Skip to content


AI in Sepsis Detection: Transforming Healthcare for Early Disease Diagnosis and Prevention

In the ever-evolving healthcare landscape, timely detection of life-threatening conditions is paramount for ensuring positive patient outcomes. The role of Artificial Intelligence (AI) in healthcare is becoming increasingly vital, particularly in the context of early disease detection and clinical decision support. This case study showcases the transformative power of AI, focusing on its application in early sepsis detection at a leading Public Hospital in Dallas, Texas. Let's delve into how AI is redefining  disease diagnosis and prevention

Before understanding Smart on FHIR, let's understand what is FHIR?

FHIR (Fast Healthcare Interoperability Resources), developed by HL7, plays a pivotal role in addressing the critical issue of data interoperability in healthcare systems. The need for FHIR arises from the inherent challenges faced when organizations with different EHR systems or departments utilizing diverse software solutions attempt to communicate and share patient information. Moreover, within a healthcare institution like a hospital, different departments often employ separate computer systems to manage various aspects of patient data, such as appointments, medical records, and billing. When patients transition between these departments, the details are often manually re-entered into the new system, leading to potential data loss and inconsistencies.

FHIR serves as a solution to these interoperability challenges by providing a standardized framework for healthcare data exchange.

How does FHIR ensure interoperability?

FHIR provides a set of data models and API’s for structuring and accessing medical data, and enabling different software applications, systems, and devices to communicate healthcare information in a standardized way.

  • Data Models: FHIR data models, termed as resources, structure healthcare data with defined attributes, relationships, and meanings for various data types like patient personal information, observations, medications, and conditions.

The FHIR Resource list published by HL7 FHIR R5 comprises 157 Resources categorized into different groups encompassing categories including administrative, clinical, financial, infrastructure, and other workflow aspects. Each resource further captures specific attributes, relationships, and metadata related to that piece of information. For instance, the Patient resource captures Patient Identifier, Patient Name, Gender, Contact Information, Active Status, also data type i.e.,. integer, Boolean, date, attachment, etc.


Fig 1: A snapshot of the Resource List

The below Unified Modelling Language (UML diagram) shows the structure, interaction, and relationship for Resource Type: Patient.


Fig 2 Unified Modeling Language Diagram

 For a detailed description and understanding of the attributes under each resource type, refer to Elements in Resource List.

What is Smart on FHIR?

SMART or Substitutable Medical Applications, Reusable Technologies was developed in 2010 with the purpose of building a standard framework that allows the development of “interchangeable healthcare applications”. It works in conjunction with and on top of FHIR hence referred to as SMART on FHIR. A huge focus is placed on the “substitutable” aspect— enabling any developer to create a healthcare application that would work across any healthcare organization, regardless of the type or version of EHR/other healthcare systems. The authentication and authorization aspect being the other pillar for SMART. It functions as a security barrier positioned over FHIR, utilizing OAuth2 and Open ID Connect to provide third-party applications with the necessary authorization to access data from any healthcare system that adheres to the SMART on FHIR standards. OAuth 2.0 operates by issuing access tokens to third-party applications. These access tokens serve as temporary keys, allowing applications to access specific healthcare resources on behalf of the user. OAuth 2.0 also defines permissions, known as "scopes", which specify what data or actions the application is permitted to access. This combination of access tokens and scopes ensures secure and controlled access to healthcare data while safeguarding user privacy. On the other hand, Open ID Connect enhances OAuth 2.0 by adding identity management capabilities. In the context of SMART on FHIR, OpenID Connect plays a critical role in providing not only secure access but also verifying identities, requesting additional user information for enhanced user experience, SSO's reducing the need for multiple logins

A schematic representation of how a Smart on FHIR app uses Oauth2 and OpenID protocol to interact with Authorisation server and FHIR server is depicted below:


Fig 4:Launch Flow from Patient

In simple terms, while FHIR focuses on the structure and format of healthcare data exchange, SMART on FHIR enables the development of third-party applications that can securely connect to EHR systems, providing clinicians and patients with better control over their healthcare information while maintaining data privacy and security.

