Ctrl+K

3. Data

import pandas as pd

3. Data#

3.1. Data sources#

All data are sourced from

Nelson. B.L. (2013). Foundations and methods of stochastic simulation. Springer.

3.2 Pre-processing#

Non additional pre-processing of data was undertaken.

3.3. Input parameters#

Time-dependent arrival rate#

The data for arrival rates varies between clinic opening at 6am and closure at 12am.

NSPP_PATH = 'https://raw.githubusercontent.com/TomMonks/' \
 + 'open-science-for-sim/main/src/notebooks/01_foss_sim/data/ed_arrivals.csv'

# visualise
ax = pd.read_csv(NSPP_PATH).plot(y='arrival_rate', x='period', rot=45,
                                 kind='bar',figsize=(9,4), legend=False)
ax.set_xlabel('hour of day')
ax.set_ylabel('mean arrivals');
../../_images/40911f9f831a65c4c3505dc96d0f9424f00ef1a8db63cfa8016b03dc2a9b8656.png

Sampling distributions#

Distributions were taken again from Nelson (2013)

Activity

Distribution

Triage

Exponential(3.0)

Registration

Log Normal(mean=5.0, var=4.0)

Examination

Normal(mean=16, var=4.0)

Stabilisation

Exponential(90.0)

Non-Trauma treatment

Lognormal(mean=13.3, var=2.0)

Trauma treatment

Lognormal(mean=30.0, var=4.0)

Prob patient requirement treatment given non-trauma

0.6

Prob patient is trauma

0.12

3.4 Assumptions#

  • In this model trauma patients are treated completely seperately from non-trauma.

Contents