SickKids is seeking a solution to leverage and enhance its current machine learning surgical scheduling model to improve surgical efficiency and throughput through a graphic user interface (GUI). 

SickKids is posting this Call for Innovation to seek out qualified Ontario* companies who can meet the desired outcomes.  SickKids and CAN Health reserves the right to not move forward with this project at its full discretion and in particular if there are no qualified Ontario companies that can reasonably meet the desired outcomes.

*Business must be registered in Ontario in order to qualify for this project.

This opportunity is closed.
Problem Statement and Objective(s)

The current surgical booking process at SickKids is a manual and siloed process with no overarching system to ensure optimal surgical scheduling and use of operating room (OR) time. OR throughput significantly effects patient flow across the continuum of care through the OR, the Post-Anesthesia Care Unit (PACU) and to post-operative destinations on the hospital inpatient units or day surgeries that are discharged home. Enhancing surgical efficiency and decreasing case fluctuations across the week (bottle necks) by facilitating and enhancing surgical patient flow has the potential to minimize case cancellations, decrease last-minute care planning, and overall reduce the length of the surgical waitlist without overwhelming hospital resources. Efficiencies in OR throughput leads to optimization in patient care and a solution to aid in surgical scheduling decision making, especially with the significant challenges at this time given the surgical waitlist backlog exacerbated by COVID-19.

SickKids is seeking to leverage their new machine learning surgical schedule model to assess, predict and optimize case bookings to maximize available OR time, inpatient beds, and post-operative care resources that would enhance surgical efficiency and throughput by facilitating surgical patient flow through a graphic user interface (GUI) surgical scheduling platform.

Objective:

  1. GUI implementation to enhance decision-making related to surgical scheduling, that connects with and displays:
    1. SickKids surgical scheduling machine learning based model;
    2. Optimized surgeon schedules;
    3. Predictions for ramp up and ramp down requirements of surgical activities.
Desired outcomes and considerations

Essential (mandatory) outcomes

The proposed solution must:

  • Be compatible with data platforms (e.g. Azure, Google Health and Epic) to run machine learning model (coded in R).
  • Create an easy to use interface to be used by clinical decision makers including:
    • Clinical Leadership Management (OR, PACU, Inpatient Units);
    • Surgeons and Surgical Division Leaders;
    • Surgical administration.
  • The GUI must aid in:
    • Identifying areas for improvement in OR scheduling including but not limited to
      • enhancing OR smoothing;
      • matching OR throughput with availability of post-operative resources both within the PACU and inpatient units.
    • hospital patient flow;
    • OR activity ramp-up or ramp-down strategies when required.
  • The GUI must:
    • meet all SickKids security requirements (use of PHI);
    • be used with our SickKids Microsoft active directory (leverage the SickKids standards for single sign on);
    • have capability to be built within SickKids existing databases and infrastructure;
    • have export capabilities (e.g. excel or csv);
    • generate figures and graphics visuals;
    • be filterable (multiple views and combinations);
    • allow for assessment of utilization rates by staff type/name.

Additional outcomes

  • In addition to creating a graphic user interface (GUI), the solution would be useful for development of case scheduling and review of surgical cases across the province, as well as customizable features that could be used in institutions across Ontario/Canada. It would be considered beneficial if the solution also had the option to utilize its own surgical scheduling platform algorithm should SickKids’ algorithm not be available or applicable.

The maximum duration for a project resulting from this Challenge is: 6 months

Background and context

The Canadian healthcare system is currently running operations in silos. SickKids is looking to break down these siloes to deal with the significant problem of the surgical backlog exacerbated by COVID-19 and its effect on post-operative resources and hospital patient flow. SickKids current Surgical Schedule Modelling includes data points that add a level of prediction and assessment to increase OR throughput and improve OR smoothing and hospital flow through post-operative destinations without overwhelming hospital inpatient resources. SickKids is seeking to enhance this modelling by adding a surgical scheduling platform (GUI) connected to the machine learning model that will allow clinical leaders, schedulers, surgeons and other key individuals to make informed, data-driven decisions with regards to OR block allocation, surgeon scheduling, post-operative destination availability and other factors pertaining to OR smoothing, throughput and patient flow.

This opportunity is closed.