CATALYST Research Project 2022

Stochastic Models of
Telemedicine Access in Minute Clinics
Faculty Director: Mark Lewis

Maxwell M. Upson Professor of Engineering
Director, School of Operations Research and Information Engineering
Cornell University



With the increase in training of medical service providers (MSPs), nurse practitioners (NPs)
and registered nurses can handle a signifcant proportion of patient medical needs. However, even when the patient issue is of low acuity, there are still instances when a physician’s expertise needs to be consulted.

It is common in clinics that arriving patients meet a licensed practical nurse (LPN). The
LPNs do basic intake by taking the patient’s height, weight, temperature, blood pressure,
etc.  If the patient visit is for something simple, like administering a vaccine, they receive
this service and leave the clinic. If they need further medical attention, they are passed to
an NP who makes the decision about whether or not the patient needs collaborative care.

Students will begin by developing a mathematical model for intake for a general number of LPNs.

Using Matlab, students will develop a computer simulation of intake with fixed inputs to
discern how many LPNs are needed at each minute clinic.

As network capabilities have increased, so has the ability to deliver quality healthcare
without requiring every patient meet in person with a medical doctor (referred to here as
general practitioners (GPs)). Together these advances have made it more prevalent for NPs
and GPs to collaborate to deliver quality health care with at least one of them doing so
virtually. Broadly speaking, virtual healthcare is called telemedicine.

As is the case with any service system, there is a question of the quality of service. In
telemedicine there are complicating factors. First, there are some patients that definitely
need the higher level of service and NPs must call a physician for collaborative help and
medical advice. However, there are other patients whose service might be improved by
collaboration, but for whom NPs will find that the delay to get a collaborating physician is
too large to warrant waiting for the consultation. In this case, the NP can decide to provide
service themselves or to call a GP. This leads to a dilemma since all patients need to be seen quickly and with high quality care.  See figure 1.1 below.

If time permits, we will
 Develop a simulation model of a minute clinic.
 We will test various control policies to see which performs the best.

The students will be divided into teams with the results presented on the final day of the

Figure 1.1 – Diagram of Generalist Collaborative Support Model