Start Your Search

J. Schwartz



Author of

  • +

    OS22 - Digital Health Big Data: Promises and Possibilities (ID 30)

    • Event: e-Health 2018 Virtual Meeting
    • Type: Oral Session
    • Track: Clinical Delivery
    • Presentations: 1
    • +

      OS22.03 - Health Data Integration to Enable System-Wide Palliative Care Analytics (ID 506)

      J. Schwartz, Cancer Care Ontario; Toronto/CA

      • Abstract
      • Slides

      Purpose/Objectives: In 2016, a provincial partnership, was formed to provide oversight on improving palliative care across the system. As this program establishes and matures, its data and analytics needs are evolving rapidly. Building a responsive analytics and informatics solution that can grow with the program is paramount. We wish to showcase our innovative approach to building a patient-centred data repository.

      Methodology/Approach: Patients at the end-of-life receive care in many settings within the provincial health care system. We developed a data repository that links routinely collected health administrative data to better understand the patient throughout this journey. The complexity and diversity of this population presents a unique opportunity to design a patient-oriented data integration solution capable of linking patients at all stages in this journey. Our approach leverages learnings from the big data revolution, specifically around the concept of data lakes We implemented a multi-layered analytical data repository (figure 1), that enables analytics across the health system. Population-level data is imported from 13 health administrative data sources and individually cleansed and standardized. Next, concepts on health system utilization, disease identification algorithms, significant health events, treatments/interventions, assessment surveys results, co-morbidity scores and other important health information is defined and collated. Cohorts, such as the decedent and palliative cohorts, are easily derived from pre-implemented algorithms and are easily linked to the derived patient information to define analytical base tables (ABTs). ABTs are the primary data product used to support all measurement. For example, the Decedent – Last Year of Life ABT supports regional reporting, scorecard development and predictive modeling. data repository design.png

      Finding/Results: This data repository design has many benefits: ? Cohorts can be quickly derived from concepts. ? Concepts are persistent, validated and centrally governed. ? Data is housed in one environment, expediting data access and manipulation tasks. ? Centralized data and analytics workflows enables better collaboration between groups ? Data Quality and Metadata is centrally maintained. ? ABTs are the single source of truth; improving consistency, accuracy, time-to-results and encourages exploratory analysis. Leveraging the data repository has enabled: ? Rapid development of the decedent and palliative cohorts. ? Release of current state assessment within 5 months of initial request. ? A tool to provide regional profiles to the regions. ? Ongoing development and implementation of risk-prognostication tools. ? Insights through exploratory analytics and data science.

      Conclusion/Implications/Recommendations: As part of a larger health system organization, we are leveraging many of the ideas developed in this proof-of-concept to modernize our information and analytics systems. The innovation and learnings are being adapted throughout our organization’s information strategy with the goal of becoming more insights-driven.

      140 Character Summary: Data driven decision making – How an innovative new data repository is providing valuable insights into the palliative care needs of a provincial health system

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.