflows in the recent literature, we decided to use a pre-specific effect size of 0.8, which amounts to assuming a within-subject standard deviation for the total time to complete both interventions in tandem of 35 seconds. (14). We have computed the sample size using a standard formula for 2-treatment, 2-period cross-over trials. Post assessment, a technology acceptance model (TAMS) post-study survey (15) was administered to assess the subjects’ perception of touchscreen utility, efficiency, portability, reliability, capacity to improve care, satisfaction, and ease of use (Table 1). Ad-hoc analysis assessing correlations between two variables was done using Spearman’s correlation coefficients.

Results

Demographics of Study Participants

A total of twenty-four practicing physicians from the MUHC volunteered to participate in this study (Table 2). Post-study questionnaire data was excluded for one participant due to insufficient time to complete the questionnaire.

Primary Results

Results from the mixed model regression analysis showed that, after adjusting for period, sequence of intervention, and the covariates of interest, there was no significant evidence to reject the hypothesis that the mean total times to complete all tasks in tandem were the same between the two interventions (p=0.08, Table 3). The adjusted total time to complete intervention means were 89.7 (SE 10.6) and 96.5 (SE 10.6) seconds. For the mouse WOW and touchscreen WOW, respectively, with a two-sided 95% CI of (-0.9, 14.5). Similar results were obtained for each task separately (Figure 1).

Results from the post-study questionnaire assessed along a 7-point Likert scale (Figure 1) demonstrated that the majority (>50%) of the participants agreed on all questions regarding attitude towards utilization (AU), behavioral intention to use (BU), perceived ease of use (PEOU), and perceived usefulness (PU) of the touchscreen WOW, except question PEOU5 and PU3. The Spearman’s correlation coefficient between experience using touchscreen computers and ease of use of a touchscreen WOW was found to be 0.6 (95% CI (0.2, 0.8)). The correlation between experience using Oacis and perceived usefulness of the touchscreen with WOW was -0.4 (95%CI (-0.7, -0.01)).

Discussion

To our knowledge, this is the first study to evaluate the impact of a touchscreen interface on workflow efficiency and perceived usability using a WOW and existing EMRs designed for non-touchscreens, mouse and keyboard interactions. Our study showed that in an environment simulating 7 commonly performed EMR tasks in the ED, there was no evidence to suggest a difference between the non-touchscreen and touchscreen WOW in terms of time to complete each task or all the tasks combined, while demonstrating an overall positive perceived benefit. On average, participants completed all the tasks (combined) faster the second time they performed them regardless of which apparatus was tested first. This is likely explained by participants’ ability to become accustomed to the apparatus and perform tasks more efficiently in the subsequent trial. Indeed, despite having an average of approximately 5 years of experience using the EMRs through the standard, non-touchscreen WOW apparatus, the workflow efficiency of participants, ranging from year 1-5 resident and attending physicians, was not significantly impacted by switching from the standard to the touchscreen WOW. This suggests that if a touchscreen WOW is implemented in the ED, there likely will not be a meaningful obstacle of adjusting to the touchscreen. This can also be influenced by the carry-over effect, where exposure to the tasks during the first trial may facilitate the subsequent trial. Despite this effect, participants demonstrated consistent task completion rates across both methods by their second exposure. This study builds on previous work on touchscreen devices by specifically evaluating the usability and efficiency of touchscreen WOWs in a controlled, simulation-based setting, differentiating from prior studies on tablets at the bedside.

Our study, which showed no isolated difference in the average time to complete the combined tasks, suggests that more tasks could be performed on a touchscreen WOW at the bedside, without losing the efficiency of standard mouse and keyboard workstations. Therefore, a touchscreen interface is feasible and does not hinder the workflow or usability. Currently, the MUHC sites, presumably along with other sites healthcare institutions, have touchscreen devices being used with EMR systems designed for traditional input (keyboard and mouse). Indeed, touchscreen WOWs do not degrade or negatively impact physician performance, even though the EMR systems used were originally designed for use with a keyboard and mouse rather than a touchscreen.

