Humana Health

TL; DR Summary

At Humana, a significant portion of medical payments is attributed to members with chronic conditions, many of which are preventable. To address this, the Humana Health engagement was initiated to better serve this member population, aiming to prevent disease progression and reduce expenses for both members and the company. I led design on a multidisciplinary product team, including product managers, engineers, and data scientists, that was assembled to conduct product discovery. I conducted remote ethnographic-style interviews, which revealed key insights into member behaviors and motivations. By developing behavioral archetypes, I was able to guide product development that was focused on internal and external motivations for behavior change. I established experience principles to inform the design process, emphasizing success, brevity, and incremental achievements. Partnering with our data scientists, we explored behavior change techniques and developed a goal-setting model for the digital product. Based on identified archetypes and principles, the team prioritized features for the Minimum Viable Product (MVP), focused on facilitating communication between members and their care teams. By employing user research methods, including card sorting, tree testing, and surveys, I tested design decisions and gathered feedback. The resulting MVP initially launched with secure messaging and member profile features, serving as a pilot for six months. Subsequently, my efforts transitioned to long-term strategic planning for the Humana Health portfolio, aiming to offer self-service health guidance to all members over a multi-year roadmap.


The Problem Space

At Humana, members with chronic conditions account for a large majority of payments that the insurance company is required to pay out for medical providers. However, folks with multiple complex chronic conditions such as cardio obstructive pulmonary disorder (COPD), diabetes, congestion heart failure(CHF), etc. constitute a majority of expense that is considered to be largely preventable. The Humana Health engagement was started as a partnership between the Experience Center (XC) and our Clinical Products Organization (CPO) to see how we might better serve this member population and help them keep their diseases from progressing, thus keeping them healthier, and saving money for themselves, Humana and CMS (The Centers for Medicare and Medicaid Services,) making it a win, win, win.

I joined a multidisciplinary product team made up of product managers, product designers, design strategists, software engineers, and data scientists to begin a product discovery in anticipation of a product build for Humana Health. We partnered with frog design as a research partner to complete ethnographic research in service of shaping up a higher-level strategy initiative that another team was responsible for, and for pointing the product team in the right direction as they got started. Secondary research that was supplied by consulting firm Deloitte included product concepts that needed to be validated through user research before being included in our efforts. My role was to serve as the link between the strategy team to the product team, delivering insights that would help better shape the product. Additionally, as a design lead on the project, my role was to work with our partner designers from the larger Humana organization and bring them up to speed on our lean design research and agile practices.

Ethnographic Discovery

Our participants and their caregivers

When doing broad exploratory research, ethnographic style discovery is considered the best way to learn very deeply about one’s subjects, but unfortunately, due to the pandemic, we were unable to visit our research participants in their homes. We opted instead to conduct remote ethnographic-style interviews, having the participants bring us into their homes over a video chat. The interviews were two hours long and included both the patients with chronic conditions and the people who help care for them. Over the course of one month, we spoke with twenty-four patients and their caregivers, three expert physicians, and fifteen Humana stakeholders. We synthesized all of the interviews as a team using affinity mapping. From here we were able to identify key themes and categories that would help us shape our design principles.

Our ethnographic research synthesis in our collaborative whiteboarding tool, Miro

These interviews allowed us to identify key sets of behaviors that our participants exhibited in certain scenarios, and from this data, we were able to develop a set of behavioral archetypes that we would be able to use to validate or invalidate our product development decisions. Differing from personas, behavioral archetypes are focused on key motivations, attitudes, and behaviors that explain WHY people act in a certain way, whereas personas are fictionalized characters that bring specific demographics to life. We chose to work with archetypes rather than personas in order to better help uncover our opportunity areas in addressing unmet needs and making design choices more intentionally. 

Behavioral Archetypes

The archetypes were identified as follows:

  • The Lone Warriors who are determined and resolute

  • The Headstrongs who are stubborn and difficult to please

  • The Truth Seekers who believe that more knowledge will lead to better health outcomes

  • The Wanderers who seek motivation and for others to hold them accountable

The archetypes were derived based on whether they were externally or internally motivated and how open to change they were. This idea was drawn from secondary research related to behavior change, referencing both the COM-B framework and the Fogg Model that we used to frame our findings. 

Experience Principles

Based on our conversations with our participants, we also identified a set of principles that would guide our work as we moved forward, ensuring that it would resonate with people struggling to keep up with their chronic conditions while balancing the needs of everyday life. 