How SMART builds on top of FHIR

The table below summarizes how SMART builds on top of FHIR:


 Smart on FHIR FHIR





Open ID connect


Data Models


FHIR Resources


SMART profiles


Data Access



Data Format



EHR UI Integration

SMART Launch Specification





EHR UI Integration: Perhaps one of the most crucial features of SMART is enabling integration with the user interface of EHR’s, thus simplifying the process of navigating between applications by bringing them together in one interface. For example, healthcare providers while using different applications would typically have to switch between multiple applications on their devices. This constant switching can disrupt their workflow, require multiple logins and potentially scatter information across tools. However, with SMART, clinicians can access all integrated applications directly within their HER, enter information view data and make informed decisions without needing to leave the EHR environment.

Authorization and Authentication: In SMART on FHIR, OAuth Authorization is employed to provide secure     access to healthcare data, allowing only authorized apps to retrieve patient information, while OpenID Connect is used to securely confirm the identity of users       accessing healthcare data. On the contrary, the focus of FHIR is on organizing and exchanging healthcare data and it does not directly incorporate OAuth Authorization or Open ID Connect for security purposes.

Data Models: SMART on FHIR leverages the predefined FHIR resources (such as patient information, medications, and lab results) when creating data models to structure and organize healthcare information to ensure consistency and interoperability. Instead of reinventing the wheel, SMART on FHIR utilizes the well-defined building blocks provided by the FHIR standard to construct its data models.

Profiles: Within the SMART on FHIR framework, profiles refer to predefined configurations or templates that specify how healthcare data should be structured and exchanged. These profiles ensure that data shared between applications follows standardized guidelines, which is essential for interoperability.

Data Access: For data access, SMART uses the FHIR REST API. This API provides a structured and standardized way for SMART on FHIR to communicate securely with different EHR systems and retrieve the necessary information ensuring interoperability across different systems.

Data Format: SMART on FHIR follows the FHIR standard's guidelines. When SMART on FHIR applications interacts with different systems, it uses either XML (extensible Markup Language) and JSON (JavaScript Object Notation) format to exchange data. These formats are specified by the FHIR standard and provide a consistent way to organize and represent healthcare information.

According to Journal of the American Medical Informatics Association (JAMIA), while the number of unique healthcare applications rose from 600 to 734 in 2023, only 22% of the applications conformed to the FHIR standard. To improve compliance to FHIR, the Lantern tool, developed by the ONC (The Office of the National Coordinator) was set up to help ONC monitor and share insights on the presence and uniformity of FHIR compliant applications. Click here to view the Lantern dashboard

As federal regulations come into play and FHIR gains more traction, the number of applications conforming to SMART on FHIR is expected to increase.


How Finarb Analytics implemented SMART in a web-based application built for a leading Texas hospital for early sepsis detection 

While building your healthcare application, its vital to choose a technology firm that understands the business need behind the application and the nuances of HL7 and SMART guidelines while building the application complying with the necessary HIPPA regulations. Recently, Finarb Analytics collaborated with a leading public hospital in Texas in developing a web-based application for early sepsis detection and prevention.

About the application: The Sepsis Monitoring application seamlessly integrates data from various sources and offers a user-friendly application displaying real-time patient data, vitals, and risk scores. The dashboard provides an overview of patients' current health status, enabling healthcare professionals monitor multiple patients simultaneously. It stores and analyses historical patient data to identify trends and patterns, which helps healthcare professionals gain a deeper understanding of patient conditions, track patient progress in real time and recognize any deteriorating conditions. The application also incorporates an alert system that notifies healthcare professionals in real-time when detecting abnormal values, trends, or combinations of vital signs/laboratory results indicative of potential risk of sepsis

EHR interoperability : SMART on FHIR is designed to be interoperable with various Electronic Health Record (EHR) systems. An EHR platform needs to be selected that supports the SMART on the FHIR standard. The process might involve researching different EHR vendors, their compatibility with SMART on FHIR, and their reputation in the healthcare industry. The EHR should offer APIs that allow your app to access patient data securely and adhere to FHIR standards.
Developing the Application: The step involves the actual development of the SMART on the FHIR app. One needs to use the FHIR resources and SMART App Launch framework to create a web-based application. There is a demand and need for developers who are familiar with FHIR standards and web application development. The SMART on FHIR framework provides libraries and tools that simplify this process and ensures that the app properly handles authentication, authorization, and data exchange using FHIR resources.


If you would like to know more or discuss our use cases in detail