Home
After Discharge Communication Program

After Discharge Communication Program

Proposed: MD Anderson wants a partner to provide a HIPAA-secure solution that will enhance a patient’s discharge transition to home.

Showcase & Media

Project Discussion

The Gunner Team breaks down the project

Wireframes

Check out wireframs of this project

Architectural Diagrams

Check out architectural diagrams of this project

Problem

What was the problem to be solved?

MD Anderson wants a partner to provide a HIPAA-Secure solution that will enhance a patient’s discharge transition to home, providing notification of clinical or service alerts to the proper MD Anderson staff, while reducing readmissions in a manner greatly acceptable to the patient.

This automated call solution should be easy and efficient for all MD Anderson staff.

Solution

What was the proposed solution?

Gunner Technology proposes a five-step solution to solve MD Anderson's problem.

Step One

Obtain patient data from the EPIC software API and minimize or negate any requirement of additional patient input.

This exchange should be a real-time or near real-time feed.

  • Patient name
  • Patient ID
  • Birth date
  • Primary patient contact (optional)
  • Contact relation (spouse, child, friend, sitter, etc.) (optional)
  • Phone #
  • Secondary phone # (optional)
  • Date of discharge
  • Facility type where discharge occurred (optional)
  • Discharge unit
  • Discharge physician
  • Discharge nurse
  • Attending physician
  • Dates of past admissions (within past 12 months) (optional)
  • Primary/secondary diagnosis
  • Best time to call (optional)
  • Option for recurring calls/follow-up calls (optional)
  • Date/times of previous completed transitional program automated calls (optional)
  • Preferred language (optional)
  • Special notes (optional)

Step Two

Automated call with default language but optional “plug and play” Algorithms can be patient specific inserted, based upon factors such as diagnosis, age, and severity.

The automated call Algorithms can be edited and personalized, if deemed appropriate by MD Anderson management staff.

  • General discharge Algorithm (questions on follow-up appointment, medications,
  • transportation, discharge instructions, etc.)
  • Symptom management Algorithms (cancer, heart failure, COPD, diabetes, etc.)
  • Home instruction education tips – can be appended to the call
  • Motivational tips

Step Three

IVR solution must be robust and capable of thousands of simultaneous calls.

Patients should have an option to either proactively call in for the automated call, respond to a text, or receive the automated phone call.

Calls should be as brief and efficient as possible.

The automated calls will begin within 24 hours of discharge.

Multiple attempts to connect will be made at an agreed-upon time of day and documented.

More than 50 lifelike voices across many 15 different languages are available.

Step Four: Management Dashboard

  • A history of all calls and responses will be available on demand
  • Any historical or present data is searchable and downloadable to Excel or other specified formats
  • Upcoming schedule calls can readily be viewed and edited
  • Any follow-up is readily specified as “Clinic Alert” or “Service Alert”
  • Current call responses can readily be seen in a color coded, green-yellow-orange-red format

    • Green – No action required; all answers were satisfactory
    • Yellow – Moderate risk, but no follow-up required
    • Orange – “Service Alert”; follow-up required
    • Red – “Clinical Alert”; follow-up required
  • Ability to see what required follow-ups are assigned, resolved, in process, or awaiting action
  • Various customizable reports will be available for analysis including:

    • Alerts by units and hospital discharge
    • Types of alert (medications, transportation, appointment follow-up, symptoms, etc.)
    • Patient to call response rate
    • Staff to patient response time frame (documenting the staff who followed-up)
  • The results of the automated call feedback will be available to be uploaded into the EHR.

Step Five: Support

Ongoing support

  • Monday – Friday; 8am to 6pm CST
  • Emergency 24/7/365
  • 99% + uptime
  • On-line tutorial and help screen for existing or new staff

Challenges

What challenges arose during the project?

MD Anderson proposed a tight timeline for project completion.

The client wants the platform implemented within approximately 20 days after kickoff, fully tested within 30 days of kickoff and launched within 120 days of kickoff.

While that is an aggressive deadline, it's certainly doable.

In fact, the only difficult aspect to this timeline is the Epic integration.

