You already know medication management sits squarely in our scope. It is an instrumental activity of daily living, it shows up on nearly every home evaluation, and it is one of the quiet reasons a client cycles back through the hospital. What has changed is the toolset. AI assistants now make it realistic to hand a client and their family a durable, low cost system that keeps doses on track long after we discharge. In this guide I walk through the exact setup you can teach in a session, and why it belongs in your intervention plan.
Why medication management is an OT opportunity, not just a nursing task
The numbers make the case for us. The World Health Organization has reported that adherence to long term therapy averages only about 50 percent among people with chronic conditions in developed countries. Among older adults living with dementia, difficulty managing medications is more common still, and that difficulty is directly linked to a higher risk of hospitalization. This is not a knowledge gap for most clients. It is a routine, memory, and habit problem, which is precisely the territory we are trained to work in.
That is a measurable functional outcome you can document, and a reason for the client to see the value of skilled OT.
How to frame it clinically: this is habit and routine training
Think of the AI system as an external cognitive support, the same way you would frame a checklist, a labeled pill organizer, or a whiteboard. The technology handles the prospective memory demand, which frees the client to focus on the motor and sequencing steps of actually taking the medication. Your clinical reasoning is what makes it safe. You are the one deciding whether the client can notice and act on a cue, whether the cognitive load is appropriate, and where a human backup still needs to sit.
Screen first: who is a good candidate, and who is not
AI medication reminders work best for a client who has a stable, well documented regimen and who can reliably notice and respond to a cue, whether that is a push notification, a text, or a voice prompt from a smart speaker. It is a strong fit for the client who lives alone or independently, manages several medications on a set schedule, and mainly needs consistent prompting plus a way for family to confirm doses were taken.
It is not the right primary solution for a client in a fast changing medical situation, a client with advanced dementia or confusion who cannot understand or act on a cue safely, or a client who tends to get distracted between the reminder and reaching the pills. It is also a poor fit as a standalone for high risk medications where a missed or doubled dose is dangerous. In those cases AI can still support the caregiver, but it should sit alongside hands on help rather than replace it. Your screen is what separates an appropriate recommendation from a risky one, and it is worth documenting the reasoning either way.
What to gather before the setup session
Have the client or caregiver bring 4 things so the session is productive.
- The complete current medication list, including over the counter items, vitamins, and supplements, confirmed against the bottles and the pharmacy rather than memory.
- The pharmacy and refill details.
- The client's real device situation, meaning the phone, tablet, or smart speaker they actually use and are comfortable with.
- The contacts who should receive missed dose alerts. The accuracy of this list is what makes the whole system safe, so it is worth the setup time.
The setup, step by step
1. Match the tool to the client, not the other way around
If the client is comfortable with a smartphone, a dedicated medication app such as Medisafe or a device like Hero Health can send push notifications and track doses. If the client relies on voice, a smart speaker such as Amazon Alexa or Google Home can announce reminders and answer simple questions. If they use a basic phone, a text based system that sends an SMS works on any device that can receive a message. Choose the channel the client already uses every day, because adherence to the reminder system follows the same rules as adherence to anything else.
2. Build the schedule with a conversational AI assistant
Take the medication list and hand it to a tool like Claude, ChatGPT, or Google Gemini, and ask it to turn the regimen into a clean, time blocked daily schedule. A prompt such as "Here are seven medications with doses and timing. Organize them into a clear morning, midday, evening, and bedtime schedule, and flag anything that needs to be taken with food or spaced apart" produces a structured routine in seconds. You then check it against the prescriptions. The AI drafts, you verify. This is also a teachable moment for the caregiver, who can repeat the process when the regimen changes.
3. Automate recurring reminders on the client's device
Translate the schedule into standing reminders that fire on their own. The path depends on the device:
- For app push notifications, download the Claude, ChatGPT, or Gemini app to the phone and ask the assistant to send reminders at each dose time.
- For text based reminders, connecting the assistant to a service like Twilio gives it the ability to send scheduled text messages.
- For a dedicated medication app, install it and enter the schedule directly.
- For a smart speaker, use the AI organized schedule and set a reminder aloud for each dose time.
Build in confirmation. A reminder that asks for a reply, such as "Reply OK once you have taken your morning pills," converts a one way cue into a closed loop you and the family can monitor.
4. Route missed dose alerts to the caregiver
Configure the system so an unconfirmed dose within a set window triggers an alert to the caregiver, and extend it to a second family member or aide so responsibility is shared. This is what gives the family passive visibility and lets them step back from daily phone calls without stepping away from oversight.
5. Add refill tracking and a safety check
Have the assistant calculate refill dates from quantity and daily dose and prompt a week before a bottle runs low. The assistant can also surface questions to bring to the pharmacist, such as common interactions to ask about. Frame this to the client clearly. AI helps them know what to ask a professional, faster. It is not the professional.
The guardrails to teach alongside the setup
Coach the client and caregiver on the same limits you work within. Verify medical information with a pharmacist or prescriber, since an AI can be wrong and does not know the full history. Protect private information by using reputable tools and sharing only what is needed. Point out that the assistant does not need a name or date of birth to build a schedule, so the medication list can go in without identifiers. And keep a human in the loop, because these tools support a caregiver, they do not replace one. Note in your plan that the setup should be revisited whenever a medication changes.
Why this belongs in your practice
Teaching an AI medication routine is high value, low cost, and repeatable. It targets a documented functional problem, it produces an outcome you can measure and defend, and it gives clients a reason to see skilled OT as forward looking rather than replaceable. For therapists building a private caseload, it is also a differentiator. Families are actively looking for someone who can make this technology safe and practical, and that someone is you.
Frequently asked questions
Is setting up AI medication reminders within OT scope?
Yes. Medication management is an instrumental activity of daily living, and using an external memory and habit support to address it is standard OT practice. Your role is the clinical reasoning around candidacy, cognitive load, and safety, not the technology itself.
How do I document this as a skilled intervention?
Frame it as training in the use of an external cognitive and environmental support to improve independence and safety with medication management. Document your candidacy screen, the client and caregiver education you provided, the client's demonstrated ability to respond to the cue, and the functional goal the system supports.
What if the client's cognition is borderline?
Then the AI system is a support, not a substitute for supervision. Pair it with a caregiver confirmation loop, simplify the regimen where possible, and set clear criteria for when a hands on backup is required. Reassess as status changes.
Do I need to be technical to teach this?
No. The conversational assistant does the heavy lifting, and each device path is a few plain steps. If you can teach a client to use a labeled pill organizer and a phone alarm, you can teach this.
Join AskSAMIE Pro
Setups like this are exactly what AskSAMIE Pro is built to support. Pro gives occupational therapists the tools, vetted equipment, and client connections to turn skills like this into a thriving private practice, so you can grow your income while widening access to care for the families who need it. If you are ready to build a practice around the value only an OT can deliver, join AskSAMIE Pro and start today. Better access, better lives, starts with therapists like you.
This guide is for educational purposes and does not replace clinical judgment. Confirm all medications, doses, and interactions with the client's pharmacist or prescriber.