OBJECTIVE: To develop a patient-reported outcome instrument for chronic gastrointestinal diseases and to select items by analyzing the data from clinical investigation. METHODS: The instrument was developed by establishing the item pool through reviewing the literature on subject of chronic gastrointestinal diseases, interviewing typical cases, focus group discussion and referring to related instruments, and then the processes of selecting and optimizing items, forming primary instrument and testing in a small sample pilot survey. A clinical investigation was conducted in 4 hospitals in Beijing. The data of 274 effective respondents with chronic gastrointestinal diseases were obtained and analyzed for selecting items using SPSS 16.0 software by applying methods such as Cronbach's α coefficient, correlation coefficient, t test, dispersion logistic regression and factor analysis. RESULTS: In the patient-reported outcome instrument, 35 items were selected from investigation. Among the 35 items, 2 items were deleted, and 6 items were modified for further investigation. CONCLUSION: The results of item selection provide a good reference for the primary instrument, which still needs further modification, and lots of work needed to be done, at the base of large sample size.
OBJECTIVE: To develop a patient-reported outcome instrument for chronic gastrointestinal diseases and to select items by analyzing the data from clinical investigation. METHODS: The instrument was developed by establishing the item pool through reviewing the literature on subject of chronic gastrointestinal diseases, interviewing typical cases, focus group discussion and referring to related instruments, and then the processes of selecting and optimizing items, forming primary instrument and testing in a small sample pilot survey. A clinical investigation was conducted in 4 hospitals in Beijing. The data of 274 effective respondents with chronic gastrointestinal diseases were obtained and analyzed for selecting items using SPSS 16.0 software by applying methods such as Cronbach's α coefficient, correlation coefficient, t test, dispersion logistic regression and factor analysis. RESULTS: In the patient-reported outcome instrument, 35 items were selected from investigation. Among the 35 items, 2 items were deleted, and 6 items were modified for further investigation. CONCLUSION: The results of item selection provide a good reference for the primary instrument, which still needs further modification, and lots of work needed to be done, at the base of large sample size.