In Gunner's experience, the firm has found that epic integration can vary from client to client and Gunner is forced to assume certain things about MD Anderson's usage of the Epic platform.

To minimize this risk, Gunner will immediately tackle this part of the project upon contract award.

Technical

What was the technical approach to the project?

Gunner Technology sees this project as three separate projects to be build as Microservices.

Microservice #1: Epic Integration

This microservice will be responsible for integrating with the Epic API.

Using the Serverless Framework, Gunner will create AWS CloudFormation templates to build out this microservice, composed of multiple AWS Lambda functions, running Node.js.

The first Lambda function will get triggered every minute by an Amazon CloudWatch event trigger and pull the most recent patient data from Epic, extract the relevant data, buttress the data with supplemental information such as next call time, last call time, default preferences and more and insert it into HIPAA-compliant, encrypted Amazon DynamoDB tables.

Microservice #2: Progressive Web App Dashboard

Gunner will use the React Framework to build a PWA for dashboard and reporting functionality.

Access to this dashboard will be secured using AWS IAM and Amazon Cognito.

All assets required for the frontend of the PWA (HTML5, CSS3, JavaScript, graphics) will be stored on Amazon S3 and served through Amazon CloudFront.

This will ensure sub-second load times.

The PWA will use a GraphQL API managed by AWS AppSync and AWS Amplify to fetch the requisite data from DynamoDB using Node.js Lambda functions served through Amazon API Gateway.

This ensures data is Secure while in transit and at REST.

Admins will have access to the data via a UI build with the Material Design, Design Framework, through the Open Source Material UI library to provide a consistent and familiar interface for admins.

To be clear, this is an automated solution and does not require admins to view or edit the data as the app is procuring defaults from Epic.

However, this portion of the dashboard will allow admins to override the defaults on a case-by-case basis.

Finally, the PWA will leverage the JavaScript D3 Open Source library to chart reporting data, which will be exportable as CSV or any other format desired.

Microservice #3: Interactive Voice Response

This is the service that will call, text or handle incoming calls from patients.

Twilio will be the crux of this service.

The microservice will interact with Twilio through Node.js Lambda functions which will trigger based on Amazon CloudWatch events that correspond with agreed-upon time of day.

The default functionality will be for Twilio to call a patient.

Using Amazon Polly, the call can be made using more than 50 different voices across more than 15 different languages, including English and Spanish.

Depending on the patient, Twilio will ask a series of questions and update the Amazon DynamoDB data based on the response to each question.

If a response requires escalation, this service will use Twilio to forward the call to an appropriate call center, use Amazon SES to email the call center a link to the patient who is calling and update the patient record to denote such a call is made.

That way, whoever answers the call in the call center can easily use Microservice #2 to quickly bring up the patient in the UI and see patient data and responses.

The actions taken by the call center can then be noted in the Dashboard and recorded for reporting and analytical purposes.

Similarly, if MD Anderson admins update a patient's contact method to "Text" or "Call-in", Twilio will behave the same way but either via text or in response to a patient dialing a specified number.

Management

What was the project management approach to the project?

Gunner Technology uses Agile Scrum for project management led by Agile and Scrum certified project managers.

Over the years, Gunner has refined the process to suit may different situations and the firm is continuously refining and tweaking it on a per client and per project basis to best suit the needs of the situation and project.

However, Gunner generally approaches each project, initially, with the same methodology.

Eliminate the Unknowns

Nearly every project starts with unclear requirements and unknowns.

More so than anything, these can derail a project, so the team's first priority is to get rid of both, using Research Sprints.

Find our Velocity

In Scrum, Velocity is how many Effort Points the entire team can knock out in a single sprint or iteration (in Gunner's case, usually one week).

Gunner's project manager estimates how long a project will take based on the summation of all Effort Points at the beginning of the project.

For example, after creating the user stories, a project may have a total of 5,000 Effort Points.

Using simple math, the PM can estimate that will take the team 100 iterations (usually weeks) to complete based on a 50-point Velocity..

However, each project is different so the team of two with an average Velocity of 50 might trend higher or lower on a given project.

