Clinical Collaboration Platform Using Anonymized EHR Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current tools for cancer care, such as multidisciplinary tumor conferences, electronic health records, and patient portals, fail to address issues of poor collaboration among healthcare teams, inadequate patient engagement, and wasteful overtreatment, leading to suboptimal patient outcomes and high financial burdens.
Innovation Solution
A comprehensive platform that includes a web-based data store interfacing with electronic medical records, a mobile clinician application for interdisciplinary collaboration, a patient portal for engagement, and an aggregated decision-driving engine using machine learning and inference engines to optimize treatment planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multidisciplinary tumor conferences are used for cancer care collaboration, then care quality may be improved, but collaboration effectiveness deteriorates due to fragmented team interaction and infrequent meetings
Solution Approach 1:
The platform enables continuous collaboration among cancer care team members through persistent digital workspaces, messaging, and document sharing that are always accessible, replacing the discontinuous weekly conference model with an ongoing collaborative environment where team members can interact in real-time regardless of schedule constraints
Solution Approach 2:
The platform acts as an intermediary digital environment that connects all cancer care team members (physicians, nurses, social workers, etc.) and their patients, providing shared workspaces, communication tools, and coordination mechanisms that facilitate seamless collaboration without requiring physical presence or scheduled meetings
2Productivity
If electronic health records are used for data collection, then documentation efficiency is improved, but collaboration support deteriorates due to data fragmentation and increased cognitive load
Solution Approach 1:
The platform segments the overwhelming EHR data into organized, role-specific workspaces and information displays tailored to each cancer care team member's needs, presenting only relevant patient information and coordination tasks in a structured format that reduces cognitive load while maintaining comprehensive documentation
Solution Approach 2:
The platform serves as an intermediary layer between the fragmented EHR system and human users, aggregating and structuring data from multiple EHR sources into unified patient care workspaces that are easy to navigate and collaborate upon, transforming raw data into actionable clinical information
3Ease of operation
If patient portals are used for patient engagement, then patient access to information is improved, but real collaboration deteriorates due to crowded interfaces and lack of interactive environment
Solution Approach 1:
The platform segments the patient portal into focused, task-oriented interfaces that present information and communication tools in organized sections rather than overwhelming single-page layouts, allowing patients to easily access specific functions such as messaging providers, viewing treatment plans, or scheduling appointments without navigating cluttered interfaces
Solution Approach 2:
The platform transforms the patient portal from a passive information display into an active collaboration intermediary, enabling patients to communicate directly with their cancer care team members through integrated messaging, participate in shared workspaces, and engage in real-time coordination of their care rather than merely viewing static information
4Reliability
If comprehensive data collection is performed in EHRs, then billing and legal protection are improved, but data analysis capability deteriorates due to unstructured format and retrieval difficulty
Solution Approach 1:
The platform extracts relevant clinical data from the comprehensive but unstructured EHR records and pulls out only the information necessary for clinical decision-making and analysis, separating structured analytical data from the full legal and billing documentation while maintaining both functions independently
Data Source
AI summary
Described are web-based clinical data stores; mobile clinician applications, web-based patient portals, and aggregated decision engines comprising: a translation module configured to express aggregated electronic medical records in a format that is human-readable and accessible to an inference engine and/ or a machine learning algorithm; a notation module configured to express relationships between treatment steps to generate human-readable diagrams and to transmit the diagrams as instructions to the inference engine or the machine learning algorithm; an encryption-based decentralized zero-trust de-identification and anonymization module to protect private patient information when releasing records for research purposes outside the treating physician's protected network; an inference engine configured to receive inputs from the translated medical records of a patient and a subset of interlinked treatment steps selected by the treating physician to generate outputs comprising predictions and probabilities; and a machine learning algorithm configured to read and aggregate the anonymized; translated medical records of multiple patients from multiple treating physicians and possibly combine them with one or more tumor registries to test hypotheses.