Uncovering Design Opportunities for Telemedicine Technologies
The rapid implementation of telemedicine tools during the COVID-19 pandemic affords us the opportunity to explore telemedicine use in daily healthcare. Our goal was to understand doctors’ perception and use of telemedicine tools.
Project Overview
We surveyed over 100 medical providers in June 2020 about how they use telemedicine and what they think of the telemedicine tools during the COVID-19 pandemic. We found that the most common challenge with telemedicine was conducting physical exams. To mitigate this challenge, doctor's developed creative workarounds. Developers of telemedicine technologies should integrate these workarounds into deliberate design features.
My Role
I was the lead data analyst on this project. My responsibilities included:
-
Cleaning the survey data (e.g. removing incomplete responses)
-
Analyzing the quantitative data and running descriptive statistics
-
Coding and analyzing qualitative data (text entry within the questionnaire)
-
Discussing and iterating the themes with the research team
-
Collaborating with a medical provider to validate the findings
-
Proposing design solutions based on the findings
-
Writing up and presenting the findings
Tools Used: Microsoft Excel, Canva
The Team
I worked with a wonderful cross-functional team:
-
Krzysztof Z. Gajos, Ph.D., Harvard University
-
Anoopum Gupta, M.D., Ph.D., Mass General Hospital
-
Vineet Pandey, Ph.D., Harvard University
-
Maia Jacobs, Ph.D., Northwestern University
Research Process and Approach
Over 100 doctors completed a survey with 35 questions about their role, attitudes towards telemedicine, and use of telemedicine technologies. The questions were a mix of multiple choice, Likert scales, and open-ended responses. I was brought in to analyze the data and then communicate the findings and provide design recommendations. I started by researching the topic and understanding scope of the study. Then, I familiarized myself with the survey questionnaire, which allowed me to more efficiently analyze and understand the data.
Analyzing the data
Once I had a grasp on the project and the research questions, I dove into analysis. I started by cleaning the data. A few things I did to clean the data was remove incomplete responses, respondents who didn't meet our criteria, and respondents who "straightline."
Next, I analyzed the quantitative data.
I used Excel functions to run descriptive statistics (mean, range, and standard deviation) of Likert-scale questions and count the frequency of answers. I made tables and graphed answers to help me understand the meaning of the data and answer our research questions.
Then, I analyzed the qualitative data. I first coded the responses by writing the main theme of each responses. After a first round of coding, I grouped the codes by themes using post-it notes in Miro to identify the high-level findings. I discussed these themes with the research team and a medical provider to get their thoughts and feedback. I iterated on the themes and then coded the data with these themes. The output of the qualitative analysis was an organized spreadsheet that described the participant's actions, perceptions, challenges, and needs.
Communicating the findings
Once the quantitive and qualitative data was analyzed, I looked at the findings of all the data together to identify the key insights. I synthesized the main challenges and workarounds that doctors were using. I then turned to other work to see how existing systems and studies were addressing these challenges. I incorporated existing strategies into my design recommendations and highlighted opportunities for novel technologies. I presented these findings with actionable next steps to the broader research team.
Insights
Through quantitative analysis of a 5-point Likert scale, I found that, on average, doctor’s were more enthusiastic about telemedicine after using the system for several months during the pandemic.
However, enthusiasm for telemedicine was still low. Why? I dug into the qualitative data to find out more.
Doctors reported many barriers to telemedicine. Through my analysis of the qualitative data, I found that 82% of doctors identified the inability to conduct proper physical exams as a major challenge with telemedicine.
They reported being unable to perform a physical exam because:
-
the patient’s inadequate setup (like bad internet connection or poor lighting); and
-
the inability to physically touch the patient.
Doctors struggle to conduct virtual physical exams so they developed creative workarounds to mitigate telemedicine limitations.
These workarounds present opportunities to improve virtual care by turning workaround strategies into deliberate telemedicine features.
My Recommendations
An automated system could improve the experience by ensuring patients have adequate audio, video, lighting, and internet connectivity. This would alleviate the setup burden on providers, allowing them to concentrate on essential patient care.
Additionally, it is difficult and time-consuming for doctors to think of strategies for conducting virtual physical exams. Thus, we are proposing that these ideas are shared and disseminated so the burden is not on the individual clinician but on the entire community. A crowdsourcing platform could facilitate this knowledge exchange and reduce the time spent on developing strategies.
Lessons Learned
Through this project, I gained valuable insights into the UX research process:
-
Developing a deep understanding of how users interact with a system in its current state, both within and outside its intended use and context, is important for supporting and designing workflows that align with the users' behaviors.
-
Collaborating with a cross-functional team (in this case, for example, a quantitative expert and a medical professional) led to a more nuanced understanding of the data and actionable insights for improving the user experience.
-
Collecting data through surveys can be beneficial as it is scalable and yields quantifiable data. Yet, surveys may not capture the intricacies of human behavior. Selecting the appropriate research method is vital for a successful project.