Case Study Banner

CASE STUDY

A global pharma company achieved uninterrupted clinical risk monitoring using custom built data aggregator delivered in 5 months

industry-iconCLIENT :Confidential
industry-iconINDUSTRY :Pharmaceutical
industry-iconDURATION :4 months
CLIENT :
Confidential
INDUSTRY :
Pharmaceutical
DURATION :
4 months

Business case

A global pharma client needed a data solution built within a quick turnaround time to ensure uninterrupted data flow to their central clinical risk monitoring team.

Previously, all the data from the various sites and the CROs was stored in a legacy data warehouse. An in-house data aggregator tool pulled the data from the data lake and uploaded it to the third-party SaaS tool, which was used by central monitoring team for their periodic checks. 

The data aggregator tool pulled the data from the data lake, grouped them into sas7bdat files - each representing one data set for a particular study. This zip file was then uploaded to the third-party SaaS tool for further analysis of the monitoring team. 

The company wanted to retire the legacy data lake to move to a unified workspace which will give them consistent and reliable insights across the organization. The new central data lake was built on top of the Data bricks platform which made the previous data aggregation tool incompatible. For certain datasets, this tool also required manual collation for some data requested by the monitoring team that fell outside the purview of the data available in the legacy data lake. 

 The client was looking to build a new and upgraded data aggregation tool, while also exploring avenues to reduce manual interventions still being needed by the older tool.

They were looking for a technology partner who not only understands technology but also understands the clinical trials data management ecosystem. They were looking for a quick turnaround time, as the deprecation of the legacy data warehouse leaves the central monitoring team with no data to review for the ongoing clinical trials. 

Our solution

Our data experts examined the current data systems being used, and after a thorough analysis of the client’s requirements they recommended the following approach. 

  • i2e built a customized data aggregator tool to pull the data from the central data lake and perform seamless data aggregation, packaging and integration, all the while following the rules and parameters set by the central monitoring team.
  • The tool was the client’s first ever internally validated tool which was built in a record time of 5 months against the initial estimated 12 months' time.  Built a streamlined testing cycle which allowed for 10-15 iterations in just a week, significantly improving the response time.
  • The previous data aggregator tool used Spotfire to build the front-end UI, which was not ideal and required an additional licensing cost for each access control. The team-built a custom web-based tool to build an intuitive form for the team to punch in requests.
  • Since there was a vision to move completely to the new data lake being built on Databricks – the team was also looking to move beyond SAS programming to Python

  • It also minimized the manual work needed by the team whenever any additional information outside the purview of the zip files is requested. – with plans to completely eliminate it in the upcoming releases. 

     

Challenges overcome

  • The project was conceptualized, architected and delivered by directly collaborating with the business users - without the direct involvement of the client’s digital team
  • Understanding the intricate data management framework of the client to narrow down the tech stack that works.
  • Understanding complex data mapping and building a validated tool within the limited time. 

Benefits

  • Availability of an upgraded data aggregator tool with a quick turnaround time of 5 months.
  • A validated tool in compliance with the client’s central validation team.
  • Average time needed to generate the zip file was brought down from 60 minutes to 30 minutes (subjected to the volume of data).
  • Reduction of any manual interventions in the data management process making it streamlined and efficient. 

Results

Results