Customized documents for health education

Present-day health-education and patient-information material is often limited in its effectiveness by the need to address it to a wide audience. What is generally produced is either a minimal, generic document that contains only the information common to everyone, or a maximal document that tries to provide all the information that might be relevant to someone (and hence much that is irrelevant to many). But material that contains irrelevant information, or omits relevant information, or that for any other reason just doesn't seem to be addressed to the particular reader, is likely to be discounted or ignored, with consequent problems in motivation for compliance with medical regimens, health-related lifestyle improvements, and so on. Recognizing this, health educators have paid much attention to identifying different segments of their audience and the needs of each, and constructing material accordingly; but at some level, the material remains generic. However, recent experiments have shown that health-education material, whether printed or on-line, can be much more effective if it is customized for the individual reader in accordance with their medical condition, demographic and personal characteristics, or other relevant factors, and can significantly enhance the patient's compliance with medical regimens. But the number of different combinations of factors can easily be in the hundreds or thousands. Template-based methods don't scale up. What is needed is a computer system for the production of tailored health-education and patient-education material that, on demand, would customize a ``master document'' to suit a particular individual.

The goal of the HealthDoc project (with Chrysanne DiMarco, University of Waterloo, and Eduard Hovy, University of Southern California), was to build such a system. Information from an on-line medical record or from a clinician would be used as the basis for deciding how best to fit the document to the patient. The Toronto role in the project was in the design of the overall framework and methodology of the project, in researching the background in health communication and relevant linguistic issues, and in overseeing the design and development of an authoring tool for master documents by Kimberley Parsons at Waterloo. Daniel Marcu worked on methods for deriving plans for coherent, well-structured, persuasive paragraphs for health-communication documents. Marcu has developed new algorithms that take the problem of coherence as one of constraint satisfaction, where the constraints are derived from rhetorical relations that hold between units of elementary information and from heuristics that our study of persuasive language has shown to be appropriate for achieving perlocutionary effects. This project has the potential not only to improve patient education, with consequent improvement in compliance and reduction in health costs, but also to make novel contributions to many aspects of research in natural language generation, particularly in sentence planning and lexical choice.

Reference:

Also: HealthDoc home page (University of Waterloo), DiMarco and Foster 1997, Hovy and Wanner 1996, Parsons thesis.

Return to Research by Graeme Hirst and students kk