Turning Medical Prescription Data Into Business Intelligence for a leading Healthcare Enterprise

and Managed Cloud Services

About The Customer


The healthcare facility, at the centre of this case story, is one of India’s largest providers of integrated world class healthcare services. It has a network of healthcare facilities across India with 2300+ leading doctors leveraging international level treatment expertise. As well as operating the hospitals, the facility offers health and wellness services at home, including Pharmacy and Pathology Services outside its central hospital network.

Industry

Healthcare Life Sciences

Type of Case Study

Custom App Development

Download Case Study PDF

The Opportunity


For the healthcare industry, the onset of the pandemic has brought about transformational, and possibly permanent, change in the use of telemedicine platforms, accelerating access to virtual consultations. With more visits shifting to an online environment, there has been a huge impact on ensuring all patient information is captured accurately, especially when patients are referred to specialist services outside the doctor’s office.
A top three Indian hospital chain was looking for a way to automatically extract and consolidate all the information regarding the patient’s entire journey, be it number and length of appointments, prescribed medications, diagnostic investigations, lab tests, etc. and accurately map it with its pricing database. Having the ability to automate the extraction and analysis of the complete data would provide critical insights to help support the patients’ continued journey, and overcome its business challenges.

Key Challenges:


This particular healthcare facility processes around 1000 e-prescriptions per day from multiple providers, and its Sales Team had the huge manual effort of calculating and estimating the expenses for each patient. Once a particular ‘fees’ threshold had been reached, this would flag up the need to contact the patient, with a view to offering additional support to continue their treatment journey within the facility. Its core business challenges being reviewed were:

  • Improving productivity: The Sales Team was spending more time on manually extracting and analyzing data than actual patient outreach. For each patient, they would be required to identify the medicines, investigations, and procedures prescribed, and then map them to the pricing database, all of which had the potential for discrepancies or inaccuracies.
  • Manual data analysis: With the manual calculation and decision-making processes, there was a lack of developing a more targeted approach, for example, outreach efforts based on particular medicines, investigations or procedures. Finding a solution that could provide automated data analytics would provide valuable business insights as well as increasing efficiencies.
  • Data extraction: Despite there being a specified form for prescriptions, more often than not healthcare professionals tend to use free text when listing instructions or advice. This requires the Sales Teams to have a comprehensive understanding of medical terminology, as well as commonly used shorthand used by the physicians. Lack of this knowledge within the team was contributing to a decreased overall patient experience.

The Strategy:


The healthcare facility first contacted SourceFuse to develop a solution that would not only read, extract and map patient data with its pricing database, but also enable them to take their first steps towards a cloud-based application. SourceFuse was its partner of choice from the word go, due to SourceFuse’s proven track record within the healthcare industry, leveraging its consulting partnership with AWS Healthcare competencies, and being one the very few companies who use AWS Comprehend Medical to address very specific medical terminology.

In order to overcome the healthcare facility’s key challenges and enable automation, without disrupting any current business flows, SourceFuse proposed building an NLP-driven microservice. The proof of concept (POC), using data from 5000 patients, was completed within a week, and once validated SourceFuse were able to deploy the solution throughout the facility within three weeks.

This speed of implementation, which is almost unheard of in the healthcare industry, was achieved through close collaboration and engagement with the business leaders and healthcare teams, focusing on the client’s needs and objectives, all while providing a bespoke analytical solution and conforming to strict HIPAA and GDPR compliances for protected health information (PHI) or personally identifiable information (PII).

Key Features of the Solution:


  • 100% Automation: The entire repetitive manual process of identifying potential cost associated with a patient is now automated via AI techniques which have helped improve efficiency, provide real-time updates and assist human decision making
  • Cost Breakdown: Apart from identifying total costs associated with a patient, by leveraging machine learning the costs can now be calculated at granular levels like medicine, radiology investigations, pathology investigations, procedures, etc. This helps gather data for further analysis like identifying trends and optimizing outreach
  • Free Text Recognition: The solution is able to efficiently classify a line of text as a valid medicine, an investigation prescription or general advice and then process the entities further. This has meant the inclusion of additional data points that were previously ignored in the manual patient data analysis
  • Cost Optimization: Deploying and running applications generally have significant costs associated with them but this solution is designed in order to have minimum impact on the current infrastructure and the costs are based on the usage
  • Data Security Assured: Since each individual record is identified using patient id, no PHI is shared ensuring patient confidentiality and without the need to build additional layers for data anonymization

To view full technical specifications, click below and download the Case Study PDF.

Conclusion:


By providing a solution that addressed the healthcare facility’s key business challenges, SourceFuse has enabled them to completely automate the data analysis and cost calculation workflows, delivering output reports with more than 91% accuracy.

The solution incorporates AI techniques specifically designed for the healthcare industry, translating medical terminology, separating PHI from non-PHI data, with an API interface that was seamlessly integrated with their electronic health records (EHR). The solution also enabled the facility to encrypt sensitive data prior to sharing with outside partners, for example, when sharing medical information for further specialist consultation.

And lastly, the healthcare facility has been able to pass on additional benefits to enhance the patient’s journey with them: being able to automatically analyze which e-prescriptions require access to a pharmacy means it can automatically link the patient with the facility’s own pharmacy, and therefore offer further discounts to the patient – it’s a win-win for everyone involved.

Gautam Ghai

Finding innovative cloud solutions to business challenges is what we do best, and the speed at which we were able to implement this particular solution was a first of its kind in the healthcare industry. The results are truly remarkable.

Gautam Ghai, Co-CEO and Co-Founder, SourceFuse

Download Case Study PDF