Connectify.AI - Oncobase

Connectify.ai is a specialized platform that supports oncologists by integrating AI into cancer care management. The OncoBase software centralizes crucial clinical data, research, and patient information, helping healthcare professionals make informed decisions in treatment planning. By leveraging AI-driven insights and data analytics, OncoBase aims to streamline workflows and improve the precision of personalized care. In this project, our team focused on redesigning both the software platform and user experience (UX) of the OncoBase modules, ensuring that the system is intuitive, efficient, and tailored to the needs of oncologists.

Connectify.AI - OncoBase Redesign: Optimizing User Experience for Enhanced Cancer Care Solutions

Goal & Timeline

The objective of the OncoBase redesign was to create a more intuitive, clinician-centered platform that streamlined workflows and addressed critical pain points. Through in-depth user research, we identified key areas where the platform could better support clinicians in managing cancer treatment data. Based on this research, we refined the information architecture, introduced tailored functionalities, and developed a user-friendly interface that integrates seamlessly with existing clinical workflows. The prototype was iteratively tested to ensure it effectively meets the needs of its users.

The Oncology Landscape

Rapid Advancements in Oncology

Ongoing evolution in cancer research, including new diagnostics, treatments, drugs and clinical trials

1

2

Dynamic Clinical Protocols

Guidelines are continuously reviewed and updated to align with best practices and evidence-based care

3

Emerging Technologies

Integration of artificial intelligence (AI), genomics and precision medicine is reshaping cancer care

Phrase 1 - Discover

User Research

Market Analysis

Ideate Concepts

Phrase 1 - Define

Research Synthesis

Information Architecture

User Personas

User Flow

Phrase 2 - Design

Style Guide

Wireframes

Prototypes

Phrase 3 - Deliver

Usability Testing

Design Iterations

Project Report

Design Handoff

Market Analysis

Core Functionalities: PathAI focuses on enhancing diagnostic accuracy in pathology using AI-powered image analysis. It offers automated solutions for detecting diseases such as cancer with high precision, assisting pathologists by providing faster and more consistent results. The platform is designed to support both clinical diagnosis and research, enabling efficient workflow integration for pathology labs.

Ease of Navigation: PathAI provides a user-friendly interface that allows pathologists to navigate easily through digital slides and diagnostic results. The streamlined layout ensures that users can access critical information, such as image analysis results and reports, with minimal effort, supporting faster decision-making in clinical workflows.

Information Access: The platform excels in providing centralized access to diagnostic information, including AI-driven insights and pathology data. Its intuitive structure consolidates relevant data in one place, reducing the need for users to switch between multiple screens or tools, making it easier for pathologists to focus on diagnostics and research.

Interface Design: PathAI offers a modern and clean interface, designed with simplicity and functionality in mind. The visual design is intuitive, with AI-generated annotations and analysis clearly highlighted, allowing pathologists to quickly interpret results without being overwhelmed by unnecessary complexity. This enhances the overall user experience by offering clarity and ease of use in a clinical setting.

We comprehensively analyzed leading AI-powered oncology platforms to benchmark OncoBase against its competitors. These comparisons helped guide our redesign of OncoBase, emphasizing intuitive data management, advanced AI capabilities, and seamless integration with clinical workflows to provide an optimized user experience. Here's a comparison of key industry players:

Core Functionalities: Artera focuses on leveraging AI to enhance oncology diagnostics and treatment planning. The platform integrates AI-powered predictive models to analyze patient data and assist in making personalized cancer treatment recommendations. It is designed to help oncologists streamline clinical decision-making by offering data-driven insights, enabling more precise treatment strategies for cancer patients.

Ease of Navigation: Artera provides a clean and organized user interface that allows oncologists to quickly access patient information, predictive models, and treatment recommendations. The platform is structured to support seamless navigation, ensuring that clinicians can efficiently retrieve data and analyze results without unnecessary complexity, helping them make faster, well-informed decisions.

Information Access: Artera excels in presenting complex clinical data in an easy-to-digest format. By centralizing information such as patient history, treatment plans, and predictive outcomes, the platform simplifies the decision-making process. Clinicians can access all relevant data in one location, minimizing the need to switch between different tools or interfaces, which improves workflow efficiency.

Interface Design: The interface design of Artera is intuitive and modern, with a strong focus on usability. Visual elements, such as AI-driven predictions and patient data summaries, are clearly displayed, helping oncologists quickly interpret key information. The interface emphasizes clarity and simplicity, reducing the cognitive load on users and enabling more accurate and timely clinical decisions.

Core Functionalities: Allscripts Sunrise is a comprehensive EHR/EMR system designed to support patient care through a variety of modules. The platform provides clinicians with access to patient summaries, pending lab orders, investigation results, clinical documentation, and more. It is equipped to handle various aspects of patient care, from diagnostics to treatment planning, helping healthcare providers streamline workflows and ensure efficient care management.

