EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
1051. Admission and early discharge policies for ICU patients during times of intense demand in a pandemic
Invited abstract in session WB-15: COVID-19, stream OR in Health Services (ORAHS).
Wednesday, 10:30-12:00Room: 18 (building: 116)
Authors (first author is the speaker)
1. | Dong Li
|
Management Science, Lancaster University | |
2. | Li Ding
|
Durham business school, Durham University | |
3. | Navid Izady
|
Bayes Business School | |
4. | Richard Wood
|
BNSSG Clinical Commissioning Group | |
5. | Christos Vasilakis
|
School of Management, University of Bath |
Abstract
This study addresses the limitations of existing intensive care unit (ICU) triage policies by developing an algorithmic model for admissions/discharges during times of intense demand in a pandemic. Embracing a multi-value ethical framework, our model aims to maximize aggregated benefits while considering the principles of fairness. The proposed model accounts for the illness pathways of individual patients, the overall mixture of patient profiles, and the uncertainty in the number of future patients. We introduce a cost for early discharges to account for psychological and ethical complexity and build a discrete time Markov Chain from actual patient data to capture disease progression of ICU patients. We conduct a survey among ICU physicians to estimate short-term benefits of intensive care. The problem is formulated as a discrete Markov Decision Process. To address the computational complexity, we relax the capacity constraint and formulate an alternative problem for each single patient. We characterise the optimal admission and discharge policy for individual patients, which allow us to develop a heuristic policy for the original problem. The performance of the proposed policy and alternative benchmarks are evaluated in a comprehensive simulation study. Our results reveal the scale of impact possible through appropriate clinical decision-making regarding patient admission and discharges, demonstrating potential benefits in both lives and life-years saved.
Keywords
- Health Care
- Stochastic Models
- Programming, Dynamic
Status: accepted
Back to the list of papers