ABUS Physician Training

A Cloud-Based Training Program for Breast Imaging Education

*I can't present the specific findings and design recommendations due to NDA, but I can share the process.

Overview

Over the course of six months, I conducted research, iterated on prototypes and executed user testing.

1. Research

• Drafted a project proposal and shared initial plans with internal teams. Attended training sessions as both an educator and learner (participant observation) in order to develop an understanding of user practices and pain points.

• Conducted informal interviews with physician learners and peer educators to help determine user needs. Conducted informal interviews with internal teams, and using that information, established guidelines to ensure any new system designs integrate well with existing workflows.

2. Exploration

• With the help of several educators on my team, I created affinity diagrams to build a comprehensive picture of Users’ Needs and Users’ current practices.

• With input from internal teams, conducted a SWOT analysis of existing programs and practices. Generated affinity diagrams to help explore potential solution strategies.

3. Prototyping

• Engaged in concept testing with physicians learners and peer educators to get early feedback on potential solutions. I generated lo-fi prototypes and iterated on designs to explore different solution strategies.

4. Analysis

• Conducted a competitive analysis to survey the landscape of education delivery solutions available within the healthcare space.

• Analyzed the steps and tasks required to complete a single training session in a task analysis, to ensure all aspects of the system were accounted for when making changes or redesigns.

• Synthesized research findings into a report, featuring a comprehensive picture of the problem at hand along with options for solutions and shared with management.

5. Requirements and Development

• Generated Product Requirement Documents to detail system requirements and constraints. Generated wireframes and lo-fi mockups to communicate the intended designs to developer.

• Interfaced with developers to manage timelines and deliverables. I designed test procedures and leveraged fellow educators to verify requirements were implemented as intended.  

6. Design

• Redesigned workflows, checklists, user instructions, and process maps to reflect changes to the system. I executed a task analysis of the redesigned system, ensuring all key events were accounted for and any changes complied with FDA-mandated guidelines.

7. User Testing

• Scheduled, recruited and moderated 3 pilot sessions (12 participants) using a prototype version of the cloud-based training platform and delivery format. During each session, I kept a log of key incidents and observations. I recorded the sessions with screen-recording software for additional analysis. I administered a post-test questionnaire after each session and conducted informal interviews.

8. Report

• A final report was generated and outcomes presented to management.

The Problem

The format used to deliver training to physicians is inefficient and will not scale up to meet predicted needs. A scalable training solution is needed. To give some context:

1. Training = Essential

Physicians who review images captured via GE Healthcare’s Invenia ABUS medical device must complete an 8 hour training curriculum as per FDA mandate. If physicians do not complete the curriculum, use of the device is considered “off label”, and insurance companies are not able to approve or pay for treatments.

2. Scheduling = Headache

Prior to 2015, all ABUS physician training was delivered in a “live” format that involved weeks of coordination with each customer (typically a hospital or private practice).

The delivery format required that a majority of physicians within a radiology group attend a single training event. Physician schedules are typically set months in advance, so coordinating schedules around a specific training date was a recurring pain point for hospital admin.

3. Legacy Solution = Unsustainable

The legacy format was also expensive and unsustainable. Executing a training session required sending cases of training equipment and at least one employee to the customer’s physical location, a practice that would not scale up as product demand increased.

4. Large File Size + Specialized Software

Physicians use Invenia ABUS workstation software to review 3D volumes of breast tissue. The file size for a single volume is large (>1GB), and the software is highly specialized for the task. For this reason, a highly specialized design solution was required.

ABUS Physician Training Delivery (Prior to Redesign)

Project Goal

Design an efficient, scalable training platform and infrastructure solution for the delivery of GE Healthcare’s Invenia Automated Breast Ultrasound Physician Training curriculum.

1. Assess Pain Points

Gather feedback regarding the legacy training program, with a focus on pain points and areas where the program is not meeting users’ needs.

2. Compliance

Keep all designs and updates in compliance with FDA-specified requirements.

3. Scheduling Solution

Alleviate scheduling and logistical pain points for users.

4. Less Hardware + Software

Reduce the amount of physical equipment and number of third-party applications required to deliver a training session.

Primary Research Questions

1. What changes in training strategy and delivery format would best address the identified problems?

2. What tools can help streamline the training process?

3. Besides the primary pain points, are there any additional pain points for physicians, peer educators or hospital administration?

4. Can the identified paint points be addressed via updates to the training program? What kind of updates are required?

Design Goals

Insights gained during research and exploration phases contributed to the recognization of a need for a two-part solution strategy. Two project goals were identified.

1. Update Training Program Platform

• Design a cloud-based version of the training software to replace the legacy desktop version.

• Build a training solution that does not rely on third party applications for essential features like bi-directional screen-sharing and video-chat.

• Provide a hassle-free experience for physician learners, hospital administration and Peer Educators.

• Build a solution that simplifies the training experience without compromising the high-quality standard established for the program.

2. Update Training Program Delivery

• Introduce the flexibility to attend training from any location (home office, etc).

• Enable physicians from the same hospital to attend training sessions on different days.

• Virtual Classroom: Participants can use their own PCs, hardware, accessories and internet connection.

Outcomes

Deliverables

• Product Requirement Documents

• Wireframes and mockups to demonstrate intended functionality and design

• A Cloud-Based Training Platform Prototype

• Documentation for responsibilities and best practices in relation to the new delivery method and training platform

• Pilot program and data from 3 pilot sessions

• Implementation Strategy

Implementation

The decentralized training delivery format went live in November 2016. The cloud-based training platform was not implemented beyond the pilot program.

ABUS Physician Training - Delivery (Redesigned)

What did I learn?

1. Balancing the needs of different user groups is a challenge

In the case of the ABUS Training software and delivery platform, physician learners were the primary user group, but other groups (peer educators, hospital admin and internal teams within GE) also used the platform in different capacities. Balancing the needs of different user groups is a challenge, as each group has different motivations and goals. In order to build a well-functioning system, it is essential to work at satisfying the needs of all user groups.

2. Flexibility is paramount

During the course of a project, priorities will naturally shift and evolve as a more complete understanding of user needs develops. It is important to remain flexible and pivot when necessary.

3. Less is more

Some of the “state-of-the-art” features I worked on — features I was most excited about — were not implemented in the final solution, as they were overly complex. Many of the features that added the most value were simple by nature. Complexity is not a measure of quality.