Many of us experienced significant challenges adapting to the COVID-19 pandemic. Some of us drove to vaccination locations only to find that appointments had not been fully transmitted through the computer system or that we misinterpreted eligibility requirements and had to return on another day. Some wasted diagnostic tests because of struggles with difficult packaging that led to spilling precious solutions. Doctors and nurses suffered from overwork due to increased patient volumes and the additional burden of infection control practices. At the same time, health systems, federal agencies, and product developers sought to develop and implement new triage, diagnostic, and treatment practices at an unprecedented pace. This posed critical challenges in engaging intended healthcare worker-users to participate in design and evaluation activities.

In a pandemic, time is of the essence. During the COVID-19 pandemic, the development and distribution of vaccines was accelerated from years to just months.^[1] The development of diagnostics and treatments similarly outpaced typical timelines.^[2][3] Ultimately, however, the outcomes of new vaccines, diagnostics, treatments, and prevention approaches rely on the ability of people—the general public, patients and families, clinicians, and public health workers—to use the provided resources effectively.

Emergency response depends on countermeasures that not only work as designed but also are used as expected. Applying human-centered design (HCD) practices can help enhance the usability and usefulness of interventions to strengthen pandemic preparedness and save lives.

Human-centered design: The big picture

Human factors engineering is the practice of applying knowledge about the capabilities and limitations of humans in the design and development of systems, technologies, or tools. Although this concept likely has been practiced throughout human history in some form, the term and formal discipline emerged during World War II in aviation, with the goal of preventing aircraft accidents.^[4] The practice has generated a significant body of design principles and guidelines that can be followed to ensure that products, technologies, and processes are easy for people to use.

For emerging technologies, there will always be some “art” in the design process. We often cannot predict how users will interpret and respond when using new systems in their own environments. Thus, formal HCD methodologies can be applied to ensure solutions are designed to be useful, comfortable, and easy to use.^[5][6] While details vary, all HCD approaches: 1) apply known design principles related to human capabilities and limitations, 2) implement formal approaches to understanding users and the environment, and 3) engage representative users early and throughout the design and development process to iteratively design, evaluate, and improve.

The costs of sub-optimal designs, such as ineffective uptake or the need for extensive training, often are not adequately weighed against the more immediately apparent resource needs required to undertake formal design and evaluation activities. Attention to how people will engage or interact with new systems may be delayed until late in the research and development process; however correcting a problem after a system is in development costs ten times as much as making the change during the design phase, and changes after release can cost as much as 100 times more than changes during design.^[7] By engaging users early and often in the design process, potential problems can be identified, and solutions implemented before it is too late.

What you really can’t skip when you are in a hurry

First, engage users early to be sure that the system offers useful functionality. Does the solution address a problem in a way that users find helpful? External funders may support exploratory activities with users to identify user needs or envision solutions to problems of interest.^[9][10][11] Developers and funders seek rapid solutions to attack urgent problems; however, they may develop solutions that are driven primarily by technological innovation. Too often the development effort is undertaken based on assumptions without sufficient early involvement with expected users. Particularly with novel technologies, initial impressions can have a long-lasting impact on future acceptance. For any new technology, it is important to confirm with the right people that they would use the solution if it was available and designed well; and to understand key features, minimum performance standards, and other, perhaps unexpected, concerns that must be addressed for efficacy and acceptance.

Second, some form of evaluation of the user experience or user interface design is almost always needed. The scope of the evaluation can vary depending on the complexity of the system, the expected context of use, prior experience with similar systems and user characteristics. For example, a modification to an existing website may simply require evaluation by a usability subject matter expert. Alternatively, the design of a medical device that is being newly introduced into a home environment for non-clinical users likely will require usability testing in the home, involving people with varying visual, literacy, and manual dexterity capabilities.

Finally, it is critical that the design and development processes include a plan for iterative improvement.^[12] Evaluation has little value if the developer is unable to improve the system to address problems.

What can we do now to prepare for the next pandemic?

‘Walk up and use’ refers to systems that can be used without any prior instruction or training. An automated teller machine is a classic example. In the context of a pandemic, when solutions such as mask wearing, at-home test kits, or new clinical decision support tools must be distributed to large numbers of people, we need more products that are ‘open and use.’ This approach requires a more comprehensive process of design and testing, for example, to design and print easy-to-understand icons and labels directly on the device versus supplementing with printed instructions or video.

