Clinic: should we add a provider?
A walk-in urgent care clinic has long visits, and the first idea on the table is to hire another provider. Use this model to test that idea before anyone signs a contract.
The business question
Your client asks: "If we add a provider, how much shorter do patient visits get, and is it worth the cost?"
A spreadsheet can tell you the average load on each role. It can't easily show you what happens when arrivals bunch up, staff are shared between stations, and only some patients need a lab test. That's what this model shows.
Open the model
In draw.io, choose Quodsi > Open Example Model…, then under Healthcare pick Example: Urgent Care System. See how to open an example.
How the clinic maps to Quodsi
| In the clinic | In Quodsi | In this model |
|---|---|---|
| Patients | Entity, the thing that flows through the process | Patient |
| Patients walking in | Generator | A patient about every 8 minutes on average, with random gaps |
| Each station | Activity | Check-In / Registration, Triage & Vitals, Provider Exam, Lab Draw, Imaging / X-ray, Checkout |
| Staff | Resource | 2 Front Desk Staff, 2 Triage Nurses, 2 Providers, 1 Lab Tech, 1 Imaging Tech |
| Rooms and space | The Activity's capacity | Up to 5 patients in Provider Exam at once, like exam rooms |
| Patient paths | Connectors with probability routing | After the exam, 30% go to Lab Draw, 15% to Imaging / X-ray and 55% go straight to Checkout |
Three parts of this model are worth noticing:
- Service times vary. Each station uses a triangular distribution: fastest, most common and slowest. For example, Provider Exam takes 10 to 25 minutes, usually about 15. This is handy when your client can say "usually about 15" but has no detailed data.
- One team covers two jobs. The same Front Desk Staff handle Check-In and Checkout. When Checkout gets busy, new arrivals wait longer to register.
- A patient needs staff and space. Service starts only when a place at the station and a member of staff are both free.
The model runs a 12-hour day, 10 times. Each run is called a replication, and each gets different random arrivals and service times. See Randomness and replications.
What to change
Click What-if Analysis and run the baseline first. Then test one change at a time, each in its own scenario.
- Add a provider. The Provider resource already has a lever for its capacity, set up to try 2 to 5 providers. Try 3.
- A busier day. Change the average time between arrivals on the Generator from 8 minutes to 6.
- More imaging. Raise the imaging share from 15% to 30%, and lower the direct share from 55% to 40%.
- Split the front desk. Give Checkout its own staff resource, and compare how long new patients wait to check in. This changes the model's structure, so make a copy of the model first.
To test several provider counts together, use a study.
What to look at
- Time in the clinic. This is the patient's cycle time, the number your client feels. Compare the baseline and each scenario.
- Queues at each station. In the busier-day scenario, watch which queue grows first. That station is your bottleneck.
- Utilization and queues together. A resource can have a utilization that looks fine and still have a persistent queue. Read both.
- Cost. Each resource has an hourly cost: $20 for front desk staff, $38 for a triage nurse, $110 for a provider, $25 for a lab tech and $30 for an imaging tech. The rate is the same whether they're busy or idle. So an extra provider's cost is easy to weigh against shorter visits. See Costing.
- Whether the answer moved the bottleneck. Adding providers can simply shift the wait to another station. Check Checkout and Triage & Vitals as well as Provider Exam.
The results window has Overview, Diagram, Flow Analytics, Dashboard and Animation tabs. Start with Overview. Use Animation to show a client what's happening.