Centene, flagship project
True Care Anywhere
Care managers assess members in their homes, and they used to carry the assessment in on paper: hundreds of pages of it, thousands of questions, all of it feeding the care plan built for that member. True Care Anywhere put it on a Chromebook. I led it, and the application was the last part of the problem rather than the first.
- What it replaced
- Hundreds of pages of paper
- Assessment time cut
- More than 80%
- Team
- Five, plus three partner orgs
- Launched
- The Florida market
A care manager visiting a member at home is there to work out what that person needs and to build the plan that gets it to them. The instrument for that is an assessment, and at the scale of a managed care organization an assessment is not a form. It is hundreds of pages and thousands of questions, branching, with whole sections that only open depending on an answer three pages back.
On paper, that costs the visit. The care manager is turning pages and writing, the member is waiting, and the care plan gets built afterwards from handwriting. Everyone involved knew what the fix was. The reason it had not happened is that the fix is not an application.
Why the hardware was the project
To put that assessment on a screen in a member's living room, the screen has to exist, it has to be a device a regulated healthcare organization is willing to let protected information sit on, it has to be manageable when it is in a car in another state, and there has to be a way to install software on it that is not an app store listing anybody can find. None of those are frontend problems, and all of them come before the first question can be answered.
We did it with Google, on Chromebooks. I had the team procure the fleet, worked with our security organization to secure the devices in Google Cloud Platform, and ran device management from there. The application was built in Angular and Ionic and installed on each device through Centene's own Play Store, so distribution was private to the organization rather than public.
The decisions
01
Digitize the branching, not the pages
The failure mode for a project like this is a beautiful reproduction of the paper: the same hundreds of pages, now scrolling. That saves nobody anything. The value is in the part the paper could not do, which is knowing which questions actually apply to the member in front of you and never showing the rest. That is where the time went, and it is why the result was more than an 80% cut in the time a care manager spent assessing a member and building their care plan rather than a marginal improvement.
02
Chromebooks and managed devices, rather than a BYOD policy
A managed fleet is more work up front than letting people use what they have, and it is the only version of this that survives contact with a compliance review. The devices were procured, secured in Google Cloud Platform with our security organization, and managed from there, which means a device in the field is a known device with known software on it, and a lost one is a solved problem rather than an incident.
03
Private distribution through Centene's own Play Store
Shipping through the organization's own Play Store meant the application reached exactly the devices it was meant to reach, updated the same way every other managed application did, and was never listed publicly. It also meant the release path was one the enterprise already understood, which is worth more than a clever alternative when the thing being released touches member health information.
04
Build it on the stack I had already got approved
Angular and Ionic were not a fresh choice here. Getting them into a Fortune 25 enterprise stack that had never carried them was work I had done years earlier, building the mobile department. By the time True Care Anywhere needed a framework, the approval, the build pipeline and the people who knew it already existed. The interesting part of that is not the framework. It is that a practice built five years earlier is what made a hard project tractable, which is the actual argument for building practices instead of products.
05
Hand it to the team that lives with it
I taught the TrueCare engineering team to take over mobile development and the release pipeline running behind it. That is the only ending I want for a project like this. A system whose changes have to route back through the group that built it is a system that stops changing, and the team closest to the care managers is the team that should be able to alter what the care managers use.
The team
Five of us: a scrum master, a product owner, three developers, and me leading it. Small for the surface area, which is why the cross-functional part mattered more than the headcount. We worked against our labs team, the care management organization whose people would actually carry the devices, and product.
Care management was the partner that made or broke it. They are the ones who know which questions matter, what a visit actually looks like, and which piece of the assessment the care manager dreads. An application built for that workflow without them in the room is a faster way to produce the wrong thing.
Stack
- Build
- Angular, Ionic
- Devices and distribution
- Chromebooks, Google Cloud Platform device management, Centene's own Play Store, Partnership with Google
- Delivery
- Cross-functional Agile team, Scrum master and product owner, Handover to the TrueCare engineering team