Skip to Main Content

Project Sparkle from Skyflow- Privacy-First, Real-Time APIs for Collecting and Curating COVID-19 Data

Project Sparkle from Skyflow provides a privacy-fist, real-time API to capture, harmonize and share COVID-19 test results and other healthcare information. The proposed solution will be able to share harmonized data with the WATERs system in near real-time. 


The Skyflow health vault, based on the extensible FHIR object model, underpins the proposed solution. Skyflow's technology leverages polymorphic encryption techniques to enable workflows and analytics on encrypted data. Automated data governance with access control, policy enforcement, and provenance tracking enable selective sharing of data while preserving privacy. 


There are three parts to the solution:



  1. Data Capture: We provide easy to use APIs, SDKs, and secure forms for capturing test results, both from labs and from IVD test kits. API integrations to LIMS systems can also be built. Our FHIR based APIs are extensible enough to ingest any medical record information, which can be shared with healthcare professionals with patient consent.




  2. Harmonization: Data from multiple sources are normalized, enriched, and de-duplicated. Data for the same patient from disparate sources can be mapped together.




  3. Data can be shared in a privacy-preserving manner, in real-time, with a variety of constituents, including the WATERS system, federal and state officials, and providers. Standard APIs and SQL interfaces can also be used to connect Analytics/AI tools such as Tableau, BigQuery, and Data Robot.



5 min Video of Capstone Project

Elevator Pitch

Skyflow - Privacy-first API for COVID-19 data - Capture, harmonize and share health information while maintaining patient privacy.

Challenge Goals

The proposed solution greatly eases the burden of data capture, harmonization and transmission by:

  • Providing developer friendly interfaces for data ingestion - mobile SDKs, REST & FHIR APIs, secure forms and SQL.

  • Normalizing, de-duplicating, and mapping data across different sources to the same patient.

  • Providing the ability to collect patient medical records, and sharing it with the relevant users based on patient consent.

  • Privacy-preserving sharing that provides different views of the same data based on need. For example, a FDA/HHS user will not be able to see specific patient information such as contact details, while a county contact tracer will be able to see contact details for any patient testing positive for COVID-19.

Feasibility

The project is scoped into three phases:

Phase 1: 

  • Capture APIs: Provide APIs and SDKs to capture the test information and store in Skyflow. This would be based on the lab & non-lab reporting formats as defined by the CARES act.

  • WATERS Connectivity: Enable near real-time connectivity to WATERS/HHS Protect through APIs/SQL. 

This leverages Skyflow's Vault technology that is already in use across various Fin-tech and Healthcare companies. 

Time horizon: 2-4 weeks.

Phase 2: 

  • Provide the ability to normalize data across multiple sources. 

Time horizon: 6-8 weeks.

Phase 3: 

  • Map data from different data sources to the same patient. 

Time horizon: 8-10 weeks.

 

Design

Protecting PHI data goes beyond compliance. Health information is stored in a highly distributed and available vault with built-in privacy preserving mechanisms - encryption, pseudonymisation, tokenization, masking, and redaction. The healthcare vault is modeled on the FHIR data model, and on the lab/non-lab reporting guidelines mandated by the CARES act. 

Access to this data is governed by a highly granular, dynamic and configurable policy based access control mechanism. The data is made available in real-time through SQL, REST and FHIR APIs

Data connectors for traditional healthcare formats such as HL7 are built on top of Mirth, and executed on the Skyflow platform. 

 

Innovation

Most data sharing solutions present a false choice - one between open exchange of information and one that protects patient privacy. This trade-off has obstructed open and timely access to health data, as has been demonstrated during this pandemic. 

We, at Skyflow, believe that it's possible to responsibly mobilize and monetize health information while preserving patient privacy. Skyflow’s privacy preservation technologies, coupled with patient consent workflows, can vastly improve access to critical health information needed to tackle a pandemic.

 

Flexibility & Scalability

Our solution is based on a highly distributed multi-region multi-cloud infrastructure that can handle 100's of millions of healthcare records. The solution can scale nationwide with minimal effort.

Sustainability & Extensibility

The Skyflow Data Grid is modeled on the extensible FHIR data model, and supports the FHIR APIs natively. This enables us to easily extend our solution to other use cases and public health needs. In addition, we also have the capability to ingest and process HL7 and other standard health information formats.

Team & Collaboration

The team at Skyflow has extensive expertise in the Enterprise SaaS, Big Data and Healthcare technology sectors. The leadership team comprises of serial entrepreneurs who have scaled companies from ideas to large enterprise products. Anshu Sharma, the co-founder & CEO, ran Platform Strategy for Salesforce.com before founding three startups. Prakash Khot, the co-founder & CTO, was previously the CTO of Athena Health. Karthik Rajan, the head of healthcare, previously ran Infrastructure for Salesforce, and established Oracle's Big Data business in the cloud.

Skyflow’s mission of privacy preservation of sensitive data aligns with the goals of this challenge to securely collect, normalize and share protected health data.

 

Additional Comments

What Team(s) contributed to this Capstone Project?

Team Skyflow.

If you are using patient data, are you using real patient data or mock data? Please use MOCK patient data only

MOCK data

edited on Nov 30, 2020 by Karthik Rajan
You will need to login to post a comment

keyur Patel 7 months ago

Are these available on GITHUB for open source usage for vendors?

Reply 0

Karthik Rajan 7 months ago

The APIs and SDKs will be published on GitHub. The bata backend is a SaaS service and is managed/hosted by Skyflow.

Reply 0

Sarah Covington 7 months ago

I love that you are preserving patient privacy whilst monetizing this valuable data. Great business model!

Reply 0

Karthik Rajan 7 months ago

thanks for you comment Sarah. Patient privacy is key, and central to our solution.

Reply 0

Sophia B Liu 7 months ago

This idea has been advanced to the current phase

People's Choice Voting Extended

Reply 0

Sophia B Liu 7 months ago

This idea has been advanced to the next phase

People's Choice Voting Extended

Reply 0

Andrea Pitkus 7 months ago

Trying to understand if there are preconditions met by other software. (In general) 1. How to collect AOEs, from ordering provider/patient/specimen collector 2. Integrate into app/LIS or other information source for patient to be married to results of IVD test device/system 3. All transmitted to public health (ELR) 4. All transmitted to HHS (may be met by 3).

Will your approach support all 4 or only portions of above for "tracking lab results"?

Do you have app, webportal or EHR based data collection and reporting? How do you handle LISs without APIs or FHIR support (as not aware of any yet)? How about patient performed testing?

Do you support LIS based reporting of ELR in HL7 v2.51 (per the MU IG)? For patient performed results are they routed to the physician who is required to report via electronic Case Reporting (eCR) by law?

How/where are LOINC, SNOMED CT and other codes systems supported in your messaging to meet requirements?

How are CLIA testing requirements met in your system?

Reply 0

Share