Study On Knowledge, Attitude And Behavior Scale Of Lifestyle Intervention In Diabetes High Risk Population

Object1. To develop "Knowledge, attitude and behavior scale of lifestyle intervention in diabetes high risk population" which will be suitable for Chinese rural areas.2. To evaluate the reliability, validity, acceptability and sensitivity of the scale.Subjects and MethodsThis study included item pool building, Delphi expert consultation and scale's performance evaluation. It established the initial item pool based on literature analysis. Then two rounds Delphi counseling was conducted to select items and determine the primary scale. After field survey data was collected, the performance of the scale was evaluated.1. This study established the item pool in compliance with the principle of objectivity, scientificity, importance and application. The three dimensions of knowledge, attitude and behavior were designed based on the characteristics of diabetes high risk population and the lifestyle intervention content. Items included were able to measure target events veritably and fully. The items were described easy to understand with maneuverability.2. The Delphi counseling selected 30 experts from the fields of clinical medicine, nursing, preventive medicine, health education and community diabetes prevention and control. Counseling papers were sent to the experts through e-mail. Two rounds of consultation were conducted. The active coefficient and authoritative coefficient of experts and the concentration and harmonious coefficients of experts'opinion were calculated based on the feedback. The primary scale was established after items being selected according to the mean importance score?4 and CV?0.25.3. A pilot survey was conducted in the Fang Zhuang Community Health Center in Beijing with a sample of 31 diabetes high risk population who were selected by convenience sampling questioned with primary scale. After pilot study, the construction of the scale was adjusted to a more reasonable structure. Shouyang County in Shanxi province was selected as the survey site.366 diabetes high risk population who were identified by an early baseline survey of diabetes prevention and control program were investigated by investigator face to face. The reliability of the scale was evaluated by internal consistency and split-half reliability. The validity of the scale was evaluated by content validity and construct validity. The acceptability of the scale was evaluated by the withdraw rate and the time finishing the scale. The sensitivity of the scale was evaluated by the difference between high score and low score group.Results1. The item pool of this study included knowledge, attitude and behavior dimensions. The knowledge dimension included 22 items of typical symptoms, risk factors, physical activity, diet control and weight control. The attitude dimension included 8 items of the necessity of health education and diabetes screening, the willing to conduct physical activity, diet control and weight control, and the responsibility of high risk people. The behavior dimension included 17 items of health education, physical activity, diet control, weight control and blood glucose monitoring.2. The withdraw rate of two Delphi consultation were 86.67% and 88.46%. The authoritative coefficient of experts was 0.79 and 0.87. After the Delphi consultation, the knowledge dimension included 12 items of basic knowledge, typical symptoms, risk factor, physical activity, diet control and weight control. The attitude dimension included 8 items of the necessary of health education and diabetes screening, the willing and self-efficacy of physical activity, diet control and weight control. The behavior dimension included 10 items of health education, physical activity, diet control, weight control and blood glucose monitoring.3. After the pilot study, the primary scale was adjusted to a more reasonable structure and was used as questionnaire in survey. Internal consistency and split-half reliability were above 0.75, which met the standard of 0.7. The correlation coefficient of items and the dimension they belong to were between 0.2521 and 0.8817, and the correlation coefficient of three dimensions and entire scale was between 0.6395and 0.9060, all with P<0.0001. Factor analysis on the scale extracted 8 common factors. The cumulative variance contribution rate was 65.23%, which met the standard of 40%. The factor loadings were between 0.3596 and 0.8971, all but one of which met the standard of 0.4. The communalities were between 0.2419 and 0.9060, all but three of which met the standard of 0.4. The withdraw rate of the scale was 100%. The completion time was 8.62±2.79 minutes. The high score group and low score group were located on the upper 27% and lower 27% score of the respondents. The difference of score in the entire scale and knowledge, attitude and behavior dimensions were statistically significant (P<0.0001).ConclusionThe reliability, validity, acceptability and responsibility of "Knowledge, attitude and behavior scale of lifestyle intervention in diabetes high risk population" are good in diabetes high risk population in rural areas. This scale can be used in surveys focused on knowledge, attitude and behavior condition of lifestyle intervention in this population.