Overview

CDC’s SimpleReport is a web application that makes it easy for under-resourced testing facilities to report COVID-19 test results to their local public health departments. Our goal for this project was to scale adoption to additional user groups by implementing a bulk upload feature.

process

1. Reviewing existing research: I began by reviewing research conducted by the other team to understand the current onboarding process and existing issues with using the product

1. Usability testing of information architecture: Since we were adding a brand new feature to our product, we conducted usability tests to ensure that potential users could find the bulk upload tool within the SimpleReport UI.

2. Onboarding and contextual inquiry: As we onboarded new users, we conducted contextual inquiry to observe how they created their CSV files, what other tools they were using in the process, and what areas in the bulk upload process had the most friction.

The problem

Background: a reporting tool for rapid COVID-19 tests

COVID-19 turned places like schools and nursing homes into testing sites. We worked with the Centers for Disease Control and Prevention (CDC) to launch and scale a free, time-saving tool that makes it easy for under-resourced organizations to record results for diagnostic tests and quickly report critical data to public health departments. SimpleReport was initially designed to be used as a workflow and reporting tool for point-of-care rapid tests, in which users can report test results one-by-one, replacing manual and burdensome methods like faxing or mailing paper results.

The challenge: Scaling SimpleReport to support new user groups

While many non-traditional testing sites like schools and nursing homes had adopted SimpleReport to ease their reporting processes, public health departments continued to receive CSV files of test results from high volume submitters like labs who could not feasibly enter results one by one, but don’t have the technical capability to send this data electronically like higher-resourced labs. For state and local public health departments, managing CSV uploads is incredibly time-consuming. Many don’t have robust validation on the CSV and spend a lot of time working with submitters to correct their files. To solve this problem, another team working with the CDC launched a basic CSV upload tool that would allow labs and other senders to upload a CSV file of test results, and began to pilot with about a dozen users. Since SimpleReport had already been developed as a tool for testing facilities and had a robust sign up and onboarding process, it was decided that our team should take on this feature.

As we adopted this feature and sought to scale it a wider audience, we had a few questions:

• Who would be using this feature? Since this feature was launched primarily to solve the needs of public health departments, we didn’t have a strong understanding of the types of facilities who were sending CSV files, and what their context of use looked like.

• How could we enable a large amount of users to successfully onboard to this tool? Formatting a CSV file correctly was a fairly technical process, and onboarding required 1:1 support.

• How could we ensure a good experience for the users of this tool? We understood that the data submission process was fairly technical and required users to format their data in a very precise way. We wanted to make this process as easy as possible for less technical users.

Usability testing

Since our first implementation of the bulk upload tool was essential a “lift and shift” of existing functionality from our partner team, our first priority was to ensure that existing pilot users and potential new users could navigate to and use the bulk upload tool. Participants included existing pilot users, non-SimpleReport users who we thought would be likely to use the tool, and existing SimpleReport users.

We conducted a task-based usability test and documented task success and failure. I also observed common patterns in user behavior and expectations around how the bulk uploader should integrate with existing functionality in SimpleReport.

Contextual inquiry

We worked with our state and local public health partners to identify additional candidates for the bulk upload tool, and we began to onboard them as we expanded our pilot. We conducted over 30 1:1 sessions with new users to observe as they created and formatted spreadsheets for upload, and used SimpleReport to upload their files. This gave us insight into the kind of data these facilities were working with, other tools they were using, and helped us identify opportunities to streamline the data submission process.

Research analysis and solution brainstorm

Key takeaways

• Users most likely to benefit from the bulk upload feature were labs and hospitals who were exporting data from existing information management systems - a very different audience than existing SimpleReport users, who were mostly non-traditional settings (like schools) who conduct rapid, point-of-care tests on site.

• Many of the existing features in SimpleReport, like a table to view test results, were not necessary for users who had other systems to manage test data - they only needed a very simple UI to upload a spreadsheet, along with guidance on how to do so.

• Formatting data to fit our specification is a very time-consuming and error-prone process. Users especially struggled to map their testing devices to the correct codes that we needed to sent to the health department, and error messages were not sufficient in helping them troubleshoot issues.

Service blueprint

To visualize the holistic user experience, including the larger ecosystem of systems and policies that shaped their experience - both seen and unseen to the user, I created a service blueprint. I mapped out steps the user took to create and submit bulk upload, interactions with people and digital touchpoints, as well as behind the scenes systems and policies (e.g., public health data requirements). I identified negative moments and points of friction as well as areas of opportunities for our team to improve the onboarding process and user experience.

Proto-personas

We drafted very lean proto-personas to document important attributes of these new user groups. We plan to flesh out more detailed user archetypes and personas as we learn more about these users and the different people involved in the testing and reporting process.

Solution brainstorm

Having identified our core challenges to scaling the bulk upload feature, we conducted a cross-disciplinary team brainstorm session to begin thinking of solutions. The brainstorm session was centered on three main “how might we” questions:

1. How might we assist Electronic Medical Record (EMR) power users in sending high quality data to receivers?

2. How might we leverage what we know about senders to streamline the uploader experience?

3. How might we integrate with senders’ current processes?

Solutions and outcomes

Improving the bulk upload guidance and error handling

To allow users to onboard with minimal support, we created a robust bulk upload guide that included a downloadable spreadsheet template, a training video, and detailed guidance with example values. Additionally, we added guidance to the bulk upload errors that gave users detailed information on where an error had occurred and how to fix it.

Adding a code lookup tool

Since we had identified that mapping devices to the correct data fields and codes was the most burdensome part of the process, we provided a tool that allowed users to quickly look up a device in our database and copy/paste values into their spreadsheet.

Results: increased adoption and reduction in support issues

In a matter of months, we were able to quickly scale adoption from 13 pilot users to 180 users. We also observed a decrease in the amount of support tickets related to the bulk uploader.

Design exploration: a separate experience for bulk upload users

As we look to the future, we have begun exploring designs for a separated bulk upload experience that would allow users to switch between “single entry” mode for the regular SimpleReport experience and “bulk upload” mode for users of the bulk uploader. We hypothesize that this will improve the efficiency of use, reduce confusion, and more closely match users’ mental models. We plan to conduct concept tests with current and prospective users in the months ahead to validate these hypotheses.

Check out the bulk upload tool on SimpleReport’s training site.