User-centred participatory design of visual cues for isolation precautions

The human factors-informed, user-centred, participatory design process employed in this project was composed of four stages, each resulting in an output that informed the subsequent activities throughout the iterative design lifecycle shown in Fig. 1 and described in the following sections [16]. The participatory process was facilitated by a psychologist with training in human factors engineering and extensive experience in the field of infection prevention and control (LC). An infection control nurse (MM) and two infectious diseases physicians (AW, HS) provided expert input throughout the process. These four individuals composed the design team.

This project was carried out at the University Hospital of Zurich (USZ), a 900-bed, university-affiliated tertiary care hospital in 2014–2016. It was initiated in response to low reported adherence to isolation precautions and numerous critical incident reports related to isolation precautions. The project aimed to introduce signage for universal use throughout the hospital. Successful implementation of the new signage would require its use by several professional categories of HCPs (e.g. nurses, physicians, allied-care professionals, assistants) and hospital staff (e.g. housekeeping) working in several hospital settings (e.g. intensive care, emergency, general wards). Participants for each phase of the design process were thus purposefully sampled by professional category and by work setting in order to include a representative sample of potential users of the system (Fig. 1).

Two primary methods were used to establish design requirements that would inform the subsequent signage design: 1) interviews with front-end users and 2) direct observations. We conducted semi-structured interviews with anticipated typical users of the isolation signage system and explored how isolation precautions are generally handled (Could you walk me through the process of how new isolated patients are handled?), how isolation precautions are communicated in their unit (How do you know what isolation status a certain patient has / which precautions to use?), and any suggestions they had to improve the current system (How do you feel about the current isolation precaution signs? Would you have any recommendations to improve?). Interviews were documented with detailed notes. Direct, unstructured observations were conducted in a representative sample of hospital settings (intensive care, emergency, general ward) with the purpose of observing multiple types of users interacting with the existing signage. Observations were documented through field notes and photos.

Following initial observations and interviews, personas representing distinct user groups were established that collectively describe the potential user population [17]. The design team established personas based on user patterns identified during interviews and observations. These personas were used by the design team and workshop participants to keep all potential users of the signage in mind and make sure that design alternatives were meeting the goals and needs of specific users. Further, a list of established design requirements was established as an evolving document that was refined throughout the design lifecycle with ongoing data collection. Throughout this process, both functional and non-functional design requirements were identified. Functional requirements described what the system should do in order to serve its purpose, whereas non-functional requirements concerned the physical, social, environmental, and technical constraints placed on the system design and its development [16].

We conducted design thinking workshops with frontline HCPs based on an adapted methodology from the Stanford University Institute of Design including modes of empathising, defining, ideating, and prototyping [18]. Participants, grouped into pairs, were first prompted to interview their partners about the current challenges surrounding communication of isolation precautions (empathising) and then to synthesise information and insights from the interview into “problem statements” that described meaningful user challenges (defining). Based on these problem statements, participants then completed several ideation activities to generate new ideas and solutions – exchanging ideas and sharing feedback with their partner after each step (ideating). As a final step, participants used provided materials to create a prototype of their solution for communicating isolation precautions (prototyping). Workshop activities were documented with photos, detailed notes, and collected worksheets. Workshop documentation was then included in a qualitative analysis whereby all data was reviewed and grouped into thematic categories.

Building on the established list of design requirements, design ideas, and initial prototypes generated during the previous design phases, a process of iterative prototyping and user-testing began. Low-fidelity prototypes were produced using paper and pencil sketches, while higher-fidelity prototypes were produced using graphics software [19‐21]. This process consisted of multiple cycles of user evaluation with continuous feedback from HCPs to make modifications to the prototype, which was then re-evaluated.

During the prototyping phase, symbols representing each isolation category (contact, droplet, and isolation), were also developed to be included in the overall isolation signage system. Multiple variants of each symbol were evaluated through symbol judgement tests and comprehensibility tests using paper-based surveys according to the methodology proposed by ISO 9186 [22]. During symbol judgement tests, participants were successively shown several variants of a symbol and its intended meaning. Participants were asked to estimate the percentage (from 0 to 100%) of HCPs and visitors to the hospital that could be expected to understand the meaning of each variant. The order with which variants of each symbol were presented was randomised for each survey to avoid order bias. The variants of each symbol with highest average score during judgement testing, meaning those that participants judged as being most likely to be understood, were then assessed through comprehensibility testing, during which a new group of participants were presented with a symbol and an image demonstrating the context in which they might find it. Participants were then asked to write exactly what they think the symbol means and what actions they should take in response to the symbol. Each participant saw one variant of each symbol and six different test versions were created with the order and combination of symbols randomised. The variant of each symbol with the highest percentage of correct responses was deemed the most comprehensible and was retained for further use during iterative prototyping and evaluation of the overall isolation signage.

The final signage prototypes underwent observations and expert review with infection prevention experts prior to being finalised by a professional graphic designer.

The existing isolation signage system in our hospital was composed of a single poster for HCPs indicating only that “special precautions” were necessary and that visitors should contact the nurse’s station (Fig. 2). Additional signs existed for housekeeping personnel, which indicated the category of isolation, which protective measures were applicable, and which disinfecting agents should be used (Fig. 2).

Observations and interviews revealed several barriers to use of the existing isolation signage system. These barriers are presented in the following sections.

Two ideation workshops were conducted. Several recurring themes were identified, as discussed in the following sections.

The process of iterative prototyping and evaluation of isolation signage occurred in parallel to the testing of symbols. During this process, several prototypes emerged with varying levels of sophistication. In the earliest stages, ideas were expressed on paper in the form of sketches. Those sketches with the most positive feedback from the target population were then transformed into more sophisticated prototypes. As prototypes were continually evaluated through informal discussions with frontline HCPs, their feedback provided insights that led to further refining of consecutive prototypes. Experts in infection prevention and control were regularly consulted to ensure accuracy of the signage content. This iterative design process, as documented in Additional file 3, continued until saturation was achieved (i.e. until no new feedback was given).

The final solutions that resulted from the iterative prototyping and evaluation process were shared with a graphic designer to produce the final signage system for the three main isolation categories (Fig. 8) as well as two signs for combined precautions (Fig. 9). This solution features a caution tape graphic and salient colours to draw attention to those entering the room that special precautions must be taken. The isolation categories are communicated dually through the colour, which is standardised throughout all isolation precaution documents, and the prominent symbol portraying the transmission pathway. The specific precautions are portrayed in the order in which they must be performed, removing method ambiguity. The use of text is limited and the signage relies largely on user-friendly graphics to support comprehension regardless of language proficiency. The signage is available for download on the intranet and can be printed with one sign per A4 sheet of paper. Beyond the scope of the physical signage, hospital guidelines have also been introduced that limit the amount and type of information that can be posted on patient doors, which increases the saliency of remaining signage. In collaboration with the infection prevention department, the housekeeping department has also revised their isolation precaution guidelines and brochures to eliminate confusion with signage intended for HCPs.

The signs were implemented by e-mailing heads of wards to inform them of the new signage, announcement and posting of the new signs on the intranet server, and physically distributing printed signs to each hospital ward according to number of beds. Additional copies were made available for order. Physical copies and links to the previous signage were removed.