Learning from our experiences with COVID-19, we are in a unique position to improve on the usability of medical devices, technologies, and information systems to support pandemic preparedness.^[13] For example, we can invest in developing at-home testing technologies that involve fewer components and user steps. When the number of user steps cannot be reduced, we can design solutions that make the process easier for the user. One simple example is the punch-out holes in the boxes included with some antigen tests that serve as a tube stand. Marks could be placed directly on nasal swabs to indicate appropriate depth of insertion. DRIVe supported the Mask Challenge to design effective masks that overcome user concerns related to comfort, communication, and maintenance.^[14][15] To address perceptions that people are being asked to wear masks when they are not needed,^[14] additional HCD activities could focus on better quantifying the value of masks in varied circumstances and designing effective communication relaying when masks are needed and why. The development of next generation of COVID-19 vaccines aiming for longer-lasting variant-proof protection presents another opportunity to embrace HCD practices early in development for better vaccine acceptance and administration.^[16]

We can promote the development of standards, guidance, or even plug-and-play user interfaces to ensure that common pandemic-related functions such as ordering tests, making appointments, evaluating eligibility for vaccines, and interpreting public health guidance are easy and clear to everyone. People place high value on personal empowerment when it comes personal health decisions and actions.^[17][18] We can design and evaluate information sharing approaches with a goal of positive impact on personal health empowerment. Through efforts like DRIVe’s Digital Health Tools for Pandemic Preparedness program and support for wearable technologies, we can work with partners to ensure that the development of public-facing applications and devices includes activities to better understand the context in which people will use these systems and to evaluate acceptance and usability before initiating larger-scale trials.

For clinicians, we can promote application of HCD methods in the development of medical devices, clinical information resources, and clinical decision support. For example, use of extracorporeal membrane oxygenation (ECMO) expanded during the COVID-19 pandemic and highlighted a need to simplify ECMO devices so that they can be used outside of specialized treatment centers. DRIVe recently announced a funding opportunity to support these needs in addition to other needs in the advancement of treatment for severe acute respiratory distress syndrome. With respect to supporting clinical information needs, COVID-19 highlighted the need for rapid compilation and evaluation of clinical research and real-world data. Now is the time to review and evaluate solutions like COVID-19 case data dashboards,^[19][20] COVID-19 evidence and treatment guideline dashboards,^[21][22] and COVID-19 data resources,^[23][24] to understand what features and design approaches enhance both use and usability and support preparation for future pandemics.

DRIVe launched a Rapidly Deployable Capabilities program at the beginning of the pandemic to rapidly deploy and clinically validate diagnostic and device technologies that were in late-stage development or already FDA cleared to empower patients and healthcare providers in healthcare decision making. Among other applications supported by this program, clinical decision support systems sought to aid in early identification of COVID-19, patient triage, and early prediction of patient deterioration, but developers faced challenges in executing these projects. There is an opportunity now to understand what worked and what did not with respect to implementation of clinical decision support tools, ^[25][26][27]and to determine whether some challenges can be overcome through greater consideration of human factors.

There are also many opportunities to apply HCD to reduce human-system process barriers to efficiency in areas such as laboratory testing and drug manufacturing and shipping processes. For example, laboratories seeking to maximize throughput of COVID-19 tests at the height of the pandemic engaged directly with lab personnel in designing a mix of automation and manual processes to speed the extraction of samples from packages into specialized equipment for analysis.^[28]

During a pandemic or other health security emergency, it can be tempting to forego steps perceived to be “nice to have” rather than “need to have.” Some may place formal HCD activities in the former category. The expectation that “people will adapt” is not necessarily wrong.^[29][30] People are creative and resilient and, as we have seen in dealing with COVID-19, will rise to the challenges they face. When we add up the negative impacts on time, well-being, and potential failures, however, the balance clearly favors the application of HCD principles and practices early and throughout the research, design, and development stages for better public health emergency preparedness.

[1] Ball P. (2021). The lightning-fast quest for COVID vaccines - and what it means for other diseases. Nature, 589(7840), 16–18.

[2] Servick, K., Couzin-Frenkel, J., & Matacic, C. (2021, March 16). Medicine’s longest year. Science. https://www.science.org/content/article/how-do-you-treat-coronavirus-here-are-physicians-best-strategies

[3] Aguiar, E., Navas, J., & Pacheco, L. (2020). The COVID-19 diagnostic technology landscape: Efficient data sharing drives diagnostic development. Frontiers in Public Health, 8, 309.

[4] Adams, A. (2006). A layman’s introduction to human factors in aircraft accident and incident investigation. Australian Transport Safety Bureau. https://www.atsb.gov.au/media/32882/b20060094.pdf

[5] Neilsen, J. (1994). Usability engineering. Morgan Kauffman.

[6] Holtzblatt, K., Wendell, J.B., & Wood, S. (2005). Rapid contextual design. Morgan Kauffman.

[7] Gilb, T. (1988). Principles of software engineering management. Pearson Education Limited.

[8] Brown University School of Public Health. (2022, May 13). New analysis shows vaccines could have prevented 318,000 deaths. https://globalepidemics.org/2022/05/13/new-analysis-shows-vaccines-could-have-prevented-318000-deaths/

[9] Wright, M. C., Dunbar, S., Macpherson, B. C., Moretti, E. W., Del Fiol, G., Bolte, J., Taekman, J. M., & Segall, N. (2016). Toward designing information display to support critical care. A qualitative contextual evaluation and visioning effort. Applied Clinical Informatics, 7(4), 912–929.

[10] Barwise, A., Caples, S., Jensen, J., Pickering, B., & Herasevich, V. (2017). Information needs for the rapid response team electronic clinical tool. BMC Medical Informatics and Decision Making, 17(1), 142.

