

Can WhatsApp Improve Medical Device Patient Support? What Teams Need First
A patient has had their new device for three days. It is nine in the evening, the pairing step isn't working, and the instructions are in a booklet somewhere. Their options are to call a support line that closed at six, open a portal they have not set up, or send a message to the one app they already use every day.
For a lot of patients, in a lot of countries, that app is WhatsApp.
For medical device companies, that raises a reasonable question: should patient support meet people there?
The short answer is yes, it can help with conditions. Messaging is fast and familiar, and it fits the small, frequent questions that make up much of device support. It also puts regulated conversations into a consumer app, which creates obligations most support teams have not had to think about. This article covers what to settle first.
Why messaging is a good way to communicate
Patients have clear habits. A 2026 survey of 1,000 U.S. patients by Sinch, a messaging vendor, found 90 percent prefer text messaging for hearing from a healthcare provider, and 68 percent want two-way texting to ask simple questions or reschedule. That survey covers SMS in the United States, not WhatsApp, and comes from a vendor that sells messaging, so read it as a signal about habits rather than a forecast for your market.
What the numbers suggest is that patients value convenience and speed over formality. Device support is full of questions that suit that style: setup help, how to clean or charge something, what a status light means, where a replacement is, when the next shipment arrives. A short answer, sent quickly, often ends the matter.
The first thing to settle: device complaints can arrive in a WhatsApp chat
This is the point that makes device support different from retail support.
A patient who messages that their device 'gave a shock' or 'stopped delivering therapy' has just made a complaint, and possibly a reportable event. FDA's training for mandatory reporters states that manufacturers must report deaths, serious injuries, and certain malfunctions within 30 calendar days of becoming aware, and certain events needing remedial action within 5 work days. FDA considers a firm aware when an employee becomes aware of the reportable event, not only when a manager confirms it.
In practice, that means every message a support agent reads is a potential regulatory input. The U.S. quality system regulation was updated to align with ISO 13485 and became effective on February 2, 2026, and TÜV SÜD notes that complaint handling and medical device reporting remain FDA-specific obligations. Teams should confirm with their regulatory and quality leaders how WhatsApp conversations feed complaint intake, and where the clock starts.
The practical requirement is simple to state. A conversation that may describe an adverse event must reach your complaint process quickly, with the full message history, and not sit in a phone or an agent's memory.
Then settle what is allowed in the conversation: consent, health data, and scope
Privacy rules come next.
Consent. Meta's platform rules say businesses can send messages only to users who have opted in. Record that consent against the patient record.
Health information. Whether and how patient information can travel over a messaging channel depends on your jurisdiction, your agreements, and the nature of the data. One compliance vendor's analysis concludes that WhatsApp is not HIPAA compliant for communications containing protected health information, but its analysis covers standard WhatsApp accounts, not the Business API, so it does not settle the question for a business deployment. Get a legal opinion for your situation. A common conservative design is to keep clinical details out of the chat and move them to a secure link or an authenticated channel.
Scope. Decide what the channel is for. Setup help, reminders, shipping updates, and basic how-to questions suit it. Clinical advice and emergencies do not. Make sure the first message from the patient receives a clear safety line: if this is an emergency, call your local emergency number.
Next, design around how WhatsApp works
WhatsApp's platform has its own mechanics. When a patient messages you, a 24-hour customer service window opens, and inside it you can reply freely. Outside it, you can send only pre-approved template messages. Meta has charged per delivered message rather than per conversation since July 1, 2025, and messages from users to businesses are free, although rates by category and country change, so check the current rate card before building a budget.
For support, this affects design. Reminders and proactive follow-ups need approved templates. Live conversations should resolve within the window, or hand off cleanly to another channel when they cannot.
And keep WhatsApp conversations in your CRM, your system of record
Many pilots stall at this point. If WhatsApp lives on a shared phone or in a separate inbox, nothing connects the conversation to the patient, the device, the serial number, or the case. Quality cannot see it. A second agent cannot pick it up. Nobody can audit it.
The sturdier approach is to treat WhatsApp as a channel into the CRM you already run, so that every message is tied to a record and follows the same permissions as everything else.
Picture the opening scenario handled well. This is illustrative, not a customer story. The patient's message creates a case against their device record. A short, approved answer fixes the pairing step. Had the patient mentioned a skin reaction instead, the same thread would have been flagged, passed to the complaint process with the full history, and answered with a safety message, all before anyone had to retype a note.
Where the tooling fits: AgentChat for WhatsApp on Salesforce
For teams already on Salesforce, this is the idea behind AgentChat, a bi-directional WhatsApp conversation app built natively on Salesforce. According to the product page, conversations sync between WhatsApp and Salesforce, can be started from any Salesforce object, are tracked against records with history and audit trails, and are routed by record type, such as cases or work orders. Access follows Salesforce's role-based permissions. Setup involves installing from AppExchange, connecting a WhatsApp Business Account, mapping record types, and setting roles.
For a device manufacturer, the value is less in the chat itself and more in what surrounds it: the conversation sits on the patient or case record, quality and support see the same thing, and an escalation to a complaint can start from the conversation. Whichever tool you choose, test it against the questions above: consent, templates, complaint routing, and what content may be exchanged.
A sensible way to start
Pick one device, one market, and one kind of conversation, such as setup and first-week questions. Write the rules before the pilot: what the channel handles, what must move elsewhere, and which words or situations trigger a complaint review. Train agents on those triggers. Run it for a few weeks, then review the transcripts with quality and regulatory.
Once the basics hold, consider automation. An AI agent can answer routine questions from approved content and hand off when a conversation touches symptoms or safety. Set its limits as carefully as you would for a human agent.
Track first-response time, time to resolution, share of conversations resolved without a call, patient satisfaction, and, most importantly, time from a potential complaint appearing in chat to its reaching the quality process.
Messaging will not fix a weak support process, and a chat window will not make a hard regulatory question easier. Done with care, though, it meets patients where they already are and gives your quality team earlier sight of what patients are experiencing.
Rialtes helps medical device companies connect patient conversations, device records, and quality processes on Salesforce. Learn more about our medical device IT solutions, or talk to us about launching your first WhatsApp support pilot.
FAQs: AI Agents for Semiconductor Technical Support
Latest Blogs
