Fast decisions, clear a qualitative study of trauma aid design

Authors: Lindsay Fountain, Kayla Corredera-Wells, Nicholas Cozzi, Jeffrey M Goodloe, Jenny M Guido, Alyssa Johnson, Christopher S Kang, Thomas McNally, Andrea Nevedal, James E Winslow, Gabriela Zavala Wong, Lacey LaGrone

Abstract

Objectives

Although trauma clinical guidance has the potential to improve patient outcomes significantly, it is applied inconsistently across the USA today. Work is needed to improve guidance application among those who treat traumatic injury, including emergency medicine (EM) clinicians. A priority step of such improvement is the visual design of clinical directives, as it can markedly influence EM clinicians’ satisfaction and subsequent usage of guidance. This qualitative study aims to analyze EM clinicians’ needs and preferences for the visual design of trauma clinical guidance.

Methods

As part of a larger exploratory sequential mixed-methods study exploring trauma clinical guidance barriers and facilitators, we conducted semistructured video conference interviews during August and September 2024 with EM clinicians. Participants were selected through purposeful criteria and snowball sampling to ensure diverse representation across professional and personal characteristics. Interviews were audio recorded and analyzed through rapid qualitative content analysis.

Results

12 EM clinicians completed interviews, providing feedback on their preferences for the visual design of trauma clinical guidance. Clinicians emphasized the need for simple guidance that can be easily used in a time-pressured setting. Dynamic features such as expandable or “clickable” content can provide deeper explanations of guidance concepts without compromising usability. Flowsheets and visual aids can clarify guidance recommendations. Clinicians asked that recommendations be specific and action-oriented.

Conclusion

When designing trauma clinical guidance, authors should promote visual design characteristics that prioritize simplicity, dynamic features, visual aids, and specificity.

Level of Evidence

Level III

Background

Contemporary clinical guidance improves patient outcomes, lowering morbidity, mortality, and patient harm and improving care quality.13 However, clinical guidance can be heterogenous and inconsistently applied throughout the spectrum of emergency departments (EDs) in the USA, with guidance adherence ranging from 40–60%.^4 5^ Emergency medicine (EM) clinicians face multifaceted barriers to using non-EM specialty clinical guidance, from difficulties with finding and accessing guidance to cumbersome-to-use materials. EM clinicians’ frustrations with clinical guidance extend to trauma care. They are often among the first to care for injured patients, particularly outside of Level I trauma centers. US trauma mortality rates rose 91% from 2000 to 2020, far outstripping the population growth rate of 17.8%, creating both urgency to reduce trauma fatalities and opportunity for EM clinicians to use guidance towards achieving better outcomes.^6^

Understanding EM clinicians’ needs from the visual design aspects of trauma guidance can enhance satisfaction with and usage of guidance. Although many attributes define clinical guidance, clarity of presentation has a high correlation with EM clinicians’ overall satisfaction with clinical guidance, emphasizing the need to provide clinicians with visually satisfactory materials.^7^ Recent research at an academic-affiliated, urban Level I trauma center into EM clinicians’ preferences for guidance design revealed a clear desire for concise guidance with clear visual cues leading users through its process.^8^ However, further research is needed into the preferences of EM clinicians working in different practice settings, as well as preferences specific to guidance concerning care for injured patients. This qualitative, interview-based study aims to analyze EM clinicians’ needs and preferences for the visual design of trauma clinical guidance.

