Applied Operations Research

In today's rapidly changing health care environment, no solution can last forever — continuous improvement must be the norm. Proactive problem-solving allows health care delivery experts to identify potential problems early and find solutions quickly. The Mayo Clinic Kern Center for the Science of Health Care Delivery's Applied Operations Research Program uses operations research, systems engineering and data science principles. Its aim is to identify inefficiencies, inform decision-making, improve resource allocation and utilization, and in general, identify optimal solutions.

The Applied Operations Research Program develops and applies industrial and management engineering principles and artificial intelligence technologies to tackle complex problems in health care. The team continuously improves systems efficiency, health care operations and staff wellness across Mayo Clinic. This leads to enhanced patient experiences through improvements in access, reduced cost of care, and innovations in quality and safety.

The program draws on the knowledge and expertise of a team of scientists with diverse backgrounds in areas including:

  • Optimization modeling.
  • Solution algorithm development.
  • Heuristic modeling.
  • Statistical modeling.
  • Mathematical modeling.
  • Simulation modeling.
  • Machine learning.

Together with clinical collaborators across Mayo Clinic, the program team delivers evidence-based solutions for practice transformation.

Focus areas

The Applied Operations Research Program focuses on redesigning outpatient scheduling and optimizing use of staff and space to improve patient access to care, treatment quality and financial performance. The program's expertise and responsibilities are:

  • Applying operations research theories and techniques.
  • Optimization modeling.
  • Developing solution algorithms.
  • Heuristic modeling.
  • Statistical modeling.
  • Mathematical modeling.
  • Machine learning.
  • Simulation modeling.
  • Helping the practice define operational issues and concerns.
  • Defining project elements and resource needs.
  • Extracting and manipulating data to report outcomes.
  • Developing data-driven optimal solutions and executing these solutions.
  • Creating tools, interfaces and software as needed.
  • Monitoring outcomes and adjusting accordingly.


Outpatient scheduling systems

The Applied Operations Research Program team is redesigning outpatient scheduling systems at Mayo Clinic by developing optimal scheduling templates.

This work seeks to:

  • Minimize wait time for patients.
  • Minimize idle time and overtime for clinicians.
  • Reduce costs and increase resource stewardship.

While developing optimal scheduling policies — including overbook-override, cancellations and urgent care needs — the project team accounts for uncertainties such as patients missing appointments without notice or other appointment mix-ups. The team also is determining optimal appointment duration for various patient groups according to patient characteristics and conditions.

An example application of this research is a new chemotherapy scheduling template design in place at all chemotherapy units across Mayo Clinic.

Chemotherapy unit patient and staffing levels before and after implementing optimal scheduling template.

Related publications:

Real-time scheduling decision tool

This research involves developing a patient callback control system that minimizes treatment resource conflict to improve service quality. The system also employs a search algorithm that minimizes patient procedure lead time by finding the most appropriate member of the health care team based on availability.

Example projects include developing proton beam therapy gatekeeper logic. This includes simulation validation and a search algorithm for optimal pairing of cardiovascular interventionalists and surgeons to perform transcatheter aortic valve replacement (TAVR) procedures.

Optimal cardiovascular interventionalist and surgeon pairing algorithm for TAVR procedure.

Related publications:

Predictive modeling

The Applied Operations Research Program team uses predictive analytics, such as statistical and machine-learning methods. The goal is to inform decision-making on staffing, schedule planning and triage appropriateness for scheduling — that is, to get the right patient to the right health care professional at the right time. This work is often part of scheduling and staffing optimization.

Example projects include:

  • Staffing cardiovascular ambulatory care nurses for clinical and nonclinical work time.
  • Predicting optimal days for bone marrow transplant stem cell collection, which allows level-loading of apheresis capacity.
  • Classifying the urgency of patients' needs for better scheduling in echocardiography laboratories.
  • Enhancing triage decision-making for pain management consultations and procedures.

Graphic representation of the "ease ensemble with adaptive boosting" machine-learning method for patient triage to pain medicine.

Time allotment and patient assignment for health care teams

The program team has developed an algorithm to track the time allotment of health care team members' assignments. This tool helps ensure that they can fulfill their scheduling commitments. It determines the optimal care team panel size, patient assignment and care team structure. The algorithm helps balance health care team workload, reduces team member burnout and improves the continuity of each patient's care.

Example projects in this area include:

  • Developing and implementing a time allotment tracer for cardiovascular medicine clinicians.
  • Redesigning family medicine health care professional panel sizes and care teams.
  • Using a simulation approach for staffing in the emergency department.

Simulation framework to evaluate health care team staffing levels in the emergency department.