Have you ever peeked behind a pharmacy counter? So many pills, syrups, ointments, and powders on shelves make up a wall of clone-like bottles. Hospital pharmacies often stock an even wider range of products than the corner Walgreens and have many IV bags and vials that look the same, but produce very different effects. Due to our extensive knowledge conducting UX research in a variety of healthcare settings, a client came to us with the challenge of redesigning medication labels on “look-alike, sound-alike” drugs to improve the accuracy of delivering the correct medications to patients.
This client recognized a need to improve the quality of use of its products. They also needed to learn more about how their products were used. They had questions like
* What tasks are done with their products?
* Who is involved in the tasks?
* What are the environmental factors that influence the tasks?
* Where can things go wrong?
* What are some strategies the user employs to complete the tasks?
Contextual inquires are an effective launching point for further research when a client has some gaps in their understanding of their users. This method is a subset of ethnographic research, in which a UX researcher collects detailed information about users’ experiences through direct observation. In our studies, “users” could be the client’s customers who buy a product or service, the client’s employees, or someone in between who comes in contact with their products.
The main goal of this study was to find out exactly how users worked with specific medications, and in the end we learned quite a bit about conducting contextual inquiries. I’d like to share some of our lessons learned in the context of a real-life example, because that’s what this method is all about. Strap on your hair net and follow me into the recesses of the hospital pharmacy…
Gain access to the right people. First, how do you even get into a hospital pharmacy? In this case, the client had sales representatives with strong relationships at different hospitals across the country. These reps got our foot in the door and contacted department supervisors far ahead of time to begin setting up the study. By talking directly to supervisors and following up with a detailed letter about the intent of the study, we were able to establish this research relationship. Once the supervisor liaison ironed out the logistics, such as security clearance and staff shift times, they needed to get us in touch with the correct users. The supervisors themselves didn’t necessarily come in contact with the medications in question. We needed to find participants from the right user groups who worked with these medications. In this case, participants weren’t just pharmacists – the group also included nurses, pharmacy technicians and purchasing agents who worked throughout hospitals in the pharmacies, ER and central storage.
Maintain participants’ confidence and trust. “Who are the workers in a hospital’s central storage?” you might ask. We needed to learn about the different user types before we were standing next to them while they worked. Prior to conducting contextual inquiries, it’s important to become familiar with common work terminology, abbreviations, job roles, equipment, and concepts that the users might reference. This helps researchers to be more efficient in their data collection and better relate to the participants. In this case we learned the medications’ names and abbreviations, how they’re stored, the processes they’re used in, and general information about how they’re administered. Even after we felt we had all the background knowledge necessary, it was important to let the participants know that we were not experts. We were observing them because theywere the experts and we were hoping to learn from them. It was also important for us to assert that we would take all measures to avoid interfering with their work, which meant knowing when to stand out of the way. Observing and talking with an ER nurse involved balancing quick reflexes and patience.
Finally, we were clear about the measures we would take to maintain participants’ confidentiality. We knew they might be concerned about supervisors judging their performance or photos winding up in odd internet places. Usually, however, participants were more than happy to share more than we’d expected, such as a cell phone photo of a medication stocking error that a nurse had kept for months. This was useful information for us, but also demonstrated the need to sometimes refocus the participant on performing tasks as they normally do, rather than using the time to participate in an exposé about their workplace or coworkers.
Come with an open mind. During the study we had a well-constructed list of objectives, potential tasks to observe and questions to ask. We discovered, however, that a blank notebook was often more useful than our guide. Some of our takeaways relating to this need for flexibility included
* What we thought we’d ask, we didn’t
* What we thought users do, they sometimes don’t
* What the client thought was the problem wasn’t the whole story
Contextual inquiries are fluid, so researchers must adjust to the situation. The method helps us to understand the cognitive processes that accompany users’ tasks, but tasks are not always linear, and are sometimes difficult to verbalize. For example, while observing pharmacists fill orders we needed to account for the differing technology and processes in different hospitals. A fully-automated system in one hospital showed robots locating and grabbing medications from mechanized bins, while a completely manual process in another hospital showed a pharmacy tech picking medications off the shelves and wheeling them on carts to the respective units.
Contextual inquiries can be combined with a number of other methods to gather richer insights. Another field-based method we used included informal follow-up interviews. These interviews allow participants to clarify what was observed earlier and uncover additional insights such as motivation and perceptions. We also used artifact analysis, which is observing cognitive aids that participants use to support task performance (cognitive aids can be anything from posted notices to a highlighted user manual). In our project, we saw many brightly-colored warning stickers on medications to alert users to potential errors. We used the information from these aids to inform the redesign of labels so that such cognitive aids would not be necessary to perform tasks successfully.
Contextual inquiries can be time-consuming and require researchers to be sharp and flexible. However, they also help to answer questions in a way that no other user-centered design method can. Our client gained a more complete understanding of workflows and associated issues, and we were able to offer more effective recommendations. In the case of the “look-alike, sound-alike” label redesigns, additional methods such as in-lab usability testing were needed to achieve the client’s goals. However, contextual inquiries created a strong foundation upon which all further research could be based.
The author is a consultant for leading user experience firm User Centric, which was recently acquired by GfK. She can be reached at firstname.lastname@example.org.