Ease of Navigation: The Allscripts Sunrise interface is organized into modules, allowing users to access different aspects of patient care quickly. The patient summary is prominently displayed at the top of the page, ensuring that critical information is easy to locate. However, the system's complexity can make navigation less intuitive for new users, with multiple clicks often required to access specific data points.

Information Access: Information is well-structured within Allscripts Sunrise, with data consolidated into relevant modules for quick access. Clinicians can easily retrieve patient records, lab results, and documentation, which supports decision-making and reduces the time spent searching for information. The system's modular approach enables a more efficient way of managing patient data compared to traditional EHR/EMR platforms.

Interface Design: While functional, the interface design of Allscripts Sunrise can feel somewhat dated, with a focus on functionality over aesthetics. The layout is clutter-free, but there is room for modernizing the visual elements to improve the overall user experience. Despite this, the system remains highly effective at providing clinicians with the information they need in a clear and structured manner.

Our primary user research revealed significant challenges faced by oncologists in keeping up with advancements. One major issue is the overwhelming volume of information available online. Oncologists frequently research cancer journals to stay updated, but filtering through this vast data to find relevant content can be both time-consuming and inefficient. Much of the information may not directly apply to their current practice. Despite our efforts to secure interviews with multiple oncologists, we only managed to speak with two. Their feedback highlighted how the sheer volume of data makes it challenging to identify and integrate useful insights into their practice.

User Research - Primary User Research

Challenges in Keeping Up with Oncology Advancements

Oncologists often conduct online research for relevant cancer journals but struggle to filter through vast amounts of information that may not be directly applicable to their practice.

1

2

Constraints Imposed by Established Clinical Protocols and Guidelines

Integrating new knowledge into daily practice is challenging due to time constraints and the relevance of research to current cases.

3

Constraints Imposed by Established Clinical Protocols and Guidelines

Oncologists rely on established protocols, limiting personalized treatment options, as prescribing is restricted to approved cancer drugs and insurance coverage further narrows drug availability.

4

Skepticism with AI

Oncologists are cautious about fully relying on AI, viewing it more as a supplementary tool, particularly useful for trainee doctors as a secondary resource.

As we have very limited primary data from interviews with our actual target group, we have supplemented our findings with secondary research through literature reviews. Our secondary research highlights a significant challenge faced by oncologists: keeping up with the rapid advancements in cancer research and treatment protocols. The sheer volume of new research, combined with their demanding workloads, makes it difficult for oncologists to stay current. This overwhelming influx of information adds to the complexity of integrating new knowledge into their practice, emphasizing the need for more efficient methods to manage and apply the latest advancements.

User Research - Secondary User Research

Difficulty in Keeping Up with Oncology Advancements

Oncologists struggle to keep up with rapidly advancing cancer research and treatment protocols due to the overwhelming volume of research and their demanding workloads.

1

2

Concerns about Reliability and Accuracy of AI Tools

Many oncologists are concerned about the accuracy of AI-powered tools, fearing they may not always align with best practices or current guidelines, leading to hesitation in fully trusting them for complex cancer care decisions.

Design Insights

The primary goal of OncoBase is to provide clinicians with reliable, up-to-date oncology information, helping them save time and focus on patient care. However, clinicians are often skeptical of AI tools due to concerns about reliability and accuracy. To address this, OncoBase's design must ensure AI-generated responses are credible and backed by relevant research, building trust and confidence in its use.

OncoBase aims to create an intuitive, user-friendly interface that requires minimal training, encouraging adoption among new users. Clinicians are hesitant to adopt new tools if they are too complex or disrupt existing workflows. Therefore, OncoBase must integrate seamlessly with existing EHR/EMR systems to ensure a smooth flow of information and enhance overall usability.

To increase OncoBase's market success, we prioritized addressing clinicians' most critical needs. Oncologists were satisfied with their current MDT systems, leading us to exclude the OncoMDT module from the prototype. Instead, we focused on refining the core modules: OncoChat, OncoDx, and OncoRx, which target the essential pain points of oncologists and specialists, enhancing OncoBase's relevance and potential impact.

User Persona

The team analyzed user interview data to develop detailed personas for the OncoBase redesign, representing distinct clinician segments such as oncologists, radiologists, and specialists. Each persona highlights unique needs and challenges, accompanied by problem statements, "How Might We" questions, and solution statements to guide design decisions and improve the user experience.

Information Architecture

The Information Architecture of OncoBase integrates its key modules—OncoChat, OncoDx, and OncoRx—designed to streamline oncology workflows and enhance patient care.