Augment if needed

Because Gunner meticulously tracks and refines Velocity, the team's project manager will know immediately if the project is off track.

For example, if the team's Velocity falls from 50 to 40, the project manager knows they need to augment the team to get the Velocity back up to 50 or greater.

The project manager can also use this score and method to alert the project stakeholder if a new feature they want is going to jeopardize the timeline and/or budget of the project.

Process workflow

Each project generally has three environments set up fo it.

Production for the live site.

Staging for the stable test site.

QA where the latest features are tested by humans.

Each step of the way includes automated testing upon each deployment, which occurs at the end of each iteration, prior to the planning meeting.

What’s on QA gets deployed to Staging.

What’s on Staging gets deployed to Production.

All code is stored in a Git repository on AWS CodeCommit.

Each user story gets its own Git branch where it is developed and tested.

Upon acceptance, that branch is merged into the master branch and deployed live.

Architectural Description

What platform was built for this project?

Gunner explained in the technical approach that it will build three Microservices using the Serverless Framework to complete this project.

A Serverless approach will allow the solution to achieve infinite Scalability without sacrificing performance at a fraction of the cost of traditional, Monolithic Architectures.

These Microservices will interact with one another via GraphQL and REST-based APIs.

That said, given the nature of the data, Security is of the utmost priority for this platform.

To ensure the platform is Secure, Gunner leverages several enterprise-level technologies provided by AWS to monitor, detect and prevent Security issues.

These technologies include AWS CloudTrail, AWS Config, AWS Shield, Amazon GuardDuty and Amazon Inspector.

To achieve maximum uptime and Availability, Gunner leverages AWS Availability Zones and AWS Regions for automatic fail-over, recovery and backup.

Gunner's solutions architect works with the developers on the team to implement a Continuous Development, Git-based, DevOps framework leveraging technologies such as AWS CodeCommit, AWS CodeDeploy, AWS CodePipeline and AWS CodeBuild.

Benefits

How did this project benefit the client?

By using Gunner's solution, MD Anderson will decrease the unplanned readmission rate and increase patient satisfaction scores.

The IVR solution will reach to 100% of MD Anderson's discharged patients, inquiring about specific matters for every discharged patients within 24 hours of discharge to ensure that the patients have transitioned home safely.

This automation will satisfy one of MD Anderson's Operational Priorities and provide patients with a better experience without consuming an excessive amount of MD Anderson Resources.

Why Gunner?

Why was Gunner selected for this project?

Gunner Technology has the technological skills to complete this project better than any other firm at a cost that can't be matched.

What Gunner is offering is a 100% cloud-based solution using the highest levels of performance, Security and protection available through Amazon Web Services.

This solution is not an off-the-shelf solution.

It is built to suit MD Anderson's goals and visions specifically.

Gunner has built many custom solutions for the healthcare industry and plans to add even more value through subcontracting with AZ Tech on this project.

AZ Tech brings a healthcare technical team with experience in building technology solutions and Algorithms for reducing hospital readmissions using IVR technology, which included a $100,000 Duke Endowment Grant.

AZ Tech closely followed the 2008 Iowa Chronic Care Consortium study using IVR technology to reduce heart failure readmissions, and the 2011 Geisinger Monitoring Program using IVRs.

The team is familiar with the Park Nicollet, Billings Clinic and UMMC’s Diabetic Telehealth programs, which also used IVRs for remote patient monitoring.

AZ Tech firmly believes that IVR technology can affordably and efficiently be utilized to make significant improvements on patient care and reductions in avoidable readmissions.

The combined team of Gunner and AZ Tech would be honored to work with MD Anderson on this after discharge communication program.

Proficiencies

What tools, techniques and methodologies were used on this project?

[object Object]

Adobe Illustrator

Create logos, icons, sketches, typography and complex illustrations

[object Object]

Adobe Photoshop

Photoshop is Adobe's photo editing, image creation and graphic design software.

[object Object]

Agile

Agile software development refers to a group of software development methodologies based on iterative development, where requirements and solutions evolve through collaboration between self-organizing cross-functional teams

[object Object]

Amazon API Gateway

Amazon API Gateway is a fully managed service that makes it easy for developers to create, publish, maintain, monitor, and secure APIs at any scale.

