Redox positions itself as a vendor-agnostic interoperability layer for healthcare organizations, health tech vendors, payers, EHRs, and life sciences teams. Its developer documentation describes a platform that converts data payloads into a consistent format so systems can integrate with one another. The company's marketing materials state that Redox handles connections, data transformations, and upkeep, while technical documentation outlines the protocols, methods, and systems the platform supports. [1][2][3]
Supported formats and communication methods
Redox's implementation guide lists multiple data formats the platform supports, including binary files, C-CDA, CSV, DICOM, FHIR, HL7v2, JSON, multipart form data, URL-encoded form data, X12, and XML. The documentation states that users can employ one or a combination of these formats depending on workflow needs. [1]
For transmitting data, the documentation lists communication methods such as CDS hooks, cloud storage, clinical networks, DICOMweb, Direct Secure Messaging, FHIR services, HL7v2 over HTTPS, HTTPS REST and SOAP, IHE profiles, MLLP, polling, webhooks, PHINMS, SFTP, and SMART on FHIR. Redox states that regardless of method and format, it converts all data payloads with the Data Model API or FHIR API so that one system can integrate with any other. [1]
Authentication and system APIs
The implementation guide documents authentication schemes that Redox supports to authorize access to partner systems. These include API key, AWS Signature Version 4, Basic auth, Basic with client certificate, Google Workload Identity Federation, JWT Bearer, mutual TLS, OAuth 2.0 two-legged, SMART Backend, three-legged OAuth, and TLS client certificate. [1]
Redox also lists EHR and other system APIs it has connected with, including Epic, Cerner, athenaHealth, eClinicalWorks, NextGen, PointClickCare, and Veradigm. The documentation cautions that scope and support may vary, meaning some workflows may be supported while others are not yet available. It directs readers to contact a Redox representative to confirm what works for a specific use case. [1]
Data Model API and field reliability
The Redox Data Model API documentation lists numerous data models, including Claim, Clinical Summary, Device, Financial, Flowsheet, Inventory, Media, Medications, Notes, Order, Organization, PatientAdmin, PatientSearch, Provider, Referral, Results, Scheduling, and SSO. Outbound Data Model API requests are sent to api.redoxengine.com/endpoint. [7]
The documentation explains that some data model fields can only contain a limited set of values called a value set. It also notes that Redox may not have access to certain fields depending on the connection's system, so some fields may not be populated. Each field in the data model schemas carries a reliability rating, and the Managed Services team works with customers to document data integration needs or plan for customization. [7]
Connections, subscriptions, and data flow
A Redox connection is defined in the documentation as a set of subscriptions, or integrations, with a connected organization that agrees to exchange healthcare data through Redox. Connections represent a business relationship, enable data flow between systems, and simplify integration because multiple subscriptions can be built for one connected organization. [6]
Each subscription must have one source and one destination. Redox states that users can typically use the same API source and destination to exchange data with all connections, though unique sources and destinations may be needed for each communication method. The documentation also describes one-party subscriptions, where an organization owns both source and destination, as useful for populating cloud repositories. Redox labels this a closed beta feature for a limited pilot group by invitation only and states that it is not currently accepting additional participants. [6]
Data exchange patterns
Redox's data exchange documentation describes three traffic types: asynchronous notifications that push data without waiting for a response, synchronous queries that request information and wait for an immediate response, and writebacks that save or push data to a connection's system. The documentation notes that polling can serve as an alternative when a connection does not support traditional event notifications. [8]
The FHIR API documentation distinguishes between lightweight notifications, which contain patient identifiers, a timestamp, and FHIR references for follow-up queries, and detailed notifications, which also include a full payload describing what happened, who was involved, and when. For writebacks, Redox states that most EHR systems do not support FHIR writeback, but the platform can translate writeback messages to standards such as CDA, HL7v2, X12, DICOM, and 360X. [5]
Implementation planning and timelines
Redox's implementation guide advises teams to review workflows in both systems, decide on an integration type, and confirm how the connection plans to handle interfaces before implementation begins. The documentation states that the average time from kickoff to go-live is about 6 to 10 weeks, depending on how quickly both parties implement and how many data exchange methods or customizations are involved. [4]
Common implementation issues listed in the documentation include process delays, competing projects, limited integration analyst experience, scope adjustments, requirements for new integration feeds, setup with a new EHR system, and new development work at Redox. Recommended practices include sharing timelines early, holding weekly check-ins, deciding on integration methods ahead of time, and keeping product and engineering teams engaged throughout the first implementation project. [4]