Lily Chen — A mobile app for cancer patients

Lily is a Product Designer based in San Francisco.
Currently, AngelList. Previously, IBM.

↑ Back

A mobile app for cancer patients

In 2016, IBM and the American Cancer Society announced a partnership to create an AI mobile app for patients fighting cancer.

The Challenge

Despite readily accessible resources on cancer today, numerous studies still show concern that lay understanding of cancer is suboptimal. Patients and their caregivers lack the medical framework to fully comprehend and manage their cancer.

In order to improve patients’ understanding of their cancer, ACS and IBM saw an opportunity to combine powerful resources from both sides: ACS’s massive resources on cancer and IBM’s artificial intelligent features, widely personified as “Watson”.

The mission was to build a consumer-facing mobile app that empowers patients and caregivers to be more educated about their cancer.

Process

I was the sole product designer of an early development team in Oct 2016 and worked alongside my project lead and three user researchers. We handed off to the production team in February 2017.

The Cancer Journey

Nearly 40% of men and women in the US will be diagnosed with cancer within their lifetime. Each patient stands to undergo the most emotionally and physically significant experience they will ever have.

To understand this experience, we interviewed 8 patients, 15 caregivers, and 43 healthcare professionals ranging between oncologists, nurses, and medical assistants.

After gaining a more sophisticated understanding of the cancer journey, we derived a few key insights:

Patients’ caregivers were typically their family members

Very few patients are able to afford a licensed caregiver. Instead, most caregivers were the patient’s spouse or children, who didn’t have a medical background and coped by becoming the primary information gatherer on the patient’s behalf.

Doctors remained the preferred source of information

Despite an increasing variety and volume of cancer resources through the internet, patients still preferred to seek answers from their doctors. Even when patients expressed concern about their doctor’s judgement, second opinions still came from other doctors.

Patients struggled to digest so much medical information

Caregivers often accompanied patients to their oncology appointments in order to take copious notes as well as record their consultations. They would then take the information home, review and research it, and come back with more questions.

The Solution

We iterated on early concepts with a consistent set of 4 patients and 3 caregivers, all of whom have experienced cancer treatment.

In an environment with a wealth of knowledge but no medical degree, both patients and caregivers struggled to take full advantage of doctor consultations. We ultimately designed a digital counterpart in the room that would make discussing complex topics more digestable.

Focus less on taking notes, more on the conversation

Doctors today are already accustomed to patients taping their sessions in order to digest the heavy amounts of information that they talk about. We piggybacked this behavior by providing live speech-to-text transcription.

Search through any conversation

Currently, patients keep archives of voice memos or a binder of notes as traceable receipts for important discussions they’re afraid of forgetting.

But as conversations are transcribed to text in the app, each word is indexed so that users can search through them at any time and no discussion is ever lost.

Search also suggests the most incidental keywords across all conversations so patients can find recordings by theme.

Rewind and review key topics

Once a patient leaves the doctor’s office, they often want to revisit the conversation and do personal research on major topics. But even with speech-to-text, reviewing hours of transcripts is still tedious.

By training on the vast ACS database, Watson can recognize and highlight terms of interest in the transcript and break down the recording into sections of major topics so patients have revisit the content easier.

The highlighted terms also serve as easier, in-context search queries by allowing users to click on them and be directed to ACS-approved related resources.

Conclusion

Although both IBM and ACS stakeholders were excited with this direction, the partnership was ultimately put on hold and we hope that it returns from its hiatus in the near future.

However, the collaborative relationship we developed with ACS and their various facilities during this project allowed us to re-connect with patients and caregivers on a regular basis for future projects.