How a Dutch GP uses ChatGPT to cut paperwork, improve clarity, and keep patient communication personal
With ChatGPT, Dr. Fleurique Franke is able to streamline and modernize routines in her health clinics, keep patient communications humane and digestible, and give her community faster, clearer care.
Franke is a general practitioner in the Netherlands, who turned to custom GPTs to reduce paperwork and raise the quality of patient communication. She runs two clinics staffed by four physicians and a team of nurses and assistants. Together, the clinics serve about 4,300 patients who represent a cross‑section of the Netherlands running from white-collar professionals and healthcare workers to members of assisted living communities.
Much of the workload arrives as so-called e‑consults, the clinic’s asynchronous message-based consultations that flow directly through its health information system. Franke and her team of doctors each handle 30 to 40 e‑consults amounting to about 2.5 hours of work per day. Many queries involve common local issues: upper airway infections, suspicious skin spots that may be melanoma, or impetigo, a bacterial skin infection frequently seen in children. Each message demands a thorough reply, however simple the diagnosis may be. “You cannot just say yes or no,” Franke says. “You need to give an informative answer.”
To do that quickly and consistently, she built a custom GPT she calls “email‑answerer.” It drafts patient‑ready responses that cite vetted national patient information pages and Dutch GP guidelines, while protecting all personal health information. Franke supplies the clinical gist to the GPT, which assembles a complete answer with next steps, then she edits before sending. She estimates the approach saves about one-third of her time, helping her respond fast and well while keeping her personal connection with patients.
Tone is an important part of the care. Franke uses ChatGPT's built-in readability slider to rewrite the same plan at different levels, from grade-school to academic, for various demographics in her community. That is, she uses it to tune the explanations she writes for patients with different levels of reading comprehension and domain expertise, so that everyone from healthcare professionals to residents in supported living can easily understand their care plan. She watches follow‑up questions as a real‑time check on patient comprehension.
Franke also built a GPT for a nurse who is learning cardiovascular care. Dutch protocols for cholesterol targets, kidney function thresholds, and medication indications are full of complex conditions and exceptions. The GPT ingests the guideline and the patient’s labs, asks clarifying questions to the nurse, and surfaces the next considerations as a checklist.
Franke’s clinic-level gains reflect a broader pattern in ChatGPT health use: health-related queries are one of the biggest categories of ChatGPT usage globally, totaling more than one billion chats each week. They span care options, understanding tests and results, and common symptoms.
