Modeling Time-Varying Productivity and Its Applications in Physician Staffing in Emergency Departments

Modeling Time-Varying Productivity and Its Applications in Physician Staffing in Emergency Departments

Stochastic modeling of a single physician

• $C(t)$: physician capacity;
  • time-dependent due to EOS effect
• Can we model it using a finite-time CTMC? Or, is it necessary?
• Arena: simulation

Statistical modeling (Poisson regression)

We can also apply statistical models to estimate the PPH: we can control MD ID, workload, learners, handovers taken, current number of patients signed up,

Medical Literature

Wang, Biao, et al. “Physician shift behavior and its impact on service performances in an emergency department.” 2013 Winter Simulations Conference (WSC). IEEE, 2013.

A plot from the above reference:

Empirical plots

I updated the estimates of arrival rates and PPH. There are some days that the information system is down. Hence, the recorded data are probamatic. The following days are removed from the estimates: 2015-01-29, 2015-02-26, 2015-03-26, 2015-04-30, 2015-05-28, 2015-06-25, 2015-07-30

All of them are 12:00 am on Thursdays, hence, I believe this is the scheduled maintenance time. Correspondingly, only the shifts S13, S14, and S15 are affected.

Probability of another round of reassess

Hence, the probability of going another round of test is 33.95%, i.e., 1/3.