Axxent Hub is a platform, easy to set up, train on and use in real time in a clinical practice that helps you manage patient volumes, decrease idle time and organize patient dose prescription.

It brings the adherence to policies, compliances and best practices under one roof on a screen throughout a patient’s treatment.


How does it help (Who needs it/ Which Industry does it belong to)

With Axxent Hub all the stakeholders during a treatment cycle can collaborate, manage workflow and increase efficiency by decreasing a chance of error and bringing transparency.

Few Features of Axxent Hub:

  • Source Management
  • Treatment Management
  • Billing
  • MD Notes
  • Reports

Reasons for adopting Cloud

Tides of the time work towards cleaning the old and adopting the new. There was a clear opportunity to try a cloud setup because existing hardware devices at on-premise environment were coming close to end of life cycle. Product warranties were going to expire and replacement costs required larger capital outlay.

The cost factors were such that the data centers and internet service providers required commitment for one year term when the customers of business were committing only for month to month terms.

While the use of application brought about a number of benefits, activities in regards to maintaining data centers, network, systems and operations took the focus away from core of business. Speed at which new servers could be deployed for scaling became a question mark when few key data center employees left the organization. All the above problems led to (a) fear of reduced availability, (b) decreased performance, (c) increased costs with multiple vendor managements.


Self Assessment before Migration

Even though the aforementioned reasons indicate strong inclination towards cloud, it is important to take stock of the current situation before migrating. Also important is the recognition of the cloud born companies competing in the same space have an edge on providing values to the customers over companies that have traditional IT operations.  There are a lot of critical decisions to be made regarding confidentiality, integrity and availability for a setup that acts as a revenue generator for the business itself. In the case of iCad, Axxent Hub is the turbine which cannot stop, compromise security or leave a loophole.

Before closing in on a migration to cloud, a healthcare product must ponder upon the following points and ask themselves whether they are ready or not:

  • Technical Feasibility: Can we actually realize the benefits in operations that we are reading on paper? Will the system move to the cloud without compromising on any feature degrading the application?
  • Customer Experience: Can the whole process of migration keep the promise that no customer would face any issues regarding usage post migration and deliver better in terms of performance?
  • No Impact on Internal Use: If the migration doesn’t foster any process by-pass or unavailability of real-time operations carried out by the internal teams and quality is ensured all along the migration and afterwards.
  • Security and Compliance: Healthcare begins with security and ends with compliance. Migration to Cloud must provide a platform that is compliant and offers durable security to the client’s healthcare data.
  • Cost: Can we save at least 10% on operations cost compared to earlier on-premises model? Can we get 100% away from fixed cost expenditures?
  • Center of Excellence for Cloud: How do we foster an environment wherein the chief innovation officer establishes a center of excellence for cloud that comprises of cloud solution engineers and architects whose expertise can be leveraged for the rest of the organization?


Building a PoC -Test Setup with Real Application and Test Data

After feeling confident about the process, its outcome and due diligence, next step is to create a Proof of Concept and exposing batches of test data to the setup.

We divided the whole process into three progressive steps as explained below, to create a test environment that would function flawlessly with the test data and real application:

Step1: Create Infrastructure on Cloud

With our past experience of migrating critical setups to cloud, we followed our tested practices while creating the required infrastructure, closest to what we were planning to build in real-time. Let’s list the resources we used:

  • Cloud formation templates to create the required IAM Roles and secure Network
  • EC2 Instances with required Packages
  • RDS Instances optimized for the test setup
  • Elastic Load Balancer to direct HTTP/HTTPs traffic to the application node
  • VPN on an EC2 Instance (required to control access to application server)

Step2: Retrieving Data Dumps and Restoring on Test Setup

According to the plan, now we needed the test data from the hub and dump of the database from the original application and restore it on the test server. Though this process doesn’t consume a lot of time, a step wise approach is required to avoid mistakes and repetitions. We broke the process into three parts:

  • Take backup of database on original server and restore the dump on the RDS Instance (we used mysql dump)
  • Copy the PHP application files and configuration files to the test setup to run on the cloud
  • Sync all the security settings like server name, IP and Database connection on both ends with configuration files.


Step3: Verification by Monitoring the test setup operation

This part was easy after deploying the infrastructure. We implemented Cloudwatch on the test setup to monitor critical metrics for the viable functioning and email notifications to get information about the trigger points and thresholds set at various points to record data.

Key Outcomes from the PoC

As we presented this setup to the iCAD team, the whole interaction resulted in building confidence over the technical feasibility and demonstrating the minimal impact on customers and inside team.

We were also able to demonstrate our authority over the domain of compliance and security with our test setup performing up on expectations with the test data and real application.

In the next part of this series, we will talk more about how we discovered the dependencies of this setup in terms of what goes in first and how data flows. We’ll also discuss in details the resources and architecture.