The PCS Generative AI is a powerful conversation engine that allows you access to the Patient Concept Authoring Tool. Combined with the EZ Mode, which offers a simple overview of the most important aspects of a PCS patient scenario, you will be able to create your own custom patients in as little as a few hours. |
Customizing patient scenarios - the concept
Your PCS Simulator comes with bundled patients that you can use for simulations straight away.
You can also use these patients as templates to create your very own patient scenarios.
The bundled patients themselves are as is, and you will not be able to make changes to them. In order to create new patients, you need to copy an existing one and use it as a template first.
Copying a bundled patient and using it as a template
Go to app.pcs.ai and navigate to the Patients tab, then Open the patient you would like to use as a template:
Click on the Copy button in the bottom right:
This will create a copy of the patient. The copy will open up in a new window, and it will also appear in your My Patients tab in the future.
Editing your custom patient in EZ Mode
Once a copy is created, it is ready to be edited.
Start by renaming the copied patient:
This is an important step, as the patient will say this name when asked during the simulation. The highlighted sections on the Basics tab will also affect the conversation, so change these too if needed.
Editing the patient's conversation
Click on the Patient Concept tab to edit the conversation of your patient:
The Patient Concept is organized into 14 sections. Click into any of the textboxes to edit or completely rewrite the section of the Patient Concept. Hover over the
icon to learn more about what each section entails.
To test the changes you make in the Patient Concept in real time, click on the buttons on the right side to open the Conversation Tester.
Tips for writing the Patient Concept:
Think of the Patient Concept as an SP training material, which contains all the information the SP ought to know about to be able to successfully simulate a scenario.
The Generative AI responsible for the conversation will accurately respond to questions covered in the Patient Concept. Since the AI is based on a large language model, the responses may sligthly differ from simulation to simulation, but the "message" behind them will be the same.
For anything NOT included in the Patient Concept, the AI will come up with believeable responses. These responses will likely be different from simulation to simulation.
For example, if multiple students ask about the name of the patient's father, the AI will come up with different names unless it is specified in the Patient Concept.
While AI technology is improving at a rapid rate, it is still not on the level of a human brain. As such, the AI is unable to make logical deductions.
Take a look at the following sentence: "Patient is concerned about pain because she fell onto her left hip this morning". Humans are able to make the logical deduction instinctively that the patient is having pain because she fell onto her hip.
The AI is unable to make this deduction and when asked if falling on her hip hurt, she will respond with no. A better wording of this sentence would be: "Patient is feeling pain and is concerned because she fell onto her hip this morning."
Create redundancies. If you see that your patient does not give the expected response, repeat the related sentence in the Patient Concept with different words.
Write the Patient Concept in chronological order, and if the patient has multiple concerns/symptoms, specify which one was first.
Avoid overly technical jargon. Write in the third person, in the language and style that the patient would use.
The patients (aside from special cases) do not know their diagnosis, or how their symptoms correlate. Keep this in mind when writing the Patient Concept.
Include the Avatar's appearance in the Patient Concept.
Setting up the Objectives
When you copy a bundled patient, your copied custom patient will come with the basic structure of the Objectives and Credits, but you will need to define what the requirements are.
The layout of the Objectives tab is as follows:
The Objectives here are the main categories on which the students are scored on.
Each Objective contains a number of Credits, which are required for the given Objective to be fully completed.
You can define what students need to do during the simulation to earn credits, and how many points each credit is worth:
Start by clicking on the dropdown on the right. Select the Conversation option if you want the credit to be conversation-based, or the Event option if you want the credit to be event-based (time passed, log entries, requested diagnostics, etc):
Click on the button to open the editing window.
For Conversation-based credits, click on the Classic Conversation Criteria tab, and select what conversation topics or log entries you would like to set as criteria.
- Speech conversation responds with: allows you to select specific conversation topics as criteria
- Log entry including: allows you to program specific keywords or phrases as criteria.
The conversation scoring system works on the basis of clinical interview topics. This allows you to select a broad range of questions and responses, or only award a point for questions asked in a very specific way.
On the screenshot above, the AI will look for any questions recognized as inquiries about the presenting complaint of the patient. To narrow this down, you can select a more specific item in the dropdown on the right side of the screen.
You can set up multiple criteria for each credit and determine if they need to be met in order or not.
4. For Event-based credits, define what events trigger the point for that specific credit:
You can set up multiple criteria for each credit and determine if they need to be met in order or not.
You can use the following criteria for Event-based credits:
Log entries - anything that can appear in the simulation logs.
Vital sign values - the value of heart rate, respiration rate, etc.
Time passed - the amount of time that needs to pass or the maximum amount of time that can pass. Best combined with other criteria.
Diagnose & Treat Orders - anything the students can request from the Diagnose & Treat.
Physical Exam - auscultations, palpations, blood pressure measurements, etc.
End-of-Simulation quiz responses - responses that students leave for the end-of-simulation quiz.
SOAP Note entries - any phrases or keywords that the students can enter into the SOAP Notes.
You can add additional Credits under the existing Objectives, or create new Objectives.
Vital Signs
You can access and edit the starting vitals of the patient scenario on the Vital Signs tab.
Click on any of the numbers or the waveforms to change them.
Support Materials
You will be able to find and edit the various supporting resources on this tab:
Hints & Tips: Opens a rich text editor that allows you to edit the Hints and Tips, add images, embed videos, etc.
Patient File: Similar to the Hints & Tips, this section also allows you to edit the text, add images and embed videos into the Patient File
Background info: You can document the changes to your custom patient scenario here for later review
Debrief Content: A rich text editor for the Debrief Content that appears to students when they finish their simulations
Diagnose & Treat
The Diagnose & Treat allows students to request various labs, imaging and other diagnostics, as well as request consultations, define the patient's diagnosis and treatment options.
The available items in the Diagnose & Treat is grouped into 4 tabs, but the editing process is identical for each tab:
Hover over an item in the list to highlight it.
Clicking on the name of the item allows you to rename it.
Click the + Add Item button to add new items to the list, and click the to delete existing items.
Clicking on an item also displays it in the right side of the window:
You can edit the text, add or remove images, and embed videos in each item.
Note: None of the Diagnose & Treat items affect the patient's vitals in any way.