Assessing reading levels of health information: uses and limitations of flesch formula
Abstract
BACKGROUND: Written health information is commonly used by health-care professionals (HCPs) to inform and assess patients in clinical practice. With growing self-management of many health conditions and increased information seeking behavior among patients, there is a greater stress on HCPs and researchers to develop and implement readable and understandable health information. Readability formulas such as Flesch Reading Ease (FRE) and Flesch-Kincaid Reading Grade Level (FKRGL) are commonly used by researchers and HCPs to assess if health information is reading grade appropriate for patients. PURPOSE: In this article, we critically analyze the role and credibility of Flesch formula in assessing the reading level of written health information. DISCUSSION: FRE and FKRGL assign a grade level by measuring semantic and syntactic difficulty. They serve as a simple tool that provides some information about the potential literacy difficulty of written health information. However, health information documents often involve complex medical words and may incorporate pictures and tables to improve the legibility. In their assessments, FRE and FKRGL do not take into account (1) document factors (layout, pictures and charts, color, font, spacing, legibility, and grammar), (2) person factors (education level, comprehension, health literacy, motivation, prior knowledge, information needs, anxiety levels), and (3) style of writing (cultural sensitivity, comprehensiveness, and appropriateness), and thus, inadequately assess reading level. New readability measures incorporate pictures and use complex algorithms to assess reading level but are only moderately used in health-care research and not in clinical practice. Future research needs to develop generic and disease-specific readability measures to evaluate comprehension of a written document based on individuals' literacy levels, cultural background, and knowledge of disease.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Machine readability scores may not fully capture nuances such as logical flow or reader familiarity with technical terms.
“Additionally, machine readability scores may not fully capture nuances such as logical flow or reader familiarity with technical terms [21].”
From Evaluating the Capability of Large Language Model Chatbots for Generating Plain Language Summaries in Radiology (2025) · cites this paper as [21]
The citation lists haven’t been fetched yet
References 47 · Cited by 275
- Authors
- Jindal, P., & MacDermid, J.
- Affiliations
- McMaster University (CA)
- Corresponding author
- Jindal, Pranay
- ORCID
- Jindal, Pranay
- Venue
- Education for Health
- Published
- 2017
- Volume
- 30
- Issue
- 1
- Pages
- 84
- ISSN
- 1357-6283 (ISSN-L)
- OpenAlex
- W2736868680
- Cited by
- 275
- References
- 47
- Citations by year
- 2026: 64 · 2025: 74 · 2024: 52 · 2023: 31 · 2022: 19 · 2021: 21 · 2020: 9 · 2019: 3 · 2018: 2
- FWCI
- 6.21
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Education for Health (publishedVersion)
- PubMed (publishedVersion)
- Scholarship@Western (Western University) (submittedVersion)
- Topics
- Health Sciences › Health Professions › General Health Professions › Health Literacy and Information Accessibility
- Physical Sciences › Computer Science › Artificial Intelligence › Text Readability and Simplification
- Health Sciences › Health Professions › General Health Professions › Patient-Provider Communication in Healthcare
- Keywords
- readability formulas, Flesch Reading Ease, Flesch-Kincaid Grade Level, health information, health literacy, written health information, readability assessment, patient understanding, readability measures, clinical practice, patient education
- MeSH terms
- Algorithms, Humans, Reading, Comprehension, Consumer Health Information, Health Literacy
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review