[object Object]

Amazon CloudFront

Amazon CloudFront is a content delivery network offered by Amazon Web Services.

[object Object]

Amazon CloudWatch

Monitor AWS resources and custom metrics generated by your applications and services

[object Object]

Amazon Cognito

Simple and Secure User Sign-Up, Sign-In, and Access Control

[object Object]

Amazon DynamoDB

DynamoDB is a nonrelational database for applications that need performance at any scale

[object Object]

Amazon Guard​Duty

Intelligent threat detection and continuous monitoring to protect your AWS accounts and workloads

[object Object]

Amazon Inspector

Automated security assessment service to help improve the security and compliance of applications deployed on AWS.

[object Object]

Amazon Polly

Text to Speech in 47 Voices and 24 Languages

[object Object]

Amazon Route 53

Amazon Route 53 is a highly available and scalable Domain Name System (DNS) web service.

[object Object]

Amazon S3

Object storage built to store and retrieve any amount of data from anywhere

[object Object]

Amazon SES

Flexible, affordable, and highly-scalable email sending and receiving platform for businesses and developers

[object Object]

AWS Amplify

AWS Amplify is a toolchain which includes a robust feature set for simplifying mobile and web application development.

[object Object]

AWS AppSync

AWS AppSync automatically updates the data in web and mobile applications in real time, and updates data for offline users as soon as they reconnect.

AA

AWS Availability Zones

An availability zone is a logical data center in a region available for use by any AWS customer.

[object Object]

AWS CloudFormation

Model and provision all your cloud infrastructure resources

[object Object]

AWS CloudTrail

Record AWS API calls for your account and have log files delivered to you

[object Object]

AWS CodeBuild

Build and test code with continuous scaling.

[object Object]

AWS CodeCommit

AWS CodeCommit is a fully-managed source control service that hosts secure Git-based repositories.

[object Object]

AWS CodeDeploy

Automate code deployments to maintain application uptime

[object Object]

AWS CodePipeline

Continuous delivery service for fast and reliable application updates

[object Object]

AWS Config

AWS Config is a service that enables developers to assess, audit and evaluate the configurations of their AWS resources.

[object Object]

AWS IAM

Securely control access to AWS services and resources for your users

[object Object]

AWS Lambda

AWS Lambda lets you run code without provisioning or managing servers.

AR

AWS Regions

A Region is a geographical area, which consists of two or more Availability Zones, which are synonymous for a Data Center.

[object Object]

AWS Shield

AWS Shield is Managed DDoS Protection for platforms and services built on AWS.

[object Object]

Babel

Babel is a compiler for writing next generation JavaScript

[object Object]

CSS3

CSS3 is the latest evolution of the Cascading Style Sheets language and aims at extending CSS2.1

[object Object]

D3

D3 is a JavaScript visualization library for HTML and SVG

[object Object]

ES6

The next version of JavaScript

[object Object]

Git

Fast, scalable, distributed revision control system

[object Object]

GraphQL

GraphQL is a query language for APIs and a runtime for fulfilling those queries with your existing data.

[object Object]

HTML5

HTML5 is the 5th major revision of the core language of the World Wide Web

[object Object]

JavaScript

JavaScript is the world's most widespread programming language, powering the Internet.

[object Object]

Material Design

Material Design is a visual language that synthesizes the classic principles of good design with the innovation of technology and science.

[object Object]

Material UI

Material UI is a CSS Framework and a Set of React Components that Implement Google's Material Design

[object Object]

Node.js

Node.js is a platform built on Chrome's JavaScript runtime for easily building fast, scalable network applications

[object Object]

React

React is a JavaScript library for building user interfaces.

[object Object]

Serverless Framework

Build web, mobile and IoT applications with serverless architectures using AWS Lambda, Azure Functions, Google CloudFunctions & more

[object Object]

Twilio

Twilio brings voice and messaging to web and mobile applications.

[object Object]

Webpack

Webpack is a module bundler

[object Object]

Yarn

Yarn is a package manager for JavaScript