TL;DR
Get health and wellness essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
In a MedPage Today opinion essay published Oct. 20, 2026, nurse and cancer patient Davis Austria describes how clinical guidelines and patient handouts failed to answer practical questions during head and neck cancer treatment. Austria argues that health systems should test patient-facing materials for clarity and usefulness—not just medical accuracy—as AI tools increasingly generate or rewrite them.
Davis Austria, an emergency and critical care nurse who has had head and neck cancer, is calling on health systems to test patient instructions for clarity and practical usefulness, not only medical accuracy. In a MedPage Today opinion essay published Oct. 20, 2026, Austria recounts how clinical guidance and patient materials left unanswered questions during treatment and argues that AI-generated information could spread those gaps unless patients help evaluate it.
Austria writes that, after a cancer diagnosis four years ago, a radiation oncologist showed them the National Comprehensive Cancer Network (NCCN) treatment recommendations for their stage: 10 chemotherapy rounds and 35 radiation sessions. The recommendations were accurate and evidence-based, Austria says, but were written for clinicians. Austria went home with the document and questions about what symptoms were normal and what to expect during recovery.
The questions included whether severe pain while swallowing was an expected effect, whether hearing would recover, and how to eat after a feeding tube was removed. Austria says the patient-facing NCCN guidance advised asking a healthcare provider about treatment for radiation burns, but did not answer the practical questions they had. They also describe using Aquaphor on the neck before radiation, then later finding another document that indicated they had used it at the wrong time.
Unable to get answers from the written materials or a patient portal Austria describes as a place to retrieve documents, they turned to a Facebook group for people with the same cancer. The essay says patients now may use AI to find and explain health information, while warning that AI can be wrong, incomplete, or confidently misleading. Austria also points to health systems’ use of AI to write or rewrite patient materials, and says those materials should be checked with patients for whether they are understandable and actionable.
Testing Instructions With Patients
The essay highlights a practical gap between information that is medically correct and information a patient can use while managing symptoms or recovery. In Austria’s account, a clinician-focused guideline and general advice did not address urgent, everyday questions. That matters because patients may seek answers elsewhere when materials and care channels do not meet their needs.
Austria cites a randomized trial of 128 patients in Germany comparing standard physician-written discharge summaries with AI-generated versions rewritten for patients. According to the essay, patients who received the patient-oriented version scored 9.6 points higher on patient activation, a measure of preparedness to manage one’s care. The essay does not provide further study details, so the result should be understood as the reported finding of that trial, not proof that every AI rewrite improves outcomes.
Austria’s central recommendation is that health systems evaluate patient materials across language, age, and social circumstances, and ask actual patients whether they can understand and act on the information. That scrutiny applies to human-written instructions as well as AI-produced material; AI raises the stakes because a flawed template could be reused at scale.
patient education materials for cancer recovery
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Clinical Guidelines to Patient Use
Austria writes from overlapping experience as a nurse, a cancer patient, and a researcher studying how health systems evaluate AI-generated patient information. The essay also notes that Austria has built hundreds of treatment order sets in electronic health records. Austria says the care team knew them as a clinician, but that this did not meet their needs as a patient receiving a cancer diagnosis.
The essay refers to a 2024 study in BMC Primary Care about doctors treating fellow doctors. Austria says the study found clinicians changed their communication with colleague-patients and made assumptions about what those patients already knew. The article uses this as context for its account, not as evidence that the study examined Austria’s own care.
Austria also identifies Epic, a major electronic health record platform, as a company selling a patient-facing virtual assistant called Emmie. The essay says one feature rewrites after-visit summaries in plainer language. It gives examples of clinical shorthand—NPO, STAT, and PRN—and explains that patients may not know these mean nothing by mouth, immediately, and as needed.
“What is happening to me, and what happens next?”
— Davis Austria, in the MedPage Today opinion essay
Limits of the Evidence and AI
The essay does not report a formal assessment of the NCCN materials Austria used, nor does it establish how often patients encounter similar gaps. It also does not provide the full methods or longer-term outcomes for the German trial beyond the reported patient-activation score difference.
Austria warns that AI-generated health information may be inaccurate or incomplete, but the essay does not quantify how frequently that occurs or compare specific AI tools. It is also unclear from the essay how widely health systems are using AI to write patient instructions, how their outputs are reviewed, or whether patients have tested the Emmie feature.
Austria’s account is a personal experience and opinion, not a clinical study or a claim that AI should replace clinicians. The essay states that its views are the author’s alone and do not represent NIH, AIM-AHEAD, Ochsner Health, or Xavier University of Louisiana.
Patient Testing Before Rollout
Austria urges health systems to ask whether materials work for the patients they serve, whether review goes beyond medical accuracy, and whether patients have tested their ability to understand and act on the instructions. Those checks, the essay argues, should occur before materials reach patients, whether a person or an AI system wrote them.
The essay does not announce a new policy, implementation deadline, or specific next step by NCCN, Epic, or a health system. Its proposed next step is for health organizations to include patients in evaluating instructions and to keep clinicians involved in care. The broader question is whether health systems will adopt that testing as AI-produced patient information becomes more common.
Key Questions
What is the main development in the essay?
Davis Austria, a nurse and cancer patient, published an opinion essay arguing that health systems should test patient-facing health materials for clarity and usefulness, especially as AI is used to generate or rewrite them.
What information did Austria say was missing during cancer treatment?
Austria wanted practical guidance about what symptoms were normal, whether hearing might recover, and how to eat after a feeding tube was removed. They say the materials they received did not answer those questions clearly.
Does the essay say AI should replace clinicians?
No. Austria says AI can help patients find and understand information but can also be wrong, incomplete, or misleading, and is not a substitute for the clinicians caring for them.
What study does Austria cite?
Austria cites a randomized trial involving 128 patients in Germany. The essay reports that patients who received AI-generated, patient-oriented discharge summaries scored 9.6 points higher on patient activation than those who received standard physician-written summaries.
What does Austria recommend health systems do?
Austria recommends checking materials for more than medical accuracy, considering whether they work for patients across different circumstances, and asking patients to test whether they can understand and act on the instructions.
Source: rss
Cold & flu season Picks
humidifiers
As an affiliate, we earn on qualifying purchases.