[11] Galliford, N., Yin, K., Blandford, A., Jung, J., & Lau, A. (2022). Patient work personas of type 2 diabetes–A data-driven approach to persona development and validation. Frontiers in Digital Health, 4, 838651.

[12] Wright, M. C., Borbolla, D., Waller, R. G., Del Fiol, G., Reese, T., Nesbitt, P., & Segall, N. (2019). Critical care information display approaches and design frameworks: A systematic review and meta-analysis. Journal of Biomedical Informatics, 100S, 100041.

[13] Farmer, S., Razin, V., Peagler, A. F., Strickler, S., Fain, W. B., Damhorst, G. L., Kempker, R. R., Pollock, N. R., Brand, O., Seitter, B., Heilman, S. S., Nehl, E. J., Levy, J. M., Gottfried, D. S., Martin, G. S., Greenleaf, M., Ku, D. N., Waggoner, J. J., Iffrig, E., Mannino, R. G., … Lam, W. A. (2022). Don't forget about human factors: Lessons learned from COVID-19 point-of-care testing. Cell Reports Methods, 2(5), 100222.

[14] Esmaeilzadeh P. (2022). Public concerns and burdens associated with face mask-wearing: Lessons learned from the COVID-19 pandemic. Progress in Disaster Science, 13, 100215.

[15] Cohen, J. (2022, May 18). Hate your face mask? There’s hope. Science. https://www.science.org/content/article/hate-your-face-mask-there-s-hope.

[16] Herper, M. & Facher, L. (2022, July 25). White House to launch effort to develop next generation of Covid vaccines. STAT. https://www.statnews.com/2022/07/25/white-house-to-launch-effort-to-develop-next-generation-of-covid-vaccines/

[17] Salgado, T., Tavares, J., & Oliveira, T. (2020). Drivers of mobile health acceptance and use from the patient perspective: Survey study and quantitative model development. JMIR mHealth and uHealth, 8(7), e17588.

[18] Schimmer, R., Orre, C., Öberg, U., Danielsson, K., & Hörnsten, Å. (2019). Digital person-centered self-management support for people with type 2 diabetes: Qualitative study exploring design challenges. JMIR Diabetes, 4(3), e10702.

[19] Centers for Disease Control and Prevention. (2022, August 4). COVID data tracker. https://covid.cdc.gov/covid-data-tracker

[20] Johns Hopkins University. (2022, August 5). COVID-19 dashboard. https://coronavirus.jhu.edu/map.html

[21] Open Critical Care. (2022, June 1). COVID-19 guidelines dashboard. https://opencriticalcare.org/covid-dashboard/

[22] COVID-19 recommendations and gateway to contextualization. (2022). COVID19 Recommendations. https://covid19.recmap.org/

[23] National Center for Advancing Translational Sciences. (2022, July 29). About the National COVID Cohort Collaborative. Accessed August 2, 2022. https://ncats.nih.gov/n3c/about

[24] The University of Texas Health Science Center at Houston & University of California at San Diego Health. (n.d.). COVID-19 data index. Accessed August 2, 2022. https://www.covid19dataindex.org/

[25] Dugdale, C. M., Rubins, D. M., Lee, H., McCluskey, S. M., Ryan, E. T., Kotton, C. N., Hurtado, R. M., Ciaranello, A. L., Barshak, M. B., McEvoy, D. S., Nelson, S. B., Basgoz, N., Lazarus, J. E., Ivers, L. C., Reedy, J. L., Hysell, K. M., Lemieux, J. E., Heller, H. M., Dutta, S., Albin, J. S., … Shenoy, E. S. (2021). Coronavirus disease 2019 (COVID-19) diagnostic clinical decision support: A pre-post implementation study of CORAL (COvid Risk cALculator). Clinical Infectious Disease, 73(12), 2248–2256.

[26] Shah, S., Switzer, S., Shippee, N. D., Wogensen, P., Kosednar, K., Jones, E., Pestka, D. L., Badlani, S., Butler, M., Wagner, B., White, K., Rhein, J., Benson, B., Reding, M., Usher, M., Melton, G. B., & Tignanelli, C. J. (2021). Implementation of an anticoagulation practice guideline for COVID-19 via a clinical decision support system in a large academic health system and its evaluation: Observational study. JMIR Medical Informatics, 9(11), e30743.

[27] Stark, N., Kerrissey, M., Grade, M., Berrean, B., & Peabody, C. (2020). Streamlining care in crisis: Rapid creation and implementation of a digital support tool for COVID-19. The Western Journal of Emergency Medicine, 21(5), 1095–1101.

[28] Sandoval, E, Nguyen, J., and Ramirez, J. Helix OpCo, LLC personal communication. June 15, 2022.

[29] Klein, G. (1944). Sources of power. (1999 edition). Massachusetts Institute of Technology Press.

[30] Boonstra, A., Jonker, T. L., van Offenbeek, M., & Vos, J. (2021). Persisting workarounds in Electronic Health Record System use: Types, risks and benefits. BMC Medical Informatics and Decision Making, 21(1), 183.