Perils of providing visual health information overviews for consumers with low health literacy or high stress.

This pilot study explores the impact of a health topics overview (HTO) on reading comprehension. The HTO is generated automatically based on the presence of Unified Medical Language System terms. In a controlled setting, we presented health texts and posed 15 questions for each. We compared performance with and without the HTO. The answers were available in the text, but not always in the HTO. Our study (n=48) showed that consumers with low health literacy or high stress performed poorly when the HTO was available without linking directly to the answer. They performed better with direct links in the HTO or when the HTO was not available at all. Consumers with high health literacy or low stress performed better regardless of the availability of the HTO. Our data suggests that vulnerable consumers relied solely on the HTO when it was available and were misled when it did not provide the answer.

Visualization of health information with predications extracted using natural language processing and filtered using the UMLS.

Increased availability of and reliance on written health information can tax the abilities of unskilled readers. We are developing a system that uses natural language processing to extract phrases, identify medical terms using the UMLS, and visualize the propositions. This system substantially reduces the amount of information a consumer must read, while providing an alternative to traditional prose based text.

Evaluating online health information: beyond readability formulas.

Although understanding health information is important, the texts provided are often difficult to understand. There are formulas to measure readability levels, but there is little understanding of how linguistic structures contribute to these difficulties. We are developing a toolkit of linguistic metrics that are validated with representative users and can be measured automatically. In this study, we provide an overview of our corpus and how readability differs by topic and source. We compare two documents for three groups of linguistic metrics. We report on a user study evaluating one of the differentiating metrics: the percentage of function words in a sentence. Our results show that this percentage correlates significantly with ease of understanding as indicated by users but not with the readability formula levels commonly used. Our study is the first to propose a user validated metric, different from readability formulas.

Dynamic generation of a table of contents with consumer-friendly labels.

Consumers increasingly look to the Internet for health information, but available resources are too difficult for the majority to understand. Interactive tables of contents (TOC) can help consumers access health information by providing an easy to understand structure. Using natural language processing and the Unified Medical Language System (UMLS), we have automatically generated TOCs for consumer health information. The TOC are categorized according to consumer-friendly labels for the UMLS semantic types and semantic groups. Categorizing phrases by semantic types is significantly more correct and relevant. Greater correctness and relevance was achieved with documents that are difficult to read than those at an easier reading level. Pruning TOCs to use categories that consumers favor further increases relevancy and correctness while reducing structural complexity.