UW Medicine

Patient Journey/Behavioral Landscape Mapping + Archetype Creation

Business Goal

The UW Medicine website serves as a key acquisition path for the health system to attract new patients or introduce existing patients to additional care pathways.

UX Objective

The objective of this effort was to obtain a better understanding of the actions taken/thoughts of the core UW Medicine patient who seeks to establish new preventative care with an organization or re-establish care in another specialty area for which they can self-refer. I sought to study the motivations, needs, goals, thought process and reported behavior of patients in the study group.

When I began the project, I sought to produce the following visual illustrations of this patient orientation toward healthcare decision-making and completed actions:

• Patient archetypes

• A behavioral landscape map of the archetypes

• A user journey map

My Impact

• I advocated for the project and secured funding for the transcription of 12 hours of moderated interviews with 10 users.

• I led the process end-to-end.

• I communicated progress to stakeholders.

• I completed instructional design for collaborative analysis workshops with 10+ attendees between both.

• I completed all thematic coding and analysis of interview transcriptions.

• I completed the visual design of the journey map and behavioral landscape map.

• I presented the final artifacts to stakeholders reviewed quarterly for the incorporation of additional insights from additional users tests or website data analysis.

Skills Showcased

• Generative user research

• Interview moderation

• User journey mapping

• Instructional design

• UX archetype creation

Study Group

The study acknowledged that there are four core patient groups. Those seeking:

• Preventative care: Primary care and wellness exams

• Episodic care: Care for a condition or symptom that is short term

• Chronic care: Care for a condition(s) or concern(s) the patient lives with ongoing

• Same-day day: Immediate care that may be accessed through urgent, emergency, virtual or primary care that may be accessed in the same day as the patient would like to receive it

This means, the patients in this group can actively choose the brand, location and the healthcare worker (doctor, etc.) providing their care. The study group included some patients in the preventative, episodic and same-day care groups.

In the scenario under study, we did not seek to understand the behavior of patients seeking care for a life-threatening or dire conditions. Additionally, we did now study the journey for patients who are referred for care. We believed patient behaviors in these states are fundamentally different and deserved separate research consideration.

Example scenarios that would be considered in the study group:

• Patients selecting a new primary care doctor or location

• Patients who access an urgent care clinic

• Patients who access a virtual clinic

• Patients selecting a maternity care provider

• Patients seeking care for a suspected condition

• Patients who may desire a second opinion

• Patients with a chronic condition who may want to change their primary care provider

Example scenarios that are not included in the study group:

• Patients who have sustained a severe injury and need immediate or life-saving care

• Patients who have a chronic condition and are not changing their health system, provider or care location

Final Deliverables

Read more about each archetype and accompanying quotes.

Process Summary

Detailed Method Steps + Agile Story Point Estimation

Process summary

• Write test plan and patient interview questions (3 points)

  • Read my interview script.

• Gather any existing data from Google Analytics about patient pathing and audience types. (4 points)

  • I began the study with a Google Analytics deep dive to understand:

    • Who was visiting the UW Medicine website

    • Primary actions taken and content consumed when they visited

    • The device for website users

    • Top affinity categories for users

      • (I looked at the top 30.)

    • Top in-market segments for website users.

      • (I looked at the top 30.)

  • This practice helped me to empathize with our users humanizing them beyond pages visited, depth and actions taken on the website.

• Review the appointing KPI dashboard to understand areas of the website performing best currently (2 points)

• Present segment data/analytics to working team (1 point)

• Review the patient journey map project plan (1 point)

• Define the overarching patient scenario for the user journey map (1 point)

• Draft the interview respondent groups (1 point)

• Review interview questions with cross-functional team collaborators (1 point)

• Socialize the plan with stakeholders (2 points)

Process summary

• Source 10 respondents through UserTesting.com (2 points)

  • It was important to me to have a balanced sample. I wanted to talk to users with/without insurance, with/without a primary care doctor, new to/established in the Seattle area and in single- vs. multi-person households.

    • 18-24, Male, 1 hour from Seattle, With commercial insurance

    • 25-34, Male, 1 hour from Seattle, Without established primary care provider

    • 25-34, Female, New to Seattle within the last year

    • 25-34, Female, 1 hour from Seattle, With commercial insurance

    • 35-44, Female, 1 hour from Seattle, With established primary care provider

    • 35-44, Male, 1 hour from Seattle, Multi-person household

    • 45-54, Male, 1 hour from Seattle, With commercial insurance

    • 45-54, Female, 1 hour from Seattle, Multi-person household

    • 45-54, Female, 1 hour from Seattle, With established primary care provider

    • 56+, Male, 1 hour from Seattle, Male

  • I also achieved gender parity in respondents.

• Conduct the interviews (10 points)

  • The interviews were conducted as live conversations that were recorded and sent for complete transcription later.

  • This enabled me to ask follow-up questions during the interviews.

  • I used a template with ~70 questions, such that each moderated conversation covered the same themes but enabled me to pivot to follow-up with users around specific areas of interest raised during each unique interview.

Process summary

• Thematically code each interview using the qualitative research software Dedoose (8 points)

  • This enabled me to identify codes used most frequently and to analyze the co-occurence of terms.

• Create a research primer that featured analysis based on the quantitative insights gathered from the coding of interview transcripts (5 points)

• Based on the quantitative insights, identify meaningful quotes for inclusion in the journey map + archetypes (3 points)

What do I mean by thematically code?

In Dedoose, I reviewed each transcript and applied codes to each discrete excerpt. Like this:

I identified 512 excerpts and applied 67 codes throughout for a total of 1,462 code applications through the 10 transcripts. This created a rich data source to review for code co-occurence.

Looking at the frequency of code application and co-occurence between parent and child codes enabled me to create quantitative data from qualitative research.

• Plan and lead two workshops to collaboratively map and analyze the journey map and behavioral landscape map for the identified archetypes (5 points)

• Draft the persona archetypes based on UW Medicine data and the interviews (2 points)

• Design the patient journey map (4 points)

• Present the journey map, archetypes and behavioral landscape (2 points)