Final Project Progress Report
Progress Report, June 3
Agent Behavior
The general conceptual model is as follows:
As from last week, I am still working on how I might implement the care process with apache probabilities.
I am awaiting response from an email from the medical director of the ICU to see if he can provide reference patterns.
Setup
- Initialize 20 beds, selectable amount of caregivers, initialize patients (apache and severity)
- Sickest patients get assigned individual caregivers
- Remaining caregivers are assigned to less acute patients. 2 patients per caregiver.
To go -
- I flow in ICU based on arrival probability list. I am either admitted or I overflow to another unit (overflow is system breakdown)
- I am cared for and have a probability of dying based on my initial apache. Length of stay data is assumed from literature (What is the most reasonable method of abstracting the care process?)
- If my apache decreases to 0, I am discharged, otherwise I die
System Behavior
The system is still not functionally flowing because the care process and outflow still isn't in place. However, there are initial measures in place that can be roughly checked against literature.9
Rationale for agent rules
- Patients in low acuity have 90% chance of surviving. Patients in medium acuity have normal distribution of survival rates. High acuity patients have 10% chance of surviving.
- Most patients discharged in a narrow time-window. Thus, outflow happens early in the morning and inflow happens all day.
- average ICU Length of stay was 69.3 hours with standard deviation of 93.3 hours
Questions
- How to best model the care process? Should patients have their apache reduced by a caregiver a certain amount each tick? Should patients with high apache be more responsive to care?
- Visualization. It seems there is a better method to model the patients?
Next Steps
- Implement care process
- Mortality probability table for apache scores.
- Verify arrival probability
- Give care-givers variable skills so that they heal at different rates.
Next Steps
- Add transfer up as a measure to track patients that have been "transferred up" to the ICU. 3% had a transfer up