Optimal Scheduling to Support Care Plans

Project Code: 1080

Faculty: Faculty of Engineering

Department: Engineering Science

Main Supervisor: Dr Michael O'Sullivan

Co supervisors: Dr Cameron Walker

Application open date: 14 Apr 2014

Application deadline: 26 Sep 2014

Enrolment information: NZ Citizens, NZ Permanent Residents, International

Introduction

The healthcare system is increasingly being asked to do more with less funding. Health planners, managers and analysts are constantly looking for where resource optimization could be realized. One of the largest areas where optimization could show significant utility and productivity is maximizing the patient’s planned care journey. Many health systems create the plan of care and operationalize the scheduling of this care using only a few of the factors that influence consistent implementation success resulting in good quality care and little waste. The failures are often seen in isolation, such as lack of patient compliance to appointments or equipment gaps (e.g. theatre room unavailable) or skill sets missing (e.g. clinician on leave). However, factors such as physical environment, human resource, evidence based care protocols, national/regional targets (i.e. minimum treatment times or wait times), cultural and social factors all influence the realization of the optimum planned care journey. By looking at as many of these factors together as possible and applying advanced modelling techniques and integer programming, these new tools could support improved planning for health organisations and better outcomes for patients.

 

What we are looking for in a successful applicant

Knowledge of optimisation, particularly scheduling. Experience with programming, particularly Java.

Objective
  1. Identification of Potential Modelling Methods for Care Plans - Identify clinical partner to prototype tool. Important considerations will be the clinical need for better care plan support and suitable historical data to benchmark solutions against. Investigate complexity of the scheduling problem to help identify modelling approach required (for example integer programming with column enumeration may be possible for smaller scope problem, or column generation may be required if the scope and combinatorial complexity is large).
  2. Prototype Models for Care Plans - Build initial model for scheduler. Initial focus will be on generating a best schedule, but consideration will also be given to customisation of the solution process to ensure solutions are generated in an acceptable timeframe. This consideration will also influence the model development, as a working system will regularly update the current schedule with new appointments, so “hot-starting” from the current schedule will be a requirement.
  3. Testing of Care Plan Models - Significant interaction with the clinical partner will be essential as the tool is tuned to accurately reflect their system, and also to ensure it can be seamlessly integrated into their systems. Cost benefit analysis will be carried out to determine the value added as a result of optimizing the scheduling of resources and appointments.
  4. Implementation/Integration of Care Plan Models - Working closely with Orion Health the optimization tool will be integrated into existing software.
Other information
Related opportunities