Methods

Study design and sample

As part of a larger exploratory sequential mixed methods study exploring trauma clinical guidance barriers and facilitators, we conducted semistructured interviews to understand EM clinicians’ preferences for the visual design aspects of trauma clinical guidance. Purposeful criterion and snowball sampling were used to recruit US-based EM clinicians who were in active practice at the time of the interview. To diversify the sample, we aimed to include EM clinicians across gender, race, ethnicity, geographic region, trauma center level, clinician profession, years in practice, and level of urbanity.^9^ Geographic regions were analyzed using Centers for Disease Control and Prevention definitions.^10^ Clinicians were recruited via email outreach, leveraging the research team’s network of professional contacts and, as the research progressed, referrals from past participants. Although some participants were known to the project team at the time, none had preexisting relationships with the interviewer, and their identities were blinded to the project team as analysis began. 34 individuals were contacted, 22 declined, and 12 completed interviews. The concept of information power, which indicates that “the more information the sample holds, relevant for the actual study, the lower number of participants is needed,” was chosen to inform sample size needs throughout the study.^11^ Although saturation is commonly discussed in qualitative methods, it has been critiqued as a universal standard, is often misused, and is not appropriate for all qualitative studies. Saturation was not appropriate in this study that did not aim to generate theory. The criterion sample size of twelve was sufficient and yielded higher information power given the specific aims (to obtain feedback on specific trauma resources), dense sample specificity (practicing EM providers with expert knowledge), and strong dialog with participants (expertize in EM topic and qualitative methods and focused discussions).^12^ (Online supplemental file 1 provides additional methods description.)

The project research team included diverse perspectives from the trauma and healthcare research communities, including two trauma surgeons (JMGu, LL), four EM physicians (CSK, JMGo, JEW, NC), one research physician (GZW), one EM physician assistant (TM), one state trauma system manager (AJ), one medical anthropologist (AN), and two medical students (KC-W, LF). No compensation was given for participation in interviews.

Data collection

The interview guide was pilot tested among coauthors and two medical students not associated with the research team. Interviews were held and recorded via video conference during August and September 2024. The average interview duration was 51 min, with a range of 31–63 min. They were conducted by a medical student (LF) who received prior training in qualitative interviewing techniques from a medical anthropologist. The interviewer encouraged participants to provide candid responses and asked clarifying questions to ensure their perspectives were understood and documented.

In addition to open-ended questions about their preferences for guidance, clinicians were asked to comment on two examples of existing trauma clinical guidance (see online supplemental file 1 for complete interview guide). The two examples were chosen to include protocols routinely used in the care of injured patients and to provide a variety of visual styles for participant reaction. The protocols shared were the 2023 American Association for the Surgery of Trauma/American College of Surgeons Committee on Trauma Protocol for Damage Control Resuscitation of the Trauma Patient (figure 1) and the University of California San Francisco Adult Massive Transfusion Protocol (figure 2, version accessed August 12, 2024).^13 14^ To minimize any potential bias against specific authors, the protocols had any indicators of authorship removed before sharing with participants (see versions below). Example protocols were emailed to participants at least 48 hours before their interview with a request to review the materials and prepare to comment on their strengths, weaknesses, and compatibility with the participant’s current practice setting.

Data analysis

Interviews were audio recorded and reviewed by two analysts (KC-W, LF) using rapid directed qualitative content analysis, with methodological expertise provided by a medical anthropologist (AN).^15^ The interviewer and primary analyst (LF) took notes on a template created in Microsoft Office Word (V.2106, Build 14131.20332), with fields to capture responses to each question. The primary analyst then used these notes to draft an interview summary immediately after each interview. The secondary analyst (KC-W) listened to the audio recordings of each interview to verify the accuracy of and edit each interview summary. The primary analyst then copied the notes into a Microsoft Office Excel (V.2106, Build 14131.20332) matrix containing rows for each participant and columns for domains for each interview question. A Microsoft Word domain summary template was populated for each domain by reviewing the full matrix for relevant quotes. Each domain was analyzed and summarized by comparing and contrasting participant responses for salience, repetition, and negative cases to understand key suggestions for improving trauma visual design.

Results

12 clinicians completed semistructured interviews (table 1). Two clinicians regularly practiced in two different settings and shared insights on their perspectives from both locations. The following sections describe the wide range of feedback about guidance shared by clinicians, which is summarized into four suggestions for improvement in guidance design (table 2).

