REAL-TIME SCHEDULING FOR OUTPATIENT CLINICS
Many outpatient clinics experience variation in their utilization with surge times and down times. We provide a real-time scheduling algorithm to better schedule the patients and improve utilization, patient wait-time, and staff satisfaction.
EFFECTS OF STAFFING PATTERNS ON PATIENT CARE
For patients who are hospitalized, it is often the case that their clinical care team changes, sometimes entirely, during their stay. An interesting question is to understand whether or not such changes in the care team affect patients’ length-of-stay in hospital.
LEVEL-LOADING IN PRIMARY CARE CLINICS
Primary care is the gateway to the healthcare system for many patients. Although primary care itself only accounts for 5% of healthcare spending, the decisions made in primary care setting influences the subsequent medical care including subspecialty referrals, imaging, medical testing, invasive procedures, and hospitalization. We use optimization and data analytics to improve the level-loading for clinical staff in primary care clinics which in turn improves patient flow, wait time, and patient care.
IMPROVING OPERATING ROOMS SCHEDULING UNDER UNCERTAINTY
Operating rooms are one of the main resources in a hospital and any disruption in their workflow can have a cascade effect on the rest of the hospital operations. We use machine-learning, data analytics, and optimization to find a stable schedule to maximize utilization while reducing the risk of operational disruptions.
AN INDUSTRY 4.0 APPROACH TO HEALTHCARE ENERGY SUSTAINABILITY
A microsimulation of every business process that requires energy within a hospital was created to build a model for prioritization of power in order from most to least critical. This is used to inform how power should be adjusted in hospitals during power shortages or power outages. A Monte Carlo simulation is used to adjust this model based on percent occupancy, number of walk-ins, and other variables. This research is especially important for areas that have experienced distress to their power grids or that regularly have limited energy capacity.
ENERGY ANALYSIS OF HEALTHCARE INFECTION CONTROL INTERVENTIONS WITHIN HOSPITALS FOR COVID-19
Covid-19 has presented countless challenges globally, one of which is the energy costs of Covid-19 interventions in hospitals. This project aims to better understand these costs and their potential benefits. By applying a de-escalation model to covid-19, researchers can evaluate the impact of High Efficiency Particulate Air (HEPA) filtered patient rooms versus the use of UV-xenon gas decontamination rooms. Modeling the energy demand and costs of these interventions and inputting these costs in a cost minimization function allows evaluation of infection impact. This can inform how funds should be distributed among these interventions.
MINIMIZING SECONDARY INFECTIONS IN HOSPITALS
We have developed an optimization framework for evaluating cost-effectiveness for built-environment infection control interventions. This is done by mathematically calculating the minimization of both secondary infections and cost for the interventions with differential equations to determine infection transmission, optimization, constraints, and objective functions.
This has been expanded upon by using cost of infection data for Methicillin-resistant Staphylococcus aureus (MRSA), carbapenem-resistant Enterobacteriaceae (CRE), and vancomycin-resistant enterococci (VRE) from 10 hospitals to evaluate two infection control interventions within hospitals. The projected number of infections from the infection de-escalation model are then validated by this data, improving the evaluation of infection control interventions during design or hospital renewal projects. As data input improves, the model becomes more precise and results improve.
RISK CALCULATOR FOR C. DIFFICILE INFECTIONS AT THE HOSPITAL UNIT LEVEL
C. difficile infections come from a toxin producing bacteria which causes an antibiotic associated diarrhea. These secondary infections result in high expense and negative patient and family impact. This research aims to build an easy to use tool for infection control teams, applying a mathematical model to prevent infections in hospitals. This will help to ensure a more proactive approach to allocating resources and preventing these infections in hospitals. Retrospective data is applied to a cox-proportional hazards model to quantify risk. Additionally, nonlinear models, cross-validation, and bootstrapping techniques assess the best tool. This research is being conducted in collaboration with Hopkins Hospital Epidemiology and Infection Control group.