Given that there were no differences in time to complete commonly performed ED tasks between the touchscreen and non-touchscreen WOW groups, further studies need to evaluate if touchscreen WOW may be an appropriate addition to facilitate ED workflow. In addition, allowing the users the option of using mouse, keyboard, and touchscreen could cater to individual preferences of which modality enables efficiency in the right context. Future studies could evaluate the impact on ED workflow of using all three modalities compared to the standard mouse and keyboard interactions. Perhaps customized interactions involving all three modalities may yield faster time to complete certain tasks in a simulated environment, or improve clinical efficiency.

Furthermore, our survey results demonstrated an overall perceived user benefit associated with the use of touchscreen WOWs based on the positive attitude towards utilization and intention to use the touchscreen WOW of resident and attending physicians. Interestingly, while the majority (>50%) of the participants agreed on the total perceived benefit of the touchscreen WOW, they did not find it is easier to use than the non-touchscreen WOW and did not feel that the touchscreen WOW would minimize errors. For all participants, this was their first interaction with the touchscreen WOW on EMRs they have been using with mouse and keyboard interactions for as long as 102 months (median MedUrge use = 36 months, median Oacis use = 50 months). Such an interaction, under a controlled, “testing” environment, can make participants more prone to errors.

The findings from the study were likely not influenced by participants’ experience with touchscreen computers, as there was a balanced distribution of experience amongst participants. Though perhaps among cohorts with more experience with touchscreen computers, participants may complete tasks faster using a touchscreen than a non-touchscreen WOW. Furthermore, ad-hoc results suggest that users with more touchscreen experience are more likely to find the touchscreen addition to the standard WOW easier to use than those with less experience, evidenced by Spearman’s correlation coefficient. We also found that users with more experience on a software (Oasis) traditionally designed for a non-touchscreen apparatus will be less likely to find the touchscreen addition useful. The correlation coefficient value (<\0.60) suggests that there may be room for customization of the software to adapt to a touchscreen WOW, which perhaps over time, will enable users to perceive that the touchscreen WOW is indeed useful.

Our results add to the existing knowledge on touchscreen modalities in clinical practice. Results by Horng et al. (12) showed that the use of touchscreen tablet computers for direct patient care reduced physicians' time spent at the workstation by 38 minutes per shift, thereby potential increasing time by the bedside. However, their inferior operating capacity compared to computers limit their generalizability across powerful software such as EMRs, in addition to the EMRs lack of touchscreen design considerations (16). Further studies could investigate the workflow efficiency impact of tablet computers vs touchscreen WOWs in busy clinical environments such as the ED.

The existing EMRs user interface, developed more than a decade ago leveraged the design considerations that may seem intuitive with a mouse, but are not equally intuitive when using a touchscreen. Key design considerations include the size of clickable boxes, and radial buttons, both of which require larger target areas to facilitate touch screen interactions. Another key design issue is the use of the mouse left double click, and the mouse right click, which require a non-intuitive learning curve of using the double touch tap and the touch and hold, respectively to achieve the required interactions. These innate differences in user interface interactions of the touchscreen and mouse with the EMR (described in the methods section) likely contributed to outcomes from the study including participants’ perceptions of ease of use and capacity to minimize errors of the touchscreen WOW. These specific differences in the functions between the touchscreen and mouse WOW were not tested and warrant further study to optimize user interactions with EMRs. In the current study, we were unable to change EMR user interfaces which were built for a mouse dominant, non-touchscreen WOW. However, the touchscreen interface interaction was found to be objectively similar to the standard non-touchscreen interface regarding workflow efficiency showing no significant differences in average time to complete tasks commonly done in the ED while demonstrating an overall positive perceived benefit. We recommend further studies to assess the impact of touchscreen WOW models on workflow efficiency and healthcare provider perceived benefit with modern EMRs who have interfaces that are designed to be touchscreen friendly.

We acknowledge some limitations of our study. The majority of the participants in the study were male, right-hand dominant, and had an average of 5 years of experience with Medurge and Oacis EMRs, likely making it relatively more challenging for experienced users accustomed to the use of the standard non-touchscreen WOW to operate a touchscreen WOW. This study was done in a simulated environment and does not reflect real clinical environment user experience. Although the testing environment used identical interfaces with different input methods (touchscreen versus non-touchscreen), this study uniquely isolates the impact of the touchscreen interface as implemented with existing EMR designs. While we did not track errors made by participants using each intervention, we encourage others to do so if essential.

