The Preliminary Revision And Application Of The Evidence-Based Practice Questionnaire And The Developing Evidence-Based Practice Questionnaire

Objective1.To translate and report the psychometric characteristics of a self-report measure of practice, attitudes and knowledge/skills towards evidence-based practice (EBPQ); and of a tool for assessing a range of factors influencing the development of evidence-based practice(EBP) among clinical nurses(DEBPQ).2. To describe clinical nurses' practices, attitudes and knowledge/skills associated with evidence-based practice, and its related affected factors so as to provide a theoretical basis for the EBP training initiatives.Methods1. Brislin translation procedure was employed to translate the EBPQ questionnaire and the DEBPQ questionnaire. Exploratory factor analysis(EFA) was used to develop the intra-structure of the questionnaire and confirmatory factor analysis(CFA) verified the factor structure of a set of observed variables. The reliability analyis included the test-retest reliability, Cronbach'sa coefficient, split-half reliability. The validity analysis included content validity, construct validity.2. A cross-sectional survey using the Chinese version of the Evidence-based practice Questionnaire (EBPQ) and the Developing Evidence-Based Practice questionnaire (DEBPQ) was conducted from July 2009 to December 2009. A toal of 1621 nurses who worked at 12 GradeⅢhospitals in Hunan province were included by multi-stage stratified cluster random sampling method. Their Knowledge/skills, attitudes and pratice and its related affected factors on Evidence-based nursing were tested. All data were input into SPSS 13.0 software for statistical analysis. Descriptive statistics, t test, F test, correlation analysis and multiple linear regression analysis were employed.Results 1. Through item analysis and exploratory factor analysis, a 24-item Chinese version of EBPQ was developed, including 3 dimensions (Knowledge, Attitude and Belief, Behavior), and DEBPQ comprised 6 dimensions grouping 48 items.2. Reliability verification:(1) The test-retest reliability coefficients for the EBPQ questionnaire and dimensions were 0.86 and 0.78 to 0.88, respectively. The test-retest reliability coefficients for the DEBPQ questionnaire and dimensions were 0.89 and 0.65 to 0.90, respectively. (2) The split-half reliability coeffient for EBPQ dimensions ranged from 0.74 to 0.89, and 0.97 for the total questionnaire. The split-half reliability coeffient for DEBPQ dimensions was 0.65 to 0.87, and 0.94 for the total questionnaire. (3) The Cronbach's a coefficients for the EBPQ questionnaire and dimensions were 0.94, and 0.79 to 0.94, respectively. The Cronbach's a coefficients for the DEBPQ questionnaire and dimensions were 0.93, and 0.73 to 0.91, respectively. (4) The correlation coefficient of EBPQ between each item and dimensions were 0.59～0.81, the correlation coefficient between dimensions were 0.50～0.67, and the correlation coeffcient between dimensions and total questionnaire was 0.72～0.94, while DEBPQ were 0.53～0.86,0.18～0.48,0.64～0.81 respectively.3. Validity verification:(1) The content validity index of EBPQ and DEBPQ were 0.83 and 0.88 respectively. (2) Structure validity:For EBPQ, first-order three factors were extracted using EFA, which accounted for 58.25% of the accumulated variance; second-order one factor were extracted which accounted for 71.75%. CFA was carried out to assess the measurement of factor structure of the questionnaire, the results showed that the first-order 3 factor model fit indices and second-order 1 factor model fit indices were at or close to the reference value. For DEBPQ, sub-level factor analysis was used, and 4 dimensions of DEBPQ was revised into 6 dimensions, the accumulated variance at all levels were 47.58%,42.03%,68.71%,62.80%, separately. In the second-order factor analysis, the variance contribution rate was 46.88%. The result of the first-order factor analysis showed thatχ2/df was greater than 5.0, RMSEA(0.07) in the acceptable range, PNFI(0.71) and PCFI (0.76) were greater than 0.50, NFI(0.76),TLI(0.79),CFI(0.80),IFI(0.80),RFI (0.75) were all less than 0.90. The result of the second-order factor analysis showed that RMSEA (0.07), PNFI (0.72), PCFI (0.75) were in the acceptable range, the rest of model fit indices were not up to psychometric standard.4. The average EBPQ score of 1621 nurses was 4.24±0.79, which was significantly lower than those of the Registered nurses in the U.S.A (P<0.05). The mean DEBPQ score of this sample was 2.89±0.43. With regard to sources of practice knowledge, experiential sources such as information from clinical nursing experience and senior clinical nurses were ranked highest, which was similar to the other study abroad. Among the barrier dimension, mean score of the subjects was higher than UK nurses. The level of support and the mean score of knowledge/skills were lower than that in studies abroad.5. Multiple linear regression analysis showed that Evidence-based practice were corelated with the relative knowledge/skills, the level of support and the attitudes towards Evidence-based practice, which could account for about 47.43% of variance.Conclusion1. The Chinese version of EBPQ is a reliable and valid assessment of EBP. But the split-half reliability coefficients, and the test-retest reliability coefficients of Chinese version of DEBPQ are lower than accepted value which need to be modified.6-factor model of DEBPQ is not a single questionnaire, but it can assist in assessing the specific "evidencing" tendencies of any given group of nurses.2. The level of knowledge/skills, attitudes, practice and support towards EBP of clinical nurses worked in Grade-Ⅲhospitals in Hunan province is lower than those of American nurses while the barrier level is higher than overseas research.3. The evidence-based practice of nurses is mainly influenced by evidence-based knowledge/skills, attitudes and supporting level.