Design clinical guidance for quick reference and simplicity

Clinicians emphasized concerns about guidance being too time-consuming to read and apply, given the unique time constraints of EM. An exemplar quote to support this concern

Although thoroughness added value to the guidance, it is likely outweighed by the need for speed in the acute care setting (eg, “It’s a lot of information…That’s great, but in a crunch situation, will I have to read all of this?”—P7). Clinicians also reported that textually dense guidance created a substantial initial hurdle for using guidance. For example, clinicians were “turned off right away” (P1) by guidance that was multiple pages long. Clinicians remarked that if offered “dense, 10-page trauma guidelines” (P8), no one would use them.

When clinicians were asked what should be avoided when creating clinical guidance, nine clinicians mentioned verbally dense guidance. The high numbers of words “make people nervous” (P1) and paragraphs of text seemed to detract from usability. Bullets were the preferred format for delivering any verbal information. Paragraphs and even longer sentences created friction for guidance users (eg, “Sentences are hard to read. ER people, we think in bullets”—P6). Other participants described redundancy and repetitiveness as characteristics to avoid.

Use dynamic features to support in-depth clinical reasoning

Despite a general aversion to detailed, longer-form guidance, clinicians acknowledged that the style of guidance presentation can add value when used outside the context of point-of-care. Examples of alternative uses for denser guidance similar to figure 1 included “looking at it as a process map between shifts, to study and know it later” (P10). The concepts described in the guidance could be applied during a shift, but clinicians would need to be familiar with the guidance before the relevant patient encounter arose.

Preferable to guidance that needed to be reviewed ahead of time was the concept of one-page, streamlined guidance that contained more detail outside of the main “quick and dirty” portion of the guidance (P9). More references and explanations may be helpful for clinicians to review, but they should be off to the side or located on a second page.

Clinicians suggested that incorporating dynamic elements into guidance design could add depth and detail to guidance without cluttering the “need-to-know” information included in a point-of-care-focused piece of guidance. Clinicians desired guidance “to be flexible and dynamic, with tables and panels that can expand with more information. But it needs to stay on the same page. Once you have multiple windows open and have to go back, it can get messy. It should be like an accordion that opens and closes” (P10). Others requested “clickable links where bullets popped up with more resources” (P6) or capabilities in the electronic medical record to hover over parts of the guidance to elaborate on a specific point.

Leverage flowsheets and visual aids to clarify clinical recommendations

All clinicians described their preferred visual format for guidance to be a flowsheet like the ones featured in figures 1 and 2. One clinician explained that “trauma when done right is an algorithm” (P5), and the “if-then” nature of boxes and arrows captured that dynamic well. Clinicians reported that flowsheets were common in EM, and that familiarity allows clinicians to read and interpret guidance more quickly. Infographic-style formatting with diagrams may have a place for clinical procedure guidance, but flowsheets are more amenable for any guidance concerning decision making.

Though clinicians favored flowsheets overall, they cautioned that crowded guidance with limited visual cues regarding workflow still proved difficult to interpret in the acute setting. Flowcharts need to be “as simple as can be” and that “trying to tamp through 60 boxes will make people flustered” (P1). Complex flowcharts such as figure 1 were found to be “overwhelming and [it is] hard to know where to start” (P7). Another clinician knew where the starting point was but found the large number of branch points made it “hard to keep track of where I’m going next” (P10). Arrows helped clinicians understand the flow of each document, but only when moving in one direction and typically from left to right.

Clinicians found many of the style elements featured in figure 2 increased navigability. Several highlighted the use of color as a quick way to draw the eye in and to delineate each phase of the workflow (eg, “Green means go, you’re at your endpoint. Red is for the unstable patient. Yellow means in process.” —P7).

However, clinicians acknowledged that overreliance on color cues would disadvantage colorblind guidance users. Other, less color-centric suggestions for improving navigability included increasing white space between boxes of text and judicious use of capitalization and bold fonts.