Design System

The initial design system was rather basic and lacked character. We introduced more icons and some vibrant illustrations at key touchpoints to make the interface more engaging. We also revamped the logo, transforming it from a plain text format into a sleek, elegant, and distinctive illustration.

Typeface

Icons

View the Oncobase prototype here!

Hi-Fi Prototype

User Testing

We conducted usability testing with 7 family physicians, we chose family physicians due to challenges securing oncologists for this phase. These physicians are familiar with cancer screening and diagnosis through their work. The tests were conducted remotely over Zoom as well as in-person. These tests aimed to evaluate the ease of use and functionality of OncoBase’s core features.

The physicians offered feedback based on their respective specialties, which may not entirely reflect the specific needs of oncology professionals. However, their insights provide valuable considerations for future development. In the next phase, we can explore implementing some of these suggestions and testing them with oncologists to ensure the tool is properly tailored to its target audience.

For now, our main priority is creating an intuitive and functional design that highlights OncoBase’s essential features. Based on our observations and the feedback from the usability testing sessions, we have identified several key areas for improvement.

Our System Usability Scale (SUS) scores averaged 78.6, indicating a generally positive response to OncoBase’s usability. Most clinicians completed their tasks within the allotted time, but a few struggled with specific actions. For example, one user skipped the query input step in OncoChat, and was confused when adding a note and collapsing the sidebar. These observations revealed areas where the interface could be refined to better align with clinician workflows.

Changes After Usability Testing

01

Finding: Users would like to be able to edit the results generated on the OncoDx and OncoRx modules before publishing it to the EHR system.

Design Iteration: Introduced an "Edit" button next to the generated results. This would provide a clear and intuitive option for oncologists to modify the information before submission.

02

Finding: One participant initially mistook the “+” icon (used to add documents) as the button to send messages in OncoChat, causing hesitation before finding the correct send button.


Design Iteration: Replace the “+” icon with a paperclip icon, which is more universally recognized for attaching documents, thereby reducing confusion and improving task flow.

03

Finding: Users had difficulty locating drafted diagnoses on the dashboard, as the cases were labeled only by case numbers (e.g., LC-001). Given the large volume of patients, oncologists are less likely to remember case numbers.



Design Iteration: Display the patient's name alongside the case number on the dashboard. This will allow oncologists to more easily identify their cases and avoid unnecessary delays.

04

Finding: Users were confused by the presence of multiple search icons—one within individual chats and another at the list of chats. When tasked with searching for the word “Treatment,” some mistakenly clicked the search icon inside the chat window.

Design Iteration: Remove the search icon within individual chats and retain the one at the list of chats. This will streamline the search functionality, allowing users to search both chat titles and their contents from a single location.

05

Finding: Physicians preferred clickable links to the sources, as manually searching for references would be too time-consuming.



Design Iteration: Add direct links to all references. This would enhance efficiency by allowing users to quickly verify sources without needing to perform additional searches.

Future Opportunities

This early prototype of OncoBase highlights key areas for future development. One critical opportunity is enhancing the depth of patient and clinical information in OncoDx and OncoRx, aligning them with the comprehensive data typically found in EHR/EMR systems. Close collaboration with oncologists is essential to ensure clinical relevance and accuracy in these modules. Additionally, feedback from usability testing suggests that OncoDx and OncoRx could potentially be merged, as overlapping information between diagnostic and treatment planning exists, especially in cases of cancer recurrence.

Clinicians emphasized the need for seamless integration between OncoBase and their current EHR/EMR systems, allowing for customization and smooth data flow. They also stressed the importance of incorporating inputs from other specialties, such as psychiatry and rehabilitation, to provide holistic patient care. Finally, some clinicians recommended adding a news feed on the dashboard to keep users informed of relevant oncology advancements, supporting OncoBase's goal of keeping specialists up-to-date with the latest research and clinical protocols.

Conclusion

My experience while leading a team in the redesign of OncoBase. While we initially faced challenges due to differing work styles, these differences ultimately proved beneficial by introducing fresh ideas and diverse perspectives into the project.

Through open communication and mutual respect, we successfully navigated these challenges, demonstrating the power of teamwork in achieving meaningful outcomes. This project, although a prototype, highlighted the immense responsibility of designing a platform that would directly impact clinicians' workflows and improve patient care. Recognizing the need for seamless integration with existing systems and addressing clinicians' most pressing needs pushed us to refine every aspect of our design.

Working on the OncoBase redesign not only allowed me to apply my UX skills but also gave me the opportunity to contribute to the future of oncology care. Although this prototype is just an early stage, it’s exciting to see the potential impact it can have on improving how clinicians stay updated on oncology research and manage patient treatment plans. Collaborating with my team on this project was incredibly rewarding and reinforced my passion for creating tools that make a positive societal impact through thoughtful design and innovation.