Choose a task you can check.
Start with a repeatable, non-clinical task whose result someone on your team can review. Drafting a staff meeting agenda from a fictional scenario is a better first exercise than asking an AI tool to make decisions about a patient.
Define the result before opening a tool.
Write down the audience, format, approved source material, and who will review it. A useful brief might ask for a short public website FAQ draft based only on the practice’s approved service list. If a fact is missing, the output should flag it instead of filling it in.
Keep the first test free of patient information.
Use public practice information or a completely fictional example. Do not paste patient names, appointment records, messages, scans, images, financial details, or other patient information into general-purpose AI tools.
Check before you use it.
Review every claim against the original source. Check names, links, services, prices, and tone. A fluent answer can still be wrong. The person responsible for the final material decides whether it is suitable to use.
Decide whether it helped.
Compare the time spent preparing, checking, and correcting the draft with doing the work yourself. Keep a simple record of useful results and common errors. Continue only if the workflow helps the team without adding unacceptable risk.
Written and reviewed by CheckMyDSO. Updated 2026-10-16.
Get new prompts by email
Free AI prompts for the front desk, Tuesday and Thursday at 7 AM ET.