Clinical data collected from patients is critical to medical practice. Collection of this data by clinicians is costly and time consuming. Automated systems based on questionnaires have successfully collected clinical information directly from patients. These systems, however, were not widely used because of high costs and failure to accommodate the diverse literacy and computer skills of individual patients. The goal of this project is to revolutionize automated data collection from patients by adapting both the questionnaire content and the content- presentation/response-entry interface to fit the skills of the patient. In the first half of this 3 year project, three parallel development efforts will create the Adaptive User Interface System (AUIS):
Aim 1 will create a generalized, Web-based program for assessing patient aptitudes in a clinical setting:
Aim 2 will define a generalized approach for developing sets of multimedia interfaces customized to user skill categories, that will be used to create the presentation of a health maintenance/risk assessment (HM/RA) questionnaire;
Aim 3 will develop the Web-based AUIS to present questionnaires through adapted user interfaces by modifying the interface presentations of an existing questionnaire-presenting system (created by this research team) to incorporate the patient assessment program and the customized multimedia interface sets developed in Aims 1 and 2, respectively. Finally (Aim 4), this project will test the hypotheses that the questionnaire presentation interface designed to collect data directly from patients by adapting to individual patient aptitudes will collect clinical data with greater completeness, higher quality, more efficiency and lower cost than traditional, paper-based methods. Specifically, completion of a HM/RA questionnaire will be evaluated in a randomized, crossover controlled trial comparing the customary paper-based questionnaire data collection method with an automated, Web-based questionnaire presented through an adaptive user interface. The trial will be performed at two primary care clinics representing diverse patient populations. The interface will be adapted primarily on patient literacy and secondarily on patient computer skills and familiarity with medical terminology. If proven effective, the AUIS will introduce a novel approach for enhancing data quality and lowering data collection costs. The AUIS can be widely disseminated because of its Web-based, generalized design and its utility in diverse patient populations. The research team is fully qualified to conduct all aspects of the proposed project because of its past work developing a Web-based, interactive clinical guideline server; refining literacy assessment tools; creating tailored patient data collection systems; and evaluating decision support systems in clinical practice.
| Sutherland, L A; Campbell, M; Ornstein, K et al. (2001) Development of an adaptive multimedia program to collect patient health data. Am J Prev Med 21:320-4 |