Conclusion

In this simulated ED environment, the use of a touchscreen WOW demonstrated adequate functionality. We were unable to find a significant difference between the non-touchscreen and touchscreen WOW in terms of time to complete tasks, while showing an overall positive perceived benefit from resident and attending physicians in the ED. A touchscreen WOW may potentially replace a non-touchscreen WOW in the ED, but this requires more research with modern EMRs that have incorporated touchscreen user interface design considerations.

References

1. Kellermann AL. Crisis in the emergency department. N Engl J Med. 2006 Sep 28;355(13):1300–3. DOI: 10.1056/NEJMp068194
2. Hoot NR, Aronsky D. Systematic review of emergency department crowding: causes, effects, and solutions. Ann Emerg Med. 2008 Aug;52(2):126–36. DOI: 10.1016/j.annemergmed.2008.03.014
3. Fatovich DM, Nagree Y, Sprivulis P. Access block causes emergency department overcrowding and ambulance diversion in Perth, Western Australia. Emerg Med J. 2005 May;22(5):351–4. DOI: 10.1136/emj.2004.018002
4. Geelhoed GC, Klerk NH. Emergency department overcrowding, mortality and the 4‐hour rule in Western Australia. Med J Aust. 2012 Mar;196(4):1–1. DOI: 10.5694/mja11.11159
5. Sprivulis PC, Da Silva J-A, Jacobs IG, Frazer ARL, Jelinek GA. The association between hospital overcrowding and mortality among patients admitted via Western Australian emergency departments. Med J Aust. 2006 Mar 6;184(5):208–12. DOI: 10.5694/j.1326-5377.2006.tb00202.x
6. Richardson DB. Increase in patient mortality at 10 days associated with emergency department overcrowding. Med J Aust. 2006 Mar 6;184(5):213–6. DOI: 10.5694/j.1326-5377.2006.tb00203.x
7. Guttmann A, Schull MJ, Vermeulen MJ, Stukel TA. Association between waiting times and short term mortality and hospital admission after departure from emergency department: population based cohort study from Ontario, Canada. BMJ. 2011 Jun 1;342:d2983. DOI: 10.1136/bmj.d2983
8. Viccellio A, Santora C, Singer AJ, Thode HC Jr, Henry MC. The association between transfer of emergency department boarders to inpatient hallways and mortality: a 4-year experience. Ann Emerg Med. 2009 Oct;54(4):487–91. DOI: 10.1016/j.annemergmed.2009.03.006
9. Innes GD, Sivilotti MLA, Ovens H, McLelland K, Dukelow A, Kwok E, et al. Emergency overcrowding and access block: A smaller problem than we think. CJEM. 2019 Mar;21(2):177–85. DOI: 10.1017/cem.2018.433
10. Wu X, Xi T. Study on Design Principle of Touch Screen with an Example of Chinese-Pinyin 10 Key Input Method in iPhone. In: Advances in Ergonomics in Design. Springer International Publishing; 2016. p. 639–50. DOI: 10.1007/978-3-319-41983-1_70
11. Travis C, Murano P. A comparative study of the usability of touch-based and mouse-based interaction. Int J Pervasive Comput Commun. 2014 Apr;10(1):115–34. DOI: 10.1108/IJPCC-01-2014-0012
12. Horng S, Goss FR, Chen RS, Nathanson LA. Prospective pilot study of a tablet computer in an Emergency Department. Int J Med Inform. 2012 May;81(5):314–9. DOI: 10.1016/j.ijmedinf.2012.01.002
13. Gates S, Ealing E. Reporting and interpretation of results from clinical trials that did not claim a treatment difference: survey of four general medical journals. BMJ Open. 2019 Sep;9(9):e024785. DOI: 10.1136/bmjopen-2018-024785
14. Cohen J. Statistical Power Analysis for the Behavioral Sciences [Internet]. 2013. Available from: http://dx.doi.org/10.4324/9780203771587
15. Adams DA, Nelson RR, Todd PA. Perceived Usefulness, Ease of Use, and Usage of Information Technology: A Replication. Miss Q. 1992;16(2):227–47. DOI: 10.2307/249577
16. Orphanides AK, Nam CS. Touchscreen interfaces in context: A systematic review of research into touchscreens across settings, populations, and implementations. Appl Ergon. 2017 May;61:116–43. DOI: 10.1016/j.apergo.2017.01.014

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.