  • Lead with success

  • Prioritize brevity

  • Promote incremental achievements

  • Utilize existing anchors

  • Leverage the community

Behavior Change Design

During our product discovery, we partnered with our data science team to identify what was possible from a technological standpoint in terms of combining user interactions with machine learning. We were trained over the course of three months in behavior change design, learning a variety of behavior change techniques, and through our partnership, we developed a model of goal setting that could be used in a digital product to keep users engaged in making progress towards their health goals. We also explored opportunities to enrich a members view of their own health, such as showing timeline of their medical history. 

Behavior change model for health goal setting

The Minimum Viable Product (MVP) Roadmap

Based on these archetypes and principles we identified where we were headed first with the product. We would build our MVP around the Wanderer, aiming to facilitate communication between our members with multiple complex chronic conditions and their Humana care team, motivating and informing the member to better manage their conditions. In a product kickoff and prioritization exercise we identified three key concepts as both having high feasibility, and high member impact:

  • Relevant Education at the Right Time

  • Member Context all in One Place

  • Personal Goal Setting & Care Plans

Our product concept prioritization exercise

At inception, Humana Health was initially conceived of as an Android app named Care Connect focused on personalized goal setting and the early user research was focused on proving out that concept. Here is a prototype of the Android app as initially conceived:

At three months in the business decided to pivot, Humana Health would now be a web application in order to be accessible to the largest amount of users, as we knew the pilot group would be limited to Humana members that were already identified as good candidates for the care management program.

Our product was to be a digital companion to an already running telephonic care management program. In this program, members who are identified as being at risk for chronic disease progression are connected with a multidisciplinary care team consisting of nurses, social workers, doctors, and pharmacists. These care managers work through a care plan with the member to help them slow or stop their disease and symptoms from getting worse by providing them education on how to change their behavior in a healthier way. 

Our three key concepts were distilled down into a core set of features to deliver:

  • A secure messaging platform, or in other words, a HIPAA compliant web application where a member could message back and forth with their care team. In addition to care team messaging, members could be onboarded and offboarded from Humana Health from the chat interface, and receive digital content such as educational pdf documents. Members' important loved ones could also participate in messaging to help keep the member on track with their care plan by logging in as a caregiver. 

  • A member profile that displays all of the members' important health information, such as medication history, list of conditions, and medical providers. 

  • A digital care plan that displays an online task list for both learning and taking action to improve one’s condition. Care plans are determined by a member’s care team based on an initial assessment that is conducted when a member first contacts their care manager. 

User Research

As a long-term project, multiple research methods were employed once we had moved beyond our discovery phase into iterative product development. While the engineering team was focused on less-risky items like identity access management, the design team focused on conceptual development to ensure the framing of our core features would resonate with our members.

For instance, while shaping the information architecture of the app we conducted unmoderated card sort activities through our research platform UserZoom to see how users would categorize the content on the site.

An example of a card sort activity to help shape information architecture

Based on that information we created a proposed structure for the site and tested it using UserZoom’s unmoderated tree test function. 

An example of a tree test activity used to validate information architecture

And we used surveys for quantitative and qualitative research. In one instance we were trying to learn more about caregiver communication preferences, specifically to validate whether caregivers would use a secure messaging platform to communicate with the care team on behalf of the member they were taking care of, and we were surprised to find out that they were quite open to being able to message back and forth with the care team, provided the member was also able to take full part in those conversations.

Surveys were also used to identify things like user intent, how often might they make use of our platform, and which features would be most useful. 

Viewing study results in the UserZoom platform

All of these unmoderated methods were balanced against a regular weekly or biweekly moderated testing cadence where we were speaking to 4 to 6 people with multiple chronic conditions and testing our software and prototypes to ensure our product was moving in the right direction. The following prototype showcases both the messaging and member profile features:

Enabling Behavior Change at Scale

Due to the complexities of including care plans in the product, our MVP launched with secure messaging and member profile features and ran as a pilot for six months. Following the launch of our MVP, I transitioned off of this effort to partner with the product manager responsible for the long-term planning of the Humana Health portfolio, and conducted a discovery and framing exercise using the insights garnered from over a year of qualitative research. Our goal was to identify which capabilities the organization needed to best support our members in changing their behavior in an effort to improve their overall health. We planned a multi-year roadmap for our partner organization to help them eventually achieve their goal of offering self-service health-based guidance to all of our members.

The roadmap that resulted from our discovery and framing

Previous
Previous

Path

Next
Next

Compass