S22 AUTOMATED PROCESSING OF QUESTIONNAIRES AS A PREREQUISITE FOR LARGE STUDIES AND QUALITY MANAGEMENT

Abstract

Introduction: Patient self-assessment plays a significant rule in the monitoring of patients within clinical studies as well as a separate quality indicator. The self-assessment of function, disease activity and quality of life is known to have a predictive value in the disease progression of rheumatoid arthritis (RA) and other orthopaedic diseases. However, all questionnaires challenge the clinical infrastructure. The questionnaire administration and their processing require still considerable manpower and is a potential source for errors. We analysed the in-house processes, identified the essential requirements and explored possible electronic solution with the aim to reduce necessary manpower and failure sources.

Materials and methods: In a first step we defined a set of questionnaires we want to administer on a regular base. We then evaluated candidate systems with respect to data handling and to further statistic processing. Two years later we re-evaluated the system and possible alternatives. We then paid special attention to scanning features and data export options. Finally we performed reliability and handling tests and a first clinical trial.

Results: The standardized set for shoulder patients comprises 144 items per patients. The set was designed as a four-color print for automatic processing with Qualicare. Four large studies with a total of more than 300 evaluations were performed using Qualicare. Our reevaluation of the system revealed major problems with the line scanner, the data processing in the system and the data export into statistics programs. After intensive search we installed a new scanning system based on an OMR reader that detects regions of interests on the questionnaire (Remark Office). This system allows simple form generation with the PC, the use of bar-code and faster processing. Reliability was more than 0.95 and handling revealed no major problems. Since first trials were successful the new system became the standard for all questionnaires in our department. Discussion: The high amount of variables in patient self-assessment requires automated processing to save manpower and to avoid failures during manual processing. During a three-year period we identified scanning and export options as the key factors for long-term success. The new system (Remark Office) accomplishes both requirements and might serve as the base for large studies or regular quality control.