Build guidance with specific, action-focused recommendations

Objectivity in patient inclusion criteria and decision-making points increased guidance value to clinicians, particularly for those with less experience. Examples included replacing subjective criteria such as “unstable pelvis” with an objective criterion such as radiographic findings that indicate pelvic instability or giving clear blood pressure ranges rather than stating “soft hemodynamics” (P3). Though overly strict criteria (eg, specific Glasgow Coma Scale scores) constrain clinician decision making, most found that providing ranges or multiple ways for patients to meet criteria strikes a balance between specificity and flexibility.

Distilling guidance to action-oriented components reduced friction for clinicians translating recommendations into practice. Clinicians identified patient population inclusion criteria, interventions, monitoring, and defined endpoints as essential components for complete guidance. Objectivity improved usability for each of these components, from definitions for hemorrhage control to specific metrics for monitoring, to deactivation criteria. Hospital-specific details for application of guidance during patient care also eased use for clinicians. Generic recommendations such as “call blood bank” frustrated clinicians seeking to act quickly (eg, “no one knows the number for the blood bank”—P4), and details such as including the current number for the blood bank on the guidance help speed action in patient care. Other measures, such as linking recommendations to order sets or providing specific dosing for medications, also decrease the time to implementation for interventions.

Discussion

Our study identifies methods for improving the visual design of trauma clinical guidance so that end users, including EM clinicians, can effectively use guidance to improve patient outcomes. Clinical guidance authors, as well as those adapting guidance to a specific hospital system or practice setting, should incorporate the following as they format their (1) design for quick reference, (2) use dynamic features, (3) leverage flowsheets and visual aids, and (4) build specific, action-oriented recommendations. Though little visual design research specific to trauma guidance in the EM setting exists to date, our findings are congruent with past findings from research into overall EM guidance needs.^8^ EM clinicians continue to advocate for clearer visual design and simplicity as means to increase usage in time-pressured practice settings. A newer finding from our work is the growing potential for dynamic, expanding features in guidance to provide optionality for clinicians seeking to deepen their clinical reasoning without sacrificing simplicity during point-of-care use. As electronic resources become more widespread, the capability grows for teams to incorporate these dynamic features into trauma clinical guidance.

The goal of this qualitative study was to obtain rich insights from EM clinicians about their needs and preferences for visual design. We will build off this work to develop a national survey to validate and further explore trauma clinical guidance needs, including but not limited to visual design. Beyond the survey, other future work could include user experience testing of guidance developed using the visual design recommendations distilled in this work. This testing would inform the development of adaptable templates that meet the needs of EM clinicians and can be easily employed by guidance authors, professional associations, and those integrating guidance into local systems.

Limitations

This work represents a focused, rapid project involving a limited number of EM clinicians within a larger mixed methods study. Future work could include a larger study focusing on a broader range of questions about trauma clinical guidance, or quantitative data collection to help prioritize recommendations for guidance authors. Participants were generally balanced across their professional and personal characteristics, but there was a disproportion of participants in the West geographic region relative to the rest of the USA and of physicians relative to other clinician professions. The snowball sampling approach may have resulted in selection bias that over-represents clinicians who are already interested in trauma guidance or who share similar professional networks to the study authors. No military-affiliated practice settings were included. Additional interviews with clinicians from other regions, professions, and practice settings could impact our understanding of the guidance needs of EM clinicians in practice.

Although this work identifies key principles to keep in mind for authors of future trauma clinical guidance, it does not provide specific templates or implementation strategies to help guidance authors. The study identifies desired visual aids and features; however, further usability evaluation is needed to validate the real-world impact of the proposed design characteristics.

Conclusion

In a qualitative study of EM clinicians, we identified four methods for improving the visual design of trauma clinical guidance. Given the importance of visual design in clinician satisfaction with guidance, implementing these changes in future versions of guidance may increase EM clinicians’ usage of and satisfaction with trauma clinical guidance. As many grapple with how to reduce the rising burden of trauma mortality in the USA, improvements in trauma clinical guidance should play an important role in those efforts.

Supplementary material