Reply to: Leveraging artificial intelligence to improve patient education for hypertensive disorders of pregnancy

To the Editor:

We thank Dr. Grünebaum for the reminder that patient educational materials should be written at a 6th–8th‐grade reading level and for the suggestion that generative artificial intelligence (GenAI) can have a role in the development of such materials [1]. Many GenAI engines are freely available online, each producing different outputs for the same input. Some of them can translate material into other languages, enabling access for non‐English speakers. When using any GenAI engine, it is essential that human readers critically review and edit the output before disseminating material to patients.

There are also numerous readability scales and free online tools for analyzing readability. A particularly useful tool is Datayze [2], which analyzes a passage of text and reports results from six readability scales (Flesch Reading Ease, Gunning Fog Scale Level, Flesch‐Kincaid Grade Level, SMOG Grade, Dale‐Chall Score, and Fry Readability Grade Level). On the Datayze web page, each scale has a “more info” question‐mark button that briefly explains how it assesses readability.

Brevity is an important attribute that is not directly analyzed by readability scales. It is self‐evident that the longer a document is, the less likely a patient will read it entirely or retain all its contents. Our sample briefing [3] has 304 words, excluding the instructions and customization statements. The sample produced by Claude has 816 words. When we asked ChatGPT to revise our sample using a 6th‐grade reading level, it returned a version with 278 words. The reduction in word count occurred despite substituting longer phrases for single words (e.g., “that you can't explain” instead of “unexplained” or “as soon as possible” instead of “today”), yet the essential elements of the content were preserved.

Graphic design is another important attribute affecting reader engagement. The selection of fonts, page layout, and addition of infographics can all influence the likelihood that a patient will read and retain the material.

Our proposed handout about long‐term cardiovascular risk is only one of many pages of information given to patients at hospital discharge. Thus, we believe it is essential that the briefing be delivered verbally by a trained healthcare professional. Direct person‐to‐person communication is often the most effective. A recorded video briefing may be a good alternative as long as it can be followed up with an opportunity for the patient to ask questions. The written handout is used to reinforce the verbal briefing. The quality of the briefing is at least as important as the content and design of the handout.

We are glad that Dr. Grünebaum followed our recommendation to adapt and customize our sample educational briefing. The briefing is at the heart of our new quality metrics on education and transition of care for postpartum patients with hypertensive disorders of pregnancy. We look forward to learning about hospitals’ experiences in implementing the quality metrics and about their progress in improving performance.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The authors have nothing to report.

CONSENT

The authors